diff --git a/CMakeLists.txt b/CMakeLists.txt index 32ed10f..993a40e 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -6,9 +6,7 @@ set(PROJECT_NAME project(${PROJECT_NAME}) -set(CMAKE_CXX_FLAGS - "${CMAKE_CXX_FLAGS} -std=c++14") - +set(CMAKE_CXX_STANDARD 14) if (WIN32) set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wa,-mbig-obj -O3") diff --git a/ext/json.hpp b/ext/json.hpp index 48d797d..2b387a8 100644 --- a/ext/json.hpp +++ b/ext/json.hpp @@ -1,11 +1,12 @@ /* __ _____ _____ _____ __| | __| | | | JSON for Modern C++ -| | |__ | | | | | | version 2.1.1 +| | |__ | | | | | | version 3.1.2 |_____|_____|_____|_|___| https://github.com/nlohmann/json Licensed under the MIT License . -Copyright (c) 2013-2017 Niels Lohmann . +SPDX-License-Identifier: MIT +Copyright (c) 2013-2018 Niels Lohmann . Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal @@ -29,39 +30,100 @@ SOFTWARE. #ifndef NLOHMANN_JSON_HPP #define NLOHMANN_JSON_HPP -#include // all_of, copy, fill, find, for_each, none_of, remove, reverse, transform -#include // array +#define NLOHMANN_JSON_VERSION_MAJOR 3 +#define NLOHMANN_JSON_VERSION_MINOR 1 +#define NLOHMANN_JSON_VERSION_PATCH 2 + +#include // all_of, find, for_each #include // assert #include // and, not, or -#include // lconv, localeconv -#include // isfinite, labs, ldexp, signbit #include // nullptr_t, ptrdiff_t, size_t -#include // int64_t, uint64_t -#include // abort, strtod, strtof, strtold, strtoul, strtoll, strtoull -#include // memcpy, strlen -#include // forward_list -#include // function, hash, less +#include // hash, less #include // initializer_list -#include // hex -#include // istream, ostream -#include // advance, begin, back_inserter, bidirectional_iterator_tag, distance, end, inserter, iterator, iterator_traits, next, random_access_iterator_tag, reverse_iterator -#include // numeric_limits -#include // locale -#include // map -#include // addressof, allocator, allocator_traits, unique_ptr +#include // istream, ostream +#include // iterator_traits, random_access_iterator_tag #include // accumulate -#include // stringstream -#include // getline, stoi, string, to_string -#include // add_pointer, conditional, decay, enable_if, false_type, integral_constant, is_arithmetic, is_base_of, is_const, is_constructible, is_convertible, is_default_constructible, is_enum, is_floating_point, is_integral, is_nothrow_move_assignable, is_nothrow_move_constructible, is_pointer, is_reference, is_same, is_scalar, is_signed, remove_const, remove_cv, remove_pointer, remove_reference, true_type, underlying_type -#include // declval, forward, make_pair, move, pair, swap +#include // string, stoi, to_string +#include // declval, forward, move, pair, swap + +// #include +#ifndef NLOHMANN_JSON_FWD_HPP +#define NLOHMANN_JSON_FWD_HPP + +#include // int64_t, uint64_t +#include // map +#include // allocator +#include // string #include // vector +/*! +@brief namespace for Niels Lohmann +@see https://github.com/nlohmann +@since version 1.0.0 +*/ +namespace nlohmann +{ +/*! +@brief default JSONSerializer template argument + +This serializer ignores the template arguments and uses ADL +([argument-dependent lookup](http://en.cppreference.com/w/cpp/language/adl)) +for serialization. +*/ +template +struct adl_serializer; + +template class ObjectType = + std::map, + template class ArrayType = std::vector, + class StringType = std::string, class BooleanType = bool, + class NumberIntegerType = std::int64_t, + class NumberUnsignedType = std::uint64_t, + class NumberFloatType = double, + template class AllocatorType = std::allocator, + template class JSONSerializer = + adl_serializer> +class basic_json; + +/*! +@brief JSON Pointer + +A JSON pointer defines a string syntax for identifying a specific value +within a JSON document. It can be used with functions `at` and +`operator[]`. Furthermore, JSON pointers are the base for JSON patches. + +@sa [RFC 6901](https://tools.ietf.org/html/rfc6901) + +@since version 2.0.0 +*/ +template +class json_pointer; + +/*! +@brief default JSON class + +This type is the default specialization of the @ref basic_json class which +uses the standard template types. + +@since version 1.0.0 +*/ +using json = basic_json<>; +} + +#endif + +// #include + + +// This file contains all internal macro definitions +// You MUST include macro_unscope.hpp at the end of json.hpp to undef all of them + // exclude unsupported compilers #if defined(__clang__) #if (__clang_major__ * 10000 + __clang_minor__ * 100 + __clang_patchlevel__) < 30400 #error "unsupported Clang version - see https://github.com/nlohmann/json#supported-compilers" #endif -#elif defined(__GNUC__) +#elif defined(__GNUC__) && !(defined(__ICC) || defined(__INTEL_COMPILER)) #if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) < 40900 #error "unsupported GCC version - see https://github.com/nlohmann/json#supported-compilers" #endif @@ -89,7 +151,7 @@ SOFTWARE. #endif // allow to disable exceptions -#if (defined(__cpp_exceptions) || defined(__EXCEPTIONS) || defined(_CPPUNWIND)) && not defined(JSON_NOEXCEPTION) +#if (defined(__cpp_exceptions) || defined(__EXCEPTIONS) || defined(_CPPUNWIND)) && !defined(JSON_NOEXCEPTION) #define JSON_THROW(exception) throw exception #define JSON_TRY try #define JSON_CATCH(exception) catch(exception) @@ -99,6 +161,20 @@ SOFTWARE. #define JSON_CATCH(exception) if(false) #endif +// override exception macros +#if defined(JSON_THROW_USER) + #undef JSON_THROW + #define JSON_THROW JSON_THROW_USER +#endif +#if defined(JSON_TRY_USER) + #undef JSON_TRY + #define JSON_TRY JSON_TRY_USER +#endif +#if defined(JSON_CATCH_USER) + #undef JSON_CATCH + #define JSON_CATCH JSON_CATCH_USER +#endif + // manual branch prediction #if defined(__clang__) || defined(__GNUC__) || defined(__GNUG__) #define JSON_LIKELY(x) __builtin_expect(!!(x), 1) @@ -108,24 +184,324 @@ SOFTWARE. #define JSON_UNLIKELY(x) x #endif +// C++ language standard detection +#if (defined(__cplusplus) && __cplusplus >= 201703L) || (defined(_HAS_CXX17) && _HAS_CXX17 == 1) // fix for issue #464 + #define JSON_HAS_CPP_17 + #define JSON_HAS_CPP_14 +#elif (defined(__cplusplus) && __cplusplus >= 201402L) || (defined(_HAS_CXX14) && _HAS_CXX14 == 1) + #define JSON_HAS_CPP_14 +#endif + +// Ugly macros to avoid uglier copy-paste when specializing basic_json. They +// may be removed in the future once the class is split. + +#define NLOHMANN_BASIC_JSON_TPL_DECLARATION \ + template class ObjectType, \ + template class ArrayType, \ + class StringType, class BooleanType, class NumberIntegerType, \ + class NumberUnsignedType, class NumberFloatType, \ + template class AllocatorType, \ + template class JSONSerializer> + +#define NLOHMANN_BASIC_JSON_TPL \ + basic_json + /*! -@brief namespace for Niels Lohmann -@see https://github.com/nlohmann -@since version 1.0.0 +@brief Helper to determine whether there's a key_type for T. + +This helper is used to tell associative containers apart from other containers +such as sequence containers. For instance, `std::map` passes the test as it +contains a `mapped_type`, whereas `std::vector` fails the test. + +@sa http://stackoverflow.com/a/7728728/266378 +@since version 1.0.0, overworked in version 2.0.6 */ +#define NLOHMANN_JSON_HAS_HELPER(type) \ + template struct has_##type { \ + private: \ + template \ + static int detect(U &&); \ + static void detect(...); \ + public: \ + static constexpr bool value = \ + std::is_integral()))>::value; \ + } + +// #include + + +#include // not +#include // size_t +#include // numeric_limits +#include // conditional, enable_if, false_type, integral_constant, is_constructible, is_integral, is_same, remove_cv, remove_reference, true_type +#include // declval + +// #include + +// #include + + namespace nlohmann { - /*! -@brief unnamed namespace with internal helper functions +@brief detail namespace with internal helper functions -This namespace collects some functions that could not be defined inside the -@ref basic_json class. +This namespace collects functions that should not be exposed, +implementations of some @ref basic_json methods, and meta-programming helpers. @since version 2.1.0 */ namespace detail { +///////////// +// helpers // +///////////// + +template struct is_basic_json : std::false_type {}; + +NLOHMANN_BASIC_JSON_TPL_DECLARATION +struct is_basic_json : std::true_type {}; + +// alias templates to reduce boilerplate +template +using enable_if_t = typename std::enable_if::type; + +template +using uncvref_t = typename std::remove_cv::type>::type; + +// implementation of C++14 index_sequence and affiliates +// source: https://stackoverflow.com/a/32223343 +template +struct index_sequence +{ + using type = index_sequence; + using value_type = std::size_t; + static constexpr std::size_t size() noexcept + { + return sizeof...(Ints); + } +}; + +template +struct merge_and_renumber; + +template +struct merge_and_renumber, index_sequence> + : index_sequence < I1..., (sizeof...(I1) + I2)... > {}; + +template +struct make_index_sequence + : merge_and_renumber < typename make_index_sequence < N / 2 >::type, + typename make_index_sequence < N - N / 2 >::type > {}; + +template<> struct make_index_sequence<0> : index_sequence<> {}; +template<> struct make_index_sequence<1> : index_sequence<0> {}; + +template +using index_sequence_for = make_index_sequence; + +/* +Implementation of two C++17 constructs: conjunction, negation. This is needed +to avoid evaluating all the traits in a condition + +For example: not std::is_same::value and has_value_type::value +will not compile when T = void (on MSVC at least). Whereas +conjunction>, has_value_type>::value will +stop evaluating if negation<...>::value == false + +Please note that those constructs must be used with caution, since symbols can +become very long quickly (which can slow down compilation and cause MSVC +internal compiler errors). Only use it when you have to (see example ahead). +*/ +template struct conjunction : std::true_type {}; +template struct conjunction : B1 {}; +template +struct conjunction : std::conditional, B1>::type {}; + +template struct negation : std::integral_constant {}; + +// dispatch utility (taken from ranges-v3) +template struct priority_tag : priority_tag < N - 1 > {}; +template<> struct priority_tag<0> {}; + +//////////////////////// +// has_/is_ functions // +//////////////////////// + +// source: https://stackoverflow.com/a/37193089/4116453 + +template +struct is_complete_type : std::false_type {}; + +template +struct is_complete_type : std::true_type {}; + +NLOHMANN_JSON_HAS_HELPER(mapped_type); +NLOHMANN_JSON_HAS_HELPER(key_type); +NLOHMANN_JSON_HAS_HELPER(value_type); +NLOHMANN_JSON_HAS_HELPER(iterator); + +template +struct is_compatible_object_type_impl : std::false_type {}; + +template +struct is_compatible_object_type_impl +{ + static constexpr auto value = + std::is_constructible::value and + std::is_constructible::value; +}; + +template +struct is_compatible_object_type +{ + static auto constexpr value = is_compatible_object_type_impl < + conjunction>, + has_mapped_type, + has_key_type>::value, + typename BasicJsonType::object_t, CompatibleObjectType >::value; +}; + +template +struct is_basic_json_nested_type +{ + static auto constexpr value = std::is_same::value or + std::is_same::value or + std::is_same::value or + std::is_same::value; +}; + +template +struct is_compatible_array_type +{ + static auto constexpr value = + conjunction>, + negation>, + negation>, + negation>, + has_value_type, + has_iterator>::value; +}; + +template +struct is_compatible_integer_type_impl : std::false_type {}; + +template +struct is_compatible_integer_type_impl +{ + // is there an assert somewhere on overflows? + using RealLimits = std::numeric_limits; + using CompatibleLimits = std::numeric_limits; + + static constexpr auto value = + std::is_constructible::value and + CompatibleLimits::is_integer and + RealLimits::is_signed == CompatibleLimits::is_signed; +}; + +template +struct is_compatible_integer_type +{ + static constexpr auto value = + is_compatible_integer_type_impl < + std::is_integral::value and + not std::is_same::value, + RealIntegerType, CompatibleNumberIntegerType > ::value; +}; + +// trait checking if JSONSerializer::from_json(json const&, udt&) exists +template +struct has_from_json +{ + private: + // also check the return type of from_json + template::from_json( + std::declval(), std::declval()))>::value>> + static int detect(U&&); + static void detect(...); + + public: + static constexpr bool value = std::is_integral>()))>::value; +}; + +// This trait checks if JSONSerializer::from_json(json const&) exists +// this overload is used for non-default-constructible user-defined-types +template +struct has_non_default_from_json +{ + private: + template < + typename U, + typename = enable_if_t::from_json(std::declval()))>::value >> + static int detect(U&&); + static void detect(...); + + public: + static constexpr bool value = std::is_integral>()))>::value; +}; + +// This trait checks if BasicJsonType::json_serializer::to_json exists +template +struct has_to_json +{ + private: + template::to_json( + std::declval(), std::declval()))> + static int detect(U&&); + static void detect(...); + + public: + static constexpr bool value = std::is_integral>()))>::value; +}; + +template +struct is_compatible_complete_type +{ + static constexpr bool value = + not std::is_base_of::value and + not is_basic_json::value and + not is_basic_json_nested_type::value and + has_to_json::value; +}; + +template +struct is_compatible_type + : conjunction, + is_compatible_complete_type> +{ +}; + +// taken from ranges-v3 +template +struct static_const +{ + static constexpr T value{}; +}; + +template +constexpr T static_const::value; +} +} + +// #include + + +#include // exception +#include // runtime_error +#include // to_string + +namespace nlohmann +{ +namespace detail +{ //////////////// // exceptions // //////////////// @@ -133,12 +509,28 @@ namespace detail /*! @brief general exception of the @ref basic_json class -Extension of std::exception objects with a member @a id for exception ids. +This class is an extension of `std::exception` objects with a member @a id for +exception ids. It is used as the base class for all exceptions thrown by the +@ref basic_json class. This class can hence be used as "wildcard" to catch +exceptions. +Subclasses: +- @ref parse_error for exceptions indicating a parse error +- @ref invalid_iterator for exceptions indicating errors with iterators +- @ref type_error for exceptions indicating executing a member function with + a wrong type +- @ref out_of_range for exceptions indicating access out of the defined range +- @ref other_error for exceptions indicating other library errors + +@internal @note To have nothrow-copy-constructible exceptions, we internally use - std::runtime_error which can cope with arbitrary-length error messages. + `std::runtime_error` which can cope with arbitrary-length error messages. Intermediate strings are built with static functions and then passed to the actual constructor. +@endinternal + +@liveexample{The following code shows how arbitrary library exceptions can be +caught.,exception} @since version 3.0.0 */ @@ -146,7 +538,7 @@ class exception : public std::exception { public: /// returns the explanatory string - virtual const char* what() const noexcept override + const char* what() const noexcept override { return m.what(); } @@ -155,13 +547,11 @@ class exception : public std::exception const int id; protected: - exception(int id_, const char* what_arg) - : id(id_), m(what_arg) - {} + exception(int id_, const char* what_arg) : id(id_), m(what_arg) {} - static std::string name(const std::string& ename, int id) + static std::string name(const std::string& ename, int id_) { - return "[json.exception." + ename + "." + std::to_string(id) + "] "; + return "[json.exception." + ename + "." + std::to_string(id_) + "] "; } private: @@ -172,36 +562,44 @@ class exception : public std::exception /*! @brief exception indicating a parse error -This excpetion is thrown by the library when a parse error occurs. Parse -errors can occur during the deserialization of JSON text as well as when -using JSON Patch. +This exception is thrown by the library when a parse error occurs. Parse errors +can occur during the deserialization of JSON text, CBOR, MessagePack, as well +as when using JSON Patch. Member @a byte holds the byte index of the last read character in the input file. -@note For an input with n bytes, 1 is the index of the first character - and n+1 is the index of the terminating null byte or the end of - file. This also holds true when reading a byte vector (CBOR or - MessagePack). - Exceptions have ids 1xx. -name / id | example massage | description +name / id | example message | description ------------------------------ | --------------- | ------------------------- json.exception.parse_error.101 | parse error at 2: unexpected end of input; expected string literal | This error indicates a syntax error while deserializing a JSON text. The error message describes that an unexpected token (character) was encountered, and the member @a byte indicates the error position. json.exception.parse_error.102 | parse error at 14: missing or wrong low surrogate | JSON uses the `\uxxxx` format to describe Unicode characters. Code points above above 0xFFFF are split into two `\uxxxx` entries ("surrogate pairs"). This error indicates that the surrogate pair is incomplete or contains an invalid code point. json.exception.parse_error.103 | parse error: code points above 0x10FFFF are invalid | Unicode supports code points up to 0x10FFFF. Code points above 0x10FFFF are invalid. json.exception.parse_error.104 | parse error: JSON patch must be an array of objects | [RFC 6902](https://tools.ietf.org/html/rfc6902) requires a JSON Patch document to be a JSON document that represents an array of objects. json.exception.parse_error.105 | parse error: operation must have string member 'op' | An operation of a JSON Patch document must contain exactly one "op" member, whose value indicates the operation to perform. Its value must be one of "add", "remove", "replace", "move", "copy", or "test"; other values are errors. -json.exception.parse_error.106 | parse error: array index '01' must not begin with '0' | An array index in a JSON Pointer ([RFC 6901](https://tools.ietf.org/html/rfc6901)) may be `0` or any number wihtout a leading `0`. +json.exception.parse_error.106 | parse error: array index '01' must not begin with '0' | An array index in a JSON Pointer ([RFC 6901](https://tools.ietf.org/html/rfc6901)) may be `0` or any number without a leading `0`. json.exception.parse_error.107 | parse error: JSON pointer must be empty or begin with '/' - was: 'foo' | A JSON Pointer must be a Unicode string containing a sequence of zero or more reference tokens, each prefixed by a `/` character. json.exception.parse_error.108 | parse error: escape character '~' must be followed with '0' or '1' | In a JSON Pointer, only `~0` and `~1` are valid escape sequences. json.exception.parse_error.109 | parse error: array index 'one' is not a number | A JSON Pointer array index must be a number. json.exception.parse_error.110 | parse error at 1: cannot read 2 bytes from vector | When parsing CBOR or MessagePack, the byte vector ends before the complete value has been read. -json.exception.parse_error.111 | parse error: bad input stream | Parsing CBOR or MessagePack from an input stream where the [`badbit` or `failbit`](http://en.cppreference.com/w/cpp/io/ios_base/iostate) is set. -json.exception.parse_error.112 | parse error at 1: error reading CBOR; last byte: 0xf8 | Not all types of CBOR or MessagePack are supported. This exception occurs if an unsupported byte was read. +json.exception.parse_error.112 | parse error at 1: error reading CBOR; last byte: 0xF8 | Not all types of CBOR or MessagePack are supported. This exception occurs if an unsupported byte was read. json.exception.parse_error.113 | parse error at 2: expected a CBOR string; last byte: 0x98 | While parsing a map key, a value that is not a string has been read. +@note For an input with n bytes, 1 is the index of the first character and n+1 + is the index of the terminating null byte or the end of file. This also + holds true when reading a byte vector (CBOR or MessagePack). + +@liveexample{The following code shows how a `parse_error` exception can be +caught.,parse_error} + +@sa @ref exception for the base class of the library exceptions +@sa @ref invalid_iterator for exceptions indicating errors with iterators +@sa @ref type_error for exceptions indicating executing a member function with + a wrong type +@sa @ref out_of_range for exceptions indicating access out of the defined range +@sa @ref other_error for exceptions indicating other library errors + @since version 3.0.0 */ class parse_error : public exception @@ -209,18 +607,18 @@ class parse_error : public exception public: /*! @brief create a parse error exception - @param[in] id the id of the exception - @param[in] byte_ the byte index where the error occured (or 0 if - the position cannot be determined) - @param[in] what_arg the explanatory string + @param[in] id_ the id of the exception + @param[in] byte_ the byte index where the error occurred (or 0 if the + position cannot be determined) + @param[in] what_arg the explanatory string @return parse_error object */ - static parse_error create(int id, size_t byte_, const std::string& what_arg) + static parse_error create(int id_, std::size_t byte_, const std::string& what_arg) { - std::string w = exception::name("parse_error", id) + "parse error" + + std::string w = exception::name("parse_error", id_) + "parse error" + (byte_ != 0 ? (" at " + std::to_string(byte_)) : "") + ": " + what_arg; - return parse_error(id, byte_, w.c_str()); + return parse_error(id_, byte_, w.c_str()); } /*! @@ -228,25 +626,26 @@ class parse_error : public exception The byte index of the last read character in the input file. - @note For an input with n bytes, 1 is the index of the first character - and n+1 is the index of the terminating null byte or the end of - file. This also holds true when reading a byte vector (CBOR or - MessagePack). + @note For an input with n bytes, 1 is the index of the first character and + n+1 is the index of the terminating null byte or the end of file. + This also holds true when reading a byte vector (CBOR or MessagePack). */ - const size_t byte; + const std::size_t byte; private: - parse_error(int id_, size_t byte_, const char* what_arg) - : exception(id_, what_arg), byte(byte_) - {} + parse_error(int id_, std::size_t byte_, const char* what_arg) + : exception(id_, what_arg), byte(byte_) {} }; /*! @brief exception indicating errors with iterators +This exception is thrown if iterators passed to a library function do not match +the expected semantics. + Exceptions have ids 2xx. -name / id | example massage | description +name / id | example message | description ----------------------------------- | --------------- | ------------------------- json.exception.invalid_iterator.201 | iterators are not compatible | The iterators passed to constructor @ref basic_json(InputIT first, InputIT last) are not compatible, meaning they do not belong to the same container. Therefore, the range (@a first, @a last) is invalid. json.exception.invalid_iterator.202 | iterator does not fit current value | In an erase or insert function, the passed iterator @a pos does not belong to the JSON value for which the function was called. It hence does not define a valid position for the deletion/insertion. @@ -260,32 +659,44 @@ json.exception.invalid_iterator.209 | cannot use offsets with object iterators | json.exception.invalid_iterator.210 | iterators do not fit | The iterator range passed to the insert function are not compatible, meaning they do not belong to the same container. Therefore, the range (@a first, @a last) is invalid. json.exception.invalid_iterator.211 | passed iterators may not belong to container | The iterator range passed to the insert function must not be a subrange of the container to insert to. json.exception.invalid_iterator.212 | cannot compare iterators of different containers | When two iterators are compared, they must belong to the same container. -json.exception.invalid_iterator.213 | cannot compare order of object iterators | The order of object iterators cannot be compated, because JSON objects are unordered. +json.exception.invalid_iterator.213 | cannot compare order of object iterators | The order of object iterators cannot be compared, because JSON objects are unordered. json.exception.invalid_iterator.214 | cannot get value | Cannot get value for iterator: Either the iterator belongs to a null value or it is an iterator to a primitive type (number, boolean, or string), but the iterator is different to @ref begin(). +@liveexample{The following code shows how an `invalid_iterator` exception can be +caught.,invalid_iterator} + +@sa @ref exception for the base class of the library exceptions +@sa @ref parse_error for exceptions indicating a parse error +@sa @ref type_error for exceptions indicating executing a member function with + a wrong type +@sa @ref out_of_range for exceptions indicating access out of the defined range +@sa @ref other_error for exceptions indicating other library errors + @since version 3.0.0 */ class invalid_iterator : public exception { public: - static invalid_iterator create(int id, const std::string& what_arg) + static invalid_iterator create(int id_, const std::string& what_arg) { - std::string w = exception::name("invalid_iterator", id) + what_arg; - return invalid_iterator(id, w.c_str()); + std::string w = exception::name("invalid_iterator", id_) + what_arg; + return invalid_iterator(id_, w.c_str()); } private: invalid_iterator(int id_, const char* what_arg) - : exception(id_, what_arg) - {} + : exception(id_, what_arg) {} }; /*! @brief exception indicating executing a member function with a wrong type +This exception is thrown in case of a type error; that is, a library function is +executed on a JSON value whose type does not match the expected semantics. + Exceptions have ids 3xx. -name / id | example massage | description +name / id | example message | description ----------------------------- | --------------- | ------------------------- json.exception.type_error.301 | cannot create object from initializer list | To create an object from an initializer list, the initializer list must consist only of a list of pairs whose first element is a string. When this constraint is violated, an array is created instead. json.exception.type_error.302 | type must be object, but is array | During implicit or explicit value conversion, the JSON type must be compatible to the target type. For instance, a JSON string can only be converted into string types, but not into numbers or boolean types. @@ -298,33 +709,46 @@ json.exception.type_error.308 | cannot use push_back() with string | The @ref pu json.exception.type_error.309 | cannot use insert() with | The @ref insert() member functions can only be executed for certain JSON types. json.exception.type_error.310 | cannot use swap() with number | The @ref swap() member functions can only be executed for certain JSON types. json.exception.type_error.311 | cannot use emplace_back() with string | The @ref emplace_back() member function can only be executed for certain JSON types. +json.exception.type_error.312 | cannot use update() with string | The @ref update() member functions can only be executed for certain JSON types. json.exception.type_error.313 | invalid value to unflatten | The @ref unflatten function converts an object whose keys are JSON Pointers back into an arbitrary nested JSON value. The JSON Pointers must not overlap, because then the resulting value would not be well defined. json.exception.type_error.314 | only objects can be unflattened | The @ref unflatten function only works for an object whose keys are JSON Pointers. json.exception.type_error.315 | values in object must be primitive | The @ref unflatten function only works for an object whose keys are JSON Pointers and whose values are primitive. +json.exception.type_error.316 | invalid UTF-8 byte at index 10: 0x7E | The @ref dump function only works with UTF-8 encoded strings; that is, if you assign a `std::string` to a JSON value, make sure it is UTF-8 encoded. | + +@liveexample{The following code shows how a `type_error` exception can be +caught.,type_error} + +@sa @ref exception for the base class of the library exceptions +@sa @ref parse_error for exceptions indicating a parse error +@sa @ref invalid_iterator for exceptions indicating errors with iterators +@sa @ref out_of_range for exceptions indicating access out of the defined range +@sa @ref other_error for exceptions indicating other library errors @since version 3.0.0 */ class type_error : public exception { public: - static type_error create(int id, const std::string& what_arg) + static type_error create(int id_, const std::string& what_arg) { - std::string w = exception::name("type_error", id) + what_arg; - return type_error(id, w.c_str()); + std::string w = exception::name("type_error", id_) + what_arg; + return type_error(id_, w.c_str()); } private: - type_error(int id_, const char* what_arg) - : exception(id_, what_arg) - {} + type_error(int id_, const char* what_arg) : exception(id_, what_arg) {} }; /*! @brief exception indicating access out of the defined range +This exception is thrown in case a library function is called on an input +parameter that exceeds the expected range, for instance in case of array +indices or nonexisting object keys. + Exceptions have ids 4xx. -name / id | example massage | description +name / id | example message | description ------------------------------- | --------------- | ------------------------- json.exception.out_of_range.401 | array index 3 is out of range | The provided array index @a i is larger than @a size-1. json.exception.out_of_range.402 | array index '-' (3) is out of range | The special array index `-` in a JSON Pointer never describes a valid element of the array, but the index past the end. That is, it can only be used to add elements at this position, but not to read it. @@ -332,52 +756,85 @@ json.exception.out_of_range.403 | key 'foo' not found | The provided key was not json.exception.out_of_range.404 | unresolved reference token 'foo' | A reference token in a JSON Pointer could not be resolved. json.exception.out_of_range.405 | JSON pointer has no parent | The JSON Patch operations 'remove' and 'add' can not be applied to the root element of the JSON value. json.exception.out_of_range.406 | number overflow parsing '10E1000' | A parsed number could not be stored as without changing it to NaN or INF. +json.exception.out_of_range.407 | number overflow serializing '9223372036854775808' | UBJSON only supports integers numbers up to 9223372036854775807. | +json.exception.out_of_range.408 | excessive array size: 8658170730974374167 | The size (following `#`) of an UBJSON array or object exceeds the maximal capacity. | + +@liveexample{The following code shows how an `out_of_range` exception can be +caught.,out_of_range} + +@sa @ref exception for the base class of the library exceptions +@sa @ref parse_error for exceptions indicating a parse error +@sa @ref invalid_iterator for exceptions indicating errors with iterators +@sa @ref type_error for exceptions indicating executing a member function with + a wrong type +@sa @ref other_error for exceptions indicating other library errors @since version 3.0.0 */ class out_of_range : public exception { public: - static out_of_range create(int id, const std::string& what_arg) + static out_of_range create(int id_, const std::string& what_arg) { - std::string w = exception::name("out_of_range", id) + what_arg; - return out_of_range(id, w.c_str()); + std::string w = exception::name("out_of_range", id_) + what_arg; + return out_of_range(id_, w.c_str()); } private: - out_of_range(int id_, const char* what_arg) - : exception(id_, what_arg) - {} + out_of_range(int id_, const char* what_arg) : exception(id_, what_arg) {} }; /*! -@brief exception indicating other errors +@brief exception indicating other library errors + +This exception is thrown in case of errors that cannot be classified with the +other exception types. Exceptions have ids 5xx. -name / id | example massage | description +name / id | example message | description ------------------------------ | --------------- | ------------------------- json.exception.other_error.501 | unsuccessful: {"op":"test","path":"/baz", "value":"bar"} | A JSON Patch operation 'test' failed. The unsuccessful operation is also printed. +@sa @ref exception for the base class of the library exceptions +@sa @ref parse_error for exceptions indicating a parse error +@sa @ref invalid_iterator for exceptions indicating errors with iterators +@sa @ref type_error for exceptions indicating executing a member function with + a wrong type +@sa @ref out_of_range for exceptions indicating access out of the defined range + +@liveexample{The following code shows how an `other_error` exception can be +caught.,other_error} + @since version 3.0.0 */ class other_error : public exception { public: - static other_error create(int id, const std::string& what_arg) + static other_error create(int id_, const std::string& what_arg) { - std::string w = exception::name("other_error", id) + what_arg; - return other_error(id, w.c_str()); + std::string w = exception::name("other_error", id_) + what_arg; + return other_error(id_, w.c_str()); } private: - other_error(int id_, const char* what_arg) - : exception(id_, what_arg) - {} + other_error(int id_, const char* what_arg) : exception(id_, what_arg) {} }; +} +} + +// #include +#include // array +#include // and +#include // size_t +#include // uint8_t +namespace nlohmann +{ +namespace detail +{ /////////////////////////// // JSON type enumeration // /////////////////////////// @@ -406,17 +863,17 @@ value with the default value for a given type @since version 1.0.0 */ -enum class value_t : uint8_t +enum class value_t : std::uint8_t { - null, ///< null value - object, ///< object (unordered set of name/value pairs) - array, ///< array (ordered collection of values) - string, ///< string value - boolean, ///< boolean value - number_integer, ///< number value (signed integer) - number_unsigned, ///< number value (unsigned integer) - number_float, ///< number value (floating-point) - discarded ///< discarded by the the parser callback function + null, ///< null value + object, ///< object (unordered set of name/value pairs) + array, ///< array (ordered collection of values) + string, ///< string value + boolean, ///< boolean value + number_integer, ///< number value (signed integer) + number_unsigned, ///< number value (unsigned integer) + number_float, ///< number value (floating-point) + discarded ///< discarded by the the parser callback function }; /*! @@ -425,70 +882,392 @@ enum class value_t : uint8_t Returns an ordering that is similar to Python: - order: null < boolean < number < object < array < string - furthermore, each type is not smaller than itself +- discarded values are not comparable @since version 1.0.0 */ inline bool operator<(const value_t lhs, const value_t rhs) noexcept { - static constexpr std::array order = {{ - 0, // null - 3, // object - 4, // array - 5, // string - 1, // boolean - 2, // integer - 2, // unsigned - 2, // float + static constexpr std::array order = {{ + 0 /* null */, 3 /* object */, 4 /* array */, 5 /* string */, + 1 /* boolean */, 2 /* integer */, 2 /* unsigned */, 2 /* float */ } }; - // discarded values are not comparable - if (lhs == value_t::discarded or rhs == value_t::discarded) - { - return false; - } - - return order[static_cast(lhs)] < - order[static_cast(rhs)]; + const auto l_index = static_cast(lhs); + const auto r_index = static_cast(rhs); + return l_index < order.size() and r_index < order.size() and order[l_index] < order[r_index]; +} +} } - -///////////// -// helpers // -///////////// - -// alias templates to reduce boilerplate -template -using enable_if_t = typename std::enable_if::type; - -template -using uncvref_t = typename std::remove_cv::type>::type; - -/* -Implementation of two C++17 constructs: conjunction, negation. This is needed -to avoid evaluating all the traits in a condition - -For example: not std::is_same::value and has_value_type::value -will not compile when T = void (on MSVC at least). Whereas -conjunction>, has_value_type>::value will -stop evaluating if negation<...>::value == false - -Please note that those constructs must be used with caution, since symbols can -become very long quickly (which can slow down compilation and cause MSVC -internal compiler errors). Only use it when you have to (see example ahead). -*/ -template struct conjunction : std::true_type {}; -template struct conjunction : B1 {}; -template -struct conjunction : std::conditional, B1>::type {}; - -template struct negation : std::integral_constant < bool, !B::value > {}; - -// dispatch utility (taken from ranges-v3) -template struct priority_tag : priority_tag < N - 1 > {}; -template<> struct priority_tag<0> {}; +// #include +#include // transform +#include // array +#include // and, not +#include // forward_list +#include // inserter, front_inserter, end +#include // map +#include // string +#include // tuple, make_tuple +#include // is_arithmetic, is_same, is_enum, underlying_type, is_convertible +#include // pair, declval +#include // valarray + +// #include + +// #include + +// #include + +// #include + + +namespace nlohmann +{ +namespace detail +{ +// overloads for basic_json template parameters +template::value and + not std::is_same::value, + int> = 0> +void get_arithmetic_value(const BasicJsonType& j, ArithmeticType& val) +{ + switch (static_cast(j)) + { + case value_t::number_unsigned: + { + val = static_cast(*j.template get_ptr()); + break; + } + case value_t::number_integer: + { + val = static_cast(*j.template get_ptr()); + break; + } + case value_t::number_float: + { + val = static_cast(*j.template get_ptr()); + break; + } + + default: + JSON_THROW(type_error::create(302, "type must be number, but is " + std::string(j.type_name()))); + } +} + +template +void from_json(const BasicJsonType& j, typename BasicJsonType::boolean_t& b) +{ + if (JSON_UNLIKELY(not j.is_boolean())) + { + JSON_THROW(type_error::create(302, "type must be boolean, but is " + std::string(j.type_name()))); + } + b = *j.template get_ptr(); +} + +template +void from_json(const BasicJsonType& j, typename BasicJsonType::string_t& s) +{ + if (JSON_UNLIKELY(not j.is_string())) + { + JSON_THROW(type_error::create(302, "type must be string, but is " + std::string(j.type_name()))); + } + s = *j.template get_ptr(); +} + +template +void from_json(const BasicJsonType& j, typename BasicJsonType::number_float_t& val) +{ + get_arithmetic_value(j, val); +} + +template +void from_json(const BasicJsonType& j, typename BasicJsonType::number_unsigned_t& val) +{ + get_arithmetic_value(j, val); +} + +template +void from_json(const BasicJsonType& j, typename BasicJsonType::number_integer_t& val) +{ + get_arithmetic_value(j, val); +} + +template::value, int> = 0> +void from_json(const BasicJsonType& j, EnumType& e) +{ + typename std::underlying_type::type val; + get_arithmetic_value(j, val); + e = static_cast(val); +} + +template +void from_json(const BasicJsonType& j, typename BasicJsonType::array_t& arr) +{ + if (JSON_UNLIKELY(not j.is_array())) + { + JSON_THROW(type_error::create(302, "type must be array, but is " + std::string(j.type_name()))); + } + arr = *j.template get_ptr(); +} + +// forward_list doesn't have an insert method +template::value, int> = 0> +void from_json(const BasicJsonType& j, std::forward_list& l) +{ + if (JSON_UNLIKELY(not j.is_array())) + { + JSON_THROW(type_error::create(302, "type must be array, but is " + std::string(j.type_name()))); + } + std::transform(j.rbegin(), j.rend(), + std::front_inserter(l), [](const BasicJsonType & i) + { + return i.template get(); + }); +} + +// valarray doesn't have an insert method +template::value, int> = 0> +void from_json(const BasicJsonType& j, std::valarray& l) +{ + if (JSON_UNLIKELY(not j.is_array())) + { + JSON_THROW(type_error::create(302, "type must be array, but is " + std::string(j.type_name()))); + } + l.resize(j.size()); + std::copy(j.m_value.array->begin(), j.m_value.array->end(), std::begin(l)); +} + +template +void from_json_array_impl(const BasicJsonType& j, CompatibleArrayType& arr, priority_tag<0> /*unused*/) +{ + using std::end; + + std::transform(j.begin(), j.end(), + std::inserter(arr, end(arr)), [](const BasicJsonType & i) + { + // get() returns *this, this won't call a from_json + // method when value_type is BasicJsonType + return i.template get(); + }); +} + +template +auto from_json_array_impl(const BasicJsonType& j, CompatibleArrayType& arr, priority_tag<1> /*unused*/) +-> decltype( + arr.reserve(std::declval()), + void()) +{ + using std::end; + + arr.reserve(j.size()); + std::transform(j.begin(), j.end(), + std::inserter(arr, end(arr)), [](const BasicJsonType & i) + { + // get() returns *this, this won't call a from_json + // method when value_type is BasicJsonType + return i.template get(); + }); +} + +template +void from_json_array_impl(const BasicJsonType& j, std::array& arr, priority_tag<2> /*unused*/) +{ + for (std::size_t i = 0; i < N; ++i) + { + arr[i] = j.at(i).template get(); + } +} + +template < + typename BasicJsonType, typename CompatibleArrayType, + enable_if_t < + is_compatible_array_type::value and + not std::is_same::value and + std::is_constructible < + BasicJsonType, typename CompatibleArrayType::value_type >::value, + int > = 0 > +void from_json(const BasicJsonType& j, CompatibleArrayType& arr) +{ + if (JSON_UNLIKELY(not j.is_array())) + { + JSON_THROW(type_error::create(302, "type must be array, but is " + + std::string(j.type_name()))); + } + + from_json_array_impl(j, arr, priority_tag<2> {}); +} + +template::value, int> = 0> +void from_json(const BasicJsonType& j, CompatibleObjectType& obj) +{ + if (JSON_UNLIKELY(not j.is_object())) + { + JSON_THROW(type_error::create(302, "type must be object, but is " + std::string(j.type_name()))); + } + + auto inner_object = j.template get_ptr(); + using value_type = typename CompatibleObjectType::value_type; + std::transform( + inner_object->begin(), inner_object->end(), + std::inserter(obj, obj.begin()), + [](typename BasicJsonType::object_t::value_type const & p) + { + return value_type(p.first, p.second.template get()); + }); +} + +// overload for arithmetic types, not chosen for basic_json template arguments +// (BooleanType, etc..); note: Is it really necessary to provide explicit +// overloads for boolean_t etc. in case of a custom BooleanType which is not +// an arithmetic type? +template::value and + not std::is_same::value and + not std::is_same::value and + not std::is_same::value and + not std::is_same::value, + int> = 0> +void from_json(const BasicJsonType& j, ArithmeticType& val) +{ + switch (static_cast(j)) + { + case value_t::number_unsigned: + { + val = static_cast(*j.template get_ptr()); + break; + } + case value_t::number_integer: + { + val = static_cast(*j.template get_ptr()); + break; + } + case value_t::number_float: + { + val = static_cast(*j.template get_ptr()); + break; + } + case value_t::boolean: + { + val = static_cast(*j.template get_ptr()); + break; + } + + default: + JSON_THROW(type_error::create(302, "type must be number, but is " + std::string(j.type_name()))); + } +} + +template +void from_json(const BasicJsonType& j, std::pair& p) +{ + p = {j.at(0).template get(), j.at(1).template get()}; +} + +template +void from_json_tuple_impl(const BasicJsonType& j, Tuple& t, index_sequence) +{ + t = std::make_tuple(j.at(Idx).template get::type>()...); +} + +template +void from_json(const BasicJsonType& j, std::tuple& t) +{ + from_json_tuple_impl(j, t, index_sequence_for {}); +} + +template ::value>> +void from_json(const BasicJsonType& j, std::map& m) +{ + if (JSON_UNLIKELY(not j.is_array())) + { + JSON_THROW(type_error::create(302, "type must be array, but is " + std::string(j.type_name()))); + } + for (const auto& p : j) + { + if (JSON_UNLIKELY(not p.is_array())) + { + JSON_THROW(type_error::create(302, "type must be array, but is " + std::string(p.type_name()))); + } + m.emplace(p.at(0).template get(), p.at(1).template get()); + } +} + +struct from_json_fn +{ + private: + template + auto call(const BasicJsonType& j, T& val, priority_tag<1> /*unused*/) const + noexcept(noexcept(from_json(j, val))) + -> decltype(from_json(j, val), void()) + { + return from_json(j, val); + } + + template + void call(const BasicJsonType& /*unused*/, T& /*unused*/, priority_tag<0> /*unused*/) const noexcept + { + static_assert(sizeof(BasicJsonType) == 0, + "could not find from_json() method in T's namespace"); +#ifdef _MSC_VER + // MSVC does not show a stacktrace for the above assert + using decayed = uncvref_t; + static_assert(sizeof(typename decayed::force_msvc_stacktrace) == 0, + "forcing MSVC stacktrace to show which T we're talking about."); +#endif + } + + public: + template + void operator()(const BasicJsonType& j, T& val) const + noexcept(noexcept(std::declval().call(j, val, priority_tag<1> {}))) + { + return call(j, val, priority_tag<1> {}); + } +}; +} + +/// namespace to hold default `from_json` function +/// to see why this is required: +/// http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2015/n4381.html +namespace +{ +constexpr const auto& from_json = detail::static_const::value; +} +} + +// #include + + +#include // or, and, not +#include // begin, end +#include // tuple, get +#include // is_same, is_constructible, is_floating_point, is_enum, underlying_type +#include // move, forward, declval, pair +#include // valarray +#include // vector + +// #include + +// #include + + +namespace nlohmann +{ +namespace detail +{ ////////////////// // constructors // ////////////////// @@ -517,6 +1296,14 @@ struct external_constructor j.m_value = s; j.assert_invariant(); } + + template + static void construct(BasicJsonType& j, typename BasicJsonType::string_t&& s) + { + j.m_type = value_t::string; + j.m_value = std::move(s); + j.assert_invariant(); + } }; template<> @@ -566,9 +1353,16 @@ struct external_constructor j.assert_invariant(); } + template + static void construct(BasicJsonType& j, typename BasicJsonType::array_t&& arr) + { + j.m_type = value_t::array; + j.m_value = std::move(arr); + j.assert_invariant(); + } + template::value, + enable_if_t::value, int> = 0> static void construct(BasicJsonType& j, const CompatibleArrayType& arr) { @@ -585,12 +1379,23 @@ struct external_constructor j.m_type = value_t::array; j.m_value = value_t::array; j.m_value.array->reserve(arr.size()); - for (bool x : arr) + for (const bool x : arr) { j.m_value.array->push_back(x); } j.assert_invariant(); } + + template::value, int> = 0> + static void construct(BasicJsonType& j, const std::valarray& arr) + { + j.m_type = value_t::array; + j.m_value = value_t::array; + j.m_value.array->resize(arr.size()); + std::copy(std::begin(arr), std::end(arr), j.m_value.array->begin()); + j.assert_invariant(); + } }; template<> @@ -604,10 +1409,16 @@ struct external_constructor j.assert_invariant(); } + template + static void construct(BasicJsonType& j, typename BasicJsonType::object_t&& obj) + { + j.m_type = value_t::object; + j.m_value = std::move(obj); + j.assert_invariant(); + } + template::value, - int> = 0> + enable_if_t::value, int> = 0> static void construct(BasicJsonType& j, const CompatibleObjectType& obj) { using std::begin; @@ -619,184 +1430,30 @@ struct external_constructor } }; - -//////////////////////// -// has_/is_ functions // -//////////////////////// - -/*! -@brief Helper to determine whether there's a key_type for T. - -This helper is used to tell associative containers apart from other containers -such as sequence containers. For instance, `std::map` passes the test as it -contains a `mapped_type`, whereas `std::vector` fails the test. - -@sa http://stackoverflow.com/a/7728728/266378 -@since version 1.0.0, overworked in version 2.0.6 -*/ -#define NLOHMANN_JSON_HAS_HELPER(type) \ - template struct has_##type { \ - private: \ - template \ - static int detect(U &&); \ - static void detect(...); \ - public: \ - static constexpr bool value = \ - std::is_integral()))>::value; \ - } - -NLOHMANN_JSON_HAS_HELPER(mapped_type); -NLOHMANN_JSON_HAS_HELPER(key_type); -NLOHMANN_JSON_HAS_HELPER(value_type); -NLOHMANN_JSON_HAS_HELPER(iterator); - -#undef NLOHMANN_JSON_HAS_HELPER - - -template -struct is_compatible_object_type_impl : std::false_type {}; - -template -struct is_compatible_object_type_impl -{ - static constexpr auto value = - std::is_constructible::value and - std::is_constructible::value; -}; - -template -struct is_compatible_object_type -{ - static auto constexpr value = is_compatible_object_type_impl < - conjunction>, - has_mapped_type, - has_key_type>::value, - typename BasicJsonType::object_t, CompatibleObjectType >::value; -}; - -template -struct is_basic_json_nested_type -{ - static auto constexpr value = std::is_same::value or - std::is_same::value or - std::is_same::value or - std::is_same::value or - std::is_same::value; -}; - -template -struct is_compatible_array_type -{ - static auto constexpr value = - conjunction>, - negation>, - negation>, - negation>, - has_value_type, - has_iterator>::value; -}; - -template -struct is_compatible_integer_type_impl : std::false_type {}; - -template -struct is_compatible_integer_type_impl -{ - // is there an assert somewhere on overflows? - using RealLimits = std::numeric_limits; - using CompatibleLimits = std::numeric_limits; - - static constexpr auto value = - std::is_constructible::value and - CompatibleLimits::is_integer and - RealLimits::is_signed == CompatibleLimits::is_signed; -}; - -template -struct is_compatible_integer_type -{ - static constexpr auto value = - is_compatible_integer_type_impl < - std::is_integral::value and - not std::is_same::value, - RealIntegerType, CompatibleNumberIntegerType > ::value; -}; - - -// trait checking if JSONSerializer::from_json(json const&, udt&) exists -template -struct has_from_json -{ - private: - // also check the return type of from_json - template::from_json( - std::declval(), std::declval()))>::value>> - static int detect(U&&); - static void detect(...); - - public: - static constexpr bool value = std::is_integral>()))>::value; -}; - -// This trait checks if JSONSerializer::from_json(json const&) exists -// this overload is used for non-default-constructible user-defined-types -template -struct has_non_default_from_json -{ - private: - template < - typename U, - typename = enable_if_t::from_json(std::declval()))>::value >> - static int detect(U&&); - static void detect(...); - - public: - static constexpr bool value = std::is_integral>()))>::value; -}; - -// This trait checks if BasicJsonType::json_serializer::to_json exists -template -struct has_to_json -{ - private: - template::to_json( - std::declval(), std::declval()))> - static int detect(U&&); - static void detect(...); - - public: - static constexpr bool value = std::is_integral>()))>::value; -}; - - ///////////// // to_json // ///////////// -template::value, int> = 0> +template::value, int> = 0> void to_json(BasicJsonType& j, T b) noexcept { external_constructor::construct(j, b); } template::value, int> = 0> + enable_if_t::value, int> = 0> void to_json(BasicJsonType& j, const CompatibleString& s) { external_constructor::construct(j, s); } +template +void to_json(BasicJsonType& j, typename BasicJsonType::string_t&& s) +{ + external_constructor::construct(j, std::move(s)); +} + template::value, int> = 0> void to_json(BasicJsonType& j, FloatType val) noexcept @@ -804,19 +1461,15 @@ void to_json(BasicJsonType& j, FloatType val) noexcept external_constructor::construct(j, static_cast(val)); } -template < - typename BasicJsonType, typename CompatibleNumberUnsignedType, - enable_if_t::value, int> = 0 > +template::value, int> = 0> void to_json(BasicJsonType& j, CompatibleNumberUnsignedType val) noexcept { external_constructor::construct(j, static_cast(val)); } -template < - typename BasicJsonType, typename CompatibleNumberIntegerType, - enable_if_t::value, int> = 0 > +template::value, int> = 0> void to_json(BasicJsonType& j, CompatibleNumberIntegerType val) noexcept { external_constructor::construct(j, static_cast(val)); @@ -836,271 +1489,88 @@ void to_json(BasicJsonType& j, const std::vector& e) external_constructor::construct(j, e); } -template < - typename BasicJsonType, typename CompatibleArrayType, - enable_if_t < - is_compatible_array_type::value or - std::is_same::value, - int > = 0 > -void to_json(BasicJsonType& j, const CompatibleArrayType& arr) +template::value or + std::is_same::value, + int> = 0> +void to_json(BasicJsonType& j, const CompatibleArrayType& arr) { external_constructor::construct(j, arr); } -template < - typename BasicJsonType, typename CompatibleObjectType, - enable_if_t::value, - int> = 0 > -void to_json(BasicJsonType& j, const CompatibleObjectType& arr) +template::value, int> = 0> +void to_json(BasicJsonType& j, std::valarray arr) { - external_constructor::construct(j, arr); + external_constructor::construct(j, std::move(arr)); } -template ::value, - int> = 0> +template +void to_json(BasicJsonType& j, typename BasicJsonType::array_t&& arr) +{ + external_constructor::construct(j, std::move(arr)); +} + +template::value, int> = 0> +void to_json(BasicJsonType& j, const CompatibleObjectType& obj) +{ + external_constructor::construct(j, obj); +} + +template +void to_json(BasicJsonType& j, typename BasicJsonType::object_t&& obj) +{ + external_constructor::construct(j, std::move(obj)); +} + +template::value, int> = 0> void to_json(BasicJsonType& j, T (&arr)[N]) { external_constructor::construct(j, arr); } -/////////////// -// from_json // -/////////////// - -// overloads for basic_json template parameters -template::value and - not std::is_same::value, - int> = 0> -void get_arithmetic_value(const BasicJsonType& j, ArithmeticType& val) +template +void to_json(BasicJsonType& j, const std::pair& p) { - switch (static_cast(j)) - { - case value_t::number_unsigned: - { - val = static_cast( - *j.template get_ptr()); - break; - } - case value_t::number_integer: - { - val = static_cast( - *j.template get_ptr()); - break; - } - case value_t::number_float: - { - val = static_cast( - *j.template get_ptr()); - break; - } - default: - { - JSON_THROW(type_error::create(302, "type must be number, but is " + j.type_name())); - } - } + j = {p.first, p.second}; } -template -void from_json(const BasicJsonType& j, typename BasicJsonType::boolean_t& b) +template +void to_json_tuple_impl(BasicJsonType& j, const Tuple& t, index_sequence) { - if (not j.is_boolean()) - { - JSON_THROW(type_error::create(302, "type must be boolean, but is " + j.type_name())); - } - b = *j.template get_ptr(); + j = {std::get(t)...}; } -template -void from_json(const BasicJsonType& j, typename BasicJsonType::string_t& s) +template +void to_json(BasicJsonType& j, const std::tuple& t) { - if (not j.is_string()) - { - JSON_THROW(type_error::create(302, "type must be string, but is " + j.type_name())); - } - s = *j.template get_ptr(); -} - -template -void from_json(const BasicJsonType& j, typename BasicJsonType::number_float_t& val) -{ - get_arithmetic_value(j, val); -} - -template -void from_json(const BasicJsonType& j, typename BasicJsonType::number_unsigned_t& val) -{ - get_arithmetic_value(j, val); -} - -template -void from_json(const BasicJsonType& j, typename BasicJsonType::number_integer_t& val) -{ - get_arithmetic_value(j, val); -} - -template::value, int> = 0> -void from_json(const BasicJsonType& j, EnumType& e) -{ - typename std::underlying_type::type val; - get_arithmetic_value(j, val); - e = static_cast(val); -} - -template -void from_json(const BasicJsonType& j, typename BasicJsonType::array_t& arr) -{ - if (not j.is_array()) - { - JSON_THROW(type_error::create(302, "type must be array, but is " + j.type_name())); - } - arr = *j.template get_ptr(); -} - -// forward_list doesn't have an insert method -template::value, int> = 0> -void from_json(const BasicJsonType& j, std::forward_list& l) -{ - if (not j.is_array()) - { - JSON_THROW(type_error::create(302, "type must be array, but is " + j.type_name())); - } - - for (auto it = j.rbegin(), end = j.rend(); it != end; ++it) - { - l.push_front(it->template get()); - } -} - -template -void from_json_array_impl(const BasicJsonType& j, CompatibleArrayType& arr, priority_tag<0>) -{ - using std::begin; - using std::end; - - std::transform(j.begin(), j.end(), - std::inserter(arr, end(arr)), [](const BasicJsonType & i) - { - // get() returns *this, this won't call a from_json - // method when value_type is BasicJsonType - return i.template get(); - }); -} - -template -auto from_json_array_impl(const BasicJsonType& j, CompatibleArrayType& arr, priority_tag<1>) --> decltype( - arr.reserve(std::declval()), - void()) -{ - using std::begin; - using std::end; - - arr.reserve(j.size()); - std::transform(j.begin(), j.end(), - std::inserter(arr, end(arr)), [](const BasicJsonType & i) - { - // get() returns *this, this won't call a from_json - // method when value_type is BasicJsonType - return i.template get(); - }); -} - -template::value and - std::is_convertible::value and - not std::is_same::value, int> = 0> -void from_json(const BasicJsonType& j, CompatibleArrayType& arr) -{ - if (not j.is_array()) - { - JSON_THROW(type_error::create(302, "type must be array, but is " + j.type_name())); - } - - from_json_array_impl(j, arr, priority_tag<1> {}); -} - -template::value, int> = 0> -void from_json(const BasicJsonType& j, CompatibleObjectType& obj) -{ - if (not j.is_object()) - { - JSON_THROW(type_error::create(302, "type must be object, but is " + j.type_name())); - } - - auto inner_object = j.template get_ptr(); - using std::begin; - using std::end; - // we could avoid the assignment, but this might require a for loop, which - // might be less efficient than the container constructor for some - // containers (would it?) - obj = CompatibleObjectType(begin(*inner_object), end(*inner_object)); -} - -// overload for arithmetic types, not chosen for basic_json template arguments -// (BooleanType, etc..); note: Is it really necessary to provide explicit -// overloads for boolean_t etc. in case of a custom BooleanType which is not -// an arithmetic type? -template::value and - not std::is_same::value and - not std::is_same::value and - not std::is_same::value and - not std::is_same::value, - int> = 0> -void from_json(const BasicJsonType& j, ArithmeticType& val) -{ - switch (static_cast(j)) - { - case value_t::number_unsigned: - { - val = static_cast(*j.template get_ptr()); - break; - } - case value_t::number_integer: - { - val = static_cast(*j.template get_ptr()); - break; - } - case value_t::number_float: - { - val = static_cast(*j.template get_ptr()); - break; - } - case value_t::boolean: - { - val = static_cast(*j.template get_ptr()); - break; - } - default: - { - JSON_THROW(type_error::create(302, "type must be number, but is " + j.type_name())); - } - } + to_json_tuple_impl(j, t, index_sequence_for {}); } struct to_json_fn { private: template - auto call(BasicJsonType& j, T&& val, priority_tag<1>) const noexcept(noexcept(to_json(j, std::forward(val)))) + auto call(BasicJsonType& j, T&& val, priority_tag<1> /*unused*/) const noexcept(noexcept(to_json(j, std::forward(val)))) -> decltype(to_json(j, std::forward(val)), void()) { return to_json(j, std::forward(val)); } template - void call(BasicJsonType&, T&&, priority_tag<0>) const noexcept + void call(BasicJsonType& /*unused*/, T&& /*unused*/, priority_tag<0> /*unused*/) const noexcept { static_assert(sizeof(BasicJsonType) == 0, "could not find to_json() method in T's namespace"); + +#ifdef _MSC_VER + // MSVC does not show a stacktrace for the above assert + using decayed = uncvref_t; + static_assert(sizeof(typename decayed::force_msvc_stacktrace) == 0, + "forcing MSVC stacktrace to show which T we're talking about."); +#endif } public: @@ -1111,62 +1581,9201 @@ struct to_json_fn return call(j, std::forward(val), priority_tag<1> {}); } }; +} -struct from_json_fn -{ - private: - template - auto call(const BasicJsonType& j, T& val, priority_tag<1>) const - noexcept(noexcept(from_json(j, val))) - -> decltype(from_json(j, val), void()) - { - return from_json(j, val); - } - - template - void call(const BasicJsonType&, T&, priority_tag<0>) const noexcept - { - static_assert(sizeof(BasicJsonType) == 0, - "could not find from_json() method in T's namespace"); - } - - public: - template - void operator()(const BasicJsonType& j, T& val) const - noexcept(noexcept(std::declval().call(j, val, priority_tag<1> {}))) - { - return call(j, val, priority_tag<1> {}); - } -}; - -// taken from ranges-v3 -template -struct static_const -{ - static constexpr T value{}; -}; - -template -constexpr T static_const::value; -} // namespace detail - - -/// namespace to hold default `to_json` / `from_json` functions +/// namespace to hold default `to_json` function namespace { constexpr const auto& to_json = detail::static_const::value; -constexpr const auto& from_json = detail::static_const::value; +} } +// #include + + +#include // assert +#include // size_t +#include // strlen +#include // istream +#include // begin, end, iterator_traits, random_access_iterator_tag, distance, next +#include // shared_ptr, make_shared, addressof +#include // accumulate +#include // string, char_traits +#include // enable_if, is_base_of, is_pointer, is_integral, remove_pointer +#include // pair, declval + +// #include + + +namespace nlohmann +{ +namespace detail +{ +/// the supported input formats +enum class input_format_t { json, cbor, msgpack, ubjson }; + +//////////////////// +// input adapters // +//////////////////// /*! -@brief default JSONSerializer template argument +@brief abstract input adapter interface -This serializer ignores the template arguments and uses ADL -([argument-dependent lookup](http://en.cppreference.com/w/cpp/language/adl)) -for serialization. +Produces a stream of std::char_traits::int_type characters from a +std::istream, a buffer, or some other input type. Accepts the return of +exactly one non-EOF character for future input. The int_type characters +returned consist of all valid char values as positive values (typically +unsigned char), plus an EOF value outside that range, specified by the value +of the function std::char_traits::eof(). This value is typically -1, but +could be any arbitrary value which is not a valid char value. */ -template +struct input_adapter_protocol +{ + /// get a character [0,255] or std::char_traits::eof(). + virtual std::char_traits::int_type get_character() = 0; + virtual ~input_adapter_protocol() = default; +}; + +/// a type to simplify interfaces +using input_adapter_t = std::shared_ptr; + +/*! +Input adapter for a (caching) istream. Ignores a UFT Byte Order Mark at +beginning of input. Does not support changing the underlying std::streambuf +in mid-input. Maintains underlying std::istream and std::streambuf to support +subsequent use of standard std::istream operations to process any input +characters following those used in parsing the JSON input. Clears the +std::istream flags; any input errors (e.g., EOF) will be detected by the first +subsequent call for input from the std::istream. +*/ +class input_stream_adapter : public input_adapter_protocol +{ + public: + ~input_stream_adapter() override + { + // clear stream flags; we use underlying streambuf I/O, do not + // maintain ifstream flags + is.clear(); + } + + explicit input_stream_adapter(std::istream& i) + : is(i), sb(*i.rdbuf()) + {} + + // delete because of pointer members + input_stream_adapter(const input_stream_adapter&) = delete; + input_stream_adapter& operator=(input_stream_adapter&) = delete; + + // std::istream/std::streambuf use std::char_traits::to_int_type, to + // ensure that std::char_traits::eof() and the character 0xFF do not + // end up as the same value, eg. 0xFFFFFFFF. + std::char_traits::int_type get_character() override + { + return sb.sbumpc(); + } + + private: + /// the associated input stream + std::istream& is; + std::streambuf& sb; +}; + +/// input adapter for buffer input +class input_buffer_adapter : public input_adapter_protocol +{ + public: + input_buffer_adapter(const char* b, const std::size_t l) + : cursor(b), limit(b + l) + {} + + // delete because of pointer members + input_buffer_adapter(const input_buffer_adapter&) = delete; + input_buffer_adapter& operator=(input_buffer_adapter&) = delete; + + std::char_traits::int_type get_character() noexcept override + { + if (JSON_LIKELY(cursor < limit)) + { + return std::char_traits::to_int_type(*(cursor++)); + } + + return std::char_traits::eof(); + } + + private: + /// pointer to the current character + const char* cursor; + /// pointer past the last character + const char* const limit; +}; + +template +class wide_string_input_adapter : public input_adapter_protocol +{ + public: + wide_string_input_adapter(const WideStringType& w) : str(w) {} + + std::char_traits::int_type get_character() noexcept override + { + // check if buffer needs to be filled + if (utf8_bytes_index == utf8_bytes_filled) + { + if (sizeof(typename WideStringType::value_type) == 2) + { + fill_buffer_utf16(); + } + else + { + fill_buffer_utf32(); + } + + assert(utf8_bytes_filled > 0); + assert(utf8_bytes_index == 0); + } + + // use buffer + assert(utf8_bytes_filled > 0); + assert(utf8_bytes_index < utf8_bytes_filled); + return utf8_bytes[utf8_bytes_index++]; + } + + private: + void fill_buffer_utf16() + { + utf8_bytes_index = 0; + + if (current_wchar == str.size()) + { + utf8_bytes[0] = std::char_traits::eof(); + utf8_bytes_filled = 1; + } + else + { + // get the current character + const int wc = static_cast(str[current_wchar++]); + + // UTF-16 to UTF-8 encoding + if (wc < 0x80) + { + utf8_bytes[0] = wc; + utf8_bytes_filled = 1; + } + else if (wc <= 0x7FF) + { + utf8_bytes[0] = 0xC0 | ((wc >> 6)); + utf8_bytes[1] = 0x80 | (wc & 0x3F); + utf8_bytes_filled = 2; + } + else if (0xD800 > wc or wc >= 0xE000) + { + utf8_bytes[0] = 0xE0 | ((wc >> 12)); + utf8_bytes[1] = 0x80 | ((wc >> 6) & 0x3F); + utf8_bytes[2] = 0x80 | (wc & 0x3F); + utf8_bytes_filled = 3; + } + else + { + if (current_wchar < str.size()) + { + const int wc2 = static_cast(str[current_wchar++]); + const int charcode = 0x10000 + (((wc & 0x3FF) << 10) | (wc2 & 0x3FF)); + utf8_bytes[0] = 0xf0 | (charcode >> 18); + utf8_bytes[1] = 0x80 | ((charcode >> 12) & 0x3F); + utf8_bytes[2] = 0x80 | ((charcode >> 6) & 0x3F); + utf8_bytes[3] = 0x80 | (charcode & 0x3F); + utf8_bytes_filled = 4; + } + else + { + // unknown character + ++current_wchar; + utf8_bytes[0] = wc; + utf8_bytes_filled = 1; + } + } + } + } + + void fill_buffer_utf32() + { + utf8_bytes_index = 0; + + if (current_wchar == str.size()) + { + utf8_bytes[0] = std::char_traits::eof(); + utf8_bytes_filled = 1; + } + else + { + // get the current character + const int wc = static_cast(str[current_wchar++]); + + // UTF-32 to UTF-8 encoding + if (wc < 0x80) + { + utf8_bytes[0] = wc; + utf8_bytes_filled = 1; + } + else if (wc <= 0x7FF) + { + utf8_bytes[0] = 0xC0 | ((wc >> 6) & 0x1F); + utf8_bytes[1] = 0x80 | (wc & 0x3F); + utf8_bytes_filled = 2; + } + else if (wc <= 0xFFFF) + { + utf8_bytes[0] = 0xE0 | ((wc >> 12) & 0x0F); + utf8_bytes[1] = 0x80 | ((wc >> 6) & 0x3F); + utf8_bytes[2] = 0x80 | (wc & 0x3F); + utf8_bytes_filled = 3; + } + else if (wc <= 0x10FFFF) + { + utf8_bytes[0] = 0xF0 | ((wc >> 18 ) & 0x07); + utf8_bytes[1] = 0x80 | ((wc >> 12) & 0x3F); + utf8_bytes[2] = 0x80 | ((wc >> 6) & 0x3F); + utf8_bytes[3] = 0x80 | (wc & 0x3F); + utf8_bytes_filled = 4; + } + else + { + // unknown character + utf8_bytes[0] = wc; + utf8_bytes_filled = 1; + } + } + } + + private: + /// the wstring to process + const WideStringType& str; + + /// index of the current wchar in str + std::size_t current_wchar = 0; + + /// a buffer for UTF-8 bytes + std::array::int_type, 4> utf8_bytes = {{0, 0, 0, 0}}; + + /// index to the utf8_codes array for the next valid byte + std::size_t utf8_bytes_index = 0; + /// number of valid bytes in the utf8_codes array + std::size_t utf8_bytes_filled = 0; +}; + +class input_adapter +{ + public: + // native support + + /// input adapter for input stream + input_adapter(std::istream& i) + : ia(std::make_shared(i)) {} + + /// input adapter for input stream + input_adapter(std::istream&& i) + : ia(std::make_shared(i)) {} + + input_adapter(const std::wstring& ws) + : ia(std::make_shared>(ws)) {} + + input_adapter(const std::u16string& ws) + : ia(std::make_shared>(ws)) {} + + input_adapter(const std::u32string& ws) + : ia(std::make_shared>(ws)) {} + + /// input adapter for buffer + template::value and + std::is_integral::type>::value and + sizeof(typename std::remove_pointer::type) == 1, + int>::type = 0> + input_adapter(CharT b, std::size_t l) + : ia(std::make_shared(reinterpret_cast(b), l)) {} + + // derived support + + /// input adapter for string literal + template::value and + std::is_integral::type>::value and + sizeof(typename std::remove_pointer::type) == 1, + int>::type = 0> + input_adapter(CharT b) + : input_adapter(reinterpret_cast(b), + std::strlen(reinterpret_cast(b))) {} + + /// input adapter for iterator range with contiguous storage + template::iterator_category, std::random_access_iterator_tag>::value, + int>::type = 0> + input_adapter(IteratorType first, IteratorType last) + { + // assertion to check that the iterator range is indeed contiguous, + // see http://stackoverflow.com/a/35008842/266378 for more discussion + assert(std::accumulate( + first, last, std::pair(true, 0), + [&first](std::pair res, decltype(*first) val) + { + res.first &= (val == *(std::next(std::addressof(*first), res.second++))); + return res; + }).first); + + // assertion to check that each element is 1 byte long + static_assert( + sizeof(typename std::iterator_traits::value_type) == 1, + "each element in the iterator range must have the size of 1 byte"); + + const auto len = static_cast(std::distance(first, last)); + if (JSON_LIKELY(len > 0)) + { + // there is at least one element: use the address of first + ia = std::make_shared(reinterpret_cast(&(*first)), len); + } + else + { + // the address of first cannot be used: use nullptr + ia = std::make_shared(nullptr, len); + } + } + + /// input adapter for array + template + input_adapter(T (&array)[N]) + : input_adapter(std::begin(array), std::end(array)) {} + + /// input adapter for contiguous container + template::value and + std::is_base_of()))>::iterator_category>::value, + int>::type = 0> + input_adapter(const ContiguousContainer& c) + : input_adapter(std::begin(c), std::end(c)) {} + + operator input_adapter_t() + { + return ia; + } + + private: + /// the actual adapter + input_adapter_t ia = nullptr; +}; +} +} + +// #include + + +#include // localeconv +#include // size_t +#include // strtof, strtod, strtold, strtoll, strtoull +#include // initializer_list +#include // hex, uppercase +#include // setw, setfill +#include // stringstream +#include // char_traits, string +#include // vector + +// #include + +// #include + + +namespace nlohmann +{ +namespace detail +{ +/////////// +// lexer // +/////////// + +/*! +@brief lexical analysis + +This class organizes the lexical analysis during JSON deserialization. +*/ +template +class lexer +{ + using number_integer_t = typename BasicJsonType::number_integer_t; + using number_unsigned_t = typename BasicJsonType::number_unsigned_t; + using number_float_t = typename BasicJsonType::number_float_t; + using string_t = typename BasicJsonType::string_t; + + public: + /// token types for the parser + enum class token_type + { + uninitialized, ///< indicating the scanner is uninitialized + literal_true, ///< the `true` literal + literal_false, ///< the `false` literal + literal_null, ///< the `null` literal + value_string, ///< a string -- use get_string() for actual value + value_unsigned, ///< an unsigned integer -- use get_number_unsigned() for actual value + value_integer, ///< a signed integer -- use get_number_integer() for actual value + value_float, ///< an floating point number -- use get_number_float() for actual value + begin_array, ///< the character for array begin `[` + begin_object, ///< the character for object begin `{` + end_array, ///< the character for array end `]` + end_object, ///< the character for object end `}` + name_separator, ///< the name separator `:` + value_separator, ///< the value separator `,` + parse_error, ///< indicating a parse error + end_of_input, ///< indicating the end of the input buffer + literal_or_value ///< a literal or the begin of a value (only for diagnostics) + }; + + /// return name of values of type token_type (only used for errors) + static const char* token_type_name(const token_type t) noexcept + { + switch (t) + { + case token_type::uninitialized: + return ""; + case token_type::literal_true: + return "true literal"; + case token_type::literal_false: + return "false literal"; + case token_type::literal_null: + return "null literal"; + case token_type::value_string: + return "string literal"; + case lexer::token_type::value_unsigned: + case lexer::token_type::value_integer: + case lexer::token_type::value_float: + return "number literal"; + case token_type::begin_array: + return "'['"; + case token_type::begin_object: + return "'{'"; + case token_type::end_array: + return "']'"; + case token_type::end_object: + return "'}'"; + case token_type::name_separator: + return "':'"; + case token_type::value_separator: + return "','"; + case token_type::parse_error: + return ""; + case token_type::end_of_input: + return "end of input"; + case token_type::literal_or_value: + return "'[', '{', or a literal"; + default: // catch non-enum values + return "unknown token"; // LCOV_EXCL_LINE + } + } + + explicit lexer(detail::input_adapter_t&& adapter) + : ia(std::move(adapter)), decimal_point_char(get_decimal_point()) {} + + // delete because of pointer members + lexer(const lexer&) = delete; + lexer& operator=(lexer&) = delete; + + private: + ///////////////////// + // locales + ///////////////////// + + /// return the locale-dependent decimal point + static char get_decimal_point() noexcept + { + const auto loc = localeconv(); + assert(loc != nullptr); + return (loc->decimal_point == nullptr) ? '.' : *(loc->decimal_point); + } + + ///////////////////// + // scan functions + ///////////////////// + + /*! + @brief get codepoint from 4 hex characters following `\u` + + For input "\u c1 c2 c3 c4" the codepoint is: + (c1 * 0x1000) + (c2 * 0x0100) + (c3 * 0x0010) + c4 + = (c1 << 12) + (c2 << 8) + (c3 << 4) + (c4 << 0) + + Furthermore, the possible characters '0'..'9', 'A'..'F', and 'a'..'f' + must be converted to the integers 0x0..0x9, 0xA..0xF, 0xA..0xF, resp. The + conversion is done by subtracting the offset (0x30, 0x37, and 0x57) + between the ASCII value of the character and the desired integer value. + + @return codepoint (0x0000..0xFFFF) or -1 in case of an error (e.g. EOF or + non-hex character) + */ + int get_codepoint() + { + // this function only makes sense after reading `\u` + assert(current == 'u'); + int codepoint = 0; + + const auto factors = { 12, 8, 4, 0 }; + for (const auto factor : factors) + { + get(); + + if (current >= '0' and current <= '9') + { + codepoint += ((current - 0x30) << factor); + } + else if (current >= 'A' and current <= 'F') + { + codepoint += ((current - 0x37) << factor); + } + else if (current >= 'a' and current <= 'f') + { + codepoint += ((current - 0x57) << factor); + } + else + { + return -1; + } + } + + assert(0x0000 <= codepoint and codepoint <= 0xFFFF); + return codepoint; + } + + /*! + @brief check if the next byte(s) are inside a given range + + Adds the current byte and, for each passed range, reads a new byte and + checks if it is inside the range. If a violation was detected, set up an + error message and return false. Otherwise, return true. + + @param[in] ranges list of integers; interpreted as list of pairs of + inclusive lower and upper bound, respectively + + @pre The passed list @a ranges must have 2, 4, or 6 elements; that is, + 1, 2, or 3 pairs. This precondition is enforced by an assertion. + + @return true if and only if no range violation was detected + */ + bool next_byte_in_range(std::initializer_list ranges) + { + assert(ranges.size() == 2 or ranges.size() == 4 or ranges.size() == 6); + add(current); + + for (auto range = ranges.begin(); range != ranges.end(); ++range) + { + get(); + if (JSON_LIKELY(*range <= current and current <= *(++range))) + { + add(current); + } + else + { + error_message = "invalid string: ill-formed UTF-8 byte"; + return false; + } + } + + return true; + } + + /*! + @brief scan a string literal + + This function scans a string according to Sect. 7 of RFC 7159. While + scanning, bytes are escaped and copied into buffer token_buffer. Then the + function returns successfully, token_buffer is *not* null-terminated (as it + may contain \0 bytes), and token_buffer.size() is the number of bytes in the + string. + + @return token_type::value_string if string could be successfully scanned, + token_type::parse_error otherwise + + @note In case of errors, variable error_message contains a textual + description. + */ + token_type scan_string() + { + // reset token_buffer (ignore opening quote) + reset(); + + // we entered the function by reading an open quote + assert(current == '\"'); + + while (true) + { + // get next character + switch (get()) + { + // end of file while parsing string + case std::char_traits::eof(): + { + error_message = "invalid string: missing closing quote"; + return token_type::parse_error; + } + + // closing quote + case '\"': + { + return token_type::value_string; + } + + // escapes + case '\\': + { + switch (get()) + { + // quotation mark + case '\"': + add('\"'); + break; + // reverse solidus + case '\\': + add('\\'); + break; + // solidus + case '/': + add('/'); + break; + // backspace + case 'b': + add('\b'); + break; + // form feed + case 'f': + add('\f'); + break; + // line feed + case 'n': + add('\n'); + break; + // carriage return + case 'r': + add('\r'); + break; + // tab + case 't': + add('\t'); + break; + + // unicode escapes + case 'u': + { + const int codepoint1 = get_codepoint(); + int codepoint = codepoint1; // start with codepoint1 + + if (JSON_UNLIKELY(codepoint1 == -1)) + { + error_message = "invalid string: '\\u' must be followed by 4 hex digits"; + return token_type::parse_error; + } + + // check if code point is a high surrogate + if (0xD800 <= codepoint1 and codepoint1 <= 0xDBFF) + { + // expect next \uxxxx entry + if (JSON_LIKELY(get() == '\\' and get() == 'u')) + { + const int codepoint2 = get_codepoint(); + + if (JSON_UNLIKELY(codepoint2 == -1)) + { + error_message = "invalid string: '\\u' must be followed by 4 hex digits"; + return token_type::parse_error; + } + + // check if codepoint2 is a low surrogate + if (JSON_LIKELY(0xDC00 <= codepoint2 and codepoint2 <= 0xDFFF)) + { + // overwrite codepoint + codepoint = + // high surrogate occupies the most significant 22 bits + (codepoint1 << 10) + // low surrogate occupies the least significant 15 bits + + codepoint2 + // there is still the 0xD800, 0xDC00 and 0x10000 noise + // in the result so we have to subtract with: + // (0xD800 << 10) + DC00 - 0x10000 = 0x35FDC00 + - 0x35FDC00; + } + else + { + error_message = "invalid string: surrogate U+DC00..U+DFFF must be followed by U+DC00..U+DFFF"; + return token_type::parse_error; + } + } + else + { + error_message = "invalid string: surrogate U+DC00..U+DFFF must be followed by U+DC00..U+DFFF"; + return token_type::parse_error; + } + } + else + { + if (JSON_UNLIKELY(0xDC00 <= codepoint1 and codepoint1 <= 0xDFFF)) + { + error_message = "invalid string: surrogate U+DC00..U+DFFF must follow U+D800..U+DBFF"; + return token_type::parse_error; + } + } + + // result of the above calculation yields a proper codepoint + assert(0x00 <= codepoint and codepoint <= 0x10FFFF); + + // translate codepoint into bytes + if (codepoint < 0x80) + { + // 1-byte characters: 0xxxxxxx (ASCII) + add(codepoint); + } + else if (codepoint <= 0x7FF) + { + // 2-byte characters: 110xxxxx 10xxxxxx + add(0xC0 | (codepoint >> 6)); + add(0x80 | (codepoint & 0x3F)); + } + else if (codepoint <= 0xFFFF) + { + // 3-byte characters: 1110xxxx 10xxxxxx 10xxxxxx + add(0xE0 | (codepoint >> 12)); + add(0x80 | ((codepoint >> 6) & 0x3F)); + add(0x80 | (codepoint & 0x3F)); + } + else + { + // 4-byte characters: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx + add(0xF0 | (codepoint >> 18)); + add(0x80 | ((codepoint >> 12) & 0x3F)); + add(0x80 | ((codepoint >> 6) & 0x3F)); + add(0x80 | (codepoint & 0x3F)); + } + + break; + } + + // other characters after escape + default: + error_message = "invalid string: forbidden character after backslash"; + return token_type::parse_error; + } + + break; + } + + // invalid control characters + case 0x00: + case 0x01: + case 0x02: + case 0x03: + case 0x04: + case 0x05: + case 0x06: + case 0x07: + case 0x08: + case 0x09: + case 0x0A: + case 0x0B: + case 0x0C: + case 0x0D: + case 0x0E: + case 0x0F: + case 0x10: + case 0x11: + case 0x12: + case 0x13: + case 0x14: + case 0x15: + case 0x16: + case 0x17: + case 0x18: + case 0x19: + case 0x1A: + case 0x1B: + case 0x1C: + case 0x1D: + case 0x1E: + case 0x1F: + { + error_message = "invalid string: control character must be escaped"; + return token_type::parse_error; + } + + // U+0020..U+007F (except U+0022 (quote) and U+005C (backspace)) + case 0x20: + case 0x21: + case 0x23: + case 0x24: + case 0x25: + case 0x26: + case 0x27: + case 0x28: + case 0x29: + case 0x2A: + case 0x2B: + case 0x2C: + case 0x2D: + case 0x2E: + case 0x2F: + case 0x30: + case 0x31: + case 0x32: + case 0x33: + case 0x34: + case 0x35: + case 0x36: + case 0x37: + case 0x38: + case 0x39: + case 0x3A: + case 0x3B: + case 0x3C: + case 0x3D: + case 0x3E: + case 0x3F: + case 0x40: + case 0x41: + case 0x42: + case 0x43: + case 0x44: + case 0x45: + case 0x46: + case 0x47: + case 0x48: + case 0x49: + case 0x4A: + case 0x4B: + case 0x4C: + case 0x4D: + case 0x4E: + case 0x4F: + case 0x50: + case 0x51: + case 0x52: + case 0x53: + case 0x54: + case 0x55: + case 0x56: + case 0x57: + case 0x58: + case 0x59: + case 0x5A: + case 0x5B: + case 0x5D: + case 0x5E: + case 0x5F: + case 0x60: + case 0x61: + case 0x62: + case 0x63: + case 0x64: + case 0x65: + case 0x66: + case 0x67: + case 0x68: + case 0x69: + case 0x6A: + case 0x6B: + case 0x6C: + case 0x6D: + case 0x6E: + case 0x6F: + case 0x70: + case 0x71: + case 0x72: + case 0x73: + case 0x74: + case 0x75: + case 0x76: + case 0x77: + case 0x78: + case 0x79: + case 0x7A: + case 0x7B: + case 0x7C: + case 0x7D: + case 0x7E: + case 0x7F: + { + add(current); + break; + } + + // U+0080..U+07FF: bytes C2..DF 80..BF + case 0xC2: + case 0xC3: + case 0xC4: + case 0xC5: + case 0xC6: + case 0xC7: + case 0xC8: + case 0xC9: + case 0xCA: + case 0xCB: + case 0xCC: + case 0xCD: + case 0xCE: + case 0xCF: + case 0xD0: + case 0xD1: + case 0xD2: + case 0xD3: + case 0xD4: + case 0xD5: + case 0xD6: + case 0xD7: + case 0xD8: + case 0xD9: + case 0xDA: + case 0xDB: + case 0xDC: + case 0xDD: + case 0xDE: + case 0xDF: + { + if (JSON_UNLIKELY(not next_byte_in_range({0x80, 0xBF}))) + { + return token_type::parse_error; + } + break; + } + + // U+0800..U+0FFF: bytes E0 A0..BF 80..BF + case 0xE0: + { + if (JSON_UNLIKELY(not (next_byte_in_range({0xA0, 0xBF, 0x80, 0xBF})))) + { + return token_type::parse_error; + } + break; + } + + // U+1000..U+CFFF: bytes E1..EC 80..BF 80..BF + // U+E000..U+FFFF: bytes EE..EF 80..BF 80..BF + case 0xE1: + case 0xE2: + case 0xE3: + case 0xE4: + case 0xE5: + case 0xE6: + case 0xE7: + case 0xE8: + case 0xE9: + case 0xEA: + case 0xEB: + case 0xEC: + case 0xEE: + case 0xEF: + { + if (JSON_UNLIKELY(not (next_byte_in_range({0x80, 0xBF, 0x80, 0xBF})))) + { + return token_type::parse_error; + } + break; + } + + // U+D000..U+D7FF: bytes ED 80..9F 80..BF + case 0xED: + { + if (JSON_UNLIKELY(not (next_byte_in_range({0x80, 0x9F, 0x80, 0xBF})))) + { + return token_type::parse_error; + } + break; + } + + // U+10000..U+3FFFF F0 90..BF 80..BF 80..BF + case 0xF0: + { + if (JSON_UNLIKELY(not (next_byte_in_range({0x90, 0xBF, 0x80, 0xBF, 0x80, 0xBF})))) + { + return token_type::parse_error; + } + break; + } + + // U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF + case 0xF1: + case 0xF2: + case 0xF3: + { + if (JSON_UNLIKELY(not (next_byte_in_range({0x80, 0xBF, 0x80, 0xBF, 0x80, 0xBF})))) + { + return token_type::parse_error; + } + break; + } + + // U+100000..U+10FFFF F4 80..8F 80..BF 80..BF + case 0xF4: + { + if (JSON_UNLIKELY(not (next_byte_in_range({0x80, 0x8F, 0x80, 0xBF, 0x80, 0xBF})))) + { + return token_type::parse_error; + } + break; + } + + // remaining bytes (80..C1 and F5..FF) are ill-formed + default: + { + error_message = "invalid string: ill-formed UTF-8 byte"; + return token_type::parse_error; + } + } + } + } + + static void strtof(float& f, const char* str, char** endptr) noexcept + { + f = std::strtof(str, endptr); + } + + static void strtof(double& f, const char* str, char** endptr) noexcept + { + f = std::strtod(str, endptr); + } + + static void strtof(long double& f, const char* str, char** endptr) noexcept + { + f = std::strtold(str, endptr); + } + + /*! + @brief scan a number literal + + This function scans a string according to Sect. 6 of RFC 7159. + + The function is realized with a deterministic finite state machine derived + from the grammar described in RFC 7159. Starting in state "init", the + input is read and used to determined the next state. Only state "done" + accepts the number. State "error" is a trap state to model errors. In the + table below, "anything" means any character but the ones listed before. + + state | 0 | 1-9 | e E | + | - | . | anything + ---------|----------|----------|----------|---------|---------|----------|----------- + init | zero | any1 | [error] | [error] | minus | [error] | [error] + minus | zero | any1 | [error] | [error] | [error] | [error] | [error] + zero | done | done | exponent | done | done | decimal1 | done + any1 | any1 | any1 | exponent | done | done | decimal1 | done + decimal1 | decimal2 | [error] | [error] | [error] | [error] | [error] | [error] + decimal2 | decimal2 | decimal2 | exponent | done | done | done | done + exponent | any2 | any2 | [error] | sign | sign | [error] | [error] + sign | any2 | any2 | [error] | [error] | [error] | [error] | [error] + any2 | any2 | any2 | done | done | done | done | done + + The state machine is realized with one label per state (prefixed with + "scan_number_") and `goto` statements between them. The state machine + contains cycles, but any cycle can be left when EOF is read. Therefore, + the function is guaranteed to terminate. + + During scanning, the read bytes are stored in token_buffer. This string is + then converted to a signed integer, an unsigned integer, or a + floating-point number. + + @return token_type::value_unsigned, token_type::value_integer, or + token_type::value_float if number could be successfully scanned, + token_type::parse_error otherwise + + @note The scanner is independent of the current locale. Internally, the + locale's decimal point is used instead of `.` to work with the + locale-dependent converters. + */ + token_type scan_number() + { + // reset token_buffer to store the number's bytes + reset(); + + // the type of the parsed number; initially set to unsigned; will be + // changed if minus sign, decimal point or exponent is read + token_type number_type = token_type::value_unsigned; + + // state (init): we just found out we need to scan a number + switch (current) + { + case '-': + { + add(current); + goto scan_number_minus; + } + + case '0': + { + add(current); + goto scan_number_zero; + } + + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + { + add(current); + goto scan_number_any1; + } + + default: + { + // all other characters are rejected outside scan_number() + assert(false); // LCOV_EXCL_LINE + } + } + +scan_number_minus: + // state: we just parsed a leading minus sign + number_type = token_type::value_integer; + switch (get()) + { + case '0': + { + add(current); + goto scan_number_zero; + } + + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + { + add(current); + goto scan_number_any1; + } + + default: + { + error_message = "invalid number; expected digit after '-'"; + return token_type::parse_error; + } + } + +scan_number_zero: + // state: we just parse a zero (maybe with a leading minus sign) + switch (get()) + { + case '.': + { + add(decimal_point_char); + goto scan_number_decimal1; + } + + case 'e': + case 'E': + { + add(current); + goto scan_number_exponent; + } + + default: + goto scan_number_done; + } + +scan_number_any1: + // state: we just parsed a number 0-9 (maybe with a leading minus sign) + switch (get()) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + { + add(current); + goto scan_number_any1; + } + + case '.': + { + add(decimal_point_char); + goto scan_number_decimal1; + } + + case 'e': + case 'E': + { + add(current); + goto scan_number_exponent; + } + + default: + goto scan_number_done; + } + +scan_number_decimal1: + // state: we just parsed a decimal point + number_type = token_type::value_float; + switch (get()) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + { + add(current); + goto scan_number_decimal2; + } + + default: + { + error_message = "invalid number; expected digit after '.'"; + return token_type::parse_error; + } + } + +scan_number_decimal2: + // we just parsed at least one number after a decimal point + switch (get()) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + { + add(current); + goto scan_number_decimal2; + } + + case 'e': + case 'E': + { + add(current); + goto scan_number_exponent; + } + + default: + goto scan_number_done; + } + +scan_number_exponent: + // we just parsed an exponent + number_type = token_type::value_float; + switch (get()) + { + case '+': + case '-': + { + add(current); + goto scan_number_sign; + } + + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + { + add(current); + goto scan_number_any2; + } + + default: + { + error_message = + "invalid number; expected '+', '-', or digit after exponent"; + return token_type::parse_error; + } + } + +scan_number_sign: + // we just parsed an exponent sign + switch (get()) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + { + add(current); + goto scan_number_any2; + } + + default: + { + error_message = "invalid number; expected digit after exponent sign"; + return token_type::parse_error; + } + } + +scan_number_any2: + // we just parsed a number after the exponent or exponent sign + switch (get()) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + { + add(current); + goto scan_number_any2; + } + + default: + goto scan_number_done; + } + +scan_number_done: + // unget the character after the number (we only read it to know that + // we are done scanning a number) + unget(); + + char* endptr = nullptr; + errno = 0; + + // try to parse integers first and fall back to floats + if (number_type == token_type::value_unsigned) + { + const auto x = std::strtoull(token_buffer.data(), &endptr, 10); + + // we checked the number format before + assert(endptr == token_buffer.data() + token_buffer.size()); + + if (errno == 0) + { + value_unsigned = static_cast(x); + if (value_unsigned == x) + { + return token_type::value_unsigned; + } + } + } + else if (number_type == token_type::value_integer) + { + const auto x = std::strtoll(token_buffer.data(), &endptr, 10); + + // we checked the number format before + assert(endptr == token_buffer.data() + token_buffer.size()); + + if (errno == 0) + { + value_integer = static_cast(x); + if (value_integer == x) + { + return token_type::value_integer; + } + } + } + + // this code is reached if we parse a floating-point number or if an + // integer conversion above failed + strtof(value_float, token_buffer.data(), &endptr); + + // we checked the number format before + assert(endptr == token_buffer.data() + token_buffer.size()); + + return token_type::value_float; + } + + /*! + @param[in] literal_text the literal text to expect + @param[in] length the length of the passed literal text + @param[in] return_type the token type to return on success + */ + token_type scan_literal(const char* literal_text, const std::size_t length, + token_type return_type) + { + assert(current == literal_text[0]); + for (std::size_t i = 1; i < length; ++i) + { + if (JSON_UNLIKELY(get() != literal_text[i])) + { + error_message = "invalid literal"; + return token_type::parse_error; + } + } + return return_type; + } + + ///////////////////// + // input management + ///////////////////// + + /// reset token_buffer; current character is beginning of token + void reset() noexcept + { + token_buffer.clear(); + token_string.clear(); + token_string.push_back(std::char_traits::to_char_type(current)); + } + + /* + @brief get next character from the input + + This function provides the interface to the used input adapter. It does + not throw in case the input reached EOF, but returns a + `std::char_traits::eof()` in that case. Stores the scanned characters + for use in error messages. + + @return character read from the input + */ + std::char_traits::int_type get() + { + ++chars_read; + if (next_unget) + { + // just reset the next_unget variable and work with current + next_unget = false; + } + else + { + current = ia->get_character(); + } + + if (JSON_LIKELY(current != std::char_traits::eof())) + { + token_string.push_back(std::char_traits::to_char_type(current)); + } + return current; + } + + /*! + @brief unget current character (read it again on next get) + + We implement unget by setting variable next_unget to true. The input is not + changed - we just simulate ungetting by modifying chars_read and + token_string. The next call to get() will behave as if the unget character + is read again. + */ + void unget() + { + next_unget = true; + --chars_read; + if (JSON_LIKELY(current != std::char_traits::eof())) + { + assert(token_string.size() != 0); + token_string.pop_back(); + } + } + + /// add a character to token_buffer + void add(int c) + { + token_buffer.push_back(std::char_traits::to_char_type(c)); + } + + public: + ///////////////////// + // value getters + ///////////////////// + + /// return integer value + constexpr number_integer_t get_number_integer() const noexcept + { + return value_integer; + } + + /// return unsigned integer value + constexpr number_unsigned_t get_number_unsigned() const noexcept + { + return value_unsigned; + } + + /// return floating-point value + constexpr number_float_t get_number_float() const noexcept + { + return value_float; + } + + /// return current string value (implicitly resets the token; useful only once) + string_t& get_string() + { + return token_buffer; + } + + ///////////////////// + // diagnostics + ///////////////////// + + /// return position of last read token + constexpr std::size_t get_position() const noexcept + { + return chars_read; + } + + /// return the last read token (for errors only). Will never contain EOF + /// (an arbitrary value that is not a valid char value, often -1), because + /// 255 may legitimately occur. May contain NUL, which should be escaped. + std::string get_token_string() const + { + // escape control characters + std::string result; + for (const auto c : token_string) + { + if ('\x00' <= c and c <= '\x1F') + { + // escape control characters + std::stringstream ss; + ss << "(c) << ">"; + result += ss.str(); + } + else + { + // add character as is + result.push_back(c); + } + } + + return result; + } + + /// return syntax error message + constexpr const char* get_error_message() const noexcept + { + return error_message; + } + + ///////////////////// + // actual scanner + ///////////////////// + + /*! + @brief skip the UTF-8 byte order mark + @return true iff there is no BOM or the correct BOM has been skipped + */ + bool skip_bom() + { + if (get() == 0xEF) + { + if (get() == 0xBB and get() == 0xBF) + { + // we completely parsed the BOM + return true; + } + else + { + // after reading 0xEF, an unexpected character followed + return false; + } + } + else + { + // the first character is not the beginning of the BOM; unget it to + // process is later + unget(); + return true; + } + } + + token_type scan() + { + // initially, skip the BOM + if (chars_read == 0 and not skip_bom()) + { + error_message = "invalid BOM; must be 0xEF 0xBB 0xBF if given"; + return token_type::parse_error; + } + + // read next character and ignore whitespace + do + { + get(); + } + while (current == ' ' or current == '\t' or current == '\n' or current == '\r'); + + switch (current) + { + // structural characters + case '[': + return token_type::begin_array; + case ']': + return token_type::end_array; + case '{': + return token_type::begin_object; + case '}': + return token_type::end_object; + case ':': + return token_type::name_separator; + case ',': + return token_type::value_separator; + + // literals + case 't': + return scan_literal("true", 4, token_type::literal_true); + case 'f': + return scan_literal("false", 5, token_type::literal_false); + case 'n': + return scan_literal("null", 4, token_type::literal_null); + + // string + case '\"': + return scan_string(); + + // number + case '-': + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + return scan_number(); + + // end of input (the null byte is needed when parsing from + // string literals) + case '\0': + case std::char_traits::eof(): + return token_type::end_of_input; + + // error + default: + error_message = "invalid literal"; + return token_type::parse_error; + } + } + + private: + /// input adapter + detail::input_adapter_t ia = nullptr; + + /// the current character + std::char_traits::int_type current = std::char_traits::eof(); + + /// whether the next get() call should just return current + bool next_unget = false; + + /// the number of characters read + std::size_t chars_read = 0; + + /// raw input token string (for error messages) + std::vector token_string {}; + + /// buffer for variable-length tokens (numbers, strings) + string_t token_buffer {}; + + /// a description of occurred lexer errors + const char* error_message = ""; + + // number values + number_integer_t value_integer = 0; + number_unsigned_t value_unsigned = 0; + number_float_t value_float = 0; + + /// the decimal point + const char decimal_point_char = '.'; +}; +} +} + +// #include + + +#include // assert +#include // isfinite +#include // uint8_t +#include // function +#include // string +#include // move + +// #include + +// #include + +// #include + +// #include + + +#include +#include +#include + +// #include + +// #include + + +namespace nlohmann +{ + +/*! +@brief SAX interface +*/ +template +struct json_sax +{ + /// type for (signed) integers + using number_integer_t = typename BasicJsonType::number_integer_t; + /// type for unsigned integers + using number_unsigned_t = typename BasicJsonType::number_unsigned_t; + /// type for floating-point numbers + using number_float_t = typename BasicJsonType::number_float_t; + /// type for strings + using string_t = typename BasicJsonType::string_t; + + /// constant to indicate that no size limit is given for array or object + static constexpr auto no_limit = std::size_t(-1); + + /*! + @brief a null value was read + @return whether parsing should proceed + */ + virtual bool null() = 0; + + /*! + @brief a boolean value was read + @param[in] val boolean value + @return whether parsing should proceed + */ + virtual bool boolean(bool val) = 0; + + /*! + @brief an integer number was read + @param[in] val integer value + @return whether parsing should proceed + */ + virtual bool number_integer(number_integer_t val) = 0; + + /*! + @brief an unsigned integer number was read + @param[in] val unsigned integer value + @return whether parsing should proceed + */ + virtual bool number_unsigned(number_unsigned_t val) = 0; + + /*! + @brief an floating-point number was read + @param[in] val floating-point value + @param[in] s raw token value + @return whether parsing should proceed + */ + virtual bool number_float(number_float_t val, const string_t& s) = 0; + + /*! + @brief a string was read + @param[in] val string value + @return whether parsing should proceed + */ + virtual bool string(string_t& val) = 0; + + /*! + @brief the beginning of an object was read + @param[in] elements number of object elements or no_limit if unknown + @return whether parsing should proceed + @note binary formats may report the number of elements + */ + virtual bool start_object(std::size_t elements = no_limit) = 0; + + /*! + @brief an object key was read + @param[in] val object key + @return whether parsing should proceed + */ + virtual bool key(string_t& val) = 0; + + /*! + @brief the end of an object was read + @return whether parsing should proceed + */ + virtual bool end_object() = 0; + + /*! + @brief the beginning of an array was read + @param[in] elements number of array elements or no_limit if unknown + @return whether parsing should proceed + @note binary formats may report the number of elements + */ + virtual bool start_array(std::size_t elements = no_limit) = 0; + + /*! + @brief the end of an array was read + @return whether parsing should proceed + */ + virtual bool end_array() = 0; + + /*! + @brief a parse error occurred + @param[in] position the position in the input where the error occurs + @param[in] last_token the last read token + @param[in] error_msg a detailed error message + @return whether parsing should proceed (must return false) + */ + virtual bool parse_error(std::size_t position, + const std::string& last_token, + const detail::exception& ex) = 0; + + virtual ~json_sax() = default; +}; + + +namespace detail +{ +/*! +@brief SAX implementation to create a JSON value from SAX events + +This class implements the @ref json_sax interface and processes the SAX events +to create a JSON value which makes it basically a DOM parser. The structure or +hierarchy of the JSON value is managed by the stack `ref_stack` which contains +a pointer to the respective array or object for each recursion depth. + +After successful parsing, the value that is passed by reference to the +constructor contains the parsed value. + +@tparam BasicJsonType the JSON type +*/ +template +class json_sax_dom_parser : public json_sax +{ + public: + using number_integer_t = typename BasicJsonType::number_integer_t; + using number_unsigned_t = typename BasicJsonType::number_unsigned_t; + using number_float_t = typename BasicJsonType::number_float_t; + using string_t = typename BasicJsonType::string_t; + + /*! + @param[in, out] r reference to a JSON value that is manipulated while + parsing + @param[in] allow_exceptions_ whether parse errors yield exceptions + */ + json_sax_dom_parser(BasicJsonType& r, const bool allow_exceptions_ = true) + : root(r), allow_exceptions(allow_exceptions_) + {} + + bool null() override + { + handle_value(nullptr); + return true; + } + + bool boolean(bool val) override + { + handle_value(val); + return true; + } + + bool number_integer(number_integer_t val) override + { + handle_value(val); + return true; + } + + bool number_unsigned(number_unsigned_t val) override + { + handle_value(val); + return true; + } + + bool number_float(number_float_t val, const string_t&) override + { + handle_value(val); + return true; + } + + bool string(string_t& val) override + { + handle_value(val); + return true; + } + + bool start_object(std::size_t len) override + { + ref_stack.push_back(handle_value(BasicJsonType::value_t::object)); + + if (JSON_UNLIKELY(len != json_sax::no_limit and len > ref_stack.back()->max_size())) + { + JSON_THROW(out_of_range::create(408, + "excessive object size: " + std::to_string(len))); + } + + return true; + } + + bool key(string_t& val) override + { + // add null at given key and store the reference for later + object_element = &(ref_stack.back()->m_value.object->operator[](val)); + return true; + } + + bool end_object() override + { + ref_stack.pop_back(); + return true; + } + + bool start_array(std::size_t len) override + { + ref_stack.push_back(handle_value(BasicJsonType::value_t::array)); + + if (JSON_UNLIKELY(len != json_sax::no_limit and len > ref_stack.back()->max_size())) + { + JSON_THROW(out_of_range::create(408, + "excessive array size: " + std::to_string(len))); + } + + return true; + } + + bool end_array() override + { + ref_stack.pop_back(); + return true; + } + + bool parse_error(std::size_t, const std::string&, + const detail::exception& ex) override + { + errored = true; + if (allow_exceptions) + { + // determine the proper exception type from the id + switch ((ex.id / 100) % 100) + { + case 1: + JSON_THROW(*reinterpret_cast(&ex)); + case 2: + JSON_THROW(*reinterpret_cast(&ex)); // LCOV_EXCL_LINE + case 3: + JSON_THROW(*reinterpret_cast(&ex)); // LCOV_EXCL_LINE + case 4: + JSON_THROW(*reinterpret_cast(&ex)); + case 5: + JSON_THROW(*reinterpret_cast(&ex)); // LCOV_EXCL_LINE + default: + assert(false); // LCOV_EXCL_LINE + } + } + return false; + } + + constexpr bool is_errored() const + { + return errored; + } + + private: + /*! + @invariant If the ref stack is empty, then the passed value will be the new + root. + @invariant If the ref stack contains a value, then it is an array or an + object to which we can add elements + */ + template + BasicJsonType* handle_value(Value&& v) + { + if (ref_stack.empty()) + { + root = BasicJsonType(std::forward(v)); + return &root; + } + else + { + assert(ref_stack.back()->is_array() or ref_stack.back()->is_object()); + if (ref_stack.back()->is_array()) + { + ref_stack.back()->m_value.array->emplace_back(std::forward(v)); + return &(ref_stack.back()->m_value.array->back()); + } + else + { + assert(object_element); + *object_element = BasicJsonType(std::forward(v)); + return object_element; + } + } + } + + /// the parsed JSON value + BasicJsonType& root; + /// stack to model hierarchy of values + std::vector ref_stack; + /// helper to hold the reference for the next object element + BasicJsonType* object_element = nullptr; + /// whether a syntax error occurred + bool errored = false; + /// whether to throw exceptions in case of errors + const bool allow_exceptions = true; +}; + +template +class json_sax_dom_callback_parser : public json_sax +{ + public: + using number_integer_t = typename BasicJsonType::number_integer_t; + using number_unsigned_t = typename BasicJsonType::number_unsigned_t; + using number_float_t = typename BasicJsonType::number_float_t; + using string_t = typename BasicJsonType::string_t; + using parser_callback_t = typename BasicJsonType::parser_callback_t; + using parse_event_t = typename BasicJsonType::parse_event_t; + + json_sax_dom_callback_parser(BasicJsonType& r, + const parser_callback_t cb, + const bool allow_exceptions_ = true) + : root(r), callback(cb), allow_exceptions(allow_exceptions_) + { + keep_stack.push_back(true); + } + + bool null() override + { + handle_value(nullptr); + return true; + } + + bool boolean(bool val) override + { + handle_value(val); + return true; + } + + bool number_integer(number_integer_t val) override + { + handle_value(val); + return true; + } + + bool number_unsigned(number_unsigned_t val) override + { + handle_value(val); + return true; + } + + bool number_float(number_float_t val, const string_t&) override + { + handle_value(val); + return true; + } + + bool string(string_t& val) override + { + handle_value(val); + return true; + } + + bool start_object(std::size_t len) override + { + // check callback for object start + const bool keep = callback(static_cast(ref_stack.size()), parse_event_t::object_start, discarded); + keep_stack.push_back(keep); + + auto val = handle_value(BasicJsonType::value_t::object, true); + ref_stack.push_back(val.second); + + // check object limit + if (ref_stack.back()) + { + if (JSON_UNLIKELY(len != json_sax::no_limit and len > ref_stack.back()->max_size())) + { + JSON_THROW(out_of_range::create(408, + "excessive object size: " + std::to_string(len))); + } + } + + return true; + } + + bool key(string_t& val) override + { + BasicJsonType k = BasicJsonType(val); + + // check callback for key + const bool keep = callback(static_cast(ref_stack.size()), parse_event_t::key, k); + key_keep_stack.push_back(keep); + + // add discarded value at given key and store the reference for later + if (keep and ref_stack.back()) + { + object_element = &(ref_stack.back()->m_value.object->operator[](val) = discarded); + } + + return true; + } + + bool end_object() override + { + bool keep = true; + if (ref_stack.back()) + { + keep = callback(static_cast(ref_stack.size()) - 1, parse_event_t::object_end, *ref_stack.back()); + if (not keep) + { + // discard object + *ref_stack.back() = discarded; + } + } + + assert(not ref_stack.empty()); + assert(not keep_stack.empty()); + ref_stack.pop_back(); + keep_stack.pop_back(); + + if (not ref_stack.empty() and ref_stack.back()) + { + // remove discarded value + if (ref_stack.back()->is_object()) + { + for (auto it = ref_stack.back()->begin(); it != ref_stack.back()->end(); ++it) + { + if (it->is_discarded()) + { + ref_stack.back()->erase(it); + break; + } + } + } + } + + return true; + } + + bool start_array(std::size_t len) override + { + const bool keep = callback(static_cast(ref_stack.size()), parse_event_t::array_start, discarded); + keep_stack.push_back(keep); + + auto val = handle_value(BasicJsonType::value_t::array, true); + ref_stack.push_back(val.second); + + // check array limit + if (ref_stack.back()) + { + if (JSON_UNLIKELY(len != json_sax::no_limit and len > ref_stack.back()->max_size())) + { + JSON_THROW(out_of_range::create(408, + "excessive array size: " + std::to_string(len))); + } + } + + return true; + } + + bool end_array() override + { + bool keep = true; + + if (ref_stack.back()) + { + keep = callback(static_cast(ref_stack.size()) - 1, parse_event_t::array_end, *ref_stack.back()); + if (not keep) + { + // discard array + *ref_stack.back() = discarded; + } + } + + assert(not ref_stack.empty()); + assert(not keep_stack.empty()); + ref_stack.pop_back(); + keep_stack.pop_back(); + + // remove discarded value + if (not keep and not ref_stack.empty()) + { + if (ref_stack.back()->is_array()) + { + ref_stack.back()->m_value.array->pop_back(); + } + } + + return true; + } + + bool parse_error(std::size_t, const std::string&, + const detail::exception& ex) override + { + errored = true; + if (allow_exceptions) + { + // determine the proper exception type from the id + switch ((ex.id / 100) % 100) + { + case 1: + JSON_THROW(*reinterpret_cast(&ex)); + case 2: + JSON_THROW(*reinterpret_cast(&ex)); // LCOV_EXCL_LINE + case 3: + JSON_THROW(*reinterpret_cast(&ex)); // LCOV_EXCL_LINE + case 4: + JSON_THROW(*reinterpret_cast(&ex)); + case 5: + JSON_THROW(*reinterpret_cast(&ex)); // LCOV_EXCL_LINE + default: + assert(false); // LCOV_EXCL_LINE + } + } + return false; + } + + constexpr bool is_errored() const + { + return errored; + } + + private: + /*! + @param[in] v value to add to the JSON value we build during parsing + @param[in] skip_callback whether we should skip calling the callback + function; this is required after start_array() and + start_object() SAX events, because otherwise we would call the + callback function with an empty array or object, respectively. + + @invariant If the ref stack is empty, then the passed value will be the new + root. + @invariant If the ref stack contains a value, then it is an array or an + object to which we can add elements + + @return pair of boolean (whether value should be kept) and pointer (to the + passed value in the ref_stack hierarchy; nullptr if not kept) + */ + template + std::pair handle_value(Value&& v, const bool skip_callback = false) + { + assert(not keep_stack.empty()); + + // do not handle this value if we know it would be added to a discarded + // container + if (not keep_stack.back()) + { + return {false, nullptr}; + } + + // create value + auto value = BasicJsonType(std::forward(v)); + + // check callback + const bool keep = skip_callback or callback(static_cast(ref_stack.size()), parse_event_t::value, value); + + // do not handle this value if we just learnt it shall be discarded + if (not keep) + { + return {false, nullptr}; + } + + if (ref_stack.empty()) + { + root = std::move(value); + return {true, &root}; + } + else + { + assert(ref_stack.back()->is_array() or ref_stack.back()->is_object()); + if (ref_stack.back()->is_array()) + { + ref_stack.back()->m_value.array->push_back(std::move(value)); + return {true, &(ref_stack.back()->m_value.array->back())}; + } + else + { + // check if we should store an element for the current key + assert(not key_keep_stack.empty()); + const bool store_element = key_keep_stack.back(); + key_keep_stack.pop_back(); + + if (not store_element) + { + return {false, nullptr}; + } + + assert(object_element); + *object_element = std::move(value); + return {true, object_element}; + } + } + } + + /// the parsed JSON value + BasicJsonType& root; + /// stack to model hierarchy of values + std::vector ref_stack; + /// stack to manage which values to keep + std::vector keep_stack; + /// stack to manage which object keys to keep + std::vector key_keep_stack; + /// helper to hold the reference for the next object element + BasicJsonType* object_element = nullptr; + /// whether a syntax error occurred + bool errored = false; + /// callback function + const parser_callback_t callback = nullptr; + /// whether to throw exceptions in case of errors + const bool allow_exceptions = true; + /// a discarded value for the callback + BasicJsonType discarded = BasicJsonType::value_t::discarded; +}; + +template +class json_sax_acceptor : public json_sax +{ + public: + using number_integer_t = typename BasicJsonType::number_integer_t; + using number_unsigned_t = typename BasicJsonType::number_unsigned_t; + using number_float_t = typename BasicJsonType::number_float_t; + using string_t = typename BasicJsonType::string_t; + + bool null() override + { + return true; + } + + bool boolean(bool) override + { + return true; + } + + bool number_integer(number_integer_t) override + { + return true; + } + + bool number_unsigned(number_unsigned_t) override + { + return true; + } + + bool number_float(number_float_t, const string_t&) override + { + return true; + } + + bool string(string_t&) override + { + return true; + } + + bool start_object(std::size_t) override + { + return true; + } + + bool key(string_t&) override + { + return true; + } + + bool end_object() override + { + return true; + } + + bool start_array(std::size_t) override + { + return true; + } + + bool end_array() override + { + return true; + } + + bool parse_error(std::size_t, const std::string&, const detail::exception&) override + { + return false; + } +}; +} + +} + +// #include + +// #include + + +namespace nlohmann +{ +namespace detail +{ +//////////// +// parser // +//////////// + +/*! +@brief syntax analysis + +This class implements a recursive decent parser. +*/ +template +class parser +{ + using number_integer_t = typename BasicJsonType::number_integer_t; + using number_unsigned_t = typename BasicJsonType::number_unsigned_t; + using number_float_t = typename BasicJsonType::number_float_t; + using string_t = typename BasicJsonType::string_t; + using lexer_t = lexer; + using token_type = typename lexer_t::token_type; + + public: + enum class parse_event_t : uint8_t + { + /// the parser read `{` and started to process a JSON object + object_start, + /// the parser read `}` and finished processing a JSON object + object_end, + /// the parser read `[` and started to process a JSON array + array_start, + /// the parser read `]` and finished processing a JSON array + array_end, + /// the parser read a key of a value in an object + key, + /// the parser finished reading a JSON value + value + }; + + using json_sax_t = json_sax; + + using parser_callback_t = + std::function; + + /// a parser reading from an input adapter + explicit parser(detail::input_adapter_t&& adapter, + const parser_callback_t cb = nullptr, + const bool allow_exceptions_ = true) + : callback(cb), m_lexer(std::move(adapter)), allow_exceptions(allow_exceptions_) + { + // read first token + get_token(); + } + + /*! + @brief public parser interface + + @param[in] strict whether to expect the last token to be EOF + @param[in,out] result parsed JSON value + + @throw parse_error.101 in case of an unexpected token + @throw parse_error.102 if to_unicode fails or surrogate error + @throw parse_error.103 if to_unicode fails + */ + void parse(const bool strict, BasicJsonType& result) + { + if (callback) + { + json_sax_dom_callback_parser sdp(result, callback, allow_exceptions); + sax_parse_internal(&sdp); + result.assert_invariant(); + + // in strict mode, input must be completely read + if (strict and (get_token() != token_type::end_of_input)) + { + sdp.parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_of_input))); + } + + // in case of an error, return discarded value + if (sdp.is_errored()) + { + result = value_t::discarded; + return; + } + + // set top-level value to null if it was discarded by the callback + // function + if (result.is_discarded()) + { + result = nullptr; + } + } + else + { + json_sax_dom_parser sdp(result, allow_exceptions); + sax_parse_internal(&sdp); + result.assert_invariant(); + + // in strict mode, input must be completely read + if (strict and (get_token() != token_type::end_of_input)) + { + sdp.parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_of_input))); + } + + // in case of an error, return discarded value + if (sdp.is_errored()) + { + result = value_t::discarded; + return; + } + } + } + + /*! + @brief public accept interface + + @param[in] strict whether to expect the last token to be EOF + @return whether the input is a proper JSON text + */ + bool accept(const bool strict = true) + { + json_sax_acceptor sax_acceptor; + return sax_parse(&sax_acceptor, strict); + } + + bool sax_parse(json_sax_t* sax, const bool strict = true) + { + const bool result = sax_parse_internal(sax); + + // strict mode: next byte must be EOF + if (result and strict and (get_token() != token_type::end_of_input)) + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_of_input))); + } + + return result; + } + + private: + bool sax_parse_internal(json_sax_t* sax) + { + // stack to remember the hieararchy of structured values we are parsing + // true = array; false = object + std::vector states; + // value to avoid a goto (see comment where set to true) + bool skip_to_state_evaluation = false; + + while (true) + { + if (not skip_to_state_evaluation) + { + // invariant: get_token() was called before each iteration + switch (last_token) + { + case token_type::begin_object: + { + if (JSON_UNLIKELY(not sax->start_object())) + { + return false; + } + + // closing } -> we are done + if (get_token() == token_type::end_object) + { + if (JSON_UNLIKELY(not sax->end_object())) + { + return false; + } + break; + } + + // parse key + if (JSON_UNLIKELY(last_token != token_type::value_string)) + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::value_string))); + } + else + { + if (JSON_UNLIKELY(not sax->key(m_lexer.get_string()))) + { + return false; + } + } + + // parse separator (:) + if (JSON_UNLIKELY(get_token() != token_type::name_separator)) + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::name_separator))); + } + + // remember we are now inside an object + states.push_back(false); + + // parse values + get_token(); + continue; + } + + case token_type::begin_array: + { + if (JSON_UNLIKELY(not sax->start_array())) + { + return false; + } + + // closing ] -> we are done + if (get_token() == token_type::end_array) + { + if (JSON_UNLIKELY(not sax->end_array())) + { + return false; + } + break; + } + + // remember we are now inside an array + states.push_back(true); + + // parse values (no need to call get_token) + continue; + } + + case token_type::value_float: + { + const auto res = m_lexer.get_number_float(); + + if (JSON_UNLIKELY(not std::isfinite(res))) + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + out_of_range::create(406, "number overflow parsing '" + m_lexer.get_token_string() + "'")); + } + else + { + if (JSON_UNLIKELY(not sax->number_float(res, m_lexer.get_string()))) + { + return false; + } + break; + } + } + + case token_type::literal_false: + { + if (JSON_UNLIKELY(not sax->boolean(false))) + { + return false; + } + break; + } + + case token_type::literal_null: + { + if (JSON_UNLIKELY(not sax->null())) + { + return false; + } + break; + } + + case token_type::literal_true: + { + if (JSON_UNLIKELY(not sax->boolean(true))) + { + return false; + } + break; + } + + case token_type::value_integer: + { + if (JSON_UNLIKELY(not sax->number_integer(m_lexer.get_number_integer()))) + { + return false; + } + break; + } + + case token_type::value_string: + { + if (JSON_UNLIKELY(not sax->string(m_lexer.get_string()))) + { + return false; + } + break; + } + + case token_type::value_unsigned: + { + if (JSON_UNLIKELY(not sax->number_unsigned(m_lexer.get_number_unsigned()))) + { + return false; + } + break; + } + + case token_type::parse_error: + { + // using "uninitialized" to avoid "expected" message + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::uninitialized))); + } + + default: // the last token was unexpected + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::literal_or_value))); + } + } + } + else + { + skip_to_state_evaluation = false; + } + + // we reached this line after we successfully parsed a value + if (states.empty()) + { + // empty stack: we reached the end of the hieararchy: done + return true; + } + else + { + if (states.back()) // array + { + // comma -> next value + if (get_token() == token_type::value_separator) + { + // parse a new value + get_token(); + continue; + } + + // closing ] + if (JSON_LIKELY(last_token == token_type::end_array)) + { + if (JSON_UNLIKELY(not sax->end_array())) + { + return false; + } + + // We are done with this array. Before we can parse a + // new value, we need to evaluate the new state first. + // By setting skip_to_state_evaluation to false, we + // are effectively jumping to the beginning of this if. + assert(not states.empty()); + states.pop_back(); + skip_to_state_evaluation = true; + continue; + } + else + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_array))); + } + } + else // object + { + // comma -> next value + if (get_token() == token_type::value_separator) + { + // parse key + if (JSON_UNLIKELY(get_token() != token_type::value_string)) + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::value_string))); + } + else + { + if (JSON_UNLIKELY(not sax->key(m_lexer.get_string()))) + { + return false; + } + } + + // parse separator (:) + if (JSON_UNLIKELY(get_token() != token_type::name_separator)) + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::name_separator))); + } + + // parse values + get_token(); + continue; + } + + // closing } + if (JSON_LIKELY(last_token == token_type::end_object)) + { + if (JSON_UNLIKELY(not sax->end_object())) + { + return false; + } + + // We are done with this object. Before we can parse a + // new value, we need to evaluate the new state first. + // By setting skip_to_state_evaluation to false, we + // are effectively jumping to the beginning of this if. + assert(not states.empty()); + states.pop_back(); + skip_to_state_evaluation = true; + continue; + } + else + { + return sax->parse_error(m_lexer.get_position(), + m_lexer.get_token_string(), + parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_object))); + } + } + } + } + } + + /// get next token from lexer + token_type get_token() + { + return (last_token = m_lexer.scan()); + } + + std::string exception_message(const token_type expected) + { + std::string error_msg = "syntax error - "; + if (last_token == token_type::parse_error) + { + error_msg += std::string(m_lexer.get_error_message()) + "; last read: '" + + m_lexer.get_token_string() + "'"; + } + else + { + error_msg += "unexpected " + std::string(lexer_t::token_type_name(last_token)); + } + + if (expected != token_type::uninitialized) + { + error_msg += "; expected " + std::string(lexer_t::token_type_name(expected)); + } + + return error_msg; + } + + private: + /// callback function + const parser_callback_t callback = nullptr; + /// the type of the last read token + token_type last_token = token_type::uninitialized; + /// the lexer + lexer_t m_lexer; + /// whether to throw exceptions in case of errors + const bool allow_exceptions = true; +}; +} +} + +// #include + + +#include // ptrdiff_t +#include // numeric_limits + +namespace nlohmann +{ +namespace detail +{ +/* +@brief an iterator for primitive JSON types + +This class models an iterator for primitive JSON types (boolean, number, +string). It's only purpose is to allow the iterator/const_iterator classes +to "iterate" over primitive values. Internally, the iterator is modeled by +a `difference_type` variable. Value begin_value (`0`) models the begin, +end_value (`1`) models past the end. +*/ +class primitive_iterator_t +{ + private: + using difference_type = std::ptrdiff_t; + static constexpr difference_type begin_value = 0; + static constexpr difference_type end_value = begin_value + 1; + + /// iterator as signed integer type + difference_type m_it = (std::numeric_limits::min)(); + + public: + constexpr difference_type get_value() const noexcept + { + return m_it; + } + + /// set iterator to a defined beginning + void set_begin() noexcept + { + m_it = begin_value; + } + + /// set iterator to a defined past the end + void set_end() noexcept + { + m_it = end_value; + } + + /// return whether the iterator can be dereferenced + constexpr bool is_begin() const noexcept + { + return m_it == begin_value; + } + + /// return whether the iterator is at end + constexpr bool is_end() const noexcept + { + return m_it == end_value; + } + + friend constexpr bool operator==(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept + { + return lhs.m_it == rhs.m_it; + } + + friend constexpr bool operator<(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept + { + return lhs.m_it < rhs.m_it; + } + + primitive_iterator_t operator+(difference_type n) noexcept + { + auto result = *this; + result += n; + return result; + } + + friend constexpr difference_type operator-(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept + { + return lhs.m_it - rhs.m_it; + } + + primitive_iterator_t& operator++() noexcept + { + ++m_it; + return *this; + } + + primitive_iterator_t const operator++(int) noexcept + { + auto result = *this; + m_it++; + return result; + } + + primitive_iterator_t& operator--() noexcept + { + --m_it; + return *this; + } + + primitive_iterator_t const operator--(int) noexcept + { + auto result = *this; + m_it--; + return result; + } + + primitive_iterator_t& operator+=(difference_type n) noexcept + { + m_it += n; + return *this; + } + + primitive_iterator_t& operator-=(difference_type n) noexcept + { + m_it -= n; + return *this; + } +}; +} +} + +// #include + + +// #include + + +namespace nlohmann +{ +namespace detail +{ +/*! +@brief an iterator value + +@note This structure could easily be a union, but MSVC currently does not allow +unions members with complex constructors, see https://github.com/nlohmann/json/pull/105. +*/ +template struct internal_iterator +{ + /// iterator for JSON objects + typename BasicJsonType::object_t::iterator object_iterator {}; + /// iterator for JSON arrays + typename BasicJsonType::array_t::iterator array_iterator {}; + /// generic iterator for all other types + primitive_iterator_t primitive_iterator {}; +}; +} +} + +// #include + + +#include // not +#include // iterator, random_access_iterator_tag, bidirectional_iterator_tag, advance, next +#include // conditional, is_const, remove_const + +// #include + +// #include + +// #include + +// #include + +// #include + +// #include + + +namespace nlohmann +{ +namespace detail +{ +// forward declare, to be able to friend it later on +template class iteration_proxy; + +/*! +@brief a template for a bidirectional iterator for the @ref basic_json class + +This class implements a both iterators (iterator and const_iterator) for the +@ref basic_json class. + +@note An iterator is called *initialized* when a pointer to a JSON value has + been set (e.g., by a constructor or a copy assignment). If the iterator is + default-constructed, it is *uninitialized* and most methods are undefined. + **The library uses assertions to detect calls on uninitialized iterators.** + +@requirement The class satisfies the following concept requirements: +- +[BidirectionalIterator](http://en.cppreference.com/w/cpp/concept/BidirectionalIterator): + The iterator that can be moved can be moved in both directions (i.e. + incremented and decremented). + +@since version 1.0.0, simplified in version 2.0.9, change to bidirectional + iterators in version 3.0.0 (see https://github.com/nlohmann/json/issues/593) +*/ +template +class iter_impl +{ + /// allow basic_json to access private members + friend iter_impl::value, typename std::remove_const::type, const BasicJsonType>::type>; + friend BasicJsonType; + friend iteration_proxy; + + using object_t = typename BasicJsonType::object_t; + using array_t = typename BasicJsonType::array_t; + // make sure BasicJsonType is basic_json or const basic_json + static_assert(is_basic_json::type>::value, + "iter_impl only accepts (const) basic_json"); + + public: + + /// The std::iterator class template (used as a base class to provide typedefs) is deprecated in C++17. + /// The C++ Standard has never required user-defined iterators to derive from std::iterator. + /// A user-defined iterator should provide publicly accessible typedefs named + /// iterator_category, value_type, difference_type, pointer, and reference. + /// Note that value_type is required to be non-const, even for constant iterators. + using iterator_category = std::bidirectional_iterator_tag; + + /// the type of the values when the iterator is dereferenced + using value_type = typename BasicJsonType::value_type; + /// a type to represent differences between iterators + using difference_type = typename BasicJsonType::difference_type; + /// defines a pointer to the type iterated over (value_type) + using pointer = typename std::conditional::value, + typename BasicJsonType::const_pointer, + typename BasicJsonType::pointer>::type; + /// defines a reference to the type iterated over (value_type) + using reference = + typename std::conditional::value, + typename BasicJsonType::const_reference, + typename BasicJsonType::reference>::type; + + /// default constructor + iter_impl() = default; + + /*! + @brief constructor for a given JSON instance + @param[in] object pointer to a JSON object for this iterator + @pre object != nullptr + @post The iterator is initialized; i.e. `m_object != nullptr`. + */ + explicit iter_impl(pointer object) noexcept : m_object(object) + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + { + m_it.object_iterator = typename object_t::iterator(); + break; + } + + case value_t::array: + { + m_it.array_iterator = typename array_t::iterator(); + break; + } + + default: + { + m_it.primitive_iterator = primitive_iterator_t(); + break; + } + } + } + + /*! + @note The conventional copy constructor and copy assignment are implicitly + defined. Combined with the following converting constructor and + assignment, they support: (1) copy from iterator to iterator, (2) + copy from const iterator to const iterator, and (3) conversion from + iterator to const iterator. However conversion from const iterator + to iterator is not defined. + */ + + /*! + @brief converting constructor + @param[in] other non-const iterator to copy from + @note It is not checked whether @a other is initialized. + */ + iter_impl(const iter_impl::type>& other) noexcept + : m_object(other.m_object), m_it(other.m_it) {} + + /*! + @brief converting assignment + @param[in,out] other non-const iterator to copy from + @return const/non-const iterator + @note It is not checked whether @a other is initialized. + */ + iter_impl& operator=(const iter_impl::type>& other) noexcept + { + m_object = other.m_object; + m_it = other.m_it; + return *this; + } + + private: + /*! + @brief set the iterator to the first value + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + void set_begin() noexcept + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + { + m_it.object_iterator = m_object->m_value.object->begin(); + break; + } + + case value_t::array: + { + m_it.array_iterator = m_object->m_value.array->begin(); + break; + } + + case value_t::null: + { + // set to end so begin()==end() is true: null is empty + m_it.primitive_iterator.set_end(); + break; + } + + default: + { + m_it.primitive_iterator.set_begin(); + break; + } + } + } + + /*! + @brief set the iterator past the last value + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + void set_end() noexcept + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + { + m_it.object_iterator = m_object->m_value.object->end(); + break; + } + + case value_t::array: + { + m_it.array_iterator = m_object->m_value.array->end(); + break; + } + + default: + { + m_it.primitive_iterator.set_end(); + break; + } + } + } + + public: + /*! + @brief return a reference to the value pointed to by the iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + reference operator*() const + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + { + assert(m_it.object_iterator != m_object->m_value.object->end()); + return m_it.object_iterator->second; + } + + case value_t::array: + { + assert(m_it.array_iterator != m_object->m_value.array->end()); + return *m_it.array_iterator; + } + + case value_t::null: + JSON_THROW(invalid_iterator::create(214, "cannot get value")); + + default: + { + if (JSON_LIKELY(m_it.primitive_iterator.is_begin())) + { + return *m_object; + } + + JSON_THROW(invalid_iterator::create(214, "cannot get value")); + } + } + } + + /*! + @brief dereference the iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + pointer operator->() const + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + { + assert(m_it.object_iterator != m_object->m_value.object->end()); + return &(m_it.object_iterator->second); + } + + case value_t::array: + { + assert(m_it.array_iterator != m_object->m_value.array->end()); + return &*m_it.array_iterator; + } + + default: + { + if (JSON_LIKELY(m_it.primitive_iterator.is_begin())) + { + return m_object; + } + + JSON_THROW(invalid_iterator::create(214, "cannot get value")); + } + } + } + + /*! + @brief post-increment (it++) + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + iter_impl const operator++(int) + { + auto result = *this; + ++(*this); + return result; + } + + /*! + @brief pre-increment (++it) + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + iter_impl& operator++() + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + { + std::advance(m_it.object_iterator, 1); + break; + } + + case value_t::array: + { + std::advance(m_it.array_iterator, 1); + break; + } + + default: + { + ++m_it.primitive_iterator; + break; + } + } + + return *this; + } + + /*! + @brief post-decrement (it--) + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + iter_impl const operator--(int) + { + auto result = *this; + --(*this); + return result; + } + + /*! + @brief pre-decrement (--it) + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + iter_impl& operator--() + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + { + std::advance(m_it.object_iterator, -1); + break; + } + + case value_t::array: + { + std::advance(m_it.array_iterator, -1); + break; + } + + default: + { + --m_it.primitive_iterator; + break; + } + } + + return *this; + } + + /*! + @brief comparison: equal + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + bool operator==(const iter_impl& other) const + { + // if objects are not the same, the comparison is undefined + if (JSON_UNLIKELY(m_object != other.m_object)) + { + JSON_THROW(invalid_iterator::create(212, "cannot compare iterators of different containers")); + } + + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + return (m_it.object_iterator == other.m_it.object_iterator); + + case value_t::array: + return (m_it.array_iterator == other.m_it.array_iterator); + + default: + return (m_it.primitive_iterator == other.m_it.primitive_iterator); + } + } + + /*! + @brief comparison: not equal + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + bool operator!=(const iter_impl& other) const + { + return not operator==(other); + } + + /*! + @brief comparison: smaller + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + bool operator<(const iter_impl& other) const + { + // if objects are not the same, the comparison is undefined + if (JSON_UNLIKELY(m_object != other.m_object)) + { + JSON_THROW(invalid_iterator::create(212, "cannot compare iterators of different containers")); + } + + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + JSON_THROW(invalid_iterator::create(213, "cannot compare order of object iterators")); + + case value_t::array: + return (m_it.array_iterator < other.m_it.array_iterator); + + default: + return (m_it.primitive_iterator < other.m_it.primitive_iterator); + } + } + + /*! + @brief comparison: less than or equal + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + bool operator<=(const iter_impl& other) const + { + return not other.operator < (*this); + } + + /*! + @brief comparison: greater than + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + bool operator>(const iter_impl& other) const + { + return not operator<=(other); + } + + /*! + @brief comparison: greater than or equal + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + bool operator>=(const iter_impl& other) const + { + return not operator<(other); + } + + /*! + @brief add to iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + iter_impl& operator+=(difference_type i) + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + JSON_THROW(invalid_iterator::create(209, "cannot use offsets with object iterators")); + + case value_t::array: + { + std::advance(m_it.array_iterator, i); + break; + } + + default: + { + m_it.primitive_iterator += i; + break; + } + } + + return *this; + } + + /*! + @brief subtract from iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + iter_impl& operator-=(difference_type i) + { + return operator+=(-i); + } + + /*! + @brief add to iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + iter_impl operator+(difference_type i) const + { + auto result = *this; + result += i; + return result; + } + + /*! + @brief addition of distance and iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + friend iter_impl operator+(difference_type i, const iter_impl& it) + { + auto result = it; + result += i; + return result; + } + + /*! + @brief subtract from iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + iter_impl operator-(difference_type i) const + { + auto result = *this; + result -= i; + return result; + } + + /*! + @brief return difference + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + difference_type operator-(const iter_impl& other) const + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + JSON_THROW(invalid_iterator::create(209, "cannot use offsets with object iterators")); + + case value_t::array: + return m_it.array_iterator - other.m_it.array_iterator; + + default: + return m_it.primitive_iterator - other.m_it.primitive_iterator; + } + } + + /*! + @brief access to successor + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + reference operator[](difference_type n) const + { + assert(m_object != nullptr); + + switch (m_object->m_type) + { + case value_t::object: + JSON_THROW(invalid_iterator::create(208, "cannot use operator[] for object iterators")); + + case value_t::array: + return *std::next(m_it.array_iterator, n); + + case value_t::null: + JSON_THROW(invalid_iterator::create(214, "cannot get value")); + + default: + { + if (JSON_LIKELY(m_it.primitive_iterator.get_value() == -n)) + { + return *m_object; + } + + JSON_THROW(invalid_iterator::create(214, "cannot get value")); + } + } + } + + /*! + @brief return the key of an object iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + const typename object_t::key_type& key() const + { + assert(m_object != nullptr); + + if (JSON_LIKELY(m_object->is_object())) + { + return m_it.object_iterator->first; + } + + JSON_THROW(invalid_iterator::create(207, "cannot use key() for non-object iterators")); + } + + /*! + @brief return the value of an iterator + @pre The iterator is initialized; i.e. `m_object != nullptr`. + */ + reference value() const + { + return operator*(); + } + + private: + /// associated JSON instance + pointer m_object = nullptr; + /// the actual iterator of the associated instance + internal_iterator::type> m_it; +}; +} +} + +// #include + + +#include // size_t +#include // string, to_string + +// #include + + +namespace nlohmann +{ +namespace detail +{ +/// proxy class for the items() function +template class iteration_proxy +{ + private: + /// helper class for iteration + class iteration_proxy_internal + { + private: + /// the iterator + IteratorType anchor; + /// an index for arrays (used to create key names) + std::size_t array_index = 0; + /// last stringified array index + mutable std::size_t array_index_last = 0; + /// a string representation of the array index + mutable std::string array_index_str = "0"; + /// an empty string (to return a reference for primitive values) + const std::string empty_str = ""; + + public: + explicit iteration_proxy_internal(IteratorType it) noexcept : anchor(it) {} + + /// dereference operator (needed for range-based for) + iteration_proxy_internal& operator*() + { + return *this; + } + + /// increment operator (needed for range-based for) + iteration_proxy_internal& operator++() + { + ++anchor; + ++array_index; + + return *this; + } + + /// inequality operator (needed for range-based for) + bool operator!=(const iteration_proxy_internal& o) const noexcept + { + return anchor != o.anchor; + } + + /// return key of the iterator + const std::string& key() const + { + assert(anchor.m_object != nullptr); + + switch (anchor.m_object->type()) + { + // use integer array index as key + case value_t::array: + { + if (array_index != array_index_last) + { + array_index_str = std::to_string(array_index); + array_index_last = array_index; + } + return array_index_str; + } + + // use key from the object + case value_t::object: + return anchor.key(); + + // use an empty key for all primitive types + default: + return empty_str; + } + } + + /// return value of the iterator + typename IteratorType::reference value() const + { + return anchor.value(); + } + }; + + /// the container to iterate + typename IteratorType::reference container; + + public: + /// construct iteration proxy from a container + explicit iteration_proxy(typename IteratorType::reference cont) noexcept + : container(cont) {} + + /// return iterator begin (needed for range-based for) + iteration_proxy_internal begin() noexcept + { + return iteration_proxy_internal(container.begin()); + } + + /// return iterator end (needed for range-based for) + iteration_proxy_internal end() noexcept + { + return iteration_proxy_internal(container.end()); + } +}; +} +} + +// #include + + +#include // ptrdiff_t +#include // reverse_iterator +#include // declval + +namespace nlohmann +{ +namespace detail +{ +////////////////////// +// reverse_iterator // +////////////////////// + +/*! +@brief a template for a reverse iterator class + +@tparam Base the base iterator type to reverse. Valid types are @ref +iterator (to create @ref reverse_iterator) and @ref const_iterator (to +create @ref const_reverse_iterator). + +@requirement The class satisfies the following concept requirements: +- +[BidirectionalIterator](http://en.cppreference.com/w/cpp/concept/BidirectionalIterator): + The iterator that can be moved can be moved in both directions (i.e. + incremented and decremented). +- [OutputIterator](http://en.cppreference.com/w/cpp/concept/OutputIterator): + It is possible to write to the pointed-to element (only if @a Base is + @ref iterator). + +@since version 1.0.0 +*/ +template +class json_reverse_iterator : public std::reverse_iterator +{ + public: + using difference_type = std::ptrdiff_t; + /// shortcut to the reverse iterator adapter + using base_iterator = std::reverse_iterator; + /// the reference type for the pointed-to element + using reference = typename Base::reference; + + /// create reverse iterator from iterator + json_reverse_iterator(const typename base_iterator::iterator_type& it) noexcept + : base_iterator(it) {} + + /// create reverse iterator from base class + json_reverse_iterator(const base_iterator& it) noexcept : base_iterator(it) {} + + /// post-increment (it++) + json_reverse_iterator const operator++(int) + { + return static_cast(base_iterator::operator++(1)); + } + + /// pre-increment (++it) + json_reverse_iterator& operator++() + { + return static_cast(base_iterator::operator++()); + } + + /// post-decrement (it--) + json_reverse_iterator const operator--(int) + { + return static_cast(base_iterator::operator--(1)); + } + + /// pre-decrement (--it) + json_reverse_iterator& operator--() + { + return static_cast(base_iterator::operator--()); + } + + /// add to iterator + json_reverse_iterator& operator+=(difference_type i) + { + return static_cast(base_iterator::operator+=(i)); + } + + /// add to iterator + json_reverse_iterator operator+(difference_type i) const + { + return static_cast(base_iterator::operator+(i)); + } + + /// subtract from iterator + json_reverse_iterator operator-(difference_type i) const + { + return static_cast(base_iterator::operator-(i)); + } + + /// return difference + difference_type operator-(const json_reverse_iterator& other) const + { + return base_iterator(*this) - base_iterator(other); + } + + /// access to successor + reference operator[](difference_type n) const + { + return *(this->operator+(n)); + } + + /// return the key of an object iterator + auto key() const -> decltype(std::declval().key()) + { + auto it = --this->base(); + return it.key(); + } + + /// return the value of an iterator + reference value() const + { + auto it = --this->base(); + return it.operator * (); + } +}; +} +} + +// #include + + +#include // copy +#include // size_t +#include // streamsize +#include // back_inserter +#include // shared_ptr, make_shared +#include // basic_ostream +#include // basic_string +#include // vector + +namespace nlohmann +{ +namespace detail +{ +/// abstract output adapter interface +template struct output_adapter_protocol +{ + virtual void write_character(CharType c) = 0; + virtual void write_characters(const CharType* s, std::size_t length) = 0; + virtual ~output_adapter_protocol() = default; +}; + +/// a type to simplify interfaces +template +using output_adapter_t = std::shared_ptr>; + +/// output adapter for byte vectors +template +class output_vector_adapter : public output_adapter_protocol +{ + public: + explicit output_vector_adapter(std::vector& vec) : v(vec) {} + + void write_character(CharType c) override + { + v.push_back(c); + } + + void write_characters(const CharType* s, std::size_t length) override + { + std::copy(s, s + length, std::back_inserter(v)); + } + + private: + std::vector& v; +}; + +/// output adapter for output streams +template +class output_stream_adapter : public output_adapter_protocol +{ + public: + explicit output_stream_adapter(std::basic_ostream& s) : stream(s) {} + + void write_character(CharType c) override + { + stream.put(c); + } + + void write_characters(const CharType* s, std::size_t length) override + { + stream.write(s, static_cast(length)); + } + + private: + std::basic_ostream& stream; +}; + +/// output adapter for basic_string +template> +class output_string_adapter : public output_adapter_protocol +{ + public: + explicit output_string_adapter(StringType& s) : str(s) {} + + void write_character(CharType c) override + { + str.push_back(c); + } + + void write_characters(const CharType* s, std::size_t length) override + { + str.append(s, length); + } + + private: + StringType& str; +}; + +template> +class output_adapter +{ + public: + output_adapter(std::vector& vec) + : oa(std::make_shared>(vec)) {} + + output_adapter(std::basic_ostream& s) + : oa(std::make_shared>(s)) {} + + output_adapter(StringType& s) + : oa(std::make_shared>(s)) {} + + operator output_adapter_t() + { + return oa; + } + + private: + output_adapter_t oa = nullptr; +}; +} +} + +// #include + + +#include // generate_n +#include // array +#include // assert +#include // ldexp +#include // size_t +#include // uint8_t, uint16_t, uint32_t, uint64_t +#include // memcpy +#include // setw, setfill +#include // hex +#include // back_inserter +#include // numeric_limits +#include // stringstream +#include // char_traits, string +#include // make_pair, move + +// #include + +// #include + +// #include + +// #include + +// #include + + +namespace nlohmann +{ +namespace detail +{ +/////////////////// +// binary reader // +/////////////////// + +/*! +@brief deserialization of CBOR, MessagePack, and UBJSON values +*/ +template +class binary_reader +{ + using number_integer_t = typename BasicJsonType::number_integer_t; + using number_unsigned_t = typename BasicJsonType::number_unsigned_t; + using string_t = typename BasicJsonType::string_t; + using json_sax_t = json_sax; + + public: + /*! + @brief create a binary reader + + @param[in] adapter input adapter to read from + */ + explicit binary_reader(input_adapter_t adapter) : ia(std::move(adapter)) + { + assert(ia); + } + + /*! + @param[in] format the binary format to parse + @param[in] sax_ a SAX event processor + @param[in] strict whether to expect the input to be consumed completed + + @return + */ + bool sax_parse(const input_format_t format, + json_sax_t* sax_, + const bool strict = true) + { + sax = sax_; + bool result; + + switch (format) + { + case input_format_t::cbor: + result = parse_cbor_internal(); + break; + + case input_format_t::msgpack: + result = parse_msgpack_internal(); + break; + + case input_format_t::ubjson: + result = parse_ubjson_internal(); + break; + + default: + assert(false); // LCOV_EXCL_LINE + } + + // strict mode: next byte must be EOF + if (result and strict) + { + if (format == input_format_t::ubjson) + { + get_ignore_noop(); + } + else + { + get(); + } + + if (JSON_UNLIKELY(current != std::char_traits::eof())) + { + return sax->parse_error(chars_read, get_token_string(), parse_error::create(110, chars_read, "expected end of input")); + } + } + + return result; + } + + /*! + @brief determine system byte order + + @return true if and only if system's byte order is little endian + + @note from http://stackoverflow.com/a/1001328/266378 + */ + static constexpr bool little_endianess(int num = 1) noexcept + { + return (*reinterpret_cast(&num) == 1); + } + + private: + /*! + @param[in] get_char whether a new character should be retrieved from the + input (true, default) or whether the last read + character should be considered instead + + @return whether a valid CBOR value was passed to the SAX parser + */ + bool parse_cbor_internal(const bool get_char = true) + { + switch (get_char ? get() : current) + { + // EOF + case std::char_traits::eof(): + return unexpect_eof(); + + // Integer 0x00..0x17 (0..23) + case 0x00: + case 0x01: + case 0x02: + case 0x03: + case 0x04: + case 0x05: + case 0x06: + case 0x07: + case 0x08: + case 0x09: + case 0x0A: + case 0x0B: + case 0x0C: + case 0x0D: + case 0x0E: + case 0x0F: + case 0x10: + case 0x11: + case 0x12: + case 0x13: + case 0x14: + case 0x15: + case 0x16: + case 0x17: + return sax->number_unsigned(static_cast(current)); + + case 0x18: // Unsigned integer (one-byte uint8_t follows) + { + uint8_t number; + return get_number(number) and sax->number_unsigned(number); + } + + case 0x19: // Unsigned integer (two-byte uint16_t follows) + { + uint16_t number; + return get_number(number) and sax->number_unsigned(number); + } + + case 0x1A: // Unsigned integer (four-byte uint32_t follows) + { + uint32_t number; + return get_number(number) and sax->number_unsigned(number); + } + + case 0x1B: // Unsigned integer (eight-byte uint64_t follows) + { + uint64_t number; + return get_number(number) and sax->number_unsigned(number); + } + + // Negative integer -1-0x00..-1-0x17 (-1..-24) + case 0x20: + case 0x21: + case 0x22: + case 0x23: + case 0x24: + case 0x25: + case 0x26: + case 0x27: + case 0x28: + case 0x29: + case 0x2A: + case 0x2B: + case 0x2C: + case 0x2D: + case 0x2E: + case 0x2F: + case 0x30: + case 0x31: + case 0x32: + case 0x33: + case 0x34: + case 0x35: + case 0x36: + case 0x37: + return sax->number_integer(static_cast(0x20 - 1 - current)); + + case 0x38: // Negative integer (one-byte uint8_t follows) + { + uint8_t number; + return get_number(number) and sax->number_integer(static_cast(-1) - number); + } + + case 0x39: // Negative integer -1-n (two-byte uint16_t follows) + { + uint16_t number; + return get_number(number) and sax->number_integer(static_cast(-1) - number); + } + + case 0x3A: // Negative integer -1-n (four-byte uint32_t follows) + { + uint32_t number; + return get_number(number) and sax->number_integer(static_cast(-1) - number); + } + + case 0x3B: // Negative integer -1-n (eight-byte uint64_t follows) + { + uint64_t number; + return get_number(number) and sax->number_integer(static_cast(-1) + - static_cast(number)); + } + + // UTF-8 string (0x00..0x17 bytes follow) + case 0x60: + case 0x61: + case 0x62: + case 0x63: + case 0x64: + case 0x65: + case 0x66: + case 0x67: + case 0x68: + case 0x69: + case 0x6A: + case 0x6B: + case 0x6C: + case 0x6D: + case 0x6E: + case 0x6F: + case 0x70: + case 0x71: + case 0x72: + case 0x73: + case 0x74: + case 0x75: + case 0x76: + case 0x77: + case 0x78: // UTF-8 string (one-byte uint8_t for n follows) + case 0x79: // UTF-8 string (two-byte uint16_t for n follow) + case 0x7A: // UTF-8 string (four-byte uint32_t for n follow) + case 0x7B: // UTF-8 string (eight-byte uint64_t for n follow) + case 0x7F: // UTF-8 string (indefinite length) + { + string_t s; + return get_cbor_string(s) and sax->string(s); + } + + // array (0x00..0x17 data items follow) + case 0x80: + case 0x81: + case 0x82: + case 0x83: + case 0x84: + case 0x85: + case 0x86: + case 0x87: + case 0x88: + case 0x89: + case 0x8A: + case 0x8B: + case 0x8C: + case 0x8D: + case 0x8E: + case 0x8F: + case 0x90: + case 0x91: + case 0x92: + case 0x93: + case 0x94: + case 0x95: + case 0x96: + case 0x97: + return get_cbor_array(static_cast(current & 0x1F)); + + case 0x98: // array (one-byte uint8_t for n follows) + { + uint8_t len; + return get_number(len) and get_cbor_array(static_cast(len)); + } + + case 0x99: // array (two-byte uint16_t for n follow) + { + uint16_t len; + return get_number(len) and get_cbor_array(static_cast(len)); + } + + case 0x9A: // array (four-byte uint32_t for n follow) + { + uint32_t len; + return get_number(len) and get_cbor_array(static_cast(len)); + } + + case 0x9B: // array (eight-byte uint64_t for n follow) + { + uint64_t len; + return get_number(len) and get_cbor_array(static_cast(len)); + } + + case 0x9F: // array (indefinite length) + return get_cbor_array(json_sax_t::no_limit); + + // map (0x00..0x17 pairs of data items follow) + case 0xA0: + case 0xA1: + case 0xA2: + case 0xA3: + case 0xA4: + case 0xA5: + case 0xA6: + case 0xA7: + case 0xA8: + case 0xA9: + case 0xAA: + case 0xAB: + case 0xAC: + case 0xAD: + case 0xAE: + case 0xAF: + case 0xB0: + case 0xB1: + case 0xB2: + case 0xB3: + case 0xB4: + case 0xB5: + case 0xB6: + case 0xB7: + return get_cbor_object(static_cast(current & 0x1F)); + + case 0xB8: // map (one-byte uint8_t for n follows) + { + uint8_t len; + return get_number(len) and get_cbor_object(static_cast(len)); + } + + case 0xB9: // map (two-byte uint16_t for n follow) + { + uint16_t len; + return get_number(len) and get_cbor_object(static_cast(len)); + } + + case 0xBA: // map (four-byte uint32_t for n follow) + { + uint32_t len; + return get_number(len) and get_cbor_object(static_cast(len)); + } + + case 0xBB: // map (eight-byte uint64_t for n follow) + { + uint64_t len; + return get_number(len) and get_cbor_object(static_cast(len)); + } + + case 0xBF: // map (indefinite length) + return get_cbor_object(json_sax_t::no_limit); + + case 0xF4: // false + return sax->boolean(false); + + case 0xF5: // true + return sax->boolean(true); + + case 0xF6: // null + return sax->null(); + + case 0xF9: // Half-Precision Float (two-byte IEEE 754) + { + const int byte1 = get(); + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + const int byte2 = get(); + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + + // code from RFC 7049, Appendix D, Figure 3: + // As half-precision floating-point numbers were only added + // to IEEE 754 in 2008, today's programming platforms often + // still only have limited support for them. It is very + // easy to include at least decoding support for them even + // without such support. An example of a small decoder for + // half-precision floating-point numbers in the C language + // is shown in Fig. 3. + const int half = (byte1 << 8) + byte2; + const int exp = (half >> 10) & 0x1F; + const int mant = half & 0x3FF; + double val; + if (exp == 0) + { + val = std::ldexp(mant, -24); + } + else if (exp != 31) + { + val = std::ldexp(mant + 1024, exp - 25); + } + else + { + val = (mant == 0) ? std::numeric_limits::infinity() + : std::numeric_limits::quiet_NaN(); + } + return sax->number_float((half & 0x8000) != 0 ? -val : val, ""); + } + + case 0xFA: // Single-Precision Float (four-byte IEEE 754) + { + float number; + return get_number(number) and sax->number_float(static_cast(number), ""); + } + + case 0xFB: // Double-Precision Float (eight-byte IEEE 754) + { + double number; + return get_number(number) and sax->number_float(number, ""); + } + + default: // anything else (0xFF is handled inside the other types) + { + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read, "error reading CBOR; last byte: 0x" + last_token)); + } + } + } + + /*! + @return whether a valid MessagePack value was passed to the SAX parser + */ + bool parse_msgpack_internal() + { + switch (get()) + { + // EOF + case std::char_traits::eof(): + return unexpect_eof(); + + // positive fixint + case 0x00: + case 0x01: + case 0x02: + case 0x03: + case 0x04: + case 0x05: + case 0x06: + case 0x07: + case 0x08: + case 0x09: + case 0x0A: + case 0x0B: + case 0x0C: + case 0x0D: + case 0x0E: + case 0x0F: + case 0x10: + case 0x11: + case 0x12: + case 0x13: + case 0x14: + case 0x15: + case 0x16: + case 0x17: + case 0x18: + case 0x19: + case 0x1A: + case 0x1B: + case 0x1C: + case 0x1D: + case 0x1E: + case 0x1F: + case 0x20: + case 0x21: + case 0x22: + case 0x23: + case 0x24: + case 0x25: + case 0x26: + case 0x27: + case 0x28: + case 0x29: + case 0x2A: + case 0x2B: + case 0x2C: + case 0x2D: + case 0x2E: + case 0x2F: + case 0x30: + case 0x31: + case 0x32: + case 0x33: + case 0x34: + case 0x35: + case 0x36: + case 0x37: + case 0x38: + case 0x39: + case 0x3A: + case 0x3B: + case 0x3C: + case 0x3D: + case 0x3E: + case 0x3F: + case 0x40: + case 0x41: + case 0x42: + case 0x43: + case 0x44: + case 0x45: + case 0x46: + case 0x47: + case 0x48: + case 0x49: + case 0x4A: + case 0x4B: + case 0x4C: + case 0x4D: + case 0x4E: + case 0x4F: + case 0x50: + case 0x51: + case 0x52: + case 0x53: + case 0x54: + case 0x55: + case 0x56: + case 0x57: + case 0x58: + case 0x59: + case 0x5A: + case 0x5B: + case 0x5C: + case 0x5D: + case 0x5E: + case 0x5F: + case 0x60: + case 0x61: + case 0x62: + case 0x63: + case 0x64: + case 0x65: + case 0x66: + case 0x67: + case 0x68: + case 0x69: + case 0x6A: + case 0x6B: + case 0x6C: + case 0x6D: + case 0x6E: + case 0x6F: + case 0x70: + case 0x71: + case 0x72: + case 0x73: + case 0x74: + case 0x75: + case 0x76: + case 0x77: + case 0x78: + case 0x79: + case 0x7A: + case 0x7B: + case 0x7C: + case 0x7D: + case 0x7E: + case 0x7F: + return sax->number_unsigned(static_cast(current)); + + // fixmap + case 0x80: + case 0x81: + case 0x82: + case 0x83: + case 0x84: + case 0x85: + case 0x86: + case 0x87: + case 0x88: + case 0x89: + case 0x8A: + case 0x8B: + case 0x8C: + case 0x8D: + case 0x8E: + case 0x8F: + return get_msgpack_object(static_cast(current & 0x0F)); + + // fixarray + case 0x90: + case 0x91: + case 0x92: + case 0x93: + case 0x94: + case 0x95: + case 0x96: + case 0x97: + case 0x98: + case 0x99: + case 0x9A: + case 0x9B: + case 0x9C: + case 0x9D: + case 0x9E: + case 0x9F: + return get_msgpack_array(static_cast(current & 0x0F)); + + // fixstr + case 0xA0: + case 0xA1: + case 0xA2: + case 0xA3: + case 0xA4: + case 0xA5: + case 0xA6: + case 0xA7: + case 0xA8: + case 0xA9: + case 0xAA: + case 0xAB: + case 0xAC: + case 0xAD: + case 0xAE: + case 0xAF: + case 0xB0: + case 0xB1: + case 0xB2: + case 0xB3: + case 0xB4: + case 0xB5: + case 0xB6: + case 0xB7: + case 0xB8: + case 0xB9: + case 0xBA: + case 0xBB: + case 0xBC: + case 0xBD: + case 0xBE: + case 0xBF: + { + string_t s; + return get_msgpack_string(s) and sax->string(s); + } + + case 0xC0: // nil + return sax->null(); + + case 0xC2: // false + return sax->boolean(false); + + case 0xC3: // true + return sax->boolean(true); + + case 0xCA: // float 32 + { + float number; + return get_number(number) and sax->number_float(static_cast(number), ""); + } + + case 0xCB: // float 64 + { + double number; + return get_number(number) and sax->number_float(number, ""); + } + + case 0xCC: // uint 8 + { + uint8_t number; + return get_number(number) and sax->number_unsigned(number); + } + + case 0xCD: // uint 16 + { + uint16_t number; + return get_number(number) and sax->number_unsigned(number); + } + + case 0xCE: // uint 32 + { + uint32_t number; + return get_number(number) and sax->number_unsigned(number); + } + + case 0xCF: // uint 64 + { + uint64_t number; + return get_number(number) and sax->number_unsigned(number); + } + + case 0xD0: // int 8 + { + int8_t number; + return get_number(number) and sax->number_integer(number); + } + + case 0xD1: // int 16 + { + int16_t number; + return get_number(number) and sax->number_integer(number); + } + + case 0xD2: // int 32 + { + int32_t number; + return get_number(number) and sax->number_integer(number); + } + + case 0xD3: // int 64 + { + int64_t number; + return get_number(number) and sax->number_integer(number); + } + + case 0xD9: // str 8 + case 0xDA: // str 16 + case 0xDB: // str 32 + { + string_t s; + return get_msgpack_string(s) and sax->string(s); + } + + case 0xDC: // array 16 + { + uint16_t len; + return get_number(len) and get_msgpack_array(static_cast(len)); + } + + case 0xDD: // array 32 + { + uint32_t len; + return get_number(len) and get_msgpack_array(static_cast(len)); + } + + case 0xDE: // map 16 + { + uint16_t len; + return get_number(len) and get_msgpack_object(static_cast(len)); + } + + case 0xDF: // map 32 + { + uint32_t len; + return get_number(len) and get_msgpack_object(static_cast(len)); + } + + // negative fixint + case 0xE0: + case 0xE1: + case 0xE2: + case 0xE3: + case 0xE4: + case 0xE5: + case 0xE6: + case 0xE7: + case 0xE8: + case 0xE9: + case 0xEA: + case 0xEB: + case 0xEC: + case 0xED: + case 0xEE: + case 0xEF: + case 0xF0: + case 0xF1: + case 0xF2: + case 0xF3: + case 0xF4: + case 0xF5: + case 0xF6: + case 0xF7: + case 0xF8: + case 0xF9: + case 0xFA: + case 0xFB: + case 0xFC: + case 0xFD: + case 0xFE: + case 0xFF: + return sax->number_integer(static_cast(current)); + + default: // anything else + { + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read, "error reading MessagePack; last byte: 0x" + last_token)); + } + } + } + + /*! + @param[in] get_char whether a new character should be retrieved from the + input (true, default) or whether the last read + character should be considered instead + + @return whether a valid UBJSON value was passed to the SAX parser + */ + bool parse_ubjson_internal(const bool get_char = true) + { + return get_ubjson_value(get_char ? get_ignore_noop() : current); + } + + /*! + @brief get next character from the input + + This function provides the interface to the used input adapter. It does + not throw in case the input reached EOF, but returns a -'ve valued + `std::char_traits::eof()` in that case. + + @return character read from the input + */ + int get() + { + ++chars_read; + return (current = ia->get_character()); + } + + /*! + @return character read from the input after ignoring all 'N' entries + */ + int get_ignore_noop() + { + do + { + get(); + } + while (current == 'N'); + + return current; + } + + /* + @brief read a number from the input + + @tparam NumberType the type of the number + @param[out] result number of type @a NumberType + + @return whether conversion completed + + @note This function needs to respect the system's endianess, because + bytes in CBOR, MessagePack, and UBJSON are stored in network order + (big endian) and therefore need reordering on little endian systems. + */ + template + bool get_number(NumberType& result) + { + // step 1: read input into array with system's byte order + std::array vec; + for (std::size_t i = 0; i < sizeof(NumberType); ++i) + { + get(); + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + + // reverse byte order prior to conversion if necessary + if (is_little_endian) + { + vec[sizeof(NumberType) - i - 1] = static_cast(current); + } + else + { + vec[i] = static_cast(current); // LCOV_EXCL_LINE + } + } + + // step 2: convert array into number of type T and return + std::memcpy(&result, vec.data(), sizeof(NumberType)); + return true; + } + + /*! + @brief create a string by reading characters from the input + + @tparam NumberType the type of the number + @param[in] len number of characters to read + @param[out] string created by reading @a len bytes + + @return whether string creation completed + + @note We can not reserve @a len bytes for the result, because @a len + may be too large. Usually, @ref unexpect_eof() detects the end of + the input before we run out of string memory. + */ + template + bool get_string(const NumberType len, string_t& result) + { + bool success = true; + std::generate_n(std::back_inserter(result), len, [this, &success]() + { + get(); + if (JSON_UNLIKELY(not unexpect_eof())) + { + success = false; + } + return static_cast(current); + }); + return success; + } + + /*! + @brief reads a CBOR string + + This function first reads starting bytes to determine the expected + string length and then copies this number of bytes into a string. + Additionally, CBOR's strings with indefinite lengths are supported. + + @param[out] result created string + + @return whether string creation completed + */ + bool get_cbor_string(string_t& result) + { + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + + switch (current) + { + // UTF-8 string (0x00..0x17 bytes follow) + case 0x60: + case 0x61: + case 0x62: + case 0x63: + case 0x64: + case 0x65: + case 0x66: + case 0x67: + case 0x68: + case 0x69: + case 0x6A: + case 0x6B: + case 0x6C: + case 0x6D: + case 0x6E: + case 0x6F: + case 0x70: + case 0x71: + case 0x72: + case 0x73: + case 0x74: + case 0x75: + case 0x76: + case 0x77: + { + return get_string(current & 0x1F, result); + } + + case 0x78: // UTF-8 string (one-byte uint8_t for n follows) + { + uint8_t len; + return get_number(len) and get_string(len, result); + } + + case 0x79: // UTF-8 string (two-byte uint16_t for n follow) + { + uint16_t len; + return get_number(len) and get_string(len, result); + } + + case 0x7A: // UTF-8 string (four-byte uint32_t for n follow) + { + uint32_t len; + return get_number(len) and get_string(len, result); + } + + case 0x7B: // UTF-8 string (eight-byte uint64_t for n follow) + { + uint64_t len; + return get_number(len) and get_string(len, result); + } + + case 0x7F: // UTF-8 string (indefinite length) + { + while (get() != 0xFF) + { + string_t chunk; + if (not get_cbor_string(chunk)) + { + return false; + } + result.append(chunk); + } + return true; + } + + default: + { + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read, "expected a CBOR string; last byte: 0x" + last_token)); + } + } + } + + /*! + @param[in] len the length of the array or json_sax_t::no_limit for an + array of indefinite size + @return whether array creation completed + */ + bool get_cbor_array(const std::size_t len) + { + if (JSON_UNLIKELY(not sax->start_array(len))) + { + return false; + } + + if (len != json_sax_t::no_limit) + for (std::size_t i = 0; i < len; ++i) + { + if (JSON_UNLIKELY(not parse_cbor_internal())) + { + return false; + } + } + else + { + while (get() != 0xFF) + { + if (JSON_UNLIKELY(not parse_cbor_internal(false))) + { + return false; + } + } + } + + return sax->end_array(); + } + + /*! + @param[in] len the length of the object or json_sax_t::no_limit for an + object of indefinite size + @return whether object creation completed + */ + bool get_cbor_object(const std::size_t len) + { + if (not JSON_UNLIKELY(sax->start_object(len))) + { + return false; + } + + string_t key; + if (len != json_sax_t::no_limit) + { + for (std::size_t i = 0; i < len; ++i) + { + get(); + if (JSON_UNLIKELY(not get_cbor_string(key) or not sax->key(key))) + { + return false; + } + + if (JSON_UNLIKELY(not parse_cbor_internal())) + { + return false; + } + key.clear(); + } + } + else + { + while (get() != 0xFF) + { + if (JSON_UNLIKELY(not get_cbor_string(key) or not sax->key(key))) + { + return false; + } + + if (JSON_UNLIKELY(not parse_cbor_internal())) + { + return false; + } + key.clear(); + } + } + + return sax->end_object(); + } + + /*! + @brief reads a MessagePack string + + This function first reads starting bytes to determine the expected + string length and then copies this number of bytes into a string. + + @param[out] result created string + + @return whether string creation completed + */ + bool get_msgpack_string(string_t& result) + { + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + + switch (current) + { + // fixstr + case 0xA0: + case 0xA1: + case 0xA2: + case 0xA3: + case 0xA4: + case 0xA5: + case 0xA6: + case 0xA7: + case 0xA8: + case 0xA9: + case 0xAA: + case 0xAB: + case 0xAC: + case 0xAD: + case 0xAE: + case 0xAF: + case 0xB0: + case 0xB1: + case 0xB2: + case 0xB3: + case 0xB4: + case 0xB5: + case 0xB6: + case 0xB7: + case 0xB8: + case 0xB9: + case 0xBA: + case 0xBB: + case 0xBC: + case 0xBD: + case 0xBE: + case 0xBF: + { + return get_string(current & 0x1F, result); + } + + case 0xD9: // str 8 + { + uint8_t len; + return get_number(len) and get_string(len, result); + } + + case 0xDA: // str 16 + { + uint16_t len; + return get_number(len) and get_string(len, result); + } + + case 0xDB: // str 32 + { + uint32_t len; + return get_number(len) and get_string(len, result); + } + + default: + { + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read, "expected a MessagePack string; last byte: 0x" + last_token)); + } + } + } + + /*! + @param[in] len the length of the array + @return whether array creation completed + */ + bool get_msgpack_array(const std::size_t len) + { + if (JSON_UNLIKELY(not sax->start_array(len))) + { + return false; + } + + for (std::size_t i = 0; i < len; ++i) + { + if (JSON_UNLIKELY(not parse_msgpack_internal())) + { + return false; + } + } + + return sax->end_array(); + } + + /*! + @param[in] len the length of the object + @return whether object creation completed + */ + bool get_msgpack_object(const std::size_t len) + { + if (JSON_UNLIKELY(not sax->start_object(len))) + { + return false; + } + + string_t key; + for (std::size_t i = 0; i < len; ++i) + { + get(); + if (JSON_UNLIKELY(not get_msgpack_string(key) or not sax->key(key))) + { + return false; + } + + if (JSON_UNLIKELY(not parse_msgpack_internal())) + { + return false; + } + key.clear(); + } + + return sax->end_object(); + } + + /*! + @brief reads a UBJSON string + + This function is either called after reading the 'S' byte explicitly + indicating a string, or in case of an object key where the 'S' byte can be + left out. + + @param[out] result created string + @param[in] get_char whether a new character should be retrieved from the + input (true, default) or whether the last read + character should be considered instead + + @return whether string creation completed + */ + bool get_ubjson_string(string_t& result, const bool get_char = true) + { + if (get_char) + { + get(); // TODO: may we ignore N here? + } + + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + + switch (current) + { + case 'U': + { + uint8_t len; + return get_number(len) and get_string(len, result); + } + + case 'i': + { + int8_t len; + return get_number(len) and get_string(len, result); + } + + case 'I': + { + int16_t len; + return get_number(len) and get_string(len, result); + } + + case 'l': + { + int32_t len; + return get_number(len) and get_string(len, result); + } + + case 'L': + { + int64_t len; + return get_number(len) and get_string(len, result); + } + + default: + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read, "expected a UBJSON string; last byte: 0x" + last_token)); + } + } + + /*! + @param[out] result determined size + @return whether size determination completed + */ + bool get_ubjson_size_value(std::size_t& result) + { + switch (get_ignore_noop()) + { + case 'U': + { + uint8_t number; + if (JSON_UNLIKELY(not get_number(number))) + { + return false; + } + result = static_cast(number); + return true; + } + + case 'i': + { + int8_t number; + if (JSON_UNLIKELY(not get_number(number))) + { + return false; + } + result = static_cast(number); + return true; + } + + case 'I': + { + int16_t number; + if (JSON_UNLIKELY(not get_number(number))) + { + return false; + } + result = static_cast(number); + return true; + } + + case 'l': + { + int32_t number; + if (JSON_UNLIKELY(not get_number(number))) + { + return false; + } + result = static_cast(number); + return true; + } + + case 'L': + { + int64_t number; + if (JSON_UNLIKELY(not get_number(number))) + { + return false; + } + result = static_cast(number); + return true; + } + + default: + { + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read, "byte after '#' must denote a number type; last byte: 0x" + last_token)); + } + } + } + + /*! + @brief determine the type and size for a container + + In the optimized UBJSON format, a type and a size can be provided to allow + for a more compact representation. + + @param[out] result pair of the size and the type + + @return whether pair creation completed + */ + bool get_ubjson_size_type(std::pair& result) + { + result.first = string_t::npos; // size + result.second = 0; // type + + get_ignore_noop(); + + if (current == '$') + { + result.second = get(); // must not ignore 'N', because 'N' maybe the type + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + + get_ignore_noop(); + if (JSON_UNLIKELY(current != '#')) + { + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read, "expected '#' after UBJSON type information; last byte: 0x" + last_token)); + } + + return get_ubjson_size_value(result.first); + } + else if (current == '#') + { + return get_ubjson_size_value(result.first); + } + return true; + } + + /*! + @param prefix the previously read or set type prefix + @return whether value creation completed + */ + bool get_ubjson_value(const int prefix) + { + switch (prefix) + { + case std::char_traits::eof(): // EOF + return unexpect_eof(); + + case 'T': // true + return sax->boolean(true); + case 'F': // false + return sax->boolean(false); + + case 'Z': // null + return sax->null(); + + case 'U': + { + uint8_t number; + return get_number(number) and sax->number_unsigned(number); + } + + case 'i': + { + int8_t number; + return get_number(number) and sax->number_integer(number); + } + + case 'I': + { + int16_t number; + return get_number(number) and sax->number_integer(number); + } + + case 'l': + { + int32_t number; + return get_number(number) and sax->number_integer(number); + } + + case 'L': + { + int64_t number; + return get_number(number) and sax->number_integer(number); + } + + case 'd': + { + float number; + return get_number(number) and sax->number_float(static_cast(number), ""); + } + + case 'D': + { + double number; + return get_number(number) and sax->number_float(number, ""); + } + + case 'C': // char + { + get(); + if (JSON_UNLIKELY(not unexpect_eof())) + { + return false; + } + if (JSON_UNLIKELY(current > 127)) + { + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read, "byte after 'C' must be in range 0x00..0x7F; last byte: 0x" + last_token)); + } + string_t s(1, static_cast(current)); + return sax->string(s); + } + + case 'S': // string + { + string_t s; + return get_ubjson_string(s) and sax->string(s); + } + + case '[': // array + return get_ubjson_array(); + + case '{': // object + return get_ubjson_object(); + + default: // anything else + { + auto last_token = get_token_string(); + return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read, "error reading UBJSON; last byte: 0x" + last_token)); + } + } + } + + /*! + @return whether array creation completed + */ + bool get_ubjson_array() + { + std::pair size_and_type; + if (JSON_UNLIKELY(not get_ubjson_size_type(size_and_type))) + { + return false; + } + + if (size_and_type.first != string_t::npos) + { + if (JSON_UNLIKELY(not sax->start_array(size_and_type.first))) + { + return false; + } + + if (size_and_type.second != 0) + { + if (size_and_type.second != 'N') + { + for (std::size_t i = 0; i < size_and_type.first; ++i) + { + if (JSON_UNLIKELY(not get_ubjson_value(size_and_type.second))) + { + return false; + } + } + } + } + else + { + for (std::size_t i = 0; i < size_and_type.first; ++i) + { + if (JSON_UNLIKELY(not parse_ubjson_internal())) + { + return false; + } + } + } + } + else + { + if (JSON_UNLIKELY(not sax->start_array())) + { + return false; + } + + while (current != ']') + { + if (JSON_UNLIKELY(not parse_ubjson_internal(false))) + { + return false; + } + get_ignore_noop(); + } + } + + return sax->end_array(); + } + + /*! + @return whether object creation completed + */ + bool get_ubjson_object() + { + std::pair size_and_type; + if (JSON_UNLIKELY(not get_ubjson_size_type(size_and_type))) + { + return false; + } + + string_t key; + if (size_and_type.first != string_t::npos) + { + if (JSON_UNLIKELY(not sax->start_object(size_and_type.first))) + { + return false; + } + + if (size_and_type.second != 0) + { + for (std::size_t i = 0; i < size_and_type.first; ++i) + { + if (JSON_UNLIKELY(not get_ubjson_string(key) or not sax->key(key))) + { + return false; + } + if (JSON_UNLIKELY(not get_ubjson_value(size_and_type.second))) + { + return false; + } + key.clear(); + } + } + else + { + for (std::size_t i = 0; i < size_and_type.first; ++i) + { + if (JSON_UNLIKELY(not get_ubjson_string(key) or not sax->key(key))) + { + return false; + } + if (JSON_UNLIKELY(not parse_ubjson_internal())) + { + return false; + } + key.clear(); + } + } + } + else + { + if (JSON_UNLIKELY(not sax->start_object())) + { + return false; + } + + while (current != '}') + { + if (JSON_UNLIKELY(not get_ubjson_string(key, false) or not sax->key(key))) + { + return false; + } + if (JSON_UNLIKELY(not parse_ubjson_internal())) + { + return false; + } + get_ignore_noop(); + key.clear(); + } + } + + return sax->end_object(); + } + + /*! + @return whether the last read character is not EOF + */ + bool unexpect_eof() const + { + if (JSON_UNLIKELY(current == std::char_traits::eof())) + { + return sax->parse_error(chars_read, "", parse_error::create(110, chars_read, "unexpected end of input")); + } + return true; + } + + /*! + @return a string representation of the last read byte + */ + std::string get_token_string() const + { + std::stringstream ss; + ss << std::setw(2) << std::uppercase << std::setfill('0') << std::hex << current; + return ss.str(); + } + + private: + /// input adapter + input_adapter_t ia = nullptr; + + /// the current character + int current = std::char_traits::eof(); + + /// the number of characters read + std::size_t chars_read = 0; + + /// whether we can assume little endianess + const bool is_little_endian = little_endianess(); + + /// the SAX parser + json_sax_t* sax = nullptr; +}; +} +} + +// #include + + +#include // reverse +#include // array +#include // uint8_t, uint16_t, uint32_t, uint64_t +#include // memcpy +#include // numeric_limits + +// #include + +// #include + + +namespace nlohmann +{ +namespace detail +{ +/////////////////// +// binary writer // +/////////////////// + +/*! +@brief serialization to CBOR and MessagePack values +*/ +template +class binary_writer +{ + public: + /*! + @brief create a binary writer + + @param[in] adapter output adapter to write to + */ + explicit binary_writer(output_adapter_t adapter) : oa(adapter) + { + assert(oa); + } + + /*! + @brief[in] j JSON value to serialize + */ + void write_cbor(const BasicJsonType& j) + { + switch (j.type()) + { + case value_t::null: + { + oa->write_character(static_cast(0xF6)); + break; + } + + case value_t::boolean: + { + oa->write_character(j.m_value.boolean + ? static_cast(0xF5) + : static_cast(0xF4)); + break; + } + + case value_t::number_integer: + { + if (j.m_value.number_integer >= 0) + { + // CBOR does not differentiate between positive signed + // integers and unsigned integers. Therefore, we used the + // code from the value_t::number_unsigned case here. + if (j.m_value.number_integer <= 0x17) + { + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_integer <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x18)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_integer <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x19)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_integer <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x1A)); + write_number(static_cast(j.m_value.number_integer)); + } + else + { + oa->write_character(static_cast(0x1B)); + write_number(static_cast(j.m_value.number_integer)); + } + } + else + { + // The conversions below encode the sign in the first + // byte, and the value is converted to a positive number. + const auto positive_number = -1 - j.m_value.number_integer; + if (j.m_value.number_integer >= -24) + { + write_number(static_cast(0x20 + positive_number)); + } + else if (positive_number <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x38)); + write_number(static_cast(positive_number)); + } + else if (positive_number <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x39)); + write_number(static_cast(positive_number)); + } + else if (positive_number <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x3A)); + write_number(static_cast(positive_number)); + } + else + { + oa->write_character(static_cast(0x3B)); + write_number(static_cast(positive_number)); + } + } + break; + } + + case value_t::number_unsigned: + { + if (j.m_value.number_unsigned <= 0x17) + { + write_number(static_cast(j.m_value.number_unsigned)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x18)); + write_number(static_cast(j.m_value.number_unsigned)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x19)); + write_number(static_cast(j.m_value.number_unsigned)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x1A)); + write_number(static_cast(j.m_value.number_unsigned)); + } + else + { + oa->write_character(static_cast(0x1B)); + write_number(static_cast(j.m_value.number_unsigned)); + } + break; + } + + case value_t::number_float: + { + oa->write_character(get_cbor_float_prefix(j.m_value.number_float)); + write_number(j.m_value.number_float); + break; + } + + case value_t::string: + { + // step 1: write control byte and the string length + const auto N = j.m_value.string->size(); + if (N <= 0x17) + { + write_number(static_cast(0x60 + N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x78)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x79)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x7A)); + write_number(static_cast(N)); + } + // LCOV_EXCL_START + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x7B)); + write_number(static_cast(N)); + } + // LCOV_EXCL_STOP + + // step 2: write the string + oa->write_characters( + reinterpret_cast(j.m_value.string->c_str()), + j.m_value.string->size()); + break; + } + + case value_t::array: + { + // step 1: write control byte and the array size + const auto N = j.m_value.array->size(); + if (N <= 0x17) + { + write_number(static_cast(0x80 + N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x98)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x99)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x9A)); + write_number(static_cast(N)); + } + // LCOV_EXCL_START + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0x9B)); + write_number(static_cast(N)); + } + // LCOV_EXCL_STOP + + // step 2: write each element + for (const auto& el : *j.m_value.array) + { + write_cbor(el); + } + break; + } + + case value_t::object: + { + // step 1: write control byte and the object size + const auto N = j.m_value.object->size(); + if (N <= 0x17) + { + write_number(static_cast(0xA0 + N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0xB8)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0xB9)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0xBA)); + write_number(static_cast(N)); + } + // LCOV_EXCL_START + else if (N <= (std::numeric_limits::max)()) + { + oa->write_character(static_cast(0xBB)); + write_number(static_cast(N)); + } + // LCOV_EXCL_STOP + + // step 2: write each element + for (const auto& el : *j.m_value.object) + { + write_cbor(el.first); + write_cbor(el.second); + } + break; + } + + default: + break; + } + } + + /*! + @brief[in] j JSON value to serialize + */ + void write_msgpack(const BasicJsonType& j) + { + switch (j.type()) + { + case value_t::null: // nil + { + oa->write_character(static_cast(0xC0)); + break; + } + + case value_t::boolean: // true and false + { + oa->write_character(j.m_value.boolean + ? static_cast(0xC3) + : static_cast(0xC2)); + break; + } + + case value_t::number_integer: + { + if (j.m_value.number_integer >= 0) + { + // MessagePack does not differentiate between positive + // signed integers and unsigned integers. Therefore, we used + // the code from the value_t::number_unsigned case here. + if (j.m_value.number_unsigned < 128) + { + // positive fixnum + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + // uint 8 + oa->write_character(static_cast(0xCC)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + // uint 16 + oa->write_character(static_cast(0xCD)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + // uint 32 + oa->write_character(static_cast(0xCE)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + // uint 64 + oa->write_character(static_cast(0xCF)); + write_number(static_cast(j.m_value.number_integer)); + } + } + else + { + if (j.m_value.number_integer >= -32) + { + // negative fixnum + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_integer >= (std::numeric_limits::min)() and + j.m_value.number_integer <= (std::numeric_limits::max)()) + { + // int 8 + oa->write_character(static_cast(0xD0)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_integer >= (std::numeric_limits::min)() and + j.m_value.number_integer <= (std::numeric_limits::max)()) + { + // int 16 + oa->write_character(static_cast(0xD1)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_integer >= (std::numeric_limits::min)() and + j.m_value.number_integer <= (std::numeric_limits::max)()) + { + // int 32 + oa->write_character(static_cast(0xD2)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_integer >= (std::numeric_limits::min)() and + j.m_value.number_integer <= (std::numeric_limits::max)()) + { + // int 64 + oa->write_character(static_cast(0xD3)); + write_number(static_cast(j.m_value.number_integer)); + } + } + break; + } + + case value_t::number_unsigned: + { + if (j.m_value.number_unsigned < 128) + { + // positive fixnum + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + // uint 8 + oa->write_character(static_cast(0xCC)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + // uint 16 + oa->write_character(static_cast(0xCD)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + // uint 32 + oa->write_character(static_cast(0xCE)); + write_number(static_cast(j.m_value.number_integer)); + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + // uint 64 + oa->write_character(static_cast(0xCF)); + write_number(static_cast(j.m_value.number_integer)); + } + break; + } + + case value_t::number_float: + { + oa->write_character(get_msgpack_float_prefix(j.m_value.number_float)); + write_number(j.m_value.number_float); + break; + } + + case value_t::string: + { + // step 1: write control byte and the string length + const auto N = j.m_value.string->size(); + if (N <= 31) + { + // fixstr + write_number(static_cast(0xA0 | N)); + } + else if (N <= (std::numeric_limits::max)()) + { + // str 8 + oa->write_character(static_cast(0xD9)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + // str 16 + oa->write_character(static_cast(0xDA)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + // str 32 + oa->write_character(static_cast(0xDB)); + write_number(static_cast(N)); + } + + // step 2: write the string + oa->write_characters( + reinterpret_cast(j.m_value.string->c_str()), + j.m_value.string->size()); + break; + } + + case value_t::array: + { + // step 1: write control byte and the array size + const auto N = j.m_value.array->size(); + if (N <= 15) + { + // fixarray + write_number(static_cast(0x90 | N)); + } + else if (N <= (std::numeric_limits::max)()) + { + // array 16 + oa->write_character(static_cast(0xDC)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + // array 32 + oa->write_character(static_cast(0xDD)); + write_number(static_cast(N)); + } + + // step 2: write each element + for (const auto& el : *j.m_value.array) + { + write_msgpack(el); + } + break; + } + + case value_t::object: + { + // step 1: write control byte and the object size + const auto N = j.m_value.object->size(); + if (N <= 15) + { + // fixmap + write_number(static_cast(0x80 | (N & 0xF))); + } + else if (N <= (std::numeric_limits::max)()) + { + // map 16 + oa->write_character(static_cast(0xDE)); + write_number(static_cast(N)); + } + else if (N <= (std::numeric_limits::max)()) + { + // map 32 + oa->write_character(static_cast(0xDF)); + write_number(static_cast(N)); + } + + // step 2: write each element + for (const auto& el : *j.m_value.object) + { + write_msgpack(el.first); + write_msgpack(el.second); + } + break; + } + + default: + break; + } + } + + /*! + @param[in] j JSON value to serialize + @param[in] use_count whether to use '#' prefixes (optimized format) + @param[in] use_type whether to use '$' prefixes (optimized format) + @param[in] add_prefix whether prefixes need to be used for this value + */ + void write_ubjson(const BasicJsonType& j, const bool use_count, + const bool use_type, const bool add_prefix = true) + { + switch (j.type()) + { + case value_t::null: + { + if (add_prefix) + { + oa->write_character(static_cast('Z')); + } + break; + } + + case value_t::boolean: + { + if (add_prefix) + oa->write_character(j.m_value.boolean + ? static_cast('T') + : static_cast('F')); + break; + } + + case value_t::number_integer: + { + write_number_with_ubjson_prefix(j.m_value.number_integer, add_prefix); + break; + } + + case value_t::number_unsigned: + { + write_number_with_ubjson_prefix(j.m_value.number_unsigned, add_prefix); + break; + } + + case value_t::number_float: + { + write_number_with_ubjson_prefix(j.m_value.number_float, add_prefix); + break; + } + + case value_t::string: + { + if (add_prefix) + { + oa->write_character(static_cast('S')); + } + write_number_with_ubjson_prefix(j.m_value.string->size(), true); + oa->write_characters( + reinterpret_cast(j.m_value.string->c_str()), + j.m_value.string->size()); + break; + } + + case value_t::array: + { + if (add_prefix) + { + oa->write_character(static_cast('[')); + } + + bool prefix_required = true; + if (use_type and not j.m_value.array->empty()) + { + assert(use_count); + const CharType first_prefix = ubjson_prefix(j.front()); + const bool same_prefix = std::all_of(j.begin() + 1, j.end(), + [this, first_prefix](const BasicJsonType & v) + { + return ubjson_prefix(v) == first_prefix; + }); + + if (same_prefix) + { + prefix_required = false; + oa->write_character(static_cast('$')); + oa->write_character(first_prefix); + } + } + + if (use_count) + { + oa->write_character(static_cast('#')); + write_number_with_ubjson_prefix(j.m_value.array->size(), true); + } + + for (const auto& el : *j.m_value.array) + { + write_ubjson(el, use_count, use_type, prefix_required); + } + + if (not use_count) + { + oa->write_character(static_cast(']')); + } + + break; + } + + case value_t::object: + { + if (add_prefix) + { + oa->write_character(static_cast('{')); + } + + bool prefix_required = true; + if (use_type and not j.m_value.object->empty()) + { + assert(use_count); + const CharType first_prefix = ubjson_prefix(j.front()); + const bool same_prefix = std::all_of(j.begin(), j.end(), + [this, first_prefix](const BasicJsonType & v) + { + return ubjson_prefix(v) == first_prefix; + }); + + if (same_prefix) + { + prefix_required = false; + oa->write_character(static_cast('$')); + oa->write_character(first_prefix); + } + } + + if (use_count) + { + oa->write_character(static_cast('#')); + write_number_with_ubjson_prefix(j.m_value.object->size(), true); + } + + for (const auto& el : *j.m_value.object) + { + write_number_with_ubjson_prefix(el.first.size(), true); + oa->write_characters( + reinterpret_cast(el.first.c_str()), + el.first.size()); + write_ubjson(el.second, use_count, use_type, prefix_required); + } + + if (not use_count) + { + oa->write_character(static_cast('}')); + } + + break; + } + + default: + break; + } + } + + private: + /* + @brief write a number to output input + + @param[in] n number of type @a NumberType + @tparam NumberType the type of the number + + @note This function needs to respect the system's endianess, because bytes + in CBOR, MessagePack, and UBJSON are stored in network order (big + endian) and therefore need reordering on little endian systems. + */ + template + void write_number(const NumberType n) + { + // step 1: write number to array of length NumberType + std::array vec; + std::memcpy(vec.data(), &n, sizeof(NumberType)); + + // step 2: write array to output (with possible reordering) + if (is_little_endian) + { + // reverse byte order prior to conversion if necessary + std::reverse(vec.begin(), vec.end()); + } + + oa->write_characters(vec.data(), sizeof(NumberType)); + } + + // UBJSON: write number (floating point) + template::value, int>::type = 0> + void write_number_with_ubjson_prefix(const NumberType n, + const bool add_prefix) + { + if (add_prefix) + { + oa->write_character(get_ubjson_float_prefix(n)); + } + write_number(n); + } + + // UBJSON: write number (unsigned integer) + template::value, int>::type = 0> + void write_number_with_ubjson_prefix(const NumberType n, + const bool add_prefix) + { + if (n <= static_cast((std::numeric_limits::max)())) + { + if (add_prefix) + { + oa->write_character(static_cast('i')); // int8 + } + write_number(static_cast(n)); + } + else if (n <= (std::numeric_limits::max)()) + { + if (add_prefix) + { + oa->write_character(static_cast('U')); // uint8 + } + write_number(static_cast(n)); + } + else if (n <= static_cast((std::numeric_limits::max)())) + { + if (add_prefix) + { + oa->write_character(static_cast('I')); // int16 + } + write_number(static_cast(n)); + } + else if (n <= static_cast((std::numeric_limits::max)())) + { + if (add_prefix) + { + oa->write_character(static_cast('l')); // int32 + } + write_number(static_cast(n)); + } + else if (n <= static_cast((std::numeric_limits::max)())) + { + if (add_prefix) + { + oa->write_character(static_cast('L')); // int64 + } + write_number(static_cast(n)); + } + else + { + JSON_THROW(out_of_range::create(407, "number overflow serializing " + std::to_string(n))); + } + } + + // UBJSON: write number (signed integer) + template::value and + not std::is_floating_point::value, int>::type = 0> + void write_number_with_ubjson_prefix(const NumberType n, + const bool add_prefix) + { + if ((std::numeric_limits::min)() <= n and n <= (std::numeric_limits::max)()) + { + if (add_prefix) + { + oa->write_character(static_cast('i')); // int8 + } + write_number(static_cast(n)); + } + else if (static_cast((std::numeric_limits::min)()) <= n and n <= static_cast((std::numeric_limits::max)())) + { + if (add_prefix) + { + oa->write_character(static_cast('U')); // uint8 + } + write_number(static_cast(n)); + } + else if ((std::numeric_limits::min)() <= n and n <= (std::numeric_limits::max)()) + { + if (add_prefix) + { + oa->write_character(static_cast('I')); // int16 + } + write_number(static_cast(n)); + } + else if ((std::numeric_limits::min)() <= n and n <= (std::numeric_limits::max)()) + { + if (add_prefix) + { + oa->write_character(static_cast('l')); // int32 + } + write_number(static_cast(n)); + } + else if ((std::numeric_limits::min)() <= n and n <= (std::numeric_limits::max)()) + { + if (add_prefix) + { + oa->write_character(static_cast('L')); // int64 + } + write_number(static_cast(n)); + } + // LCOV_EXCL_START + else + { + JSON_THROW(out_of_range::create(407, "number overflow serializing " + std::to_string(n))); + } + // LCOV_EXCL_STOP + } + + /*! + @brief determine the type prefix of container values + + @note This function does not need to be 100% accurate when it comes to + integer limits. In case a number exceeds the limits of int64_t, + this will be detected by a later call to function + write_number_with_ubjson_prefix. Therefore, we return 'L' for any + value that does not fit the previous limits. + */ + CharType ubjson_prefix(const BasicJsonType& j) const noexcept + { + switch (j.type()) + { + case value_t::null: + return 'Z'; + + case value_t::boolean: + return j.m_value.boolean ? 'T' : 'F'; + + case value_t::number_integer: + { + if ((std::numeric_limits::min)() <= j.m_value.number_integer and j.m_value.number_integer <= (std::numeric_limits::max)()) + { + return 'i'; + } + else if ((std::numeric_limits::min)() <= j.m_value.number_integer and j.m_value.number_integer <= (std::numeric_limits::max)()) + { + return 'U'; + } + else if ((std::numeric_limits::min)() <= j.m_value.number_integer and j.m_value.number_integer <= (std::numeric_limits::max)()) + { + return 'I'; + } + else if ((std::numeric_limits::min)() <= j.m_value.number_integer and j.m_value.number_integer <= (std::numeric_limits::max)()) + { + return 'l'; + } + else // no check and assume int64_t (see note above) + { + return 'L'; + } + } + + case value_t::number_unsigned: + { + if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + return 'i'; + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + return 'U'; + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + return 'I'; + } + else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) + { + return 'l'; + } + else // no check and assume int64_t (see note above) + { + return 'L'; + } + } + + case value_t::number_float: + return get_ubjson_float_prefix(j.m_value.number_float); + + case value_t::string: + return 'S'; + + case value_t::array: + return '['; + + case value_t::object: + return '{'; + + default: // discarded values + return 'N'; + } + } + + static constexpr CharType get_cbor_float_prefix(float) + { + return static_cast(0xFA); // Single-Precision Float + } + + static constexpr CharType get_cbor_float_prefix(double) + { + return static_cast(0xFB); // Double-Precision Float + } + + static constexpr CharType get_msgpack_float_prefix(float) + { + return static_cast(0xCA); // float 32 + } + + static constexpr CharType get_msgpack_float_prefix(double) + { + return static_cast(0xCB); // float 64 + } + + static constexpr CharType get_ubjson_float_prefix(float) + { + return 'd'; // float 32 + } + + static constexpr CharType get_ubjson_float_prefix(double) + { + return 'D'; // float 64 + } + + private: + /// whether we can assume little endianess + const bool is_little_endian = binary_reader::little_endianess(); + + /// the output + output_adapter_t oa = nullptr; +}; +} +} + +// #include + + +#include // reverse, remove, fill, find, none_of +#include // array +#include // assert +#include // and, or +#include // localeconv, lconv +#include // labs, isfinite, isnan, signbit +#include // size_t, ptrdiff_t +#include // uint8_t +#include // snprintf +#include // setfill +#include // next +#include // numeric_limits +#include // string +#include // stringstream +#include // is_same + +// #include + +// #include + + +#include // assert +#include // or, and, not +#include // signbit, isfinite +#include // intN_t, uintN_t +#include // memcpy, memmove + +namespace nlohmann +{ +namespace detail +{ + +/*! +@brief implements the Grisu2 algorithm for binary to decimal floating-point +conversion. + +This implementation is a slightly modified version of the reference +implementation which may be obtained from +http://florian.loitsch.com/publications (bench.tar.gz). + +The code is distributed under the MIT license, Copyright (c) 2009 Florian Loitsch. + +For a detailed description of the algorithm see: + +[1] Loitsch, "Printing Floating-Point Numbers Quickly and Accurately with + Integers", Proceedings of the ACM SIGPLAN 2010 Conference on Programming + Language Design and Implementation, PLDI 2010 +[2] Burger, Dybvig, "Printing Floating-Point Numbers Quickly and Accurately", + Proceedings of the ACM SIGPLAN 1996 Conference on Programming Language + Design and Implementation, PLDI 1996 +*/ +namespace dtoa_impl +{ + +template +Target reinterpret_bits(const Source source) +{ + static_assert(sizeof(Target) == sizeof(Source), "size mismatch"); + + Target target; + std::memcpy(&target, &source, sizeof(Source)); + return target; +} + +struct diyfp // f * 2^e +{ + static constexpr int kPrecision = 64; // = q + + uint64_t f; + int e; + + constexpr diyfp() noexcept : f(0), e(0) {} + constexpr diyfp(uint64_t f_, int e_) noexcept : f(f_), e(e_) {} + + /*! + @brief returns x - y + @pre x.e == y.e and x.f >= y.f + */ + static diyfp sub(const diyfp& x, const diyfp& y) noexcept + { + assert(x.e == y.e); + assert(x.f >= y.f); + + return diyfp(x.f - y.f, x.e); + } + + /*! + @brief returns x * y + @note The result is rounded. (Only the upper q bits are returned.) + */ + static diyfp mul(const diyfp& x, const diyfp& y) noexcept + { + static_assert(kPrecision == 64, "internal error"); + + // Computes: + // f = round((x.f * y.f) / 2^q) + // e = x.e + y.e + q + + // Emulate the 64-bit * 64-bit multiplication: + // + // p = u * v + // = (u_lo + 2^32 u_hi) (v_lo + 2^32 v_hi) + // = (u_lo v_lo ) + 2^32 ((u_lo v_hi ) + (u_hi v_lo )) + 2^64 (u_hi v_hi ) + // = (p0 ) + 2^32 ((p1 ) + (p2 )) + 2^64 (p3 ) + // = (p0_lo + 2^32 p0_hi) + 2^32 ((p1_lo + 2^32 p1_hi) + (p2_lo + 2^32 p2_hi)) + 2^64 (p3 ) + // = (p0_lo ) + 2^32 (p0_hi + p1_lo + p2_lo ) + 2^64 (p1_hi + p2_hi + p3) + // = (p0_lo ) + 2^32 (Q ) + 2^64 (H ) + // = (p0_lo ) + 2^32 (Q_lo + 2^32 Q_hi ) + 2^64 (H ) + // + // (Since Q might be larger than 2^32 - 1) + // + // = (p0_lo + 2^32 Q_lo) + 2^64 (Q_hi + H) + // + // (Q_hi + H does not overflow a 64-bit int) + // + // = p_lo + 2^64 p_hi + + const uint64_t u_lo = x.f & 0xFFFFFFFF; + const uint64_t u_hi = x.f >> 32; + const uint64_t v_lo = y.f & 0xFFFFFFFF; + const uint64_t v_hi = y.f >> 32; + + const uint64_t p0 = u_lo * v_lo; + const uint64_t p1 = u_lo * v_hi; + const uint64_t p2 = u_hi * v_lo; + const uint64_t p3 = u_hi * v_hi; + + const uint64_t p0_hi = p0 >> 32; + const uint64_t p1_lo = p1 & 0xFFFFFFFF; + const uint64_t p1_hi = p1 >> 32; + const uint64_t p2_lo = p2 & 0xFFFFFFFF; + const uint64_t p2_hi = p2 >> 32; + + uint64_t Q = p0_hi + p1_lo + p2_lo; + + // The full product might now be computed as + // + // p_hi = p3 + p2_hi + p1_hi + (Q >> 32) + // p_lo = p0_lo + (Q << 32) + // + // But in this particular case here, the full p_lo is not required. + // Effectively we only need to add the highest bit in p_lo to p_hi (and + // Q_hi + 1 does not overflow). + + Q += uint64_t{1} << (64 - 32 - 1); // round, ties up + + const uint64_t h = p3 + p2_hi + p1_hi + (Q >> 32); + + return diyfp(h, x.e + y.e + 64); + } + + /*! + @brief normalize x such that the significand is >= 2^(q-1) + @pre x.f != 0 + */ + static diyfp normalize(diyfp x) noexcept + { + assert(x.f != 0); + + while ((x.f >> 63) == 0) + { + x.f <<= 1; + x.e--; + } + + return x; + } + + /*! + @brief normalize x such that the result has the exponent E + @pre e >= x.e and the upper e - x.e bits of x.f must be zero. + */ + static diyfp normalize_to(const diyfp& x, const int target_exponent) noexcept + { + const int delta = x.e - target_exponent; + + assert(delta >= 0); + assert(((x.f << delta) >> delta) == x.f); + + return diyfp(x.f << delta, target_exponent); + } +}; + +struct boundaries +{ + diyfp w; + diyfp minus; + diyfp plus; +}; + +/*! +Compute the (normalized) diyfp representing the input number 'value' and its +boundaries. + +@pre value must be finite and positive +*/ +template +boundaries compute_boundaries(FloatType value) +{ + assert(std::isfinite(value)); + assert(value > 0); + + // Convert the IEEE representation into a diyfp. + // + // If v is denormal: + // value = 0.F * 2^(1 - bias) = ( F) * 2^(1 - bias - (p-1)) + // If v is normalized: + // value = 1.F * 2^(E - bias) = (2^(p-1) + F) * 2^(E - bias - (p-1)) + + static_assert(std::numeric_limits::is_iec559, + "internal error: dtoa_short requires an IEEE-754 floating-point implementation"); + + constexpr int kPrecision = std::numeric_limits::digits; // = p (includes the hidden bit) + constexpr int kBias = std::numeric_limits::max_exponent - 1 + (kPrecision - 1); + constexpr int kMinExp = 1 - kBias; + constexpr uint64_t kHiddenBit = uint64_t{1} << (kPrecision - 1); // = 2^(p-1) + + using bits_type = typename std::conditional< kPrecision == 24, uint32_t, uint64_t >::type; + + const uint64_t bits = reinterpret_bits(value); + const uint64_t E = bits >> (kPrecision - 1); + const uint64_t F = bits & (kHiddenBit - 1); + + const bool is_denormal = (E == 0); + const diyfp v = is_denormal + ? diyfp(F, kMinExp) + : diyfp(F + kHiddenBit, static_cast(E) - kBias); + + // Compute the boundaries m- and m+ of the floating-point value + // v = f * 2^e. + // + // Determine v- and v+, the floating-point predecessor and successor if v, + // respectively. + // + // v- = v - 2^e if f != 2^(p-1) or e == e_min (A) + // = v - 2^(e-1) if f == 2^(p-1) and e > e_min (B) + // + // v+ = v + 2^e + // + // Let m- = (v- + v) / 2 and m+ = (v + v+) / 2. All real numbers _strictly_ + // between m- and m+ round to v, regardless of how the input rounding + // algorithm breaks ties. + // + // ---+-------------+-------------+-------------+-------------+--- (A) + // v- m- v m+ v+ + // + // -----------------+------+------+-------------+-------------+--- (B) + // v- m- v m+ v+ + + const bool lower_boundary_is_closer = (F == 0 and E > 1); + const diyfp m_plus = diyfp(2 * v.f + 1, v.e - 1); + const diyfp m_minus = lower_boundary_is_closer + ? diyfp(4 * v.f - 1, v.e - 2) // (B) + : diyfp(2 * v.f - 1, v.e - 1); // (A) + + // Determine the normalized w+ = m+. + const diyfp w_plus = diyfp::normalize(m_plus); + + // Determine w- = m- such that e_(w-) = e_(w+). + const diyfp w_minus = diyfp::normalize_to(m_minus, w_plus.e); + + return {diyfp::normalize(v), w_minus, w_plus}; +} + +// Given normalized diyfp w, Grisu needs to find a (normalized) cached +// power-of-ten c, such that the exponent of the product c * w = f * 2^e lies +// within a certain range [alpha, gamma] (Definition 3.2 from [1]) +// +// alpha <= e = e_c + e_w + q <= gamma +// +// or +// +// f_c * f_w * 2^alpha <= f_c 2^(e_c) * f_w 2^(e_w) * 2^q +// <= f_c * f_w * 2^gamma +// +// Since c and w are normalized, i.e. 2^(q-1) <= f < 2^q, this implies +// +// 2^(q-1) * 2^(q-1) * 2^alpha <= c * w * 2^q < 2^q * 2^q * 2^gamma +// +// or +// +// 2^(q - 2 + alpha) <= c * w < 2^(q + gamma) +// +// The choice of (alpha,gamma) determines the size of the table and the form of +// the digit generation procedure. Using (alpha,gamma)=(-60,-32) works out well +// in practice: +// +// The idea is to cut the number c * w = f * 2^e into two parts, which can be +// processed independently: An integral part p1, and a fractional part p2: +// +// f * 2^e = ( (f div 2^-e) * 2^-e + (f mod 2^-e) ) * 2^e +// = (f div 2^-e) + (f mod 2^-e) * 2^e +// = p1 + p2 * 2^e +// +// The conversion of p1 into decimal form requires a series of divisions and +// modulos by (a power of) 10. These operations are faster for 32-bit than for +// 64-bit integers, so p1 should ideally fit into a 32-bit integer. This can be +// achieved by choosing +// +// -e >= 32 or e <= -32 := gamma +// +// In order to convert the fractional part +// +// p2 * 2^e = p2 / 2^-e = d[-1] / 10^1 + d[-2] / 10^2 + ... +// +// into decimal form, the fraction is repeatedly multiplied by 10 and the digits +// d[-i] are extracted in order: +// +// (10 * p2) div 2^-e = d[-1] +// (10 * p2) mod 2^-e = d[-2] / 10^1 + ... +// +// The multiplication by 10 must not overflow. It is sufficient to choose +// +// 10 * p2 < 16 * p2 = 2^4 * p2 <= 2^64. +// +// Since p2 = f mod 2^-e < 2^-e, +// +// -e <= 60 or e >= -60 := alpha + +constexpr int kAlpha = -60; +constexpr int kGamma = -32; + +struct cached_power // c = f * 2^e ~= 10^k +{ + uint64_t f; + int e; + int k; +}; + +/*! +For a normalized diyfp w = f * 2^e, this function returns a (normalized) cached +power-of-ten c = f_c * 2^e_c, such that the exponent of the product w * c +satisfies (Definition 3.2 from [1]) + + alpha <= e_c + e + q <= gamma. +*/ +inline cached_power get_cached_power_for_binary_exponent(int e) +{ + // Now + // + // alpha <= e_c + e + q <= gamma (1) + // ==> f_c * 2^alpha <= c * 2^e * 2^q + // + // and since the c's are normalized, 2^(q-1) <= f_c, + // + // ==> 2^(q - 1 + alpha) <= c * 2^(e + q) + // ==> 2^(alpha - e - 1) <= c + // + // If c were an exakt power of ten, i.e. c = 10^k, one may determine k as + // + // k = ceil( log_10( 2^(alpha - e - 1) ) ) + // = ceil( (alpha - e - 1) * log_10(2) ) + // + // From the paper: + // "In theory the result of the procedure could be wrong since c is rounded, + // and the computation itself is approximated [...]. In practice, however, + // this simple function is sufficient." + // + // For IEEE double precision floating-point numbers converted into + // normalized diyfp's w = f * 2^e, with q = 64, + // + // e >= -1022 (min IEEE exponent) + // -52 (p - 1) + // -52 (p - 1, possibly normalize denormal IEEE numbers) + // -11 (normalize the diyfp) + // = -1137 + // + // and + // + // e <= +1023 (max IEEE exponent) + // -52 (p - 1) + // -11 (normalize the diyfp) + // = 960 + // + // This binary exponent range [-1137,960] results in a decimal exponent + // range [-307,324]. One does not need to store a cached power for each + // k in this range. For each such k it suffices to find a cached power + // such that the exponent of the product lies in [alpha,gamma]. + // This implies that the difference of the decimal exponents of adjacent + // table entries must be less than or equal to + // + // floor( (gamma - alpha) * log_10(2) ) = 8. + // + // (A smaller distance gamma-alpha would require a larger table.) + + // NB: + // Actually this function returns c, such that -60 <= e_c + e + 64 <= -34. + + constexpr int kCachedPowersSize = 79; + constexpr int kCachedPowersMinDecExp = -300; + constexpr int kCachedPowersDecStep = 8; + + static constexpr cached_power kCachedPowers[] = + { + { 0xAB70FE17C79AC6CA, -1060, -300 }, + { 0xFF77B1FCBEBCDC4F, -1034, -292 }, + { 0xBE5691EF416BD60C, -1007, -284 }, + { 0x8DD01FAD907FFC3C, -980, -276 }, + { 0xD3515C2831559A83, -954, -268 }, + { 0x9D71AC8FADA6C9B5, -927, -260 }, + { 0xEA9C227723EE8BCB, -901, -252 }, + { 0xAECC49914078536D, -874, -244 }, + { 0x823C12795DB6CE57, -847, -236 }, + { 0xC21094364DFB5637, -821, -228 }, + { 0x9096EA6F3848984F, -794, -220 }, + { 0xD77485CB25823AC7, -768, -212 }, + { 0xA086CFCD97BF97F4, -741, -204 }, + { 0xEF340A98172AACE5, -715, -196 }, + { 0xB23867FB2A35B28E, -688, -188 }, + { 0x84C8D4DFD2C63F3B, -661, -180 }, + { 0xC5DD44271AD3CDBA, -635, -172 }, + { 0x936B9FCEBB25C996, -608, -164 }, + { 0xDBAC6C247D62A584, -582, -156 }, + { 0xA3AB66580D5FDAF6, -555, -148 }, + { 0xF3E2F893DEC3F126, -529, -140 }, + { 0xB5B5ADA8AAFF80B8, -502, -132 }, + { 0x87625F056C7C4A8B, -475, -124 }, + { 0xC9BCFF6034C13053, -449, -116 }, + { 0x964E858C91BA2655, -422, -108 }, + { 0xDFF9772470297EBD, -396, -100 }, + { 0xA6DFBD9FB8E5B88F, -369, -92 }, + { 0xF8A95FCF88747D94, -343, -84 }, + { 0xB94470938FA89BCF, -316, -76 }, + { 0x8A08F0F8BF0F156B, -289, -68 }, + { 0xCDB02555653131B6, -263, -60 }, + { 0x993FE2C6D07B7FAC, -236, -52 }, + { 0xE45C10C42A2B3B06, -210, -44 }, + { 0xAA242499697392D3, -183, -36 }, + { 0xFD87B5F28300CA0E, -157, -28 }, + { 0xBCE5086492111AEB, -130, -20 }, + { 0x8CBCCC096F5088CC, -103, -12 }, + { 0xD1B71758E219652C, -77, -4 }, + { 0x9C40000000000000, -50, 4 }, + { 0xE8D4A51000000000, -24, 12 }, + { 0xAD78EBC5AC620000, 3, 20 }, + { 0x813F3978F8940984, 30, 28 }, + { 0xC097CE7BC90715B3, 56, 36 }, + { 0x8F7E32CE7BEA5C70, 83, 44 }, + { 0xD5D238A4ABE98068, 109, 52 }, + { 0x9F4F2726179A2245, 136, 60 }, + { 0xED63A231D4C4FB27, 162, 68 }, + { 0xB0DE65388CC8ADA8, 189, 76 }, + { 0x83C7088E1AAB65DB, 216, 84 }, + { 0xC45D1DF942711D9A, 242, 92 }, + { 0x924D692CA61BE758, 269, 100 }, + { 0xDA01EE641A708DEA, 295, 108 }, + { 0xA26DA3999AEF774A, 322, 116 }, + { 0xF209787BB47D6B85, 348, 124 }, + { 0xB454E4A179DD1877, 375, 132 }, + { 0x865B86925B9BC5C2, 402, 140 }, + { 0xC83553C5C8965D3D, 428, 148 }, + { 0x952AB45CFA97A0B3, 455, 156 }, + { 0xDE469FBD99A05FE3, 481, 164 }, + { 0xA59BC234DB398C25, 508, 172 }, + { 0xF6C69A72A3989F5C, 534, 180 }, + { 0xB7DCBF5354E9BECE, 561, 188 }, + { 0x88FCF317F22241E2, 588, 196 }, + { 0xCC20CE9BD35C78A5, 614, 204 }, + { 0x98165AF37B2153DF, 641, 212 }, + { 0xE2A0B5DC971F303A, 667, 220 }, + { 0xA8D9D1535CE3B396, 694, 228 }, + { 0xFB9B7CD9A4A7443C, 720, 236 }, + { 0xBB764C4CA7A44410, 747, 244 }, + { 0x8BAB8EEFB6409C1A, 774, 252 }, + { 0xD01FEF10A657842C, 800, 260 }, + { 0x9B10A4E5E9913129, 827, 268 }, + { 0xE7109BFBA19C0C9D, 853, 276 }, + { 0xAC2820D9623BF429, 880, 284 }, + { 0x80444B5E7AA7CF85, 907, 292 }, + { 0xBF21E44003ACDD2D, 933, 300 }, + { 0x8E679C2F5E44FF8F, 960, 308 }, + { 0xD433179D9C8CB841, 986, 316 }, + { 0x9E19DB92B4E31BA9, 1013, 324 }, + }; + + // This computation gives exactly the same results for k as + // k = ceil((kAlpha - e - 1) * 0.30102999566398114) + // for |e| <= 1500, but doesn't require floating-point operations. + // NB: log_10(2) ~= 78913 / 2^18 + assert(e >= -1500); + assert(e <= 1500); + const int f = kAlpha - e - 1; + const int k = (f * 78913) / (1 << 18) + (f > 0); + + const int index = (-kCachedPowersMinDecExp + k + (kCachedPowersDecStep - 1)) / kCachedPowersDecStep; + assert(index >= 0); + assert(index < kCachedPowersSize); + static_cast(kCachedPowersSize); // Fix warning. + + const cached_power cached = kCachedPowers[index]; + assert(kAlpha <= cached.e + e + 64); + assert(kGamma >= cached.e + e + 64); + + return cached; +} + +/*! +For n != 0, returns k, such that pow10 := 10^(k-1) <= n < 10^k. +For n == 0, returns 1 and sets pow10 := 1. +*/ +inline int find_largest_pow10(const uint32_t n, uint32_t& pow10) +{ + // LCOV_EXCL_START + if (n >= 1000000000) + { + pow10 = 1000000000; + return 10; + } + // LCOV_EXCL_STOP + else if (n >= 100000000) + { + pow10 = 100000000; + return 9; + } + else if (n >= 10000000) + { + pow10 = 10000000; + return 8; + } + else if (n >= 1000000) + { + pow10 = 1000000; + return 7; + } + else if (n >= 100000) + { + pow10 = 100000; + return 6; + } + else if (n >= 10000) + { + pow10 = 10000; + return 5; + } + else if (n >= 1000) + { + pow10 = 1000; + return 4; + } + else if (n >= 100) + { + pow10 = 100; + return 3; + } + else if (n >= 10) + { + pow10 = 10; + return 2; + } + else + { + pow10 = 1; + return 1; + } +} + +inline void grisu2_round(char* buf, int len, uint64_t dist, uint64_t delta, + uint64_t rest, uint64_t ten_k) +{ + assert(len >= 1); + assert(dist <= delta); + assert(rest <= delta); + assert(ten_k > 0); + + // <--------------------------- delta ----> + // <---- dist ---------> + // --------------[------------------+-------------------]-------------- + // M- w M+ + // + // ten_k + // <------> + // <---- rest ----> + // --------------[------------------+----+--------------]-------------- + // w V + // = buf * 10^k + // + // ten_k represents a unit-in-the-last-place in the decimal representation + // stored in buf. + // Decrement buf by ten_k while this takes buf closer to w. + + // The tests are written in this order to avoid overflow in unsigned + // integer arithmetic. + + while (rest < dist + and delta - rest >= ten_k + and (rest + ten_k < dist or dist - rest > rest + ten_k - dist)) + { + assert(buf[len - 1] != '0'); + buf[len - 1]--; + rest += ten_k; + } +} + +/*! +Generates V = buffer * 10^decimal_exponent, such that M- <= V <= M+. +M- and M+ must be normalized and share the same exponent -60 <= e <= -32. +*/ +inline void grisu2_digit_gen(char* buffer, int& length, int& decimal_exponent, + diyfp M_minus, diyfp w, diyfp M_plus) +{ + static_assert(kAlpha >= -60, "internal error"); + static_assert(kGamma <= -32, "internal error"); + + // Generates the digits (and the exponent) of a decimal floating-point + // number V = buffer * 10^decimal_exponent in the range [M-, M+]. The diyfp's + // w, M- and M+ share the same exponent e, which satisfies alpha <= e <= gamma. + // + // <--------------------------- delta ----> + // <---- dist ---------> + // --------------[------------------+-------------------]-------------- + // M- w M+ + // + // Grisu2 generates the digits of M+ from left to right and stops as soon as + // V is in [M-,M+]. + + assert(M_plus.e >= kAlpha); + assert(M_plus.e <= kGamma); + + uint64_t delta = diyfp::sub(M_plus, M_minus).f; // (significand of (M+ - M-), implicit exponent is e) + uint64_t dist = diyfp::sub(M_plus, w ).f; // (significand of (M+ - w ), implicit exponent is e) + + // Split M+ = f * 2^e into two parts p1 and p2 (note: e < 0): + // + // M+ = f * 2^e + // = ((f div 2^-e) * 2^-e + (f mod 2^-e)) * 2^e + // = ((p1 ) * 2^-e + (p2 )) * 2^e + // = p1 + p2 * 2^e + + const diyfp one(uint64_t{1} << -M_plus.e, M_plus.e); + + uint32_t p1 = static_cast(M_plus.f >> -one.e); // p1 = f div 2^-e (Since -e >= 32, p1 fits into a 32-bit int.) + uint64_t p2 = M_plus.f & (one.f - 1); // p2 = f mod 2^-e + + // 1) + // + // Generate the digits of the integral part p1 = d[n-1]...d[1]d[0] + + assert(p1 > 0); + + uint32_t pow10; + const int k = find_largest_pow10(p1, pow10); + + // 10^(k-1) <= p1 < 10^k, pow10 = 10^(k-1) + // + // p1 = (p1 div 10^(k-1)) * 10^(k-1) + (p1 mod 10^(k-1)) + // = (d[k-1] ) * 10^(k-1) + (p1 mod 10^(k-1)) + // + // M+ = p1 + p2 * 2^e + // = d[k-1] * 10^(k-1) + (p1 mod 10^(k-1)) + p2 * 2^e + // = d[k-1] * 10^(k-1) + ((p1 mod 10^(k-1)) * 2^-e + p2) * 2^e + // = d[k-1] * 10^(k-1) + ( rest) * 2^e + // + // Now generate the digits d[n] of p1 from left to right (n = k-1,...,0) + // + // p1 = d[k-1]...d[n] * 10^n + d[n-1]...d[0] + // + // but stop as soon as + // + // rest * 2^e = (d[n-1]...d[0] * 2^-e + p2) * 2^e <= delta * 2^e + + int n = k; + while (n > 0) + { + // Invariants: + // M+ = buffer * 10^n + (p1 + p2 * 2^e) (buffer = 0 for n = k) + // pow10 = 10^(n-1) <= p1 < 10^n + // + const uint32_t d = p1 / pow10; // d = p1 div 10^(n-1) + const uint32_t r = p1 % pow10; // r = p1 mod 10^(n-1) + // + // M+ = buffer * 10^n + (d * 10^(n-1) + r) + p2 * 2^e + // = (buffer * 10 + d) * 10^(n-1) + (r + p2 * 2^e) + // + assert(d <= 9); + buffer[length++] = static_cast('0' + d); // buffer := buffer * 10 + d + // + // M+ = buffer * 10^(n-1) + (r + p2 * 2^e) + // + p1 = r; + n--; + // + // M+ = buffer * 10^n + (p1 + p2 * 2^e) + // pow10 = 10^n + // + + // Now check if enough digits have been generated. + // Compute + // + // p1 + p2 * 2^e = (p1 * 2^-e + p2) * 2^e = rest * 2^e + // + // Note: + // Since rest and delta share the same exponent e, it suffices to + // compare the significands. + const uint64_t rest = (uint64_t{p1} << -one.e) + p2; + if (rest <= delta) + { + // V = buffer * 10^n, with M- <= V <= M+. + + decimal_exponent += n; + + // We may now just stop. But instead look if the buffer could be + // decremented to bring V closer to w. + // + // pow10 = 10^n is now 1 ulp in the decimal representation V. + // The rounding procedure works with diyfp's with an implicit + // exponent of e. + // + // 10^n = (10^n * 2^-e) * 2^e = ulp * 2^e + // + const uint64_t ten_n = uint64_t{pow10} << -one.e; + grisu2_round(buffer, length, dist, delta, rest, ten_n); + + return; + } + + pow10 /= 10; + // + // pow10 = 10^(n-1) <= p1 < 10^n + // Invariants restored. + } + + // 2) + // + // The digits of the integral part have been generated: + // + // M+ = d[k-1]...d[1]d[0] + p2 * 2^e + // = buffer + p2 * 2^e + // + // Now generate the digits of the fractional part p2 * 2^e. + // + // Note: + // No decimal point is generated: the exponent is adjusted instead. + // + // p2 actually represents the fraction + // + // p2 * 2^e + // = p2 / 2^-e + // = d[-1] / 10^1 + d[-2] / 10^2 + ... + // + // Now generate the digits d[-m] of p1 from left to right (m = 1,2,...) + // + // p2 * 2^e = d[-1]d[-2]...d[-m] * 10^-m + // + 10^-m * (d[-m-1] / 10^1 + d[-m-2] / 10^2 + ...) + // + // using + // + // 10^m * p2 = ((10^m * p2) div 2^-e) * 2^-e + ((10^m * p2) mod 2^-e) + // = ( d) * 2^-e + ( r) + // + // or + // 10^m * p2 * 2^e = d + r * 2^e + // + // i.e. + // + // M+ = buffer + p2 * 2^e + // = buffer + 10^-m * (d + r * 2^e) + // = (buffer * 10^m + d) * 10^-m + 10^-m * r * 2^e + // + // and stop as soon as 10^-m * r * 2^e <= delta * 2^e + + assert(p2 > delta); + + int m = 0; + for (;;) + { + // Invariant: + // M+ = buffer * 10^-m + 10^-m * (d[-m-1] / 10 + d[-m-2] / 10^2 + ...) * 2^e + // = buffer * 10^-m + 10^-m * (p2 ) * 2^e + // = buffer * 10^-m + 10^-m * (1/10 * (10 * p2) ) * 2^e + // = buffer * 10^-m + 10^-m * (1/10 * ((10*p2 div 2^-e) * 2^-e + (10*p2 mod 2^-e)) * 2^e + // + assert(p2 <= UINT64_MAX / 10); + p2 *= 10; + const uint64_t d = p2 >> -one.e; // d = (10 * p2) div 2^-e + const uint64_t r = p2 & (one.f - 1); // r = (10 * p2) mod 2^-e + // + // M+ = buffer * 10^-m + 10^-m * (1/10 * (d * 2^-e + r) * 2^e + // = buffer * 10^-m + 10^-m * (1/10 * (d + r * 2^e)) + // = (buffer * 10 + d) * 10^(-m-1) + 10^(-m-1) * r * 2^e + // + assert(d <= 9); + buffer[length++] = static_cast('0' + d); // buffer := buffer * 10 + d + // + // M+ = buffer * 10^(-m-1) + 10^(-m-1) * r * 2^e + // + p2 = r; + m++; + // + // M+ = buffer * 10^-m + 10^-m * p2 * 2^e + // Invariant restored. + + // Check if enough digits have been generated. + // + // 10^-m * p2 * 2^e <= delta * 2^e + // p2 * 2^e <= 10^m * delta * 2^e + // p2 <= 10^m * delta + delta *= 10; + dist *= 10; + if (p2 <= delta) + { + break; + } + } + + // V = buffer * 10^-m, with M- <= V <= M+. + + decimal_exponent -= m; + + // 1 ulp in the decimal representation is now 10^-m. + // Since delta and dist are now scaled by 10^m, we need to do the + // same with ulp in order to keep the units in sync. + // + // 10^m * 10^-m = 1 = 2^-e * 2^e = ten_m * 2^e + // + const uint64_t ten_m = one.f; + grisu2_round(buffer, length, dist, delta, p2, ten_m); + + // By construction this algorithm generates the shortest possible decimal + // number (Loitsch, Theorem 6.2) which rounds back to w. + // For an input number of precision p, at least + // + // N = 1 + ceil(p * log_10(2)) + // + // decimal digits are sufficient to identify all binary floating-point + // numbers (Matula, "In-and-Out conversions"). + // This implies that the algorithm does not produce more than N decimal + // digits. + // + // N = 17 for p = 53 (IEEE double precision) + // N = 9 for p = 24 (IEEE single precision) +} + +/*! +v = buf * 10^decimal_exponent +len is the length of the buffer (number of decimal digits) +The buffer must be large enough, i.e. >= max_digits10. +*/ +inline void grisu2(char* buf, int& len, int& decimal_exponent, + diyfp m_minus, diyfp v, diyfp m_plus) +{ + assert(m_plus.e == m_minus.e); + assert(m_plus.e == v.e); + + // --------(-----------------------+-----------------------)-------- (A) + // m- v m+ + // + // --------------------(-----------+-----------------------)-------- (B) + // m- v m+ + // + // First scale v (and m- and m+) such that the exponent is in the range + // [alpha, gamma]. + + const cached_power cached = get_cached_power_for_binary_exponent(m_plus.e); + + const diyfp c_minus_k(cached.f, cached.e); // = c ~= 10^-k + + // The exponent of the products is = v.e + c_minus_k.e + q and is in the range [alpha,gamma] + const diyfp w = diyfp::mul(v, c_minus_k); + const diyfp w_minus = diyfp::mul(m_minus, c_minus_k); + const diyfp w_plus = diyfp::mul(m_plus, c_minus_k); + + // ----(---+---)---------------(---+---)---------------(---+---)---- + // w- w w+ + // = c*m- = c*v = c*m+ + // + // diyfp::mul rounds its result and c_minus_k is approximated too. w, w- and + // w+ are now off by a small amount. + // In fact: + // + // w - v * 10^k < 1 ulp + // + // To account for this inaccuracy, add resp. subtract 1 ulp. + // + // --------+---[---------------(---+---)---------------]---+-------- + // w- M- w M+ w+ + // + // Now any number in [M-, M+] (bounds included) will round to w when input, + // regardless of how the input rounding algorithm breaks ties. + // + // And digit_gen generates the shortest possible such number in [M-, M+]. + // Note that this does not mean that Grisu2 always generates the shortest + // possible number in the interval (m-, m+). + const diyfp M_minus(w_minus.f + 1, w_minus.e); + const diyfp M_plus (w_plus.f - 1, w_plus.e ); + + decimal_exponent = -cached.k; // = -(-k) = k + + grisu2_digit_gen(buf, len, decimal_exponent, M_minus, w, M_plus); +} + +/*! +v = buf * 10^decimal_exponent +len is the length of the buffer (number of decimal digits) +The buffer must be large enough, i.e. >= max_digits10. +*/ +template +void grisu2(char* buf, int& len, int& decimal_exponent, FloatType value) +{ + static_assert(diyfp::kPrecision >= std::numeric_limits::digits + 3, + "internal error: not enough precision"); + + assert(std::isfinite(value)); + assert(value > 0); + + // If the neighbors (and boundaries) of 'value' are always computed for double-precision + // numbers, all float's can be recovered using strtod (and strtof). However, the resulting + // decimal representations are not exactly "short". + // + // The documentation for 'std::to_chars' (http://en.cppreference.com/w/cpp/utility/to_chars) + // says "value is converted to a string as if by std::sprintf in the default ("C") locale" + // and since sprintf promotes float's to double's, I think this is exactly what 'std::to_chars' + // does. + // On the other hand, the documentation for 'std::to_chars' requires that "parsing the + // representation using the corresponding std::from_chars function recovers value exactly". That + // indicates that single precision floating-point numbers should be recovered using + // 'std::strtof'. + // + // NB: If the neighbors are computed for single-precision numbers, there is a single float + // (7.0385307e-26f) which can't be recovered using strtod. The resulting double precision + // value is off by 1 ulp. +#if 0 + const boundaries w = compute_boundaries(static_cast(value)); +#else + const boundaries w = compute_boundaries(value); +#endif + + grisu2(buf, len, decimal_exponent, w.minus, w.w, w.plus); +} + +/*! +@brief appends a decimal representation of e to buf +@return a pointer to the element following the exponent. +@pre -1000 < e < 1000 +*/ +inline char* append_exponent(char* buf, int e) +{ + assert(e > -1000); + assert(e < 1000); + + if (e < 0) + { + e = -e; + *buf++ = '-'; + } + else + { + *buf++ = '+'; + } + + uint32_t k = static_cast(e); + if (k < 10) + { + // Always print at least two digits in the exponent. + // This is for compatibility with printf("%g"). + *buf++ = '0'; + *buf++ = static_cast('0' + k); + } + else if (k < 100) + { + *buf++ = static_cast('0' + k / 10); + k %= 10; + *buf++ = static_cast('0' + k); + } + else + { + *buf++ = static_cast('0' + k / 100); + k %= 100; + *buf++ = static_cast('0' + k / 10); + k %= 10; + *buf++ = static_cast('0' + k); + } + + return buf; +} + +/*! +@brief prettify v = buf * 10^decimal_exponent + +If v is in the range [10^min_exp, 10^max_exp) it will be printed in fixed-point +notation. Otherwise it will be printed in exponential notation. + +@pre min_exp < 0 +@pre max_exp > 0 +*/ +inline char* format_buffer(char* buf, int len, int decimal_exponent, + int min_exp, int max_exp) +{ + assert(min_exp < 0); + assert(max_exp > 0); + + const int k = len; + const int n = len + decimal_exponent; + + // v = buf * 10^(n-k) + // k is the length of the buffer (number of decimal digits) + // n is the position of the decimal point relative to the start of the buffer. + + if (k <= n and n <= max_exp) + { + // digits[000] + // len <= max_exp + 2 + + std::memset(buf + k, '0', static_cast(n - k)); + // Make it look like a floating-point number (#362, #378) + buf[n + 0] = '.'; + buf[n + 1] = '0'; + return buf + (n + 2); + } + + if (0 < n and n <= max_exp) + { + // dig.its + // len <= max_digits10 + 1 + + assert(k > n); + + std::memmove(buf + (n + 1), buf + n, static_cast(k - n)); + buf[n] = '.'; + return buf + (k + 1); + } + + if (min_exp < n and n <= 0) + { + // 0.[000]digits + // len <= 2 + (-min_exp - 1) + max_digits10 + + std::memmove(buf + (2 + -n), buf, static_cast(k)); + buf[0] = '0'; + buf[1] = '.'; + std::memset(buf + 2, '0', static_cast(-n)); + return buf + (2 + (-n) + k); + } + + if (k == 1) + { + // dE+123 + // len <= 1 + 5 + + buf += 1; + } + else + { + // d.igitsE+123 + // len <= max_digits10 + 1 + 5 + + std::memmove(buf + 2, buf + 1, static_cast(k - 1)); + buf[1] = '.'; + buf += 1 + k; + } + + *buf++ = 'e'; + return append_exponent(buf, n - 1); +} + +} // namespace dtoa_impl + +/*! +@brief generates a decimal representation of the floating-point number value in [first, last). + +The format of the resulting decimal representation is similar to printf's %g +format. Returns an iterator pointing past-the-end of the decimal representation. + +@note The input number must be finite, i.e. NaN's and Inf's are not supported. +@note The buffer must be large enough. +@note The result is NOT null-terminated. +*/ +template +char* to_chars(char* first, char* last, FloatType value) +{ + static_cast(last); // maybe unused - fix warning + assert(std::isfinite(value)); + + // Use signbit(value) instead of (value < 0) since signbit works for -0. + if (std::signbit(value)) + { + value = -value; + *first++ = '-'; + } + + if (value == 0) // +-0 + { + *first++ = '0'; + // Make it look like a floating-point number (#362, #378) + *first++ = '.'; + *first++ = '0'; + return first; + } + + assert(last - first >= std::numeric_limits::max_digits10); + + // Compute v = buffer * 10^decimal_exponent. + // The decimal digits are stored in the buffer, which needs to be interpreted + // as an unsigned decimal integer. + // len is the length of the buffer, i.e. the number of decimal digits. + int len = 0; + int decimal_exponent = 0; + dtoa_impl::grisu2(first, len, decimal_exponent, value); + + assert(len <= std::numeric_limits::max_digits10); + + // Format the buffer like printf("%.*g", prec, value) + constexpr int kMinExp = -4; + // Use digits10 here to increase compatibility with version 2. + constexpr int kMaxExp = std::numeric_limits::digits10; + + assert(last - first >= kMaxExp + 2); + assert(last - first >= 2 + (-kMinExp - 1) + std::numeric_limits::max_digits10); + assert(last - first >= std::numeric_limits::max_digits10 + 6); + + return dtoa_impl::format_buffer(first, len, decimal_exponent, kMinExp, kMaxExp); +} + +} // namespace detail +} // namespace nlohmann + +// #include + +// #include + +// #include + +// #include + + +namespace nlohmann +{ +namespace detail +{ +/////////////////// +// serialization // +/////////////////// + +template +class serializer +{ + using string_t = typename BasicJsonType::string_t; + using number_float_t = typename BasicJsonType::number_float_t; + using number_integer_t = typename BasicJsonType::number_integer_t; + using number_unsigned_t = typename BasicJsonType::number_unsigned_t; + static constexpr uint8_t UTF8_ACCEPT = 0; + static constexpr uint8_t UTF8_REJECT = 1; + + public: + /*! + @param[in] s output stream to serialize to + @param[in] ichar indentation character to use + */ + serializer(output_adapter_t s, const char ichar) + : o(std::move(s)), loc(std::localeconv()), + thousands_sep(loc->thousands_sep == nullptr ? '\0' : * (loc->thousands_sep)), + decimal_point(loc->decimal_point == nullptr ? '\0' : * (loc->decimal_point)), + indent_char(ichar), indent_string(512, indent_char) + {} + + // delete because of pointer members + serializer(const serializer&) = delete; + serializer& operator=(const serializer&) = delete; + + /*! + @brief internal implementation of the serialization function + + This function is called by the public member function dump and organizes + the serialization internally. The indentation level is propagated as + additional parameter. In case of arrays and objects, the function is + called recursively. + + - strings and object keys are escaped using `escape_string()` + - integer numbers are converted implicitly via `operator<<` + - floating-point numbers are converted to a string using `"%g"` format + + @param[in] val value to serialize + @param[in] pretty_print whether the output shall be pretty-printed + @param[in] indent_step the indent level + @param[in] current_indent the current indent level (only used internally) + */ + void dump(const BasicJsonType& val, const bool pretty_print, + const bool ensure_ascii, + const unsigned int indent_step, + const unsigned int current_indent = 0) + { + switch (val.m_type) + { + case value_t::object: + { + if (val.m_value.object->empty()) + { + o->write_characters("{}", 2); + return; + } + + if (pretty_print) + { + o->write_characters("{\n", 2); + + // variable to hold indentation for recursive calls + const auto new_indent = current_indent + indent_step; + if (JSON_UNLIKELY(indent_string.size() < new_indent)) + { + indent_string.resize(indent_string.size() * 2, ' '); + } + + // first n-1 elements + auto i = val.m_value.object->cbegin(); + for (std::size_t cnt = 0; cnt < val.m_value.object->size() - 1; ++cnt, ++i) + { + o->write_characters(indent_string.c_str(), new_indent); + o->write_character('\"'); + dump_escaped(i->first, ensure_ascii); + o->write_characters("\": ", 3); + dump(i->second, true, ensure_ascii, indent_step, new_indent); + o->write_characters(",\n", 2); + } + + // last element + assert(i != val.m_value.object->cend()); + assert(std::next(i) == val.m_value.object->cend()); + o->write_characters(indent_string.c_str(), new_indent); + o->write_character('\"'); + dump_escaped(i->first, ensure_ascii); + o->write_characters("\": ", 3); + dump(i->second, true, ensure_ascii, indent_step, new_indent); + + o->write_character('\n'); + o->write_characters(indent_string.c_str(), current_indent); + o->write_character('}'); + } + else + { + o->write_character('{'); + + // first n-1 elements + auto i = val.m_value.object->cbegin(); + for (std::size_t cnt = 0; cnt < val.m_value.object->size() - 1; ++cnt, ++i) + { + o->write_character('\"'); + dump_escaped(i->first, ensure_ascii); + o->write_characters("\":", 2); + dump(i->second, false, ensure_ascii, indent_step, current_indent); + o->write_character(','); + } + + // last element + assert(i != val.m_value.object->cend()); + assert(std::next(i) == val.m_value.object->cend()); + o->write_character('\"'); + dump_escaped(i->first, ensure_ascii); + o->write_characters("\":", 2); + dump(i->second, false, ensure_ascii, indent_step, current_indent); + + o->write_character('}'); + } + + return; + } + + case value_t::array: + { + if (val.m_value.array->empty()) + { + o->write_characters("[]", 2); + return; + } + + if (pretty_print) + { + o->write_characters("[\n", 2); + + // variable to hold indentation for recursive calls + const auto new_indent = current_indent + indent_step; + if (JSON_UNLIKELY(indent_string.size() < new_indent)) + { + indent_string.resize(indent_string.size() * 2, ' '); + } + + // first n-1 elements + for (auto i = val.m_value.array->cbegin(); + i != val.m_value.array->cend() - 1; ++i) + { + o->write_characters(indent_string.c_str(), new_indent); + dump(*i, true, ensure_ascii, indent_step, new_indent); + o->write_characters(",\n", 2); + } + + // last element + assert(not val.m_value.array->empty()); + o->write_characters(indent_string.c_str(), new_indent); + dump(val.m_value.array->back(), true, ensure_ascii, indent_step, new_indent); + + o->write_character('\n'); + o->write_characters(indent_string.c_str(), current_indent); + o->write_character(']'); + } + else + { + o->write_character('['); + + // first n-1 elements + for (auto i = val.m_value.array->cbegin(); + i != val.m_value.array->cend() - 1; ++i) + { + dump(*i, false, ensure_ascii, indent_step, current_indent); + o->write_character(','); + } + + // last element + assert(not val.m_value.array->empty()); + dump(val.m_value.array->back(), false, ensure_ascii, indent_step, current_indent); + + o->write_character(']'); + } + + return; + } + + case value_t::string: + { + o->write_character('\"'); + dump_escaped(*val.m_value.string, ensure_ascii); + o->write_character('\"'); + return; + } + + case value_t::boolean: + { + if (val.m_value.boolean) + { + o->write_characters("true", 4); + } + else + { + o->write_characters("false", 5); + } + return; + } + + case value_t::number_integer: + { + dump_integer(val.m_value.number_integer); + return; + } + + case value_t::number_unsigned: + { + dump_integer(val.m_value.number_unsigned); + return; + } + + case value_t::number_float: + { + dump_float(val.m_value.number_float); + return; + } + + case value_t::discarded: + { + o->write_characters("", 11); + return; + } + + case value_t::null: + { + o->write_characters("null", 4); + return; + } + } + } + + private: + /*! + @brief dump escaped string + + Escape a string by replacing certain special characters by a sequence of an + escape character (backslash) and another character and other control + characters by a sequence of "\u" followed by a four-digit hex + representation. The escaped string is written to output stream @a o. + + @param[in] s the string to escape + @param[in] ensure_ascii whether to escape non-ASCII characters with + \uXXXX sequences + + @complexity Linear in the length of string @a s. + */ + void dump_escaped(const string_t& s, const bool ensure_ascii) + { + uint32_t codepoint; + uint8_t state = UTF8_ACCEPT; + std::size_t bytes = 0; // number of bytes written to string_buffer + + for (std::size_t i = 0; i < s.size(); ++i) + { + const auto byte = static_cast(s[i]); + + switch (decode(state, codepoint, byte)) + { + case UTF8_ACCEPT: // decode found a new code point + { + switch (codepoint) + { + case 0x08: // backspace + { + string_buffer[bytes++] = '\\'; + string_buffer[bytes++] = 'b'; + break; + } + + case 0x09: // horizontal tab + { + string_buffer[bytes++] = '\\'; + string_buffer[bytes++] = 't'; + break; + } + + case 0x0A: // newline + { + string_buffer[bytes++] = '\\'; + string_buffer[bytes++] = 'n'; + break; + } + + case 0x0C: // formfeed + { + string_buffer[bytes++] = '\\'; + string_buffer[bytes++] = 'f'; + break; + } + + case 0x0D: // carriage return + { + string_buffer[bytes++] = '\\'; + string_buffer[bytes++] = 'r'; + break; + } + + case 0x22: // quotation mark + { + string_buffer[bytes++] = '\\'; + string_buffer[bytes++] = '\"'; + break; + } + + case 0x5C: // reverse solidus + { + string_buffer[bytes++] = '\\'; + string_buffer[bytes++] = '\\'; + break; + } + + default: + { + // escape control characters (0x00..0x1F) or, if + // ensure_ascii parameter is used, non-ASCII characters + if ((codepoint <= 0x1F) or (ensure_ascii and (codepoint >= 0x7F))) + { + if (codepoint <= 0xFFFF) + { + std::snprintf(string_buffer.data() + bytes, 7, "\\u%04x", + static_cast(codepoint)); + bytes += 6; + } + else + { + std::snprintf(string_buffer.data() + bytes, 13, "\\u%04x\\u%04x", + static_cast(0xD7C0 + (codepoint >> 10)), + static_cast(0xDC00 + (codepoint & 0x3FF))); + bytes += 12; + } + } + else + { + // copy byte to buffer (all previous bytes + // been copied have in default case above) + string_buffer[bytes++] = s[i]; + } + break; + } + } + + // write buffer and reset index; there must be 13 bytes + // left, as this is the maximal number of bytes to be + // written ("\uxxxx\uxxxx\0") for one code point + if (string_buffer.size() - bytes < 13) + { + o->write_characters(string_buffer.data(), bytes); + bytes = 0; + } + break; + } + + case UTF8_REJECT: // decode found invalid UTF-8 byte + { + std::stringstream ss; + ss << std::setw(2) << std::uppercase << std::setfill('0') << std::hex << static_cast(byte); + JSON_THROW(type_error::create(316, "invalid UTF-8 byte at index " + std::to_string(i) + ": 0x" + ss.str())); + } + + default: // decode found yet incomplete multi-byte code point + { + if (not ensure_ascii) + { + // code point will not be escaped - copy byte to buffer + string_buffer[bytes++] = s[i]; + } + break; + } + } + } + + if (JSON_LIKELY(state == UTF8_ACCEPT)) + { + // write buffer + if (bytes > 0) + { + o->write_characters(string_buffer.data(), bytes); + } + } + else + { + // we finish reading, but do not accept: string was incomplete + std::stringstream ss; + ss << std::setw(2) << std::uppercase << std::setfill('0') << std::hex << static_cast(static_cast(s.back())); + JSON_THROW(type_error::create(316, "incomplete UTF-8 string; last byte: 0x" + ss.str())); + } + } + + /*! + @brief dump an integer + + Dump a given integer to output stream @a o. Works internally with + @a number_buffer. + + @param[in] x integer number (signed or unsigned) to dump + @tparam NumberType either @a number_integer_t or @a number_unsigned_t + */ + template::value or + std::is_same::value, + int> = 0> + void dump_integer(NumberType x) + { + // special case for "0" + if (x == 0) + { + o->write_character('0'); + return; + } + + const bool is_negative = (x <= 0) and (x != 0); // see issue #755 + std::size_t i = 0; + + while (x != 0) + { + // spare 1 byte for '\0' + assert(i < number_buffer.size() - 1); + + const auto digit = std::labs(static_cast(x % 10)); + number_buffer[i++] = static_cast('0' + digit); + x /= 10; + } + + if (is_negative) + { + // make sure there is capacity for the '-' + assert(i < number_buffer.size() - 2); + number_buffer[i++] = '-'; + } + + std::reverse(number_buffer.begin(), number_buffer.begin() + i); + o->write_characters(number_buffer.data(), i); + } + + /*! + @brief dump a floating-point number + + Dump a given floating-point number to output stream @a o. Works internally + with @a number_buffer. + + @param[in] x floating-point number to dump + */ + void dump_float(number_float_t x) + { + // NaN / inf + if (not std::isfinite(x)) + { + o->write_characters("null", 4); + return; + } + + // If number_float_t is an IEEE-754 single or double precision number, + // use the Grisu2 algorithm to produce short numbers which are + // guaranteed to round-trip, using strtof and strtod, resp. + // + // NB: The test below works if == . + static constexpr bool is_ieee_single_or_double + = (std::numeric_limits::is_iec559 and std::numeric_limits::digits == 24 and std::numeric_limits::max_exponent == 128) or + (std::numeric_limits::is_iec559 and std::numeric_limits::digits == 53 and std::numeric_limits::max_exponent == 1024); + + dump_float(x, std::integral_constant()); + } + + void dump_float(number_float_t x, std::true_type /*is_ieee_single_or_double*/) + { + char* begin = number_buffer.data(); + char* end = ::nlohmann::detail::to_chars(begin, begin + number_buffer.size(), x); + + o->write_characters(begin, static_cast(end - begin)); + } + + void dump_float(number_float_t x, std::false_type /*is_ieee_single_or_double*/) + { + // get number of digits for a float -> text -> float round-trip + static constexpr auto d = std::numeric_limits::max_digits10; + + // the actual conversion + std::ptrdiff_t len = snprintf(number_buffer.data(), number_buffer.size(), "%.*g", d, x); + + // negative value indicates an error + assert(len > 0); + // check if buffer was large enough + assert(static_cast(len) < number_buffer.size()); + + // erase thousands separator + if (thousands_sep != '\0') + { + const auto end = std::remove(number_buffer.begin(), + number_buffer.begin() + len, thousands_sep); + std::fill(end, number_buffer.end(), '\0'); + assert((end - number_buffer.begin()) <= len); + len = (end - number_buffer.begin()); + } + + // convert decimal point to '.' + if (decimal_point != '\0' and decimal_point != '.') + { + const auto dec_pos = std::find(number_buffer.begin(), number_buffer.end(), decimal_point); + if (dec_pos != number_buffer.end()) + { + *dec_pos = '.'; + } + } + + o->write_characters(number_buffer.data(), static_cast(len)); + + // determine if need to append ".0" + const bool value_is_int_like = + std::none_of(number_buffer.begin(), number_buffer.begin() + len + 1, + [](char c) + { + return (c == '.' or c == 'e'); + }); + + if (value_is_int_like) + { + o->write_characters(".0", 2); + } + } + + /*! + @brief check whether a string is UTF-8 encoded + + The function checks each byte of a string whether it is UTF-8 encoded. The + result of the check is stored in the @a state parameter. The function must + be called initially with state 0 (accept). State 1 means the string must + be rejected, because the current byte is not allowed. If the string is + completely processed, but the state is non-zero, the string ended + prematurely; that is, the last byte indicated more bytes should have + followed. + + @param[in,out] state the state of the decoding + @param[in,out] codep codepoint (valid only if resulting state is UTF8_ACCEPT) + @param[in] byte next byte to decode + @return new state + + @note The function has been edited: a std::array is used. + + @copyright Copyright (c) 2008-2009 Bjoern Hoehrmann + @sa http://bjoern.hoehrmann.de/utf-8/decoder/dfa/ + */ + static uint8_t decode(uint8_t& state, uint32_t& codep, const uint8_t byte) noexcept + { + static const std::array utf8d = + { + { + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 00..1F + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 20..3F + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 40..5F + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 60..7F + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, // 80..9F + 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, // A0..BF + 8, 8, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, // C0..DF + 0xA, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x4, 0x3, 0x3, // E0..EF + 0xB, 0x6, 0x6, 0x6, 0x5, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, // F0..FF + 0x0, 0x1, 0x2, 0x3, 0x5, 0x8, 0x7, 0x1, 0x1, 0x1, 0x4, 0x6, 0x1, 0x1, 0x1, 0x1, // s0..s0 + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, // s1..s2 + 1, 2, 1, 1, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, // s3..s4 + 1, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 1, 3, 1, 1, 1, 1, 1, 1, // s5..s6 + 1, 3, 1, 1, 1, 1, 1, 3, 1, 3, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 // s7..s8 + } + }; + + const uint8_t type = utf8d[byte]; + + codep = (state != UTF8_ACCEPT) + ? (byte & 0x3fu) | (codep << 6) + : static_cast(0xff >> type) & (byte); + + state = utf8d[256u + state * 16u + type]; + return state; + } + + private: + /// the output of the serializer + output_adapter_t o = nullptr; + + /// a (hopefully) large enough character buffer + std::array number_buffer{{}}; + + /// the locale + const std::lconv* loc = nullptr; + /// the locale's thousand separator character + const char thousands_sep = '\0'; + /// the locale's decimal point character + const char decimal_point = '\0'; + + /// string buffer + std::array string_buffer{{}}; + + /// the indentation character + const char indent_char; + /// the indentation string + string_t indent_string; +}; +} +} + +// #include + + +#include +#include + +namespace nlohmann +{ +namespace detail +{ +template +class json_ref +{ + public: + using value_type = BasicJsonType; + + json_ref(value_type&& value) + : owned_value(std::move(value)), value_ref(&owned_value), is_rvalue(true) + {} + + json_ref(const value_type& value) + : value_ref(const_cast(&value)), is_rvalue(false) + {} + + json_ref(std::initializer_list init) + : owned_value(init), value_ref(&owned_value), is_rvalue(true) + {} + + template + json_ref(Args&& ... args) + : owned_value(std::forward(args)...), value_ref(&owned_value), is_rvalue(true) + {} + + // class should be movable only + json_ref(json_ref&&) = default; + json_ref(const json_ref&) = delete; + json_ref& operator=(const json_ref&) = delete; + + value_type moved_or_copied() const + { + if (is_rvalue) + { + return std::move(*value_ref); + } + return *value_ref; + } + + value_type const& operator*() const + { + return *static_cast(value_ref); + } + + value_type const* operator->() const + { + return static_cast(value_ref); + } + + private: + mutable value_type owned_value = nullptr; + value_type* value_ref = nullptr; + const bool is_rvalue; +}; +} +} + +// #include + + +#include // assert +#include // accumulate +#include // string +#include // vector + +// #include + +// #include + +// #include + + +namespace nlohmann +{ +template +class json_pointer +{ + // allow basic_json to access private members + NLOHMANN_BASIC_JSON_TPL_DECLARATION + friend class basic_json; + + public: + /*! + @brief create JSON pointer + + Create a JSON pointer according to the syntax described in + [Section 3 of RFC6901](https://tools.ietf.org/html/rfc6901#section-3). + + @param[in] s string representing the JSON pointer; if omitted, the empty + string is assumed which references the whole JSON value + + @throw parse_error.107 if the given JSON pointer @a s is nonempty and does + not begin with a slash (`/`); see example below + + @throw parse_error.108 if a tilde (`~`) in the given JSON pointer @a s is + not followed by `0` (representing `~`) or `1` (representing `/`); see + example below + + @liveexample{The example shows the construction several valid JSON pointers + as well as the exceptional behavior.,json_pointer} + + @since version 2.0.0 + */ + explicit json_pointer(const std::string& s = "") + : reference_tokens(split(s)) + {} + + /*! + @brief return a string representation of the JSON pointer + + @invariant For each JSON pointer `ptr`, it holds: + @code {.cpp} + ptr == json_pointer(ptr.to_string()); + @endcode + + @return a string representation of the JSON pointer + + @liveexample{The example shows the result of `to_string`., + json_pointer__to_string} + + @since version 2.0.0 + */ + std::string to_string() const noexcept + { + return std::accumulate(reference_tokens.begin(), reference_tokens.end(), + std::string{}, + [](const std::string & a, const std::string & b) + { + return a + "/" + escape(b); + }); + } + + /// @copydoc to_string() + operator std::string() const + { + return to_string(); + } + + /*! + @param[in] s reference token to be converted into an array index + + @return integer representation of @a s + + @throw out_of_range.404 if string @a s could not be converted to an integer + */ + static int array_index(const std::string& s) + { + std::size_t processed_chars = 0; + const int res = std::stoi(s, &processed_chars); + + // check if the string was completely read + if (JSON_UNLIKELY(processed_chars != s.size())) + { + JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + s + "'")); + } + + return res; + } + + private: + /*! + @brief remove and return last reference pointer + @throw out_of_range.405 if JSON pointer has no parent + */ + std::string pop_back() + { + if (JSON_UNLIKELY(is_root())) + { + JSON_THROW(detail::out_of_range::create(405, "JSON pointer has no parent")); + } + + auto last = reference_tokens.back(); + reference_tokens.pop_back(); + return last; + } + + /// return whether pointer points to the root document + bool is_root() const + { + return reference_tokens.empty(); + } + + json_pointer top() const + { + if (JSON_UNLIKELY(is_root())) + { + JSON_THROW(detail::out_of_range::create(405, "JSON pointer has no parent")); + } + + json_pointer result = *this; + result.reference_tokens = {reference_tokens[0]}; + return result; + } + + /*! + @brief create and return a reference to the pointed to value + + @complexity Linear in the number of reference tokens. + + @throw parse_error.109 if array index is not a number + @throw type_error.313 if value cannot be unflattened + */ + BasicJsonType& get_and_create(BasicJsonType& j) const + { + using size_type = typename BasicJsonType::size_type; + auto result = &j; + + // in case no reference tokens exist, return a reference to the JSON value + // j which will be overwritten by a primitive value + for (const auto& reference_token : reference_tokens) + { + switch (result->m_type) + { + case detail::value_t::null: + { + if (reference_token == "0") + { + // start a new array if reference token is 0 + result = &result->operator[](0); + } + else + { + // start a new object otherwise + result = &result->operator[](reference_token); + } + break; + } + + case detail::value_t::object: + { + // create an entry in the object + result = &result->operator[](reference_token); + break; + } + + case detail::value_t::array: + { + // create an entry in the array + JSON_TRY + { + result = &result->operator[](static_cast(array_index(reference_token))); + } + JSON_CATCH(std::invalid_argument&) + { + JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); + } + break; + } + + /* + The following code is only reached if there exists a reference + token _and_ the current value is primitive. In this case, we have + an error situation, because primitive values may only occur as + single value; that is, with an empty list of reference tokens. + */ + default: + JSON_THROW(detail::type_error::create(313, "invalid value to unflatten")); + } + } + + return *result; + } + + /*! + @brief return a reference to the pointed to value + + @note This version does not throw if a value is not present, but tries to + create nested values instead. For instance, calling this function + with pointer `"/this/that"` on a null value is equivalent to calling + `operator[]("this").operator[]("that")` on that value, effectively + changing the null value to an object. + + @param[in] ptr a JSON value + + @return reference to the JSON value pointed to by the JSON pointer + + @complexity Linear in the length of the JSON pointer. + + @throw parse_error.106 if an array index begins with '0' + @throw parse_error.109 if an array index was not a number + @throw out_of_range.404 if the JSON pointer can not be resolved + */ + BasicJsonType& get_unchecked(BasicJsonType* ptr) const + { + using size_type = typename BasicJsonType::size_type; + for (const auto& reference_token : reference_tokens) + { + // convert null values to arrays or objects before continuing + if (ptr->m_type == detail::value_t::null) + { + // check if reference token is a number + const bool nums = + std::all_of(reference_token.begin(), reference_token.end(), + [](const char x) + { + return (x >= '0' and x <= '9'); + }); + + // change value to array for numbers or "-" or to object otherwise + *ptr = (nums or reference_token == "-") + ? detail::value_t::array + : detail::value_t::object; + } + + switch (ptr->m_type) + { + case detail::value_t::object: + { + // use unchecked object access + ptr = &ptr->operator[](reference_token); + break; + } + + case detail::value_t::array: + { + // error condition (cf. RFC 6901, Sect. 4) + if (JSON_UNLIKELY(reference_token.size() > 1 and reference_token[0] == '0')) + { + JSON_THROW(detail::parse_error::create(106, 0, + "array index '" + reference_token + + "' must not begin with '0'")); + } + + if (reference_token == "-") + { + // explicitly treat "-" as index beyond the end + ptr = &ptr->operator[](ptr->m_value.array->size()); + } + else + { + // convert array index to number; unchecked access + JSON_TRY + { + ptr = &ptr->operator[]( + static_cast(array_index(reference_token))); + } + JSON_CATCH(std::invalid_argument&) + { + JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); + } + } + break; + } + + default: + JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + reference_token + "'")); + } + } + + return *ptr; + } + + /*! + @throw parse_error.106 if an array index begins with '0' + @throw parse_error.109 if an array index was not a number + @throw out_of_range.402 if the array index '-' is used + @throw out_of_range.404 if the JSON pointer can not be resolved + */ + BasicJsonType& get_checked(BasicJsonType* ptr) const + { + using size_type = typename BasicJsonType::size_type; + for (const auto& reference_token : reference_tokens) + { + switch (ptr->m_type) + { + case detail::value_t::object: + { + // note: at performs range check + ptr = &ptr->at(reference_token); + break; + } + + case detail::value_t::array: + { + if (JSON_UNLIKELY(reference_token == "-")) + { + // "-" always fails the range check + JSON_THROW(detail::out_of_range::create(402, + "array index '-' (" + std::to_string(ptr->m_value.array->size()) + + ") is out of range")); + } + + // error condition (cf. RFC 6901, Sect. 4) + if (JSON_UNLIKELY(reference_token.size() > 1 and reference_token[0] == '0')) + { + JSON_THROW(detail::parse_error::create(106, 0, + "array index '" + reference_token + + "' must not begin with '0'")); + } + + // note: at performs range check + JSON_TRY + { + ptr = &ptr->at(static_cast(array_index(reference_token))); + } + JSON_CATCH(std::invalid_argument&) + { + JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); + } + break; + } + + default: + JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + reference_token + "'")); + } + } + + return *ptr; + } + + /*! + @brief return a const reference to the pointed to value + + @param[in] ptr a JSON value + + @return const reference to the JSON value pointed to by the JSON + pointer + + @throw parse_error.106 if an array index begins with '0' + @throw parse_error.109 if an array index was not a number + @throw out_of_range.402 if the array index '-' is used + @throw out_of_range.404 if the JSON pointer can not be resolved + */ + const BasicJsonType& get_unchecked(const BasicJsonType* ptr) const + { + using size_type = typename BasicJsonType::size_type; + for (const auto& reference_token : reference_tokens) + { + switch (ptr->m_type) + { + case detail::value_t::object: + { + // use unchecked object access + ptr = &ptr->operator[](reference_token); + break; + } + + case detail::value_t::array: + { + if (JSON_UNLIKELY(reference_token == "-")) + { + // "-" cannot be used for const access + JSON_THROW(detail::out_of_range::create(402, + "array index '-' (" + std::to_string(ptr->m_value.array->size()) + + ") is out of range")); + } + + // error condition (cf. RFC 6901, Sect. 4) + if (JSON_UNLIKELY(reference_token.size() > 1 and reference_token[0] == '0')) + { + JSON_THROW(detail::parse_error::create(106, 0, + "array index '" + reference_token + + "' must not begin with '0'")); + } + + // use unchecked array access + JSON_TRY + { + ptr = &ptr->operator[]( + static_cast(array_index(reference_token))); + } + JSON_CATCH(std::invalid_argument&) + { + JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); + } + break; + } + + default: + JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + reference_token + "'")); + } + } + + return *ptr; + } + + /*! + @throw parse_error.106 if an array index begins with '0' + @throw parse_error.109 if an array index was not a number + @throw out_of_range.402 if the array index '-' is used + @throw out_of_range.404 if the JSON pointer can not be resolved + */ + const BasicJsonType& get_checked(const BasicJsonType* ptr) const + { + using size_type = typename BasicJsonType::size_type; + for (const auto& reference_token : reference_tokens) + { + switch (ptr->m_type) + { + case detail::value_t::object: + { + // note: at performs range check + ptr = &ptr->at(reference_token); + break; + } + + case detail::value_t::array: + { + if (JSON_UNLIKELY(reference_token == "-")) + { + // "-" always fails the range check + JSON_THROW(detail::out_of_range::create(402, + "array index '-' (" + std::to_string(ptr->m_value.array->size()) + + ") is out of range")); + } + + // error condition (cf. RFC 6901, Sect. 4) + if (JSON_UNLIKELY(reference_token.size() > 1 and reference_token[0] == '0')) + { + JSON_THROW(detail::parse_error::create(106, 0, + "array index '" + reference_token + + "' must not begin with '0'")); + } + + // note: at performs range check + JSON_TRY + { + ptr = &ptr->at(static_cast(array_index(reference_token))); + } + JSON_CATCH(std::invalid_argument&) + { + JSON_THROW(detail::parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); + } + break; + } + + default: + JSON_THROW(detail::out_of_range::create(404, "unresolved reference token '" + reference_token + "'")); + } + } + + return *ptr; + } + + /*! + @brief split the string input to reference tokens + + @note This function is only called by the json_pointer constructor. + All exceptions below are documented there. + + @throw parse_error.107 if the pointer is not empty or begins with '/' + @throw parse_error.108 if character '~' is not followed by '0' or '1' + */ + static std::vector split(const std::string& reference_string) + { + std::vector result; + + // special case: empty reference string -> no reference tokens + if (reference_string.empty()) + { + return result; + } + + // check if nonempty reference string begins with slash + if (JSON_UNLIKELY(reference_string[0] != '/')) + { + JSON_THROW(detail::parse_error::create(107, 1, + "JSON pointer must be empty or begin with '/' - was: '" + + reference_string + "'")); + } + + // extract the reference tokens: + // - slash: position of the last read slash (or end of string) + // - start: position after the previous slash + for ( + // search for the first slash after the first character + std::size_t slash = reference_string.find_first_of('/', 1), + // set the beginning of the first reference token + start = 1; + // we can stop if start == string::npos+1 = 0 + start != 0; + // set the beginning of the next reference token + // (will eventually be 0 if slash == std::string::npos) + start = slash + 1, + // find next slash + slash = reference_string.find_first_of('/', start)) + { + // use the text between the beginning of the reference token + // (start) and the last slash (slash). + auto reference_token = reference_string.substr(start, slash - start); + + // check reference tokens are properly escaped + for (std::size_t pos = reference_token.find_first_of('~'); + pos != std::string::npos; + pos = reference_token.find_first_of('~', pos + 1)) + { + assert(reference_token[pos] == '~'); + + // ~ must be followed by 0 or 1 + if (JSON_UNLIKELY(pos == reference_token.size() - 1 or + (reference_token[pos + 1] != '0' and + reference_token[pos + 1] != '1'))) + { + JSON_THROW(detail::parse_error::create(108, 0, "escape character '~' must be followed with '0' or '1'")); + } + } + + // finally, store the reference token + unescape(reference_token); + result.push_back(reference_token); + } + + return result; + } + + /*! + @brief replace all occurrences of a substring by another string + + @param[in,out] s the string to manipulate; changed so that all + occurrences of @a f are replaced with @a t + @param[in] f the substring to replace with @a t + @param[in] t the string to replace @a f + + @pre The search string @a f must not be empty. **This precondition is + enforced with an assertion.** + + @since version 2.0.0 + */ + static void replace_substring(std::string& s, const std::string& f, + const std::string& t) + { + assert(not f.empty()); + for (auto pos = s.find(f); // find first occurrence of f + pos != std::string::npos; // make sure f was found + s.replace(pos, f.size(), t), // replace with t, and + pos = s.find(f, pos + t.size())) // find next occurrence of f + {} + } + + /// escape "~"" to "~0" and "/" to "~1" + static std::string escape(std::string s) + { + replace_substring(s, "~", "~0"); + replace_substring(s, "/", "~1"); + return s; + } + + /// unescape "~1" to tilde and "~0" to slash (order is important!) + static void unescape(std::string& s) + { + replace_substring(s, "~1", "/"); + replace_substring(s, "~0", "~"); + } + + /*! + @param[in] reference_string the reference string to the current value + @param[in] value the value to consider + @param[in,out] result the result object to insert values to + + @note Empty objects or arrays are flattened to `null`. + */ + static void flatten(const std::string& reference_string, + const BasicJsonType& value, + BasicJsonType& result) + { + switch (value.m_type) + { + case detail::value_t::array: + { + if (value.m_value.array->empty()) + { + // flatten empty array as null + result[reference_string] = nullptr; + } + else + { + // iterate array and use index as reference string + for (std::size_t i = 0; i < value.m_value.array->size(); ++i) + { + flatten(reference_string + "/" + std::to_string(i), + value.m_value.array->operator[](i), result); + } + } + break; + } + + case detail::value_t::object: + { + if (value.m_value.object->empty()) + { + // flatten empty object as null + result[reference_string] = nullptr; + } + else + { + // iterate object and use keys as reference string + for (const auto& element : *value.m_value.object) + { + flatten(reference_string + "/" + escape(element.first), element.second, result); + } + } + break; + } + + default: + { + // add primitive value with its reference string + result[reference_string] = value; + break; + } + } + } + + /*! + @param[in] value flattened JSON + + @return unflattened JSON + + @throw parse_error.109 if array index is not a number + @throw type_error.314 if value is not an object + @throw type_error.315 if object values are not primitive + @throw type_error.313 if value cannot be unflattened + */ + static BasicJsonType + unflatten(const BasicJsonType& value) + { + if (JSON_UNLIKELY(not value.is_object())) + { + JSON_THROW(detail::type_error::create(314, "only objects can be unflattened")); + } + + BasicJsonType result; + + // iterate the JSON object values + for (const auto& element : *value.m_value.object) + { + if (JSON_UNLIKELY(not element.second.is_primitive())) + { + JSON_THROW(detail::type_error::create(315, "values in object must be primitive")); + } + + // assign value to reference pointed to by JSON pointer; Note that if + // the JSON pointer is "" (i.e., points to the whole value), function + // get_and_create returns a reference to result itself. An assignment + // will then create a primitive value. + json_pointer(element.first).get_and_create(result) = element.second; + } + + return result; + } + + friend bool operator==(json_pointer const& lhs, + json_pointer const& rhs) noexcept + { + return (lhs.reference_tokens == rhs.reference_tokens); + } + + friend bool operator!=(json_pointer const& lhs, + json_pointer const& rhs) noexcept + { + return not (lhs == rhs); + } + + /// the reference tokens + std::vector reference_tokens; +}; +} + +// #include + + +#include + +// #include + +// #include + + +namespace nlohmann +{ +template struct adl_serializer { /*! @@ -1201,8 +10810,17 @@ struct adl_serializer ::nlohmann::to_json(j, std::forward(val)); } }; +} +/*! +@brief namespace for Niels Lohmann +@see https://github.com/nlohmann +@since version 1.0.0 +*/ +namespace nlohmann +{ + /*! @brief a class to store JSON values @@ -1284,35 +10902,59 @@ Format](http://rfc7159.net/rfc7159) @nosubgrouping */ -template < - template class ObjectType = std::map, - template class ArrayType = std::vector, - class StringType = std::string, - class BooleanType = bool, - class NumberIntegerType = std::int64_t, - class NumberUnsignedType = std::uint64_t, - class NumberFloatType = double, - template class AllocatorType = std::allocator, - template class JSONSerializer = adl_serializer - > +NLOHMANN_BASIC_JSON_TPL_DECLARATION class basic_json { private: template friend struct detail::external_constructor; + friend ::nlohmann::json_pointer; + friend ::nlohmann::detail::parser; + friend ::nlohmann::detail::serializer; + template + friend class ::nlohmann::detail::iter_impl; + template + friend class ::nlohmann::detail::binary_writer; + template + friend class ::nlohmann::detail::binary_reader; + template + friend class ::nlohmann::detail::json_sax_dom_parser; + template + friend class ::nlohmann::detail::json_sax_dom_callback_parser; + /// workaround type for MSVC - using basic_json_t = basic_json; + using basic_json_t = NLOHMANN_BASIC_JSON_TPL; + + // convenience aliases for types residing in namespace detail; + using lexer = ::nlohmann::detail::lexer; + using parser = ::nlohmann::detail::parser; + + using primitive_iterator_t = ::nlohmann::detail::primitive_iterator_t; + template + using internal_iterator = ::nlohmann::detail::internal_iterator; + template + using iter_impl = ::nlohmann::detail::iter_impl; + template + using iteration_proxy = ::nlohmann::detail::iteration_proxy; + template using json_reverse_iterator = ::nlohmann::detail::json_reverse_iterator; + + template + using output_adapter_t = ::nlohmann::detail::output_adapter_t; + + using binary_reader = ::nlohmann::detail::binary_reader; + template using binary_writer = ::nlohmann::detail::binary_writer; + + using serializer = ::nlohmann::detail::serializer; public: using value_t = detail::value_t; - // forward declarations - template class iter_impl; - template class json_reverse_iterator; - class json_pointer; + /// @copydoc nlohmann::json_pointer + using json_pointer = ::nlohmann::json_pointer; template using json_serializer = JSONSerializer; + /// helper type for initializer lists of basic_json values + using initializer_list_t = std::initializer_list>; + using input_format_t = detail::input_format_t; //////////////// // exceptions // @@ -1407,6 +11049,9 @@ class basic_json @liveexample{The following code shows an example output of the `meta()` function.,meta} + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes to any JSON value. + @complexity Constant. @since 2.1.0 @@ -1418,10 +11063,13 @@ class basic_json result["copyright"] = "(C) 2013-2017 Niels Lohmann"; result["name"] = "JSON for Modern C++"; result["url"] = "https://github.com/nlohmann/json"; - result["version"] = - { - {"string", "2.1.1"}, {"major", 2}, {"minor", 1}, {"patch", 1} - }; + result["version"]["string"] = + std::to_string(NLOHMANN_JSON_VERSION_MAJOR) + "." + + std::to_string(NLOHMANN_JSON_VERSION_MINOR) + "." + + std::to_string(NLOHMANN_JSON_VERSION_PATCH); + result["version"]["major"] = NLOHMANN_JSON_VERSION_MAJOR; + result["version"]["minor"] = NLOHMANN_JSON_VERSION_MINOR; + result["version"]["patch"] = NLOHMANN_JSON_VERSION_PATCH; #ifdef _WIN32 result["platform"] = "win32"; @@ -1435,10 +11083,10 @@ class basic_json result["platform"] = "unknown"; #endif -#if defined(__clang__) - result["compiler"] = {{"family", "clang"}, {"version", __clang_version__}}; -#elif defined(__ICC) || defined(__INTEL_COMPILER) +#if defined(__ICC) || defined(__INTEL_COMPILER) result["compiler"] = {{"family", "icc"}, {"version", __INTEL_COMPILER}}; +#elif defined(__clang__) + result["compiler"] = {{"family", "clang"}, {"version", __clang_version__}}; #elif defined(__GNUC__) || defined(__GNUG__) result["compiler"] = {{"family", "gcc"}, {"version", std::to_string(__GNUC__) + "." + std::to_string(__GNUC_MINOR__) + "." + std::to_string(__GNUC_PATCHLEVEL__)}}; #elif defined(__HP_cc) || defined(__HP_aCC) @@ -1473,6 +11121,14 @@ class basic_json /// the template arguments passed to class @ref basic_json. /// @{ +#if defined(JSON_HAS_CPP_14) + // Use transparent comparator if possible, combined with perfect forwarding + // on find() and count() calls prevents unnecessary string construction. + using object_comparator_t = std::less<>; +#else + using object_comparator_t = std::less; +#endif + /*! @brief a type for an object @@ -1515,10 +11171,10 @@ class basic_json - When all names are unique, objects will be interoperable in the sense that all software implementations receiving that object will agree on the name-value mappings. - - When the names within an object are not unique, later stored name/value - pairs overwrite previously stored name/value pairs, leaving the used - names unique. For instance, `{"key": 1}` and `{"key": 2, "key": 1}` will - be treated as equal and both stored as `{"key": 1}`. + - When the names within an object are not unique, it is unspecified which + one of the values for a given key will be chosen. For instance, + `{"key": 2, "key": 1}` could be equal to either `{"key": 1}` or + `{"key": 2}`. - Internally, name/value pairs are stored in lexicographical order of the names. Objects will also be serialized (see @ref dump) in this order. For instance, `{"b": 1, "a": 2}` and `{"a": 2, "b": 1}` will be stored @@ -1533,7 +11189,7 @@ class basic_json [RFC 7159](http://rfc7159.net/rfc7159) specifies: > An implementation may set limits on the maximum depth of nesting. - In this class, the object's limit of nesting is not constraint explicitly. + In this class, the object's limit of nesting is not explicitly constrained. However, a maximum depth of nesting may be introduced by the compiler or runtime environment. A theoretical limit can be queried by calling the @ref max_size function of a JSON object. @@ -1558,7 +11214,7 @@ class basic_json */ using object_t = ObjectType, + object_comparator_t, AllocatorType>>; @@ -1592,7 +11248,7 @@ class basic_json [RFC 7159](http://rfc7159.net/rfc7159) specifies: > An implementation may set limits on the maximum depth of nesting. - In this class, the array's limit of nesting is not constraint explicitly. + In this class, the array's limit of nesting is not explicitly constrained. However, a maximum depth of nesting may be introduced by the compiler or runtime environment. A theoretical limit can be queried by calling the @ref max_size function of a JSON array. @@ -1907,12 +11563,14 @@ class basic_json static T* create(Args&& ... args) { AllocatorType alloc; + using AllocatorTraits = std::allocator_traits>; + auto deleter = [&](T * object) { - alloc.deallocate(object, 1); + AllocatorTraits::deallocate(alloc, object, 1); }; - std::unique_ptr object(alloc.allocate(1), deleter); - alloc.construct(object.get(), std::forward(args)...); + std::unique_ptr object(AllocatorTraits::allocate(alloc, 1), deleter); + AllocatorTraits::construct(alloc, object.get(), std::forward(args)...); assert(object != nullptr); return object.release(); } @@ -2021,14 +11679,16 @@ class basic_json case value_t::null: { + object = nullptr; // silence warning, see #821 break; } default: { + object = nullptr; // silence warning, see #821 if (JSON_UNLIKELY(t == value_t::null)) { - JSON_THROW(other_error::create(500, "961c151d2e87f2686a955a9be24d316f1362bf21 2.1.1")); // LCOV_EXCL_LINE + JSON_THROW(other_error::create(500, "961c151d2e87f2686a955a9be24d316f1362bf21 3.1.2")); // LCOV_EXCL_LINE } break; } @@ -2041,17 +11701,70 @@ class basic_json string = create(value); } + /// constructor for rvalue strings + json_value(string_t&& value) + { + string = create(std::move(value)); + } + /// constructor for objects json_value(const object_t& value) { object = create(value); } + /// constructor for rvalue objects + json_value(object_t&& value) + { + object = create(std::move(value)); + } + /// constructor for arrays json_value(const array_t& value) { array = create(value); } + + /// constructor for rvalue arrays + json_value(array_t&& value) + { + array = create(std::move(value)); + } + + void destroy(value_t t) noexcept + { + switch (t) + { + case value_t::object: + { + AllocatorType alloc; + std::allocator_traits::destroy(alloc, object); + std::allocator_traits::deallocate(alloc, object, 1); + break; + } + + case value_t::array: + { + AllocatorType alloc; + std::allocator_traits::destroy(alloc, array); + std::allocator_traits::deallocate(alloc, array, 1); + break; + } + + case value_t::string: + { + AllocatorType alloc; + std::allocator_traits::destroy(alloc, string); + std::allocator_traits::deallocate(alloc, string, 1); + break; + } + + default: + { + break; + } + } + } }; /*! @@ -2063,7 +11776,7 @@ class basic_json value is changed, because the invariant expresses a relationship between @a m_type and @a m_value. */ - void assert_invariant() const + void assert_invariant() const noexcept { assert(m_type != value_t::object or m_value.object != nullptr); assert(m_type != value_t::array or m_value.array != nullptr); @@ -2076,42 +11789,31 @@ class basic_json ////////////////////////// /*! - @brief JSON callback events + @brief parser event types - This enumeration lists the parser events that can trigger calling a - callback function of type @ref parser_callback_t during parsing. + The parser callback distinguishes the following events: + - `object_start`: the parser read `{` and started to process a JSON object + - `key`: the parser read a key of a value in an object + - `object_end`: the parser read `}` and finished processing a JSON object + - `array_start`: the parser read `[` and started to process a JSON array + - `array_end`: the parser read `]` and finished processing a JSON array + - `value`: the parser finished reading a JSON value @image html callback_events.png "Example when certain parse events are triggered" - @since version 1.0.0 + @sa @ref parser_callback_t for more information and examples */ - enum class parse_event_t : uint8_t - { - /// the parser read `{` and started to process a JSON object - object_start, - /// the parser read `}` and finished processing a JSON object - object_end, - /// the parser read `[` and started to process a JSON array - array_start, - /// the parser read `]` and finished processing a JSON array - array_end, - /// the parser read a key of a value in an object - key, - /// the parser finished reading a JSON value - value - }; + using parse_event_t = typename parser::parse_event_t; /*! @brief per-element parser callback type With a parser callback function, the result of parsing a JSON text can be - influenced. When passed to @ref parse(std::istream&, const - parser_callback_t) or @ref parse(const CharT, const parser_callback_t), - it is called on certain events (passed as @ref parse_event_t via parameter - @a event) with a set recursion depth @a depth and context JSON value - @a parsed. The return value of the callback function is a boolean - indicating whether the element that emitted the callback shall be kept or - not. + influenced. When passed to @ref parse, it is called on certain events + (passed as @ref parse_event_t via parameter @a event) with a set recursion + depth @a depth and context JSON value @a parsed. The return value of the + callback function is a boolean indicating whether the element that emitted + the callback shall be kept or not. We distinguish six scenarios (determined by the event type) in which the callback function can be called. The following table describes the values @@ -2148,15 +11850,13 @@ class basic_json should be kept (`true`) or not (`false`). In the latter case, it is either skipped completely or replaced by an empty discarded object. - @sa @ref parse(std::istream&, parser_callback_t) or - @ref parse(const CharT, const parser_callback_t) for examples + @sa @ref parse for examples @since version 1.0.0 */ - using parser_callback_t = std::function; + using parser_callback_t = typename parser::parser_callback_t; + using json_sax_t = typename parser::json_sax_t; ////////////////// // constructors // @@ -2182,17 +11882,22 @@ class basic_json object | `{}` array | `[]` - @param[in] value_type the type of the value to create + @param[in] v the type of the value to create @complexity Constant. + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes to any JSON value. + @liveexample{The following code shows the constructor for different @ref value_t values,basic_json__value_t} + @sa @ref clear() -- restores the postcondition of this constructor + @since version 1.0.0 */ - basic_json(const value_t value_type) - : m_type(value_type), m_value(value_type) + basic_json(const value_t v) + : m_type(v), m_value(v) { assert_invariant(); } @@ -2225,7 +11930,7 @@ class basic_json @brief create a JSON value This is a "catch all" constructor for all compatible JSON types; that is, - types for which a `to_json()` method exsits. The constructor forwards the + types for which a `to_json()` method exists. The constructor forwards the parameter @a val to that method (to `json_serializer::to_json` method with `U = uncvref_t`, to be exact). @@ -2233,9 +11938,9 @@ class basic_json following types: - **arrays**: @ref array_t and all kinds of compatible containers such as `std::vector`, `std::deque`, `std::list`, `std::forward_list`, - `std::array`, `std::set`, `std::unordered_set`, `std::multiset`, and - `unordered_multiset` with a `value_type` from which a @ref basic_json - value can be constructed. + `std::array`, `std::valarray`, `std::set`, `std::unordered_set`, + `std::multiset`, and `std::unordered_multiset` with a `value_type` from + which a @ref basic_json value can be constructed. - **objects**: @ref object_t and all kinds of compatible associative containers such as `std::map`, `std::unordered_map`, `std::multimap`, and `std::unordered_multimap` with a `key_type` compatible to @@ -2254,6 +11959,7 @@ class basic_json - @a CompatibleType is not derived from `std::istream`, - @a CompatibleType is not @ref basic_json (to avoid hijacking copy/move constructors), + - @a CompatibleType is not a different @ref basic_json type (i.e. with different template arguments) - @a CompatibleType is not a @ref basic_json nested type (e.g., @ref json_pointer, @ref iterator, etc ...) - @ref @ref json_serializer has a @@ -2261,33 +11967,106 @@ class basic_json @tparam U = `uncvref_t` - @param[in] val the value to be forwarded + @param[in] val the value to be forwarded to the respective constructor @complexity Usually linear in the size of the passed @a val, also depending on the implementation of the called `to_json()` method. - @throw what `json_serializer::to_json()` throws + @exceptionsafety Depends on the called constructor. For types directly + supported by the library (i.e., all types for which no `to_json()` function + was provided), strong guarantee holds: if an exception is thrown, there are + no changes to any JSON value. @liveexample{The following code shows the constructor with several compatible types.,basic_json__CompatibleType} @since version 2.1.0 */ - template, - detail::enable_if_t::value and - not std::is_same::value and - not detail::is_basic_json_nested_type< - basic_json_t, U>::value and - detail::has_to_json::value, - int> = 0> - basic_json(CompatibleType && val) noexcept(noexcept(JSONSerializer::to_json( - std::declval(), std::forward(val)))) + template , + detail::enable_if_t< + detail::is_compatible_type::value, int> = 0> + basic_json(CompatibleType && val) noexcept(noexcept( + JSONSerializer::to_json(std::declval(), + std::forward(val)))) { JSONSerializer::to_json(*this, std::forward(val)); assert_invariant(); } + /*! + @brief create a JSON value from an existing one + + This is a constructor for existing @ref basic_json types. + It does not hijack copy/move constructors, since the parameter has different + template arguments than the current ones. + + The constructor tries to convert the internal @ref m_value of the parameter. + + @tparam BasicJsonType a type such that: + - @a BasicJsonType is a @ref basic_json type. + - @a BasicJsonType has different template arguments than @ref basic_json_t. + + @param[in] val the @ref basic_json value to be converted. + + @complexity Usually linear in the size of the passed @a val, also + depending on the implementation of the called `to_json()` + method. + + @exceptionsafety Depends on the called constructor. For types directly + supported by the library (i.e., all types for which no `to_json()` function + was provided), strong guarantee holds: if an exception is thrown, there are + no changes to any JSON value. + + @since version 3.1.2 + */ + template ::value and not std::is_same::value, int> = 0> + basic_json(const BasicJsonType& val) + { + using other_boolean_t = typename BasicJsonType::boolean_t; + using other_number_float_t = typename BasicJsonType::number_float_t; + using other_number_integer_t = typename BasicJsonType::number_integer_t; + using other_number_unsigned_t = typename BasicJsonType::number_unsigned_t; + using other_string_t = typename BasicJsonType::string_t; + using other_object_t = typename BasicJsonType::object_t; + using other_array_t = typename BasicJsonType::array_t; + + switch (val.type()) + { + case value_t::boolean: + JSONSerializer::to_json(*this, val.template get()); + break; + case value_t::number_float: + JSONSerializer::to_json(*this, val.template get()); + break; + case value_t::number_integer: + JSONSerializer::to_json(*this, val.template get()); + break; + case value_t::number_unsigned: + JSONSerializer::to_json(*this, val.template get()); + break; + case value_t::string: + JSONSerializer::to_json(*this, val.template get_ref()); + break; + case value_t::object: + JSONSerializer::to_json(*this, val.template get_ref()); + break; + case value_t::array: + JSONSerializer::to_json(*this, val.template get_ref()); + break; + case value_t::null: + *this = nullptr; + break; + case value_t::discarded: + m_type = value_t::discarded; + break; + } + assert_invariant(); + } + /*! @brief create a container (array or object) from an initializer list @@ -2307,7 +12086,7 @@ class basic_json 1. The empty initializer list is written as `{}` which is exactly an empty JSON object. - 2. C++ has now way of describing mapped types other than to list a list of + 2. C++ has no way of describing mapped types other than to list a list of pairs. As JSON requires that keys must be of type string, rule 2 is the weakest constraint one can pose on initializer lists to interpret them as an object. @@ -2317,10 +12096,10 @@ class basic_json With the rules described above, the following JSON values cannot be expressed by an initializer list: - - the empty array (`[]`): use @ref array(std::initializer_list) + - the empty array (`[]`): use @ref array(initializer_list_t) with an empty initializer list in this case - arrays whose elements satisfy rule 2: use @ref - array(std::initializer_list) with the same initializer list + array(initializer_list_t) with the same initializer list in this case @note When used without parentheses around an empty initializer list, @ref @@ -2332,8 +12111,8 @@ class basic_json @param[in] type_deduction internal parameter; when set to `true`, the type of the JSON value is deducted from the initializer list @a init; when set to `false`, the type provided via @a manual_type is forced. This mode is - used by the functions @ref array(std::initializer_list) and - @ref object(std::initializer_list). + used by the functions @ref array(initializer_list_t) and + @ref object(initializer_list_t). @param[in] manual_type internal parameter; when @a type_deduction is set to `false`, the created JSON value will use the provided type (only @ref @@ -2344,31 +12123,34 @@ class basic_json `value_t::object`, but @a init contains an element which is not a pair whose first element is a string. In this case, the constructor could not create an object. If @a type_deduction would have be `true`, an array - would have been created. See @ref object(std::initializer_list) + would have been created. See @ref object(initializer_list_t) for an example. @complexity Linear in the size of the initializer list @a init. + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes to any JSON value. + @liveexample{The example below shows how JSON values are created from initializer lists.,basic_json__list_init_t} - @sa @ref array(std::initializer_list) -- create a JSON array + @sa @ref array(initializer_list_t) -- create a JSON array value from an initializer list - @sa @ref object(std::initializer_list) -- create a JSON object + @sa @ref object(initializer_list_t) -- create a JSON object value from an initializer list @since version 1.0.0 */ - basic_json(std::initializer_list init, + basic_json(initializer_list_t init, bool type_deduction = true, value_t manual_type = value_t::array) { // check if each element is an array with two elements whose first // element is a string bool is_an_object = std::all_of(init.begin(), init.end(), - [](const basic_json & element) + [](const detail::json_ref& element_ref) { - return element.is_array() and element.size() == 2 and element[0].is_string(); + return (element_ref->is_array() and element_ref->size() == 2 and (*element_ref)[0].is_string()); }); // adjust type if type deduction is not wanted @@ -2381,7 +12163,7 @@ class basic_json } // if object is wanted but impossible, throw an exception - if (manual_type == value_t::object and not is_an_object) + if (JSON_UNLIKELY(manual_type == value_t::object and not is_an_object)) { JSON_THROW(type_error::create(301, "cannot create object from initializer list")); } @@ -2393,16 +12175,19 @@ class basic_json m_type = value_t::object; m_value = value_t::object; - std::for_each(init.begin(), init.end(), [this](const basic_json & element) + std::for_each(init.begin(), init.end(), [this](const detail::json_ref& element_ref) { - m_value.object->emplace(*(element[0].m_value.string), element[1]); + auto element = element_ref.moved_or_copied(); + m_value.object->emplace( + std::move(*((*element.m_value.array)[0].m_value.string)), + std::move((*element.m_value.array)[1])); }); } else { // the initializer list describes an array -> create array m_type = value_t::array; - m_value.array = create(init); + m_value.array = create(init.begin(), init.end()); } assert_invariant(); @@ -2417,7 +12202,7 @@ class basic_json @note This function is only needed to express two edge cases that cannot be realized with the initializer list constructor (@ref - basic_json(std::initializer_list, bool, value_t)). These cases + basic_json(initializer_list_t, bool, value_t)). These cases are: 1. creating an array whose elements are all pairs whose first element is a string -- in this case, the initializer list constructor would create an @@ -2432,18 +12217,20 @@ class basic_json @complexity Linear in the size of @a init. + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes to any JSON value. + @liveexample{The following code shows an example for the `array` function.,array} - @sa @ref basic_json(std::initializer_list, bool, value_t) -- + @sa @ref basic_json(initializer_list_t, bool, value_t) -- create a JSON value from an initializer list - @sa @ref object(std::initializer_list) -- create a JSON object + @sa @ref object(initializer_list_t) -- create a JSON object value from an initializer list @since version 1.0.0 */ - static basic_json array(std::initializer_list init = - std::initializer_list()) + static basic_json array(initializer_list_t init = {}) { return basic_json(init, false, value_t::array); } @@ -2456,10 +12243,10 @@ class basic_json the initializer list is empty, the empty object `{}` is created. @note This function is only added for symmetry reasons. In contrast to the - related function @ref array(std::initializer_list), there are + related function @ref array(initializer_list_t), there are no cases which can only be expressed by this function. That is, any initializer list @a init can also be passed to the initializer list - constructor @ref basic_json(std::initializer_list, bool, value_t). + constructor @ref basic_json(initializer_list_t, bool, value_t). @param[in] init initializer list to create an object from (optional) @@ -2467,24 +12254,26 @@ class basic_json @throw type_error.301 if @a init is not a list of pairs whose first elements are strings. In this case, no object can be created. When such a - value is passed to @ref basic_json(std::initializer_list, bool, value_t), + value is passed to @ref basic_json(initializer_list_t, bool, value_t), an array would have been created from the passed initializer list @a init. See example below. @complexity Linear in the size of @a init. + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes to any JSON value. + @liveexample{The following code shows an example for the `object` function.,object} - @sa @ref basic_json(std::initializer_list, bool, value_t) -- + @sa @ref basic_json(initializer_list_t, bool, value_t) -- create a JSON value from an initializer list - @sa @ref array(std::initializer_list) -- create a JSON array + @sa @ref array(initializer_list_t) -- create a JSON array value from an initializer list @since version 1.0.0 */ - static basic_json object(std::initializer_list init = - std::initializer_list()) + static basic_json object(initializer_list_t init = {}) { return basic_json(init, false, value_t::object); } @@ -2493,14 +12282,18 @@ class basic_json @brief construct an array with count copies of given value Constructs a JSON array value by creating @a cnt copies of a passed value. - In case @a cnt is `0`, an empty array is created. As postcondition, - `std::distance(begin(),end()) == cnt` holds. + In case @a cnt is `0`, an empty array is created. @param[in] cnt the number of JSON copies of @a val to create @param[in] val the JSON value to copy + @post `std::distance(begin(),end()) == cnt` holds. + @complexity Linear in @a cnt. + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes to any JSON value. + @liveexample{The following code shows examples for the @ref basic_json(size_type\, const basic_json&) constructor.,basic_json__size_type_basic_json} @@ -2519,12 +12312,13 @@ class basic_json Constructs the JSON value with the contents of the range `[first, last)`. The semantics depends on the different types a JSON value can have: - - In case of primitive types (number, boolean, or string), @a first must - be `begin()` and @a last must be `end()`. In this case, the value is + - In case of a null type, invalid_iterator.206 is thrown. + - In case of other primitive types (number, boolean, or string), @a first + must be `begin()` and @a last must be `end()`. In this case, the value is copied. Otherwise, invalid_iterator.204 is thrown. - In case of structured types (array, object), the constructor behaves as - similar versions for `std::vector`. - - In case of a null type, invalid_iterator.206 is thrown. + similar versions for `std::vector` or `std::map`; that is, a JSON array + or object is constructed from the values in the range. @tparam InputIT an input iterator type (@ref iterator or @ref const_iterator) @@ -2533,11 +12327,20 @@ class basic_json @param[in] last end of the range to copy from (excluded) @pre Iterators @a first and @a last must be initialized. **This - precondition is enforced with an assertion.** + precondition is enforced with an assertion (see warning).** If + assertions are switched off, a violation of this precondition yields + undefined behavior. @pre Range `[first, last)` is valid. Usually, this precondition cannot be checked efficiently. Only certain edge cases are detected; see the - description of the exceptions below. + description of the exceptions below. A violation of this precondition + yields undefined behavior. + + @warning A precondition is enforced with a runtime assertion that will + result in calling `std::abort` if this precondition is not met. + Assertions can be disabled by defining `NDEBUG` at compile time. + See http://en.cppreference.com/w/cpp/error/assert for more + information. @throw invalid_iterator.201 if iterators @a first and @a last are not compatible (i.e., do not belong to the same JSON value). In this case, @@ -2551,6 +12354,9 @@ class basic_json @complexity Linear in distance between @a first and @a last. + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes to any JSON value. + @liveexample{The example below shows several ways to create JSON values by specifying a subrange with iterators.,basic_json__InputIt_InputIt} @@ -2565,7 +12371,7 @@ class basic_json assert(last.m_object != nullptr); // make sure iterator fits the current value - if (first.m_object != last.m_object) + if (JSON_UNLIKELY(first.m_object != last.m_object)) { JSON_THROW(invalid_iterator::create(201, "iterators are not compatible")); } @@ -2582,7 +12388,8 @@ class basic_json case value_t::number_unsigned: case value_t::string: { - if (not first.m_it.primitive_iterator.is_begin() or not last.m_it.primitive_iterator.is_end()) + if (JSON_UNLIKELY(not first.m_it.primitive_iterator.is_begin() + or not last.m_it.primitive_iterator.is_end())) { JSON_THROW(invalid_iterator::create(204, "iterators out of range")); } @@ -2590,9 +12397,7 @@ class basic_json } default: - { break; - } } switch (m_type) @@ -2642,10 +12447,8 @@ class basic_json } default: - { JSON_THROW(invalid_iterator::create(206, "cannot construct with iterators from " + - first.m_object->type_name())); - } + std::string(first.m_object->type_name()))); } assert_invariant(); @@ -2656,6 +12459,11 @@ class basic_json // other constructors and destructor // /////////////////////////////////////// + /// @private + basic_json(const detail::json_ref& ref) + : basic_json(ref.moved_or_copied()) + {} + /*! @brief copy constructor @@ -2663,8 +12471,13 @@ class basic_json @param[in] other the JSON value to copy + @post `*this == other` + @complexity Linear in the size of @a other. + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes to any JSON value. + @requirement This function helps `basic_json` satisfying the [Container](http://en.cppreference.com/w/cpp/concept/Container) requirements: @@ -2727,9 +12540,7 @@ class basic_json } default: - { break; - } } assert_invariant(); @@ -2744,10 +12555,18 @@ class basic_json @param[in,out] other value to move to this object - @post @a other is a JSON null value + @post `*this` has the same value as @a other before the call. + @post @a other is a JSON null value. @complexity Constant. + @exceptionsafety No-throw guarantee: this constructor never throws + exceptions. + + @requirement This function helps `basic_json` satisfying the + [MoveConstructible](http://en.cppreference.com/w/cpp/concept/MoveConstructible) + requirements. + @liveexample{The code below shows the move constructor explicitly called via std::move.,basic_json__moveconstructor} @@ -2772,7 +12591,7 @@ class basic_json Copy assignment operator. Copies a JSON value via the "copy and swap" strategy: It is expressed in terms of the copy constructor, destructor, - and the swap() member function. + and the `swap()` member function. @param[in] other value to copy from @@ -2823,42 +12642,10 @@ class basic_json @since version 1.0.0 */ - ~basic_json() + ~basic_json() noexcept { assert_invariant(); - - switch (m_type) - { - case value_t::object: - { - AllocatorType alloc; - alloc.destroy(m_value.object); - alloc.deallocate(m_value.object, 1); - break; - } - - case value_t::array: - { - AllocatorType alloc; - alloc.destroy(m_value.array); - alloc.deallocate(m_value.array, 1); - break; - } - - case value_t::string: - { - AllocatorType alloc; - alloc.destroy(m_value.string); - alloc.deallocate(m_value.string, 1); - break; - } - - default: - { - // all other types need no specific destructor - break; - } - } + m_value.destroy(m_type); } /// @} @@ -2877,38 +12664,50 @@ class basic_json Serialization function for JSON values. The function tries to mimic Python's `json.dumps()` function, and currently supports its @a indent - parameter. + and @a ensure_ascii parameters. @param[in] indent If indent is nonnegative, then array elements and object members will be pretty-printed with that indent level. An indent level of `0` will only insert newlines. `-1` (the default) selects the most compact representation. - @param[in] indent_char The character to use for indentation of @a indent is - greate than `0`. The default is ` ` (space). + @param[in] indent_char The character to use for indentation if @a indent is + greater than `0`. The default is ` ` (space). + @param[in] ensure_ascii If @a ensure_ascii is true, all non-ASCII characters + in the output are escaped with `\uXXXX` sequences, and the result consists + of ASCII characters only. @return string containing the serialization of the JSON value + @throw type_error.316 if a string stored inside the JSON value is not + UTF-8 encoded + @complexity Linear. - @liveexample{The following example shows the effect of different @a indent - parameters to the result of the serialization.,dump} + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes in the JSON value. + + @liveexample{The following example shows the effect of different @a indent\, + @a indent_char\, and @a ensure_ascii parameters to the result of the + serialization.,dump} @see https://docs.python.org/2/library/json.html#json.dump - @since version 1.0.0; indentaction character added in version 3.0.0 + @since version 1.0.0; indentation character @a indent_char, option + @a ensure_ascii and exceptions added in version 3.0.0 */ - string_t dump(const int indent = -1, const char indent_char = ' ') const + string_t dump(const int indent = -1, const char indent_char = ' ', + const bool ensure_ascii = false) const { string_t result; - serializer s(output_adapter::create(result), indent_char); + serializer s(detail::output_adapter(result), indent_char); if (indent >= 0) { - s.dump(*this, true, static_cast(indent)); + s.dump(*this, true, ensure_ascii, static_cast(indent)); } else { - s.dump(*this, false, 0); + s.dump(*this, false, ensure_ascii, 0); } return result; @@ -2921,6 +12720,17 @@ class basic_json enumeration. @return the type of the JSON value + Value type | return value + ------------------------- | ------------------------- + null | value_t::null + boolean | value_t::boolean + string | value_t::string + number (integer) | value_t::number_integer + number (unsigned integer) | value_t::number_unsigned + number (floating-point) | value_t::number_float + object | value_t::object + array | value_t::array + discarded | value_t::discarded @complexity Constant. @@ -2930,6 +12740,9 @@ class basic_json @liveexample{The following code exemplifies `type()` for all JSON types.,type} + @sa @ref operator value_t() -- return the type of the JSON value (implicit) + @sa @ref type_name() -- return the type as string + @since version 1.0.0 */ constexpr value_t type() const noexcept @@ -2940,8 +12753,8 @@ class basic_json /*! @brief return whether type is primitive - This function returns true iff the JSON type is primitive (string, number, - boolean, or null). + This function returns true if and only if the JSON type is primitive + (string, number, boolean, or null). @return `true` if type is primitive (string, number, boolean, or null), `false` otherwise. @@ -2970,8 +12783,8 @@ class basic_json /*! @brief return whether type is structured - This function returns true iff the JSON type is structured (array or - object). + This function returns true if and only if the JSON type is structured + (array or object). @return `true` if type is structured (array or object), `false` otherwise. @@ -2997,7 +12810,7 @@ class basic_json /*! @brief return whether value is null - This function returns true iff the JSON value is null. + This function returns true if and only if the JSON value is null. @return `true` if type is null, `false` otherwise. @@ -3013,13 +12826,13 @@ class basic_json */ constexpr bool is_null() const noexcept { - return m_type == value_t::null; + return (m_type == value_t::null); } /*! @brief return whether value is a boolean - This function returns true iff the JSON value is a boolean. + This function returns true if and only if the JSON value is a boolean. @return `true` if type is boolean, `false` otherwise. @@ -3035,14 +12848,14 @@ class basic_json */ constexpr bool is_boolean() const noexcept { - return m_type == value_t::boolean; + return (m_type == value_t::boolean); } /*! @brief return whether value is a number - This function returns true iff the JSON value is a number. This includes - both integer and floating-point values. + This function returns true if and only if the JSON value is a number. This + includes both integer (signed and unsigned) and floating-point values. @return `true` if type is number (regardless whether integer, unsigned integer or floating-type), `false` otherwise. @@ -3071,8 +12884,8 @@ class basic_json /*! @brief return whether value is an integer number - This function returns true iff the JSON value is an integer or unsigned - integer number. This excludes floating-point values. + This function returns true if and only if the JSON value is a signed or + unsigned integer number. This excludes floating-point values. @return `true` if type is an integer or unsigned integer number, `false` otherwise. @@ -3094,14 +12907,14 @@ class basic_json */ constexpr bool is_number_integer() const noexcept { - return m_type == value_t::number_integer or m_type == value_t::number_unsigned; + return (m_type == value_t::number_integer or m_type == value_t::number_unsigned); } /*! @brief return whether value is an unsigned integer number - This function returns true iff the JSON value is an unsigned integer - number. This excludes floating-point and (signed) integer values. + This function returns true if and only if the JSON value is an unsigned + integer number. This excludes floating-point and signed integer values. @return `true` if type is an unsigned integer number, `false` otherwise. @@ -3122,14 +12935,14 @@ class basic_json */ constexpr bool is_number_unsigned() const noexcept { - return m_type == value_t::number_unsigned; + return (m_type == value_t::number_unsigned); } /*! @brief return whether value is a floating-point number - This function returns true iff the JSON value is a floating-point number. - This excludes integer and unsigned integer values. + This function returns true if and only if the JSON value is a + floating-point number. This excludes signed and unsigned integer values. @return `true` if type is a floating-point number, `false` otherwise. @@ -3150,13 +12963,13 @@ class basic_json */ constexpr bool is_number_float() const noexcept { - return m_type == value_t::number_float; + return (m_type == value_t::number_float); } /*! @brief return whether value is an object - This function returns true iff the JSON value is an object. + This function returns true if and only if the JSON value is an object. @return `true` if type is object, `false` otherwise. @@ -3172,13 +12985,13 @@ class basic_json */ constexpr bool is_object() const noexcept { - return m_type == value_t::object; + return (m_type == value_t::object); } /*! @brief return whether value is an array - This function returns true iff the JSON value is an array. + This function returns true if and only if the JSON value is an array. @return `true` if type is array, `false` otherwise. @@ -3194,13 +13007,13 @@ class basic_json */ constexpr bool is_array() const noexcept { - return m_type == value_t::array; + return (m_type == value_t::array); } /*! @brief return whether value is a string - This function returns true iff the JSON value is a string. + This function returns true if and only if the JSON value is a string. @return `true` if type is string, `false` otherwise. @@ -3216,14 +13029,14 @@ class basic_json */ constexpr bool is_string() const noexcept { - return m_type == value_t::string; + return (m_type == value_t::string); } /*! @brief return whether value is discarded - This function returns true iff the JSON value was discarded during parsing - with a callback function (see @ref parser_callback_t). + This function returns true if and only if the JSON value was discarded + during parsing with a callback function (see @ref parser_callback_t). @note This function will always be `false` for JSON values after parsing. That is, discarded values can only occur during parsing, but will be @@ -3243,7 +13056,7 @@ class basic_json */ constexpr bool is_discarded() const noexcept { - return m_type == value_t::discarded; + return (m_type == value_t::discarded); } /*! @@ -3262,6 +13075,9 @@ class basic_json @liveexample{The following code exemplifies the @ref value_t operator for all JSON types.,operator__value_t} + @sa @ref type() -- return the type of the JSON value (explicit) + @sa @ref type_name() -- return the type as string + @since version 1.0.0 */ constexpr operator value_t() const noexcept @@ -3279,12 +13095,12 @@ class basic_json /// get a boolean (explicit) boolean_t get_impl(boolean_t* /*unused*/) const { - if (is_boolean()) + if (JSON_LIKELY(is_boolean())) { return m_value.boolean; } - JSON_THROW(type_error::create(302, "type must be boolean, but is " + type_name())); + JSON_THROW(type_error::create(302, "type must be boolean, but is " + std::string(type_name()))); } /// get a pointer to the value (object) @@ -3374,7 +13190,7 @@ class basic_json /*! @brief helper function to implement get_ref() - This funcion helps to implement get_ref() without code duplication for + This function helps to implement get_ref() without code duplication for const and non-const overloads @tparam ThisType will be deduced as `basic_json` or `const basic_json` @@ -3385,18 +13201,15 @@ class basic_json template static ReferenceType get_ref_impl(ThisType& obj) { - // helper type - using PointerType = typename std::add_pointer::type; - // delegate the call to get_ptr<>() - auto ptr = obj.template get_ptr(); + auto ptr = obj.template get_ptr::type>(); - if (ptr != nullptr) + if (JSON_LIKELY(ptr != nullptr)) { return *ptr; } - JSON_THROW(type_error::create(303, "incompatible ReferenceType for get_ref, actual type is " + obj.type_name())); + JSON_THROW(type_error::create(303, "incompatible ReferenceType for get_ref, actual type is " + std::string(obj.type_name()))); } public: @@ -3418,16 +13231,37 @@ class basic_json @since version 2.1.0 */ - template < - typename BasicJsonType, - detail::enable_if_t::type, - basic_json_t>::value, - int> = 0 > + template::type, basic_json_t>::value, + int> = 0> basic_json get() const { return *this; } + /*! + @brief get special-case overload + + This overloads converts the current @ref basic_json in a different + @ref basic_json type + + @tparam BasicJsonType == @ref basic_json + + @return a copy of *this, converted into @tparam BasicJsonType + + @complexity Depending on the implementation of the called `from_json()` + method. + + @since version 3.1.2 + */ + template::value and + detail::is_basic_json::value, int> = 0> + BasicJsonType get() const + { + return *this; + } + /*! @brief get a value (explicit) @@ -3467,14 +13301,12 @@ class basic_json @since version 2.1.0 */ - template < - typename ValueTypeCV, - typename ValueType = detail::uncvref_t, - detail::enable_if_t < - not std::is_same::value and - detail::has_from_json::value and - not detail::has_non_default_from_json::value, - int > = 0 > + template, + detail::enable_if_t < + not detail::is_basic_json::value and + detail::has_from_json::value and + not detail::has_non_default_from_json::value, + int> = 0> ValueType get() const noexcept(noexcept( JSONSerializer::from_json(std::declval(), std::declval()))) { @@ -3522,12 +13354,10 @@ class basic_json @since version 2.1.0 */ - template < - typename ValueTypeCV, - typename ValueType = detail::uncvref_t, - detail::enable_if_t::value and - detail::has_non_default_from_json::value, int> = 0 > + template, + detail::enable_if_t::value and + detail::has_non_default_from_json::value, + int> = 0> ValueType get() const noexcept(noexcept( JSONSerializer::from_json(std::declval()))) { @@ -3657,7 +13487,7 @@ class basic_json , "incompatible pointer type"); // delegate the call to get_impl_ptr<>() const - return get_impl_ptr(static_cast(nullptr)); + return get_impl_ptr(static_cast(nullptr)); } /*! @@ -3738,11 +13568,13 @@ class basic_json */ template < typename ValueType, typename std::enable_if < not std::is_pointer::value and - not std::is_same::value + not std::is_same>::value and + not std::is_same::value and + not detail::is_basic_json::value #ifndef _MSC_VER // fix for issue #167 operator<< ambiguity under VS2015 and not std::is_same>::value #endif -#if (defined(__cplusplus) && __cplusplus >= 201703L) || (defined(_MSC_VER) && _MSC_VER >1900 && defined(_HAS_CXX17) && _HAS_CXX17 == 1) // fix for issue #464 +#if defined(JSON_HAS_CPP_17) and not std::is_same::value #endif , int >::type = 0 > @@ -3792,7 +13624,7 @@ class basic_json reference at(size_type idx) { // at only works for arrays - if (is_array()) + if (JSON_LIKELY(is_array())) { JSON_TRY { @@ -3806,7 +13638,7 @@ class basic_json } else { - JSON_THROW(type_error::create(304, "cannot use at() with " + type_name())); + JSON_THROW(type_error::create(304, "cannot use at() with " + std::string(type_name()))); } } @@ -3839,7 +13671,7 @@ class basic_json const_reference at(size_type idx) const { // at only works for arrays - if (is_array()) + if (JSON_LIKELY(is_array())) { JSON_TRY { @@ -3853,7 +13685,7 @@ class basic_json } else { - JSON_THROW(type_error::create(304, "cannot use at() with " + type_name())); + JSON_THROW(type_error::create(304, "cannot use at() with " + std::string(type_name()))); } } @@ -3890,7 +13722,7 @@ class basic_json reference at(const typename object_t::key_type& key) { // at only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { JSON_TRY { @@ -3904,7 +13736,7 @@ class basic_json } else { - JSON_THROW(type_error::create(304, "cannot use at() with " + type_name())); + JSON_THROW(type_error::create(304, "cannot use at() with " + std::string(type_name()))); } } @@ -3941,7 +13773,7 @@ class basic_json const_reference at(const typename object_t::key_type& key) const { // at only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { JSON_TRY { @@ -3955,7 +13787,7 @@ class basic_json } else { - JSON_THROW(type_error::create(304, "cannot use at() with " + type_name())); + JSON_THROW(type_error::create(304, "cannot use at() with " + std::string(type_name()))); } } @@ -3995,7 +13827,7 @@ class basic_json } // operator[] only works for arrays - if (is_array()) + if (JSON_LIKELY(is_array())) { // fill up array with null values if given idx is outside range if (idx >= m_value.array->size()) @@ -4008,7 +13840,7 @@ class basic_json return m_value.array->operator[](idx); } - JSON_THROW(type_error::create(305, "cannot use operator[] with " + type_name())); + JSON_THROW(type_error::create(305, "cannot use operator[] with " + std::string(type_name()))); } /*! @@ -4020,7 +13852,7 @@ class basic_json @return const reference to the element at index @a idx - @throw type_error.305 if the JSON value is not an array; in that cases, + @throw type_error.305 if the JSON value is not an array; in that case, using the [] operator with an index makes no sense. @complexity Constant. @@ -4033,12 +13865,12 @@ class basic_json const_reference operator[](size_type idx) const { // const operator[] only works for arrays - if (is_array()) + if (JSON_LIKELY(is_array())) { return m_value.array->operator[](idx); } - JSON_THROW(type_error::create(305, "cannot use operator[] with " + type_name())); + JSON_THROW(type_error::create(305, "cannot use operator[] with " + std::string(type_name()))); } /*! @@ -4079,12 +13911,12 @@ class basic_json } // operator[] only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { return m_value.object->operator[](key); } - JSON_THROW(type_error::create(305, "cannot use operator[] with " + type_name())); + JSON_THROW(type_error::create(305, "cannot use operator[] with " + std::string(type_name()))); } /*! @@ -4103,7 +13935,7 @@ class basic_json @pre The element with key @a key must exist. **This precondition is enforced with an assertion.** - @throw type_error.305 if the JSON value is not an object; in that cases, + @throw type_error.305 if the JSON value is not an object; in that case, using the [] operator with a key makes no sense. @complexity Logarithmic in the size of the container. @@ -4120,81 +13952,13 @@ class basic_json const_reference operator[](const typename object_t::key_type& key) const { // const operator[] only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { assert(m_value.object->find(key) != m_value.object->end()); return m_value.object->find(key)->second; } - JSON_THROW(type_error::create(305, "cannot use operator[] with " + type_name())); - } - - /*! - @brief access specified object element - - Returns a reference to the element at with specified key @a key. - - @note If @a key is not found in the object, then it is silently added to - the object and filled with a `null` value to make `key` a valid reference. - In case the value was `null` before, it is converted to an object. - - @param[in] key key of the element to access - - @return reference to the element at key @a key - - @throw type_error.305 if the JSON value is not an object or null; in that - cases, using the [] operator with a key makes no sense. - - @complexity Logarithmic in the size of the container. - - @liveexample{The example below shows how object elements can be read and - written using the `[]` operator.,operatorarray__key_type} - - @sa @ref at(const typename object_t::key_type&) for access by reference - with range checking - @sa @ref value() for access by value with a default value - - @since version 1.0.0 - */ - template - reference operator[](T * (&key)[n]) - { - return operator[](static_cast(key)); - } - - /*! - @brief read-only access specified object element - - Returns a const reference to the element at with specified key @a key. No - bounds checking is performed. - - @warning If the element with key @a key does not exist, the behavior is - undefined. - - @note This function is required for compatibility reasons with Clang. - - @param[in] key key of the element to access - - @return const reference to the element at key @a key - - @throw type_error.305 if the JSON value is not an object; in that cases, - using the [] operator with a key makes no sense. - - @complexity Logarithmic in the size of the container. - - @liveexample{The example below shows how object elements can be read using - the `[]` operator.,operatorarray__key_type_const} - - @sa @ref at(const typename object_t::key_type&) for access by reference - with range checking - @sa @ref value() for access by value with a default value - - @since version 1.0.0 - */ - template - const_reference operator[](T * (&key)[n]) const - { - return operator[](static_cast(key)); + JSON_THROW(type_error::create(305, "cannot use operator[] with " + std::string(type_name()))); } /*! @@ -4236,12 +14000,12 @@ class basic_json } // at only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { return m_value.object->operator[](key); } - JSON_THROW(type_error::create(305, "cannot use operator[] with " + type_name())); + JSON_THROW(type_error::create(305, "cannot use operator[] with " + std::string(type_name()))); } /*! @@ -4260,7 +14024,7 @@ class basic_json @pre The element with key @a key must exist. **This precondition is enforced with an assertion.** - @throw type_error.305 if the JSON value is not an object; in that cases, + @throw type_error.305 if the JSON value is not an object; in that case, using the [] operator with a key makes no sense. @complexity Logarithmic in the size of the container. @@ -4278,13 +14042,13 @@ class basic_json const_reference operator[](T* key) const { // at only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { assert(m_value.object->find(key) != m_value.object->end()); return m_value.object->find(key)->second; } - JSON_THROW(type_error::create(305, "cannot use operator[] with " + type_name())); + JSON_THROW(type_error::create(305, "cannot use operator[] with " + std::string(type_name()))); } /*! @@ -4320,7 +14084,7 @@ class basic_json @return copy of the element at key @a key or @a default_value if @a key is not found - @throw type_error.306 if the JSON value is not an objec; in that cases, + @throw type_error.306 if the JSON value is not an object; in that case, using `value()` with a key makes no sense. @complexity Logarithmic in the size of the container. @@ -4337,10 +14101,10 @@ class basic_json */ template::value, int>::type = 0> - ValueType value(const typename object_t::key_type& key, ValueType default_value) const + ValueType value(const typename object_t::key_type& key, const ValueType& default_value) const { // at only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { // if key is found, return value and given default value otherwise const auto it = find(key); @@ -4351,10 +14115,8 @@ class basic_json return default_value; } - else - { - JSON_THROW(type_error::create(306, "cannot use value() with " + type_name())); - } + + JSON_THROW(type_error::create(306, "cannot use value() with " + std::string(type_name()))); } /*! @@ -4395,7 +14157,7 @@ class basic_json @return copy of the element at key @a key or @a default_value if @a key is not found - @throw type_error.306 if the JSON value is not an objec; in that cases, + @throw type_error.306 if the JSON value is not an object; in that case, using `value()` with a key makes no sense. @complexity Logarithmic in the size of the container. @@ -4409,10 +14171,10 @@ class basic_json */ template::value, int>::type = 0> - ValueType value(const json_pointer& ptr, ValueType default_value) const + ValueType value(const json_pointer& ptr, const ValueType& default_value) const { // at only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { // if pointer resolves a value, return it or use default value JSON_TRY @@ -4425,7 +14187,7 @@ class basic_json } } - JSON_THROW(type_error::create(306, "cannot use value() with " + type_name())); + JSON_THROW(type_error::create(306, "cannot use value() with " + std::string(type_name()))); } /*! @@ -4576,7 +14338,7 @@ class basic_json IteratorType erase(IteratorType pos) { // make sure iterator fits the current value - if (this != pos.m_object) + if (JSON_UNLIKELY(this != pos.m_object)) { JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value")); } @@ -4591,7 +14353,7 @@ class basic_json case value_t::number_unsigned: case value_t::string: { - if (not pos.m_it.primitive_iterator.is_begin()) + if (JSON_UNLIKELY(not pos.m_it.primitive_iterator.is_begin())) { JSON_THROW(invalid_iterator::create(205, "iterator out of range")); } @@ -4599,8 +14361,8 @@ class basic_json if (is_string()) { AllocatorType alloc; - alloc.destroy(m_value.string); - alloc.deallocate(m_value.string, 1); + std::allocator_traits::destroy(alloc, m_value.string); + std::allocator_traits::deallocate(alloc, m_value.string, 1); m_value.string = nullptr; } @@ -4622,9 +14384,7 @@ class basic_json } default: - { - JSON_THROW(type_error::create(307, "cannot use erase() with " + type_name())); - } + JSON_THROW(type_error::create(307, "cannot use erase() with " + std::string(type_name()))); } return result; @@ -4683,7 +14443,7 @@ class basic_json IteratorType erase(IteratorType first, IteratorType last) { // make sure iterator fits the current value - if (this != first.m_object or this != last.m_object) + if (JSON_UNLIKELY(this != first.m_object or this != last.m_object)) { JSON_THROW(invalid_iterator::create(203, "iterators do not fit current value")); } @@ -4698,7 +14458,8 @@ class basic_json case value_t::number_unsigned: case value_t::string: { - if (not first.m_it.primitive_iterator.is_begin() or not last.m_it.primitive_iterator.is_end()) + if (JSON_LIKELY(not first.m_it.primitive_iterator.is_begin() + or not last.m_it.primitive_iterator.is_end())) { JSON_THROW(invalid_iterator::create(204, "iterators out of range")); } @@ -4706,8 +14467,8 @@ class basic_json if (is_string()) { AllocatorType alloc; - alloc.destroy(m_value.string); - alloc.deallocate(m_value.string, 1); + std::allocator_traits::destroy(alloc, m_value.string); + std::allocator_traits::deallocate(alloc, m_value.string, 1); m_value.string = nullptr; } @@ -4731,9 +14492,7 @@ class basic_json } default: - { - JSON_THROW(type_error::create(307, "cannot use erase() with " + type_name())); - } + JSON_THROW(type_error::create(307, "cannot use erase() with " + std::string(type_name()))); } return result; @@ -4771,12 +14530,12 @@ class basic_json size_type erase(const typename object_t::key_type& key) { // this erase only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { return m_value.object->erase(key); } - JSON_THROW(type_error::create(307, "cannot use erase() with " + type_name())); + JSON_THROW(type_error::create(307, "cannot use erase() with " + std::string(type_name()))); } /*! @@ -4806,9 +14565,9 @@ class basic_json void erase(const size_type idx) { // this erase only works for arrays - if (is_array()) + if (JSON_LIKELY(is_array())) { - if (idx >= size()) + if (JSON_UNLIKELY(idx >= size())) { JSON_THROW(out_of_range::create(401, "array index " + std::to_string(idx) + " is out of range")); } @@ -4817,7 +14576,7 @@ class basic_json } else { - JSON_THROW(type_error::create(307, "cannot use erase() with " + type_name())); + JSON_THROW(type_error::create(307, "cannot use erase() with " + std::string(type_name()))); } } @@ -4841,7 +14600,7 @@ class basic_json @note This method always returns @ref end() when executed on a JSON type that is not an object. - @param[in] key key value of the element to search for + @param[in] key key value of the element to search for. @return Iterator to an element with key equivalent to @a key. If no such element is found or the JSON value is not an object, past-the-end (see @@ -4853,13 +14612,14 @@ class basic_json @since version 1.0.0 */ - iterator find(typename object_t::key_type key) + template + iterator find(KeyT&& key) { auto result = end(); if (is_object()) { - result.m_it.object_iterator = m_value.object->find(key); + result.m_it.object_iterator = m_value.object->find(std::forward(key)); } return result; @@ -4867,15 +14627,16 @@ class basic_json /*! @brief find an element in a JSON object - @copydoc find(typename object_t::key_type) + @copydoc find(KeyT&&) */ - const_iterator find(typename object_t::key_type key) const + template + const_iterator find(KeyT&& key) const { auto result = cend(); if (is_object()) { - result.m_it.object_iterator = m_value.object->find(key); + result.m_it.object_iterator = m_value.object->find(std::forward(key)); } return result; @@ -4902,10 +14663,11 @@ class basic_json @since version 1.0.0 */ - size_type count(typename object_t::key_type key) const + template + size_type count(KeyT&& key) const { // return 0 for all nonobject types - return is_object() ? m_value.object->count(key) : 0; + return is_object() ? m_value.object->count(std::forward(key)) : 0; } /// @} @@ -5191,10 +14953,6 @@ class basic_json return const_reverse_iterator(cbegin()); } - private: - // forward declaration - template class iteration_proxy; - public: /*! @brief wrapper to access iterator member functions in range-based for @@ -5204,22 +14962,133 @@ class basic_json reference to the JSON values is returned, so there is no access to the underlying iterator. + For loop without iterator_wrapper: + + @code{cpp} + for (auto it = j_object.begin(); it != j_object.end(); ++it) + { + std::cout << "key: " << it.key() << ", value:" << it.value() << '\n'; + } + @endcode + + Range-based for loop without iterator proxy: + + @code{cpp} + for (auto it : j_object) + { + // "it" is of type json::reference and has no key() member + std::cout << "value: " << it << '\n'; + } + @endcode + + Range-based for loop with iterator proxy: + + @code{cpp} + for (auto it : json::iterator_wrapper(j_object)) + { + std::cout << "key: " << it.key() << ", value:" << it.value() << '\n'; + } + @endcode + + @note When iterating over an array, `key()` will return the index of the + element as string (see example). + + @param[in] ref reference to a JSON value + @return iteration proxy object wrapping @a ref with an interface to use in + range-based for loops + @liveexample{The following code shows how the wrapper is used,iterator_wrapper} + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes in the JSON value. + + @complexity Constant. + @note The name of this function is not yet final and may change in the future. + + @deprecated This stream operator is deprecated and will be removed in + future 4.0.0 of the library. Please use @ref items() instead; + that is, replace `json::iterator_wrapper(j)` with `j.items()`. */ - static iteration_proxy iterator_wrapper(reference cont) + JSON_DEPRECATED + static iteration_proxy iterator_wrapper(reference ref) noexcept { - return iteration_proxy(cont); + return ref.items(); } /*! @copydoc iterator_wrapper(reference) */ - static iteration_proxy iterator_wrapper(const_reference cont) + JSON_DEPRECATED + static iteration_proxy iterator_wrapper(const_reference ref) noexcept { - return iteration_proxy(cont); + return ref.items(); + } + + /*! + @brief helper to access iterator member functions in range-based for + + This function allows to access @ref iterator::key() and @ref + iterator::value() during range-based for loops. In these loops, a + reference to the JSON values is returned, so there is no access to the + underlying iterator. + + For loop without `items()` function: + + @code{cpp} + for (auto it = j_object.begin(); it != j_object.end(); ++it) + { + std::cout << "key: " << it.key() << ", value:" << it.value() << '\n'; + } + @endcode + + Range-based for loop without `items()` function: + + @code{cpp} + for (auto it : j_object) + { + // "it" is of type json::reference and has no key() member + std::cout << "value: " << it << '\n'; + } + @endcode + + Range-based for loop with `items()` function: + + @code{cpp} + for (auto it : j_object.items()) + { + std::cout << "key: " << it.key() << ", value:" << it.value() << '\n'; + } + @endcode + + @note When iterating over an array, `key()` will return the index of the + element as string (see example). For primitive types (e.g., numbers), + `key()` returns an empty string. + + @return iteration proxy object wrapping @a ref with an interface to use in + range-based for loops + + @liveexample{The following code shows how the function is used.,items} + + @exceptionsafety Strong guarantee: if an exception is thrown, there are no + changes in the JSON value. + + @complexity Constant. + + @since version 3.x.x. + */ + iteration_proxy items() noexcept + { + return iteration_proxy(*this); + } + + /*! + @copydoc items() + */ + iteration_proxy items() const noexcept + { + return iteration_proxy(*this); } /// @} @@ -5233,9 +15102,9 @@ class basic_json /// @{ /*! - @brief checks whether the container is empty + @brief checks whether the container is empty. - Checks if a JSON value has no elements. + Checks if a JSON value has no elements (i.e. whether its @ref size is `0`). @return The return value depends on the different types and is defined as follows: @@ -5248,23 +15117,27 @@ class basic_json object | result of function `object_t::empty()` array | result of function `array_t::empty()` - @note This function does not return whether a string stored as JSON value - is empty - it returns whether the JSON container itself is empty which is - false in the case of a string. + @liveexample{The following code uses `empty()` to check if a JSON + object contains any elements.,empty} @complexity Constant, as long as @ref array_t and @ref object_t satisfy the Container concept; that is, their `empty()` functions have constant complexity. + @iterators No changes. + + @exceptionsafety No-throw guarantee: this function never throws exceptions. + + @note This function does not return whether a string stored as JSON value + is empty - it returns whether the JSON container itself is empty which is + false in the case of a string. + @requirement This function helps `basic_json` satisfying the [Container](http://en.cppreference.com/w/cpp/concept/Container) requirements: - The complexity is constant. - Has the semantics of `begin() == end()`. - @liveexample{The following code uses `empty()` to check if a JSON - object contains any elements.,empty} - @sa @ref size() -- returns the number of elements @since version 1.0.0 @@ -5315,23 +15188,27 @@ class basic_json object | result of function object_t::size() array | result of function array_t::size() - @note This function does not return the length of a string stored as JSON - value - it returns the number of elements in the JSON value which is 1 in - the case of a string. + @liveexample{The following code calls `size()` on the different value + types.,size} @complexity Constant, as long as @ref array_t and @ref object_t satisfy the Container concept; that is, their size() functions have constant complexity. + @iterators No changes. + + @exceptionsafety No-throw guarantee: this function never throws exceptions. + + @note This function does not return the length of a string stored as JSON + value - it returns the number of elements in the JSON value which is 1 in + the case of a string. + @requirement This function helps `basic_json` satisfying the [Container](http://en.cppreference.com/w/cpp/concept/Container) requirements: - The complexity is constant. - Has the semantics of `std::distance(begin(), end())`. - @liveexample{The following code calls `size()` on the different value - types.,size} - @sa @ref empty() -- checks whether the container is empty @sa @ref max_size() -- returns the maximal number of elements @@ -5385,10 +15262,17 @@ class basic_json object | result of function `object_t::max_size()` array | result of function `array_t::max_size()` + @liveexample{The following code calls `max_size()` on the different value + types. Note the output is implementation specific.,max_size} + @complexity Constant, as long as @ref array_t and @ref object_t satisfy the Container concept; that is, their `max_size()` functions have constant complexity. + @iterators No changes. + + @exceptionsafety No-throw guarantee: this function never throws exceptions. + @requirement This function helps `basic_json` satisfying the [Container](http://en.cppreference.com/w/cpp/concept/Container) requirements: @@ -5396,9 +15280,6 @@ class basic_json - Has the semantics of returning `b.size()` where `b` is the largest possible JSON value. - @liveexample{The following code calls `max_size()` on the different value - types. Note the output is implementation specific.,max_size} - @sa @ref size() -- returns the number of elements @since version 1.0.0 @@ -5441,7 +15322,8 @@ class basic_json @brief clears the contents Clears the content of a JSON value and resets it to the default value as - if @ref basic_json(value_t) would have been called: + if @ref basic_json(value_t) would have been called with the current value + type from @ref type(): Value type | initial value ----------- | ------------- @@ -5452,11 +15334,24 @@ class basic_json object | `{}` array | `[]` - @complexity Linear in the size of the JSON value. + @post Has the same effect as calling + @code {.cpp} + *this = basic_json(type()); + @endcode @liveexample{The example below shows the effect of `clear()` to different JSON types.,clear} + @complexity Linear in the size of the JSON value. + + @iterators All iterators, pointers and references related to this container + are invalidated. + + @exceptionsafety No-throw guarantee: this function never throws exceptions. + + @sa @ref basic_json(value_t) -- constructor that creates an object with the + same value than calling `clear()` + @since version 1.0.0 */ void clear() noexcept @@ -5506,9 +15401,7 @@ class basic_json } default: - { break; - } } } @@ -5535,9 +15428,9 @@ class basic_json void push_back(basic_json&& val) { // push_back only works for null objects or arrays - if (not(is_null() or is_array())) + if (JSON_UNLIKELY(not(is_null() or is_array()))) { - JSON_THROW(type_error::create(308, "cannot use push_back() with " + type_name())); + JSON_THROW(type_error::create(308, "cannot use push_back() with " + std::string(type_name()))); } // transform null object into an array @@ -5571,9 +15464,9 @@ class basic_json void push_back(const basic_json& val) { // push_back only works for null objects or arrays - if (not(is_null() or is_array())) + if (JSON_UNLIKELY(not(is_null() or is_array()))) { - JSON_THROW(type_error::create(308, "cannot use push_back() with " + type_name())); + JSON_THROW(type_error::create(308, "cannot use push_back() with " + std::string(type_name()))); } // transform null object into an array @@ -5621,9 +15514,9 @@ class basic_json void push_back(const typename object_t::value_type& val) { // push_back only works for null objects or objects - if (not(is_null() or is_object())) + if (JSON_UNLIKELY(not(is_null() or is_object()))) { - JSON_THROW(type_error::create(308, "cannot use push_back() with " + type_name())); + JSON_THROW(type_error::create(308, "cannot use push_back() with " + std::string(type_name()))); } // transform null object into an object @@ -5673,12 +15566,13 @@ class basic_json @liveexample{The example shows how initializer lists are treated as objects when possible.,push_back__initializer_list} */ - void push_back(std::initializer_list init) + void push_back(initializer_list_t init) { - if (is_object() and init.size() == 2 and init.begin()->is_string()) + if (is_object() and init.size() == 2 and (*init.begin())->is_string()) { - const string_t key = *init.begin(); - push_back(typename object_t::value_type(key, *(init.begin() + 1))); + basic_json&& key = init.begin()->moved_or_copied(); + push_back(typename object_t::value_type( + std::move(key.get_ref()), (init.begin() + 1)->moved_or_copied())); } else { @@ -5688,9 +15582,9 @@ class basic_json /*! @brief add an object to an object - @copydoc push_back(std::initializer_list) + @copydoc push_back(initializer_list_t) */ - reference operator+=(std::initializer_list init) + reference operator+=(initializer_list_t init) { push_back(init); return *this; @@ -5721,9 +15615,9 @@ class basic_json void emplace_back(Args&& ... args) { // emplace_back only works for null objects or arrays - if (not(is_null() or is_array())) + if (JSON_UNLIKELY(not(is_null() or is_array()))) { - JSON_THROW(type_error::create(311, "cannot use emplace_back() with " + type_name())); + JSON_THROW(type_error::create(311, "cannot use emplace_back() with " + std::string(type_name()))); } // transform null object into an array @@ -5769,9 +15663,9 @@ class basic_json std::pair emplace(Args&& ... args) { // emplace only works for null objects or arrays - if (not(is_null() or is_object())) + if (JSON_UNLIKELY(not(is_null() or is_object()))) { - JSON_THROW(type_error::create(311, "cannot use emplace() with " + type_name())); + JSON_THROW(type_error::create(311, "cannot use emplace() with " + std::string(type_name()))); } // transform null object into an object @@ -5817,10 +15711,10 @@ class basic_json iterator insert(const_iterator pos, const basic_json& val) { // insert only works for arrays - if (is_array()) + if (JSON_LIKELY(is_array())) { // check if iterator pos fits to this JSON value - if (pos.m_object != this) + if (JSON_UNLIKELY(pos.m_object != this)) { JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value")); } @@ -5831,7 +15725,7 @@ class basic_json return result; } - JSON_THROW(type_error::create(309, "cannot use insert() with " + type_name())); + JSON_THROW(type_error::create(309, "cannot use insert() with " + std::string(type_name()))); } /*! @@ -5870,10 +15764,10 @@ class basic_json iterator insert(const_iterator pos, size_type cnt, const basic_json& val) { // insert only works for arrays - if (is_array()) + if (JSON_LIKELY(is_array())) { // check if iterator pos fits to this JSON value - if (pos.m_object != this) + if (JSON_UNLIKELY(pos.m_object != this)) { JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value")); } @@ -5884,7 +15778,7 @@ class basic_json return result; } - JSON_THROW(type_error::create(309, "cannot use insert() with " + type_name())); + JSON_THROW(type_error::create(309, "cannot use insert() with " + std::string(type_name()))); } /*! @@ -5920,24 +15814,24 @@ class basic_json iterator insert(const_iterator pos, const_iterator first, const_iterator last) { // insert only works for arrays - if (not is_array()) + if (JSON_UNLIKELY(not is_array())) { - JSON_THROW(type_error::create(309, "cannot use insert() with " + type_name())); + JSON_THROW(type_error::create(309, "cannot use insert() with " + std::string(type_name()))); } // check if iterator pos fits to this JSON value - if (pos.m_object != this) + if (JSON_UNLIKELY(pos.m_object != this)) { JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value")); } // check if range iterators belong to the same JSON object - if (first.m_object != last.m_object) + if (JSON_UNLIKELY(first.m_object != last.m_object)) { JSON_THROW(invalid_iterator::create(210, "iterators do not fit")); } - if (first.m_object == this or last.m_object == this) + if (JSON_UNLIKELY(first.m_object == this)) { JSON_THROW(invalid_iterator::create(211, "passed iterators may not belong to container")); } @@ -5975,23 +15869,23 @@ class basic_json @since version 1.0.0 */ - iterator insert(const_iterator pos, std::initializer_list ilist) + iterator insert(const_iterator pos, initializer_list_t ilist) { // insert only works for arrays - if (not is_array()) + if (JSON_UNLIKELY(not is_array())) { - JSON_THROW(type_error::create(309, "cannot use insert() with " + type_name())); + JSON_THROW(type_error::create(309, "cannot use insert() with " + std::string(type_name()))); } // check if iterator pos fits to this JSON value - if (pos.m_object != this) + if (JSON_UNLIKELY(pos.m_object != this)) { JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value")); } // insert to array and return iterator iterator result(this); - result.m_it.array_iterator = m_value.array->insert(pos.m_it.array_iterator, ilist); + result.m_it.array_iterator = m_value.array->insert(pos.m_it.array_iterator, ilist.begin(), ilist.end()); return result; } @@ -6021,19 +15915,19 @@ class basic_json void insert(const_iterator first, const_iterator last) { // insert only works for objects - if (not is_object()) + if (JSON_UNLIKELY(not is_object())) { - JSON_THROW(type_error::create(309, "cannot use insert() with " + type_name())); + JSON_THROW(type_error::create(309, "cannot use insert() with " + std::string(type_name()))); } // check if range iterators belong to the same JSON object - if (first.m_object != last.m_object) + if (JSON_UNLIKELY(first.m_object != last.m_object)) { JSON_THROW(invalid_iterator::create(210, "iterators do not fit")); } // passed iterators must belong to objects - if (not first.m_object->is_object() or not first.m_object->is_object()) + if (JSON_UNLIKELY(not first.m_object->is_object())) { JSON_THROW(invalid_iterator::create(202, "iterators first and last must point to objects")); } @@ -6041,6 +15935,110 @@ class basic_json m_value.object->insert(first.m_it.object_iterator, last.m_it.object_iterator); } + /*! + @brief updates a JSON object from another object, overwriting existing keys + + Inserts all values from JSON object @a j and overwrites existing keys. + + @param[in] j JSON object to read values from + + @throw type_error.312 if called on JSON values other than objects; example: + `"cannot use update() with string"` + + @complexity O(N*log(size() + N)), where N is the number of elements to + insert. + + @liveexample{The example shows how `update()` is used.,update} + + @sa https://docs.python.org/3.6/library/stdtypes.html#dict.update + + @since version 3.0.0 + */ + void update(const_reference j) + { + // implicitly convert null value to an empty object + if (is_null()) + { + m_type = value_t::object; + m_value.object = create(); + assert_invariant(); + } + + if (JSON_UNLIKELY(not is_object())) + { + JSON_THROW(type_error::create(312, "cannot use update() with " + std::string(type_name()))); + } + if (JSON_UNLIKELY(not j.is_object())) + { + JSON_THROW(type_error::create(312, "cannot use update() with " + std::string(j.type_name()))); + } + + for (auto it = j.cbegin(); it != j.cend(); ++it) + { + m_value.object->operator[](it.key()) = it.value(); + } + } + + /*! + @brief updates a JSON object from another object, overwriting existing keys + + Inserts all values from from range `[first, last)` and overwrites existing + keys. + + @param[in] first begin of the range of elements to insert + @param[in] last end of the range of elements to insert + + @throw type_error.312 if called on JSON values other than objects; example: + `"cannot use update() with string"` + @throw invalid_iterator.202 if iterator @a first or @a last does does not + point to an object; example: `"iterators first and last must point to + objects"` + @throw invalid_iterator.210 if @a first and @a last do not belong to the + same JSON value; example: `"iterators do not fit"` + + @complexity O(N*log(size() + N)), where N is the number of elements to + insert. + + @liveexample{The example shows how `update()` is used__range.,update} + + @sa https://docs.python.org/3.6/library/stdtypes.html#dict.update + + @since version 3.0.0 + */ + void update(const_iterator first, const_iterator last) + { + // implicitly convert null value to an empty object + if (is_null()) + { + m_type = value_t::object; + m_value.object = create(); + assert_invariant(); + } + + if (JSON_UNLIKELY(not is_object())) + { + JSON_THROW(type_error::create(312, "cannot use update() with " + std::string(type_name()))); + } + + // check if range iterators belong to the same JSON object + if (JSON_UNLIKELY(first.m_object != last.m_object)) + { + JSON_THROW(invalid_iterator::create(210, "iterators do not fit")); + } + + // passed iterators must belong to objects + if (JSON_UNLIKELY(not first.m_object->is_object() + or not last.m_object->is_object())) + { + JSON_THROW(invalid_iterator::create(202, "iterators first and last must point to objects")); + } + + for (auto it = first; it != last; ++it) + { + m_value.object->operator[](it.key()) = it.value(); + } + } + /*! @brief exchanges the values @@ -6093,13 +16091,13 @@ class basic_json void swap(array_t& other) { // swap only works for arrays - if (is_array()) + if (JSON_LIKELY(is_array())) { std::swap(*(m_value.array), other); } else { - JSON_THROW(type_error::create(310, "cannot use swap() with " + type_name())); + JSON_THROW(type_error::create(310, "cannot use swap() with " + std::string(type_name()))); } } @@ -6126,13 +16124,13 @@ class basic_json void swap(object_t& other) { // swap only works for objects - if (is_object()) + if (JSON_LIKELY(is_object())) { std::swap(*(m_value.object), other); } else { - JSON_THROW(type_error::create(310, "cannot use swap() with " + type_name())); + JSON_THROW(type_error::create(310, "cannot use swap() with " + std::string(type_name()))); } } @@ -6159,13 +16157,13 @@ class basic_json void swap(string_t& other) { // swap only works for strings - if (is_string()) + if (JSON_LIKELY(is_string())) { std::swap(*(m_value.string), other); } else { - JSON_THROW(type_error::create(310, "cannot use swap() with " + type_name())); + JSON_THROW(type_error::create(310, "cannot use swap() with " + std::string(type_name()))); } } @@ -6187,18 +16185,30 @@ class basic_json their stored values are the same according to their respective `operator==`. - Integer and floating-point numbers are automatically converted before - comparison. Floating-point numbers are compared indirectly: two - floating-point numbers `f1` and `f2` are considered equal if neither - `f1 > f2` nor `f2 > f1` holds. Note than two NaN values are always - treated as unequal. + comparison. Note than two NaN values are always treated as unequal. - Two JSON null values are equal. + @note Floating-point inside JSON values numbers are compared with + `json::number_float_t::operator==` which is `double::operator==` by + default. To compare floating-point while respecting an epsilon, an alternative + [comparison function](https://github.com/mariokonrad/marnav/blob/master/src/marnav/math/floatingpoint.hpp#L34-#L39) + could be used, for instance + @code {.cpp} + template::value, T>::type> + inline bool is_same(T a, T b, T epsilon = std::numeric_limits::epsilon()) noexcept + { + return std::abs(a - b) <= epsilon; + } + @endcode + @note NaN values never compare equal to themselves or to other NaN values. @param[in] lhs first JSON value to consider @param[in] rhs second JSON value to consider @return whether the values @a lhs and @a rhs are equal + @exceptionsafety No-throw guarantee: this function never throws exceptions. + @complexity Linear. @liveexample{The example demonstrates comparing several JSON @@ -6216,66 +16226,56 @@ class basic_json switch (lhs_type) { case value_t::array: - { - return *lhs.m_value.array == *rhs.m_value.array; - } + return (*lhs.m_value.array == *rhs.m_value.array); + case value_t::object: - { - return *lhs.m_value.object == *rhs.m_value.object; - } + return (*lhs.m_value.object == *rhs.m_value.object); + case value_t::null: - { return true; - } + case value_t::string: - { - return *lhs.m_value.string == *rhs.m_value.string; - } + return (*lhs.m_value.string == *rhs.m_value.string); + case value_t::boolean: - { - return lhs.m_value.boolean == rhs.m_value.boolean; - } + return (lhs.m_value.boolean == rhs.m_value.boolean); + case value_t::number_integer: - { - return lhs.m_value.number_integer == rhs.m_value.number_integer; - } + return (lhs.m_value.number_integer == rhs.m_value.number_integer); + case value_t::number_unsigned: - { - return lhs.m_value.number_unsigned == rhs.m_value.number_unsigned; - } + return (lhs.m_value.number_unsigned == rhs.m_value.number_unsigned); + case value_t::number_float: - { - return lhs.m_value.number_float == rhs.m_value.number_float; - } + return (lhs.m_value.number_float == rhs.m_value.number_float); + default: - { return false; - } } } else if (lhs_type == value_t::number_integer and rhs_type == value_t::number_float) { - return static_cast(lhs.m_value.number_integer) == rhs.m_value.number_float; + return (static_cast(lhs.m_value.number_integer) == rhs.m_value.number_float); } else if (lhs_type == value_t::number_float and rhs_type == value_t::number_integer) { - return lhs.m_value.number_float == static_cast(rhs.m_value.number_integer); + return (lhs.m_value.number_float == static_cast(rhs.m_value.number_integer)); } else if (lhs_type == value_t::number_unsigned and rhs_type == value_t::number_float) { - return static_cast(lhs.m_value.number_unsigned) == rhs.m_value.number_float; + return (static_cast(lhs.m_value.number_unsigned) == rhs.m_value.number_float); } else if (lhs_type == value_t::number_float and rhs_type == value_t::number_unsigned) { - return lhs.m_value.number_float == static_cast(rhs.m_value.number_unsigned); + return (lhs.m_value.number_float == static_cast(rhs.m_value.number_unsigned)); } else if (lhs_type == value_t::number_unsigned and rhs_type == value_t::number_integer) { - return static_cast(lhs.m_value.number_unsigned) == rhs.m_value.number_integer; + return (static_cast(lhs.m_value.number_unsigned) == rhs.m_value.number_integer); } else if (lhs_type == value_t::number_integer and rhs_type == value_t::number_unsigned) { - return lhs.m_value.number_integer == static_cast(rhs.m_value.number_unsigned); + return (lhs.m_value.number_integer == static_cast(rhs.m_value.number_unsigned)); } return false; @@ -6314,6 +16314,8 @@ class basic_json @complexity Linear. + @exceptionsafety No-throw guarantee: this function never throws exceptions. + @liveexample{The example demonstrates comparing several JSON types.,operator__notequal} @@ -6365,6 +16367,8 @@ class basic_json @complexity Linear. + @exceptionsafety No-throw guarantee: this function never throws exceptions. + @liveexample{The example demonstrates comparing several JSON types.,operator__less} @@ -6380,44 +16384,31 @@ class basic_json switch (lhs_type) { case value_t::array: - { - //return *lhs.m_value.array < *rhs.m_value.array; - //return *lhs.m_value.array.operator<(rhs); - return (*lhs.m_value.array) < *rhs.m_value.array; - // return nlohmann::detail::operator<(lhs, rhs); - } + return (*lhs.m_value.array) < (*rhs.m_value.array); + case value_t::object: - { return *lhs.m_value.object < *rhs.m_value.object; - } + case value_t::null: - { return false; - } + case value_t::string: - { return *lhs.m_value.string < *rhs.m_value.string; - } + case value_t::boolean: - { return lhs.m_value.boolean < rhs.m_value.boolean; - } + case value_t::number_integer: - { return lhs.m_value.number_integer < rhs.m_value.number_integer; - } + case value_t::number_unsigned: - { return lhs.m_value.number_unsigned < rhs.m_value.number_unsigned; - } + case value_t::number_float: - { return lhs.m_value.number_float < rhs.m_value.number_float; - } + default: - { return false; - } } } else if (lhs_type == value_t::number_integer and rhs_type == value_t::number_float) @@ -6485,6 +16476,8 @@ class basic_json @complexity Linear. + @exceptionsafety No-throw guarantee: this function never throws exceptions. + @liveexample{The example demonstrates comparing several JSON types.,operator__greater} @@ -6529,6 +16522,8 @@ class basic_json @complexity Linear. + @exceptionsafety No-throw guarantee: this function never throws exceptions. + @liveexample{The example demonstrates comparing several JSON types.,operator__lessequal} @@ -6573,6 +16568,8 @@ class basic_json @complexity Linear. + @exceptionsafety No-throw guarantee: this function never throws exceptions. + @liveexample{The example demonstrates comparing several JSON types.,operator__greaterequal} @@ -6607,110 +16604,6 @@ class basic_json /// @} - private: - ///////////////////// - // output adapters // - ///////////////////// - - /// abstract output adapter interface - template - class output_adapter - { - public: - virtual void write_character(CharType c) = 0; - virtual void write_characters(const CharType* s, size_t length) = 0; - virtual ~output_adapter() {} - - static std::shared_ptr> create(std::vector& vec) - { - return std::shared_ptr(new output_vector_adapter(vec)); - } - - static std::shared_ptr> create(std::ostream& s) - { - return std::shared_ptr(new output_stream_adapter(s)); - } - - static std::shared_ptr> create(std::string& s) - { - return std::shared_ptr(new output_string_adapter(s)); - } - }; - - /// a type to simplify interfaces - template - using output_adapter_t = std::shared_ptr>; - - /// output adapter for byte vectors - template - class output_vector_adapter : public output_adapter - { - public: - output_vector_adapter(std::vector& vec) - : v(vec) - {} - - void write_character(CharType c) override - { - v.push_back(c); - } - - void write_characters(const CharType* s, size_t length) override - { - std::copy(s, s + length, std::back_inserter(v)); - } - - private: - std::vector& v; - }; - - /// putput adatpter for output streams - template - class output_stream_adapter : public output_adapter - { - public: - output_stream_adapter(std::basic_ostream& s) - : stream(s) - {} - - void write_character(CharType c) override - { - stream.put(c); - } - - void write_characters(const CharType* s, size_t length) override - { - stream.write(s, static_cast(length)); - } - - private: - std::basic_ostream& stream; - }; - - /// output adapter for basic_string - template - class output_string_adapter : public output_adapter - { - public: - output_string_adapter(std::string& s) - : str(s) - {} - - void write_character(CharType c) override - { - str.push_back(c); - } - - void write_characters(const CharType* s, size_t length) override - { - str.append(s, length); - } - - private: - std::basic_string& str; - }; - - /////////////////// // serialization // /////////////////// @@ -6718,598 +16611,6 @@ class basic_json /// @name serialization /// @{ - private: - /*! - @brief wrapper around the serialization functions - */ - class serializer - { - private: - serializer(const serializer&) = delete; - serializer& operator=(const serializer&) = delete; - - public: - /*! - @param[in] s output stream to serialize to - @param[in] ichar indentation character to use - */ - serializer(output_adapter_t s, const char ichar) - : o(s), loc(std::localeconv()), - thousands_sep(!loc->thousands_sep ? '\0' : loc->thousands_sep[0]), - decimal_point(!loc->decimal_point ? '\0' : loc->decimal_point[0]), - indent_char(ichar), indent_string(512, indent_char) - {} - - /*! - @brief internal implementation of the serialization function - - This function is called by the public member function dump and - organizes the serialization internally. The indentation level is - propagated as additional parameter. In case of arrays and objects, the - function is called recursively. - - - strings and object keys are escaped using `escape_string()` - - integer numbers are converted implicitly via `operator<<` - - floating-point numbers are converted to a string using `"%g"` format - - @param[in] val value to serialize - @param[in] pretty_print whether the output shall be pretty-printed - @param[in] indent_step the indent level - @param[in] current_indent the current indent level (only used internally) - */ - void dump(const basic_json& val, - const bool pretty_print, - const unsigned int indent_step, - const unsigned int current_indent = 0) - { - switch (val.m_type) - { - case value_t::object: - { - if (val.m_value.object->empty()) - { - o->write_characters("{}", 2); - return; - } - - if (pretty_print) - { - o->write_characters("{\n", 2); - - // variable to hold indentation for recursive calls - const auto new_indent = current_indent + indent_step; - if (indent_string.size() < new_indent) - { - indent_string.resize(new_indent, ' '); - } - - // first n-1 elements - auto i = val.m_value.object->cbegin(); - for (size_t cnt = 0; cnt < val.m_value.object->size() - 1; ++cnt, ++i) - { - o->write_characters(indent_string.c_str(), new_indent); - o->write_character('\"'); - dump_escaped(i->first); - o->write_characters("\": ", 3); - dump(i->second, true, indent_step, new_indent); - o->write_characters(",\n", 2); - } - - // last element - assert(i != val.m_value.object->cend()); - o->write_characters(indent_string.c_str(), new_indent); - o->write_character('\"'); - dump_escaped(i->first); - o->write_characters("\": ", 3); - dump(i->second, true, indent_step, new_indent); - - o->write_character('\n'); - o->write_characters(indent_string.c_str(), current_indent); - o->write_character('}'); - } - else - { - o->write_character('{'); - - // first n-1 elements - auto i = val.m_value.object->cbegin(); - for (size_t cnt = 0; cnt < val.m_value.object->size() - 1; ++cnt, ++i) - { - o->write_character('\"'); - dump_escaped(i->first); - o->write_characters("\":", 2); - dump(i->second, false, indent_step, current_indent); - o->write_character(','); - } - - // last element - assert(i != val.m_value.object->cend()); - o->write_character('\"'); - dump_escaped(i->first); - o->write_characters("\":", 2); - dump(i->second, false, indent_step, current_indent); - - o->write_character('}'); - } - - return; - } - - case value_t::array: - { - if (val.m_value.array->empty()) - { - o->write_characters("[]", 2); - return; - } - - if (pretty_print) - { - o->write_characters("[\n", 2); - - // variable to hold indentation for recursive calls - const auto new_indent = current_indent + indent_step; - if (indent_string.size() < new_indent) - { - indent_string.resize(new_indent, ' '); - } - - // first n-1 elements - for (auto i = val.m_value.array->cbegin(); i != val.m_value.array->cend() - 1; ++i) - { - o->write_characters(indent_string.c_str(), new_indent); - dump(*i, true, indent_step, new_indent); - o->write_characters(",\n", 2); - } - - // last element - assert(not val.m_value.array->empty()); - o->write_characters(indent_string.c_str(), new_indent); - dump(val.m_value.array->back(), true, indent_step, new_indent); - - o->write_character('\n'); - o->write_characters(indent_string.c_str(), current_indent); - o->write_character(']'); - } - else - { - o->write_character('['); - - // first n-1 elements - for (auto i = val.m_value.array->cbegin(); i != val.m_value.array->cend() - 1; ++i) - { - dump(*i, false, indent_step, current_indent); - o->write_character(','); - } - - // last element - assert(not val.m_value.array->empty()); - dump(val.m_value.array->back(), false, indent_step, current_indent); - - o->write_character(']'); - } - - return; - } - - case value_t::string: - { - o->write_character('\"'); - dump_escaped(*val.m_value.string); - o->write_character('\"'); - return; - } - - case value_t::boolean: - { - if (val.m_value.boolean) - { - o->write_characters("true", 4); - } - else - { - o->write_characters("false", 5); - } - return; - } - - case value_t::number_integer: - { - dump_integer(val.m_value.number_integer); - return; - } - - case value_t::number_unsigned: - { - dump_integer(val.m_value.number_unsigned); - return; - } - - case value_t::number_float: - { - dump_float(val.m_value.number_float); - return; - } - - case value_t::discarded: - { - o->write_characters("", 11); - return; - } - - case value_t::null: - { - o->write_characters("null", 4); - return; - } - } - } - - private: - /*! - @brief calculates the extra space to escape a JSON string - - @param[in] s the string to escape - @return the number of characters required to escape string @a s - - @complexity Linear in the length of string @a s. - */ - static std::size_t extra_space(const string_t& s) noexcept - { - return std::accumulate(s.begin(), s.end(), size_t{}, - [](size_t res, typename string_t::value_type c) - { - switch (c) - { - case '"': - case '\\': - case '\b': - case '\f': - case '\n': - case '\r': - case '\t': - { - // from c (1 byte) to \x (2 bytes) - return res + 1; - } - - case 0x00: - case 0x01: - case 0x02: - case 0x03: - case 0x04: - case 0x05: - case 0x06: - case 0x07: - case 0x0b: - case 0x0e: - case 0x0f: - case 0x10: - case 0x11: - case 0x12: - case 0x13: - case 0x14: - case 0x15: - case 0x16: - case 0x17: - case 0x18: - case 0x19: - case 0x1a: - case 0x1b: - case 0x1c: - case 0x1d: - case 0x1e: - case 0x1f: - { - // from c (1 byte) to \uxxxx (6 bytes) - return res + 5; - } - - default: - { - return res; - } - } - }); - } - - /*! - @brief dump escaped string - - Escape a string by replacing certain special characters by a sequence - of an escape character (backslash) and another character and other - control characters by a sequence of "\u" followed by a four-digit hex - representation. The escaped string is written to output stream @a o. - - @param[in] s the string to escape - - @complexity Linear in the length of string @a s. - */ - void dump_escaped(const string_t& s) const - { - const auto space = extra_space(s); - if (space == 0) - { - o->write_characters(s.c_str(), s.size()); - return; - } - - // create a result string of necessary size - string_t result(s.size() + space, '\\'); - std::size_t pos = 0; - - for (const auto& c : s) - { - switch (c) - { - // quotation mark (0x22) - case '"': - { - result[pos + 1] = '"'; - pos += 2; - break; - } - - // reverse solidus (0x5c) - case '\\': - { - // nothing to change - pos += 2; - break; - } - - // backspace (0x08) - case '\b': - { - result[pos + 1] = 'b'; - pos += 2; - break; - } - - // formfeed (0x0c) - case '\f': - { - result[pos + 1] = 'f'; - pos += 2; - break; - } - - // newline (0x0a) - case '\n': - { - result[pos + 1] = 'n'; - pos += 2; - break; - } - - // carriage return (0x0d) - case '\r': - { - result[pos + 1] = 'r'; - pos += 2; - break; - } - - // horizontal tab (0x09) - case '\t': - { - result[pos + 1] = 't'; - pos += 2; - break; - } - - case 0x00: - case 0x01: - case 0x02: - case 0x03: - case 0x04: - case 0x05: - case 0x06: - case 0x07: - case 0x0b: - case 0x0e: - case 0x0f: - case 0x10: - case 0x11: - case 0x12: - case 0x13: - case 0x14: - case 0x15: - case 0x16: - case 0x17: - case 0x18: - case 0x19: - case 0x1a: - case 0x1b: - case 0x1c: - case 0x1d: - case 0x1e: - case 0x1f: - { - // convert a number 0..15 to its hex representation - // (0..f) - static const char hexify[16] = - { - '0', '1', '2', '3', '4', '5', '6', '7', - '8', '9', 'a', 'b', 'c', 'd', 'e', 'f' - }; - - // print character c as \uxxxx - for (const char m : - { 'u', '0', '0', hexify[c >> 4], hexify[c & 0x0f] - }) - { - result[++pos] = m; - } - - ++pos; - break; - } - - default: - { - // all other characters are added as-is - result[pos++] = c; - break; - } - } - } - - assert(pos == s.size() + space); - o->write_characters(result.c_str(), result.size()); - } - - /*! - @brief dump an integer - - Dump a given integer to output stream @a o. Works internally with - @a number_buffer. - - @param[in] x integer number (signed or unsigned) to dump - @tparam NumberType either @a number_integer_t or @a number_unsigned_t - */ - template::value or - std::is_same::value, int> = 0> - void dump_integer(NumberType x) - { - // special case for "0" - if (x == 0) - { - o->write_character('0'); - return; - } - - const bool is_negative = x < 0; - size_t i = 0; - - // spare 1 byte for '\0' - while (x != 0 and i < number_buffer.size() - 1) - { - const auto digit = std::labs(static_cast(x % 10)); - number_buffer[i++] = static_cast('0' + digit); - x /= 10; - } - - // make sure the number has been processed completely - assert(x == 0); - - if (is_negative) - { - // make sure there is capacity for the '-' - assert(i < number_buffer.size() - 2); - number_buffer[i++] = '-'; - } - - std::reverse(number_buffer.begin(), number_buffer.begin() + i); - o->write_characters(number_buffer.data(), i); - } - - /*! - @brief dump a floating-point number - - Dump a given floating-point number to output stream @a o. Works - internally with @a number_buffer. - - @param[in] x floating-point number to dump - */ - void dump_float(number_float_t x) - { - // NaN / inf - if (not std::isfinite(x) or std::isnan(x)) - { - o->write_characters("null", 4); - return; - } - - // special case for 0.0 and -0.0 - if (x == 0) - { - if (std::signbit(x)) - { - o->write_characters("-0.0", 4); - } - else - { - o->write_characters("0.0", 3); - } - return; - } - - // get number of digits for a text -> float -> text round-trip - static constexpr auto d = std::numeric_limits::digits10; - - // the actual conversion - std::ptrdiff_t len = snprintf(number_buffer.data(), number_buffer.size(), - "%.*g", d, x); - - // negative value indicates an error - assert(len > 0); - // check if buffer was large enough - assert(static_cast(len) < number_buffer.size()); - - // erase thousands separator - if (thousands_sep != '\0') - { - const auto end = std::remove(number_buffer.begin(), - number_buffer.begin() + len, - thousands_sep); - std::fill(end, number_buffer.end(), '\0'); - assert((end - number_buffer.begin()) <= len); - len = (end - number_buffer.begin()); - } - - // convert decimal point to '.' - if (decimal_point != '\0' and decimal_point != '.') - { - for (auto& c : number_buffer) - { - if (c == decimal_point) - { - c = '.'; - break; - } - } - } - - o->write_characters(number_buffer.data(), static_cast(len)); - - // determine if need to append ".0" - const bool value_is_int_like = std::none_of(number_buffer.begin(), - number_buffer.begin() + len + 1, - [](char c) - { - return c == '.' or c == 'e'; - }); - - if (value_is_int_like) - { - o->write_characters(".0", 2); - } - } - - private: - /// the output of the serializer - output_adapter_t o = nullptr; - - /// a (hopefully) large enough character buffer - std::array number_buffer{{}}; - - /// the locale - const std::lconv* loc = nullptr; - /// the locale's thousand separator character - const char thousands_sep = '\0'; - /// the locale's decimal point character - const char decimal_point = '\0'; - - /// the indentation character - const char indent_char; - - /// the indentation string - string_t indent_string; - }; - - public: /*! @brief serialize to stream @@ -7321,7 +16622,7 @@ class basic_json `std::setw(4)` on @a o sets the indentation level to `4` and the serialization result is the same as calling `dump(4)`. - - The indentation characrer can be controlled with the member variable + - The indentation character can be controlled with the member variable `fill` of the output stream @a o. For instance, the manipulator `std::setfill('\\t')` sets indentation to use a tab character rather than the default space character. @@ -7331,12 +16632,15 @@ class basic_json @return the stream @a o + @throw type_error.316 if a string stored inside the JSON value is not + UTF-8 encoded + @complexity Linear. @liveexample{The example below shows the serialization with different parameters to `width` to adjust the indentation level.,operator_serialize} - @since version 1.0.0; indentaction character added in version 3.0.0 + @since version 1.0.0; indentation character added in version 3.0.0 */ friend std::ostream& operator<<(std::ostream& o, const basic_json& j) { @@ -7348,17 +16652,18 @@ class basic_json o.width(0); // do the actual serialization - serializer s(output_adapter::create(o), o.fill()); - s.dump(j, pretty_print, static_cast(indentation)); + serializer s(detail::output_adapter(o), o.fill()); + s.dump(j, pretty_print, false, static_cast(indentation)); return o; } /*! @brief serialize to stream - @deprecated This stream operator is deprecated and will be removed in a - future version of the library. Please use - @ref std::ostream& operator<<(std::ostream&, const basic_json&) + @deprecated This stream operator is deprecated and will be removed in + future 4.0.0 of the library. Please use + @ref operator<<(std::ostream&, const basic_json&) instead; that is, replace calls like `j >> o;` with `o << j;`. + @since version 1.0.0; deprecated since version 3.0.0 */ JSON_DEPRECATED friend std::ostream& operator>>(const basic_json& j, std::ostream& o) @@ -7377,15 +16682,36 @@ class basic_json /// @{ /*! - @brief deserialize from an array + @brief deserialize from a compatible input - This function reads from an array of 1-byte values. + This function reads from a compatible input. Examples are: + - an array of 1-byte values + - strings with character/literal type with size of 1 byte + - input streams + - container with contiguous storage of 1-byte values. Compatible container + types include `std::vector`, `std::string`, `std::array`, + `std::valarray`, and `std::initializer_list`. Furthermore, C-style + arrays can be used with `std::begin()`/`std::end()`. User-defined + containers can be used as long as they implement random-access iterators + and a contiguous storage. @pre Each element of the container has a size of 1 byte. Violating this precondition yields undefined behavior. **This precondition is enforced with a static assertion.** - @param[in] array array to read from + @pre The container storage is contiguous. Violating this precondition + yields undefined behavior. **This precondition is enforced with an + assertion.** + @pre Each element of the container has a size of 1 byte. Violating this + precondition yields undefined behavior. **This precondition is enforced + with a static assertion.** + + @warning There is no way to enforce all preconditions at compile-time. If + the function is called with a noncompliant container and with + assertions switched off, the behavior is undefined and will most + likely yield segmentation violation. + + @param[in] i input to read from @param[in] cb a parser callback function of type @ref parser_callback_t which is used to control the deserialization by filtering unwanted values (optional) @@ -7406,99 +16732,43 @@ class basic_json @liveexample{The example below demonstrates the `parse()` function reading from an array.,parse__array__parser_callback_t} - @since version 2.0.3 - */ - template - static basic_json parse(T (&array)[N], - const parser_callback_t cb = nullptr) - { - // delegate the call to the iterator-range parse overload - return parse(std::begin(array), std::end(array), cb); - } - - /*! - @brief deserialize from string literal - - @tparam CharT character/literal type with size of 1 byte - @param[in] s string literal to read a serialized JSON value from - @param[in] cb a parser callback function of type @ref parser_callback_t - which is used to control the deserialization by filtering unwanted values - (optional) - - @return result of the deserialization - - @throw parse_error.101 in case of an unexpected token - @throw parse_error.102 if to_unicode fails or surrogate error - @throw parse_error.103 if to_unicode fails - - @complexity Linear in the length of the input. The parser is a predictive - LL(1) parser. The complexity can be higher if the parser callback function - @a cb has a super-linear complexity. - - @note A UTF-8 byte order mark is silently ignored. - @note String containers like `std::string` or @ref string_t can be parsed - with @ref parse(const ContiguousContainer&, const parser_callback_t) - @liveexample{The example below demonstrates the `parse()` function with and without callback function.,parse__string__parser_callback_t} - @sa @ref parse(std::istream&, const parser_callback_t) for a version that - reads from an input stream - - @since version 1.0.0 (originally for @ref string_t) - */ - template::value and - std::is_integral::type>::value and - sizeof(typename std::remove_pointer::type) == 1, int>::type = 0> - static basic_json parse(const CharT s, - const parser_callback_t cb = nullptr) - { - return parser(input_adapter::create(s), cb).parse(true); - } - - /*! - @brief deserialize from stream - - @param[in,out] i stream to read a serialized JSON value from - @param[in] cb a parser callback function of type @ref parser_callback_t - which is used to control the deserialization by filtering unwanted values - (optional) - - @return result of the deserialization - - @throw parse_error.101 in case of an unexpected token - @throw parse_error.102 if to_unicode fails or surrogate error - @throw parse_error.103 if to_unicode fails - @throw parse_error.111 if input stream is in a bad state - - @complexity Linear in the length of the input. The parser is a predictive - LL(1) parser. The complexity can be higher if the parser callback function - @a cb has a super-linear complexity. - - @note A UTF-8 byte order mark is silently ignored. - @liveexample{The example below demonstrates the `parse()` function with and without callback function.,parse__istream__parser_callback_t} - @sa @ref parse(const CharT, const parser_callback_t) for a version - that reads from a string + @liveexample{The example below demonstrates the `parse()` function reading + from a contiguous container.,parse__contiguouscontainer__parser_callback_t} - @since version 1.0.0 + @since version 2.0.3 (contiguous containers) */ - static basic_json parse(std::istream& i, - const parser_callback_t cb = nullptr) + static basic_json parse(detail::input_adapter&& i, + const parser_callback_t cb = nullptr, + const bool allow_exceptions = true) { - return parser(input_adapter::create(i), cb).parse(true); + basic_json result; + parser(i, cb, allow_exceptions).parse(true, result); + return result; } - /*! - @copydoc parse(std::istream&, const parser_callback_t) - */ - static basic_json parse(std::istream&& i, - const parser_callback_t cb = nullptr) + static bool accept(detail::input_adapter&& i) { - return parser(input_adapter::create(i), cb).parse(true); + return parser(i).accept(true); + } + + static bool sax_parse(detail::input_adapter&& i, json_sax_t* sax, + input_format_t format = input_format_t::json, + const bool strict = true) + { + assert(sax); + switch (format) + { + case input_format_t::json: + return parser(std::move(i)).sax_parse(sax, strict); + default: + return binary_reader(std::move(i)).sax_parse(format, sax, strict); + } } /*! @@ -7528,6 +16798,8 @@ class basic_json @param[in] cb a parser callback function of type @ref parser_callback_t which is used to control the deserialization by filtering unwanted values (optional) + @param[in] allow_exceptions whether to throw exceptions in case of a + parse error (optional, true by default) @return result of the deserialization @@ -7551,80 +16823,44 @@ class basic_json std::random_access_iterator_tag, typename std::iterator_traits::iterator_category>::value, int>::type = 0> static basic_json parse(IteratorType first, IteratorType last, - const parser_callback_t cb = nullptr) + const parser_callback_t cb = nullptr, + const bool allow_exceptions = true) { - return parser(input_adapter::create(first, last), cb).parse(true); + basic_json result; + parser(detail::input_adapter(first, last), cb, allow_exceptions).parse(true, result); + return result; } - /*! - @brief deserialize from a container with contiguous storage - - This function reads from a container with contiguous storage of 1-byte - values. Compatible container types include `std::vector`, `std::string`, - `std::array`, and `std::initializer_list`. User-defined containers can be - used as long as they implement random-access iterators and a contiguous - storage. - - @pre The container storage is contiguous. Violating this precondition - yields undefined behavior. **This precondition is enforced with an - assertion.** - @pre Each element of the container has a size of 1 byte. Violating this - precondition yields undefined behavior. **This precondition is enforced - with a static assertion.** - - @warning There is no way to enforce all preconditions at compile-time. If - the function is called with a noncompliant container and with - assertions switched off, the behavior is undefined and will most - likely yield segmentation violation. - - @tparam ContiguousContainer container type with contiguous storage - @param[in] c container to read from - @param[in] cb a parser callback function of type @ref parser_callback_t - which is used to control the deserialization by filtering unwanted values - (optional) - - @return result of the deserialization - - @throw parse_error.101 in case of an unexpected token - @throw parse_error.102 if to_unicode fails or surrogate error - @throw parse_error.103 if to_unicode fails - - @complexity Linear in the length of the input. The parser is a predictive - LL(1) parser. The complexity can be higher if the parser callback function - @a cb has a super-linear complexity. - - @note A UTF-8 byte order mark is silently ignored. - - @liveexample{The example below demonstrates the `parse()` function reading - from a contiguous container.,parse__contiguouscontainer__parser_callback_t} - - @since version 2.0.3 - */ - template::value and + template()))>::iterator_category>::value - , int>::type = 0> - static basic_json parse(const ContiguousContainer& c, - const parser_callback_t cb = nullptr) + typename std::iterator_traits::iterator_category>::value, int>::type = 0> + static bool accept(IteratorType first, IteratorType last) { - // delegate the call to the iterator-range parse overload - return parse(std::begin(c), std::end(c), cb); + return parser(detail::input_adapter(first, last)).accept(true); + } + + template::iterator_category>::value, int>::type = 0> + static bool sax_parse(IteratorType first, IteratorType last, json_sax_t* sax) + { + return parser(detail::input_adapter(first, last)).sax_parse(sax); } /*! @brief deserialize from stream - @deprecated This stream operator is deprecated and will be removed in a - future version of the library. Please use - @ref std::istream& operator>>(std::istream&, basic_json&) + @deprecated This stream operator is deprecated and will be removed in + version 4.0.0 of the library. Please use + @ref operator>>(std::istream&, basic_json&) instead; that is, replace calls like `j << i;` with `i >> j;`. + @since version 1.0.0; deprecated since version 3.0.0 */ JSON_DEPRECATED friend std::istream& operator<<(basic_json& j, std::istream& i) { - j = parser(input_adapter::create(i)).parse(false); - return i; + return operator>>(i, j); } /*! @@ -7638,7 +16874,6 @@ class basic_json @throw parse_error.101 in case of an unexpected token @throw parse_error.102 if to_unicode fails or surrogate error @throw parse_error.103 if to_unicode fails - @throw parse_error.111 if input stream is in a bad state @complexity Linear in the length of the input. The parser is a predictive LL(1) parser. @@ -7655,7 +16890,7 @@ class basic_json */ friend std::istream& operator>>(std::istream& i, basic_json& j) { - j = parser(input_adapter::create(i)).parse(false); + parser(detail::input_adapter(i)).parse(false, j); return i; } @@ -7671,16 +16906,31 @@ class basic_json Returns the type name as string to be used in error messages - usually to indicate that a function was called on a wrong JSON type. - @return basically a string representation of a the @a m_type member + @return a string representation of a the @a m_type member: + Value type | return value + ----------- | ------------- + null | `"null"` + boolean | `"boolean"` + string | `"string"` + number | `"number"` (for all number types) + object | `"object"` + array | `"array"` + discarded | `"discarded"` + + @exceptionsafety No-throw guarantee: this function never throws exceptions. @complexity Constant. @liveexample{The following code exemplifies `type_name()` for all JSON types.,type_name} - @since version 1.0.0, public since 2.1.0 + @sa @ref type() -- return the type of the JSON value + @sa @ref operator value_t() -- return the type of the JSON value (implicit) + + @since version 1.0.0, public since 2.1.0, `const char*` and `noexcept` + since 3.0.0 */ - std::string type_name() const + const char* type_name() const noexcept { { switch (m_type) @@ -7715,1245 +16965,6 @@ class basic_json /// the value of the current element json_value m_value = {}; - - private: - /////////////// - // iterators // - /////////////// - - /*! - @brief an iterator for primitive JSON types - - This class models an iterator for primitive JSON types (boolean, number, - string). It's only purpose is to allow the iterator/const_iterator classes - to "iterate" over primitive values. Internally, the iterator is modeled by - a `difference_type` variable. Value begin_value (`0`) models the begin, - end_value (`1`) models past the end. - */ - class primitive_iterator_t - { - public: - - difference_type get_value() const noexcept - { - return m_it; - } - /// set iterator to a defined beginning - void set_begin() noexcept - { - m_it = begin_value; - } - - /// set iterator to a defined past the end - void set_end() noexcept - { - m_it = end_value; - } - - /// return whether the iterator can be dereferenced - constexpr bool is_begin() const noexcept - { - return (m_it == begin_value); - } - - /// return whether the iterator is at end - constexpr bool is_end() const noexcept - { - return (m_it == end_value); - } - - friend constexpr bool operator==(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept - { - return lhs.m_it == rhs.m_it; - } - - friend constexpr bool operator!=(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept - { - return !(lhs == rhs); - } - - friend constexpr bool operator<(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept - { - return lhs.m_it < rhs.m_it; - } - - friend constexpr bool operator<=(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept - { - return lhs.m_it <= rhs.m_it; - } - - friend constexpr bool operator>(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept - { - return lhs.m_it > rhs.m_it; - } - - friend constexpr bool operator>=(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept - { - return lhs.m_it >= rhs.m_it; - } - - primitive_iterator_t operator+(difference_type i) - { - auto result = *this; - result += i; - return result; - } - - friend constexpr difference_type operator-(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept - { - return lhs.m_it - rhs.m_it; - } - - friend std::ostream& operator<<(std::ostream& os, primitive_iterator_t it) - { - return os << it.m_it; - } - - primitive_iterator_t& operator++() - { - ++m_it; - return *this; - } - - primitive_iterator_t operator++(int) - { - auto result = *this; - m_it++; - return result; - } - - primitive_iterator_t& operator--() - { - --m_it; - return *this; - } - - primitive_iterator_t operator--(int) - { - auto result = *this; - m_it--; - return result; - } - - primitive_iterator_t& operator+=(difference_type n) - { - m_it += n; - return *this; - } - - primitive_iterator_t& operator-=(difference_type n) - { - m_it -= n; - return *this; - } - - private: - static constexpr difference_type begin_value = 0; - static constexpr difference_type end_value = begin_value + 1; - - /// iterator as signed integer type - difference_type m_it = std::numeric_limits::denorm_min(); - }; - - /*! - @brief an iterator value - - @note This structure could easily be a union, but MSVC currently does not - allow unions members with complex constructors, see - https://github.com/nlohmann/json/pull/105. - */ - struct internal_iterator - { - /// iterator for JSON objects - typename object_t::iterator object_iterator; - /// iterator for JSON arrays - typename array_t::iterator array_iterator; - /// generic iterator for all other types - primitive_iterator_t primitive_iterator; - - /// create an uninitialized internal_iterator - internal_iterator() noexcept - : object_iterator(), array_iterator(), primitive_iterator() - {} - }; - - /// proxy class for the iterator_wrapper functions - template - class iteration_proxy - { - private: - /// helper class for iteration - class iteration_proxy_internal - { - private: - /// the iterator - IteratorType anchor; - /// an index for arrays (used to create key names) - size_t array_index = 0; - - public: - explicit iteration_proxy_internal(IteratorType it) noexcept - : anchor(it) - {} - - /// dereference operator (needed for range-based for) - iteration_proxy_internal& operator*() - { - return *this; - } - - /// increment operator (needed for range-based for) - iteration_proxy_internal& operator++() - { - ++anchor; - ++array_index; - - return *this; - } - - /// inequality operator (needed for range-based for) - bool operator!= (const iteration_proxy_internal& o) const - { - return anchor != o.anchor; - } - - /// return key of the iterator - typename basic_json::string_t key() const - { - assert(anchor.m_object != nullptr); - - switch (anchor.m_object->type()) - { - // use integer array index as key - case value_t::array: - { - return std::to_string(array_index); - } - - // use key from the object - case value_t::object: - { - return anchor.key(); - } - - // use an empty key for all primitive types - default: - { - return ""; - } - } - } - - /// return value of the iterator - typename IteratorType::reference value() const - { - return anchor.value(); - } - }; - - /// the container to iterate - typename IteratorType::reference container; - - public: - /// construct iteration proxy from a container - explicit iteration_proxy(typename IteratorType::reference cont) - : container(cont) - {} - - /// return iterator begin (needed for range-based for) - iteration_proxy_internal begin() noexcept - { - return iteration_proxy_internal(container.begin()); - } - - /// return iterator end (needed for range-based for) - iteration_proxy_internal end() noexcept - { - return iteration_proxy_internal(container.end()); - } - }; - - public: - /*! - @brief a template for a random access iterator for the @ref basic_json class - - This class implements a both iterators (iterator and const_iterator) for the - @ref basic_json class. - - @note An iterator is called *initialized* when a pointer to a JSON value - has been set (e.g., by a constructor or a copy assignment). If the - iterator is default-constructed, it is *uninitialized* and most - methods are undefined. **The library uses assertions to detect calls - on uninitialized iterators.** - - @requirement The class satisfies the following concept requirements: - - [RandomAccessIterator](http://en.cppreference.com/w/cpp/concept/RandomAccessIterator): - The iterator that can be moved to point (forward and backward) to any - element in constant time. - - @since version 1.0.0, simplified in version 2.0.9 - */ - template - class iter_impl : public std::iterator - { - /// allow basic_json to access private members - friend class basic_json; - - // make sure U is basic_json or const basic_json - static_assert(std::is_same::value - or std::is_same::value, - "iter_impl only accepts (const) basic_json"); - - public: - /// the type of the values when the iterator is dereferenced - using value_type = typename basic_json::value_type; - /// a type to represent differences between iterators - using difference_type = typename basic_json::difference_type; - /// defines a pointer to the type iterated over (value_type) - using pointer = typename std::conditional::value, - typename basic_json::const_pointer, - typename basic_json::pointer>::type; - /// defines a reference to the type iterated over (value_type) - using reference = typename std::conditional::value, - typename basic_json::const_reference, - typename basic_json::reference>::type; - /// the category of the iterator - using iterator_category = std::bidirectional_iterator_tag; - - /// default constructor - iter_impl() = default; - - /*! - @brief constructor for a given JSON instance - @param[in] object pointer to a JSON object for this iterator - @pre object != nullptr - @post The iterator is initialized; i.e. `m_object != nullptr`. - */ - explicit iter_impl(pointer object) noexcept - : m_object(object) - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - m_it.object_iterator = typename object_t::iterator(); - break; - } - - case basic_json::value_t::array: - { - m_it.array_iterator = typename array_t::iterator(); - break; - } - - default: - { - m_it.primitive_iterator = primitive_iterator_t(); - break; - } - } - } - - /* - Use operator `const_iterator` instead of `const_iterator(const iterator& - other) noexcept` to avoid two class definitions for @ref iterator and - @ref const_iterator. - - This function is only called if this class is an @ref iterator. If this - class is a @ref const_iterator this function is not called. - */ - operator const_iterator() const - { - const_iterator ret; - - if (m_object) - { - ret.m_object = m_object; - ret.m_it = m_it; - } - - return ret; - } - - /*! - @brief copy constructor - @param[in] other iterator to copy from - @note It is not checked whether @a other is initialized. - */ - iter_impl(const iter_impl& other) noexcept - : m_object(other.m_object), m_it(other.m_it) - {} - - /*! - @brief copy assignment - @param[in,out] other iterator to copy from - @note It is not checked whether @a other is initialized. - */ - iter_impl& operator=(iter_impl other) noexcept( - std::is_nothrow_move_constructible::value and - std::is_nothrow_move_assignable::value and - std::is_nothrow_move_constructible::value and - std::is_nothrow_move_assignable::value - ) - { - std::swap(m_object, other.m_object); - std::swap(m_it, other.m_it); - return *this; - } - - private: - /*! - @brief set the iterator to the first value - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - void set_begin() noexcept - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - m_it.object_iterator = m_object->m_value.object->begin(); - break; - } - - case basic_json::value_t::array: - { - m_it.array_iterator = m_object->m_value.array->begin(); - break; - } - - case basic_json::value_t::null: - { - // set to end so begin()==end() is true: null is empty - m_it.primitive_iterator.set_end(); - break; - } - - default: - { - m_it.primitive_iterator.set_begin(); - break; - } - } - } - - /*! - @brief set the iterator past the last value - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - void set_end() noexcept - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - m_it.object_iterator = m_object->m_value.object->end(); - break; - } - - case basic_json::value_t::array: - { - m_it.array_iterator = m_object->m_value.array->end(); - break; - } - - default: - { - m_it.primitive_iterator.set_end(); - break; - } - } - } - - public: - /*! - @brief return a reference to the value pointed to by the iterator - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - reference operator*() const - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - assert(m_it.object_iterator != m_object->m_value.object->end()); - return m_it.object_iterator->second; - } - - case basic_json::value_t::array: - { - assert(m_it.array_iterator != m_object->m_value.array->end()); - return *m_it.array_iterator; - } - - case basic_json::value_t::null: - { - JSON_THROW(invalid_iterator::create(214, "cannot get value")); - } - - default: - { - if (m_it.primitive_iterator.is_begin()) - { - return *m_object; - } - - JSON_THROW(invalid_iterator::create(214, "cannot get value")); - } - } - } - - /*! - @brief dereference the iterator - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - pointer operator->() const - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - assert(m_it.object_iterator != m_object->m_value.object->end()); - return &(m_it.object_iterator->second); - } - - case basic_json::value_t::array: - { - assert(m_it.array_iterator != m_object->m_value.array->end()); - return &*m_it.array_iterator; - } - - default: - { - if (m_it.primitive_iterator.is_begin()) - { - return m_object; - } - - JSON_THROW(invalid_iterator::create(214, "cannot get value")); - } - } - } - - /*! - @brief post-increment (it++) - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - iter_impl operator++(int) - { - auto result = *this; - ++(*this); - return result; - } - - /*! - @brief pre-increment (++it) - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - iter_impl& operator++() - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - std::advance(m_it.object_iterator, 1); - break; - } - - case basic_json::value_t::array: - { - std::advance(m_it.array_iterator, 1); - break; - } - - default: - { - ++m_it.primitive_iterator; - break; - } - } - - return *this; - } - - /*! - @brief post-decrement (it--) - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - iter_impl operator--(int) - { - auto result = *this; - --(*this); - return result; - } - - /*! - @brief pre-decrement (--it) - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - iter_impl& operator--() - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - std::advance(m_it.object_iterator, -1); - break; - } - - case basic_json::value_t::array: - { - std::advance(m_it.array_iterator, -1); - break; - } - - default: - { - --m_it.primitive_iterator; - break; - } - } - - return *this; - } - - /*! - @brief comparison: equal - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - bool operator==(const iter_impl& other) const - { - // if objects are not the same, the comparison is undefined - if (m_object != other.m_object) - { - JSON_THROW(invalid_iterator::create(212, "cannot compare iterators of different containers")); - } - - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - return (m_it.object_iterator == other.m_it.object_iterator); - } - - case basic_json::value_t::array: - { - return (m_it.array_iterator == other.m_it.array_iterator); - } - - default: - { - return (m_it.primitive_iterator == other.m_it.primitive_iterator); - } - } - } - - /*! - @brief comparison: not equal - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - bool operator!=(const iter_impl& other) const - { - return not operator==(other); - } - - /*! - @brief comparison: smaller - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - bool operator<(const iter_impl& other) const - { - // if objects are not the same, the comparison is undefined - if (m_object != other.m_object) - { - JSON_THROW(invalid_iterator::create(212, "cannot compare iterators of different containers")); - } - - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - JSON_THROW(invalid_iterator::create(213, "cannot compare order of object iterators")); - } - - case basic_json::value_t::array: - { - return (m_it.array_iterator < other.m_it.array_iterator); - } - - default: - { - return (m_it.primitive_iterator < other.m_it.primitive_iterator); - } - } - } - - /*! - @brief comparison: less than or equal - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - bool operator<=(const iter_impl& other) const - { - return not other.operator < (*this); - } - - /*! - @brief comparison: greater than - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - bool operator>(const iter_impl& other) const - { - return not operator<=(other); - } - - /*! - @brief comparison: greater than or equal - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - bool operator>=(const iter_impl& other) const - { - return not operator<(other); - } - - /*! - @brief add to iterator - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - iter_impl& operator+=(difference_type i) - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - JSON_THROW(invalid_iterator::create(209, "cannot use offsets with object iterators")); - } - - case basic_json::value_t::array: - { - std::advance(m_it.array_iterator, i); - break; - } - - default: - { - m_it.primitive_iterator += i; - break; - } - } - - return *this; - } - - /*! - @brief subtract from iterator - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - iter_impl& operator-=(difference_type i) - { - return operator+=(-i); - } - - /*! - @brief add to iterator - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - iter_impl operator+(difference_type i) - { - auto result = *this; - result += i; - return result; - } - - /*! - @brief subtract from iterator - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - iter_impl operator-(difference_type i) - { - auto result = *this; - result -= i; - return result; - } - - /*! - @brief return difference - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - difference_type operator-(const iter_impl& other) const - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - JSON_THROW(invalid_iterator::create(209, "cannot use offsets with object iterators")); - } - - case basic_json::value_t::array: - { - return m_it.array_iterator - other.m_it.array_iterator; - } - - default: - { - return m_it.primitive_iterator - other.m_it.primitive_iterator; - } - } - } - - /*! - @brief access to successor - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - reference operator[](difference_type n) const - { - assert(m_object != nullptr); - - switch (m_object->m_type) - { - case basic_json::value_t::object: - { - JSON_THROW(invalid_iterator::create(208, "cannot use operator[] for object iterators")); - } - - case basic_json::value_t::array: - { - return *std::next(m_it.array_iterator, n); - } - - case basic_json::value_t::null: - { - JSON_THROW(invalid_iterator::create(214, "cannot get value")); - } - - default: - { - if (m_it.primitive_iterator.get_value() == -n) - { - return *m_object; - } - - JSON_THROW(invalid_iterator::create(214, "cannot get value")); - } - } - } - - /*! - @brief return the key of an object iterator - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - typename object_t::key_type key() const - { - assert(m_object != nullptr); - - if (m_object->is_object()) - { - return m_it.object_iterator->first; - } - - JSON_THROW(invalid_iterator::create(207, "cannot use key() for non-object iterators")); - } - - /*! - @brief return the value of an iterator - @pre The iterator is initialized; i.e. `m_object != nullptr`. - */ - reference value() const - { - return operator*(); - } - - private: - /// associated JSON instance - pointer m_object = nullptr; - /// the actual iterator of the associated instance - internal_iterator m_it = internal_iterator(); - }; - - /*! - @brief a template for a reverse iterator class - - @tparam Base the base iterator type to reverse. Valid types are @ref - iterator (to create @ref reverse_iterator) and @ref const_iterator (to - create @ref const_reverse_iterator). - - @requirement The class satisfies the following concept requirements: - - [RandomAccessIterator](http://en.cppreference.com/w/cpp/concept/RandomAccessIterator): - The iterator that can be moved to point (forward and backward) to any - element in constant time. - - [OutputIterator](http://en.cppreference.com/w/cpp/concept/OutputIterator): - It is possible to write to the pointed-to element (only if @a Base is - @ref iterator). - - @since version 1.0.0 - */ - template - class json_reverse_iterator : public std::reverse_iterator - { - public: - /// shortcut to the reverse iterator adaptor - using base_iterator = std::reverse_iterator; - /// the reference type for the pointed-to element - using reference = typename Base::reference; - - /// create reverse iterator from iterator - json_reverse_iterator(const typename base_iterator::iterator_type& it) noexcept - : base_iterator(it) - {} - - /// create reverse iterator from base class - json_reverse_iterator(const base_iterator& it) noexcept - : base_iterator(it) - {} - - /// post-increment (it++) - json_reverse_iterator operator++(int) - { - return base_iterator::operator++(1); - } - - /// pre-increment (++it) - json_reverse_iterator& operator++() - { - base_iterator::operator++(); - return *this; - } - - /// post-decrement (it--) - json_reverse_iterator operator--(int) - { - return base_iterator::operator--(1); - } - - /// pre-decrement (--it) - json_reverse_iterator& operator--() - { - base_iterator::operator--(); - return *this; - } - - /// add to iterator - json_reverse_iterator& operator+=(difference_type i) - { - base_iterator::operator+=(i); - return *this; - } - - /// add to iterator - json_reverse_iterator operator+(difference_type i) const - { - auto result = *this; - result += i; - return result; - } - - /// subtract from iterator - json_reverse_iterator operator-(difference_type i) const - { - auto result = *this; - result -= i; - return result; - } - - /// return difference - difference_type operator-(const json_reverse_iterator& other) const - { - return this->base() - other.base(); - } - - /// access to successor - reference operator[](difference_type n) const - { - return *(this->operator+(n)); - } - - /// return the key of an object iterator - typename object_t::key_type key() const - { - auto it = --this->base(); - return it.key(); - } - - /// return the value of an iterator - reference value() const - { - auto it = --this->base(); - return it.operator * (); - } - }; - - - private: - //////////////////// - // input adapters // - //////////////////// - - /// abstract input adapter interface - class input_adapter - { - public: - virtual int get_character() = 0; - virtual std::string read(size_t offset, size_t length) = 0; - virtual ~input_adapter() {} - - // native support - - /// input adapter for input stream - static std::shared_ptr create(std::istream& i, const size_t buffer_size = 16384) - { - return std::shared_ptr(new cached_input_stream_adapter(i, buffer_size)); - } - - /// input adapter for input stream - static std::shared_ptr create(std::istream&& i, const size_t buffer_size = 16384) - { - return std::shared_ptr(new cached_input_stream_adapter(i, buffer_size)); - } - - /// input adapter for buffer - static std::shared_ptr create(const char* b, size_t l) - { - return std::shared_ptr(new input_buffer_adapter(b, l)); - } - - // derived support - - /// input adapter for string literal - template::value and - std::is_integral::type>::value and - sizeof(typename std::remove_pointer::type) == 1, int>::type = 0> - static std::shared_ptr create(CharT b) - { - return create(reinterpret_cast(b), - std::strlen(reinterpret_cast(b))); - } - - /// input adapter for iterator range with contiguous storage - template::iterator_category, std::random_access_iterator_tag>::value - , int>::type - = 0> - static std::shared_ptr create(IteratorType first, IteratorType last) - { - // assertion to check that the iterator range is indeed contiguous, - // see http://stackoverflow.com/a/35008842/266378 for more discussion - assert(std::accumulate(first, last, std::pair(true, 0), - [&first](std::pair res, decltype(*first) val) - { - res.first &= (val == *(std::next(std::addressof(*first), res.second++))); - return res; - }).first); - - // assertion to check that each element is 1 byte long - static_assert(sizeof(typename std::iterator_traits::value_type) == 1, - "each element in the iterator range must have the size of 1 byte"); - - return create(reinterpret_cast(&(*first)), - static_cast(std::distance(first, last))); - } - - /// input adapter for array - template - static std::shared_ptr create(T (&array)[N]) - { - // delegate the call to the iterator-range overload - return create(std::begin(array), std::end(array)); - } - - /// input adapter for contiguous container - template::value and - std::is_base_of< - std::random_access_iterator_tag, - typename std::iterator_traits()))>::iterator_category>::value - , int>::type = 0> - static std::shared_ptr create(const ContiguousContainer& c) - { - // delegate the call to the iterator-range overload - return create(std::begin(c), std::end(c)); - } - }; - - /// a type to simplify interfaces - using input_adapter_t = std::shared_ptr; - - /// input adapter for cached stream input - class cached_input_stream_adapter : public input_adapter - { - public: - cached_input_stream_adapter(std::istream& i, const size_t buffer_size) - : is(i), start_position(is.tellg()), buffer(buffer_size, '\0') - { - // immediately abort if stream is erroneous - if (JSON_UNLIKELY(i.fail())) - { - JSON_THROW(parse_error::create(111, 0, "bad input stream")); - } - - // initial fill - is.read(buffer.data(), static_cast(buffer.size())); - // store number of bytes in the buffer - fill_size = static_cast(is.gcount()); - - // skip byte order mark - if (fill_size >= 3 and buffer[0] == '\xEF' and buffer[1] == '\xBB' and buffer[2] == '\xBF') - { - buffer_pos += 3; - processed_chars += 3; - } - } - - ~cached_input_stream_adapter() override - { - // clear stream flags - is.clear(); - // We initially read a lot of characters into the buffer, and we - // may not have processed all of them. Therefore, we need to - // "rewind" the stream after the last processed char. - is.seekg(start_position + static_cast(processed_chars)); - // clear stream flags - is.clear(); - } - - int get_character() override - { - // check if refilling is necessary and possible - if (buffer_pos == fill_size and not eof) - { - // refill - is.read(buffer.data(), static_cast(buffer.size())); - // store number of bytes in the buffer - fill_size = static_cast(is.gcount()); - - // the buffer is ready - buffer_pos = 0; - - // remember that filling did not yield new input - if (fill_size == 0) - { - eof = true; - return std::char_traits::eof(); - } - } - - ++processed_chars; - return buffer[buffer_pos++] & 0xFF;; - } - - std::string read(size_t offset, size_t length) override - { - // create buffer - std::string result(length, '\0'); - - // save stream position - auto current_pos = is.tellg(); - // save stream flags - auto flags = is.rdstate(); - - // clear stream flags - is.clear(); - // set stream position - is.seekg(static_cast(offset)); - // read bytes - is.read(&result[0], static_cast(length)); - - // reset stream position - is.seekg(current_pos); - // reset stream flags - is.setstate(flags); - - return result; - } - - private: - /// the associated input stream - std::istream& is; - - /// chars returned via get_character() - size_t processed_chars = 0; - /// chars processed in the current buffer - size_t buffer_pos = 0; - - /// whether stream reached eof - bool eof = false; - /// how many chars have been copied to the buffer by last (re)fill - size_t fill_size = 0; - - /// position of the stream when we started - const std::streampos start_position; - - /// internal buffer - std::vector buffer; - }; - - /// input adapter for buffer input - class input_buffer_adapter : public input_adapter - { - public: - input_buffer_adapter(const char* b, size_t l) - : input_adapter(), cursor(b), limit(b + l), start(b) - { - // skip byte order mark - if (l >= 3 and b[0] == '\xEF' and b[1] == '\xBB' and b[2] == '\xBF') - { - cursor += 3; - } - } - - // delete because of pointer members - input_buffer_adapter(const input_buffer_adapter&) = delete; - input_buffer_adapter& operator=(input_buffer_adapter&) = delete; - - int get_character() override - { - if (JSON_LIKELY(cursor < limit)) - { - return *(cursor++) & 0xFF; - } - else - { - return std::char_traits::eof(); - } - } - - std::string read(size_t offset, size_t length) override - { - // avoid reading too many characters - const size_t max_length = static_cast(limit - start); - return std::string(start + offset, std::min(length, max_length - offset)); - } - - private: - /// pointer to the current character - const char* cursor; - /// pointer past the last character - const char* limit; - /// pointer to the first character - const char* start; - }; - ////////////////////////////////////////// // binary serialization/deserialization // ////////////////////////////////////////// @@ -8961,1678 +16972,6 @@ class basic_json /// @name binary serialization/deserialization support /// @{ - private: - /*! - @brief deserialization of CBOR and MessagePack values - */ - class binary_reader - { - public: - /*! - @brief create a binary reader - - @param[in] adapter input adapter to read from - */ - explicit binary_reader(input_adapter_t adapter) - : ia(adapter), is_little_endian(little_endianess()) - { - assert(ia); - } - - /*! - @brief create a JSON value from CBOR input - - @param[in] get_char whether a new character should be retrieved from - the input (true, default) or whether the last - read character should be considered instead - - @return JSON value created from CBOR input - - @throw parse_error.110 if input ended unexpectedly - @throw parse_error.112 if unsupported byte was read - */ - basic_json parse_cbor(const bool get_char = true) - { - switch (get_char ? get() : current) - { - // EOF - case std::char_traits::eof(): - { - JSON_THROW(parse_error::create(110, chars_read, "unexpected end of input")); - } - - // Integer 0x00..0x17 (0..23) - case 0x00: - case 0x01: - case 0x02: - case 0x03: - case 0x04: - case 0x05: - case 0x06: - case 0x07: - case 0x08: - case 0x09: - case 0x0a: - case 0x0b: - case 0x0c: - case 0x0d: - case 0x0e: - case 0x0f: - case 0x10: - case 0x11: - case 0x12: - case 0x13: - case 0x14: - case 0x15: - case 0x16: - case 0x17: - { - return static_cast(current); - } - - case 0x18: // Unsigned integer (one-byte uint8_t follows) - { - return get_number(); - } - - case 0x19: // Unsigned integer (two-byte uint16_t follows) - { - return get_number(); - } - - case 0x1a: // Unsigned integer (four-byte uint32_t follows) - { - return get_number(); - } - - case 0x1b: // Unsigned integer (eight-byte uint64_t follows) - { - return get_number(); - } - - // Negative integer -1-0x00..-1-0x17 (-1..-24) - case 0x20: - case 0x21: - case 0x22: - case 0x23: - case 0x24: - case 0x25: - case 0x26: - case 0x27: - case 0x28: - case 0x29: - case 0x2a: - case 0x2b: - case 0x2c: - case 0x2d: - case 0x2e: - case 0x2f: - case 0x30: - case 0x31: - case 0x32: - case 0x33: - case 0x34: - case 0x35: - case 0x36: - case 0x37: - { - return static_cast(0x20 - 1 - current); - } - - case 0x38: // Negative integer (one-byte uint8_t follows) - { - // must be uint8_t ! - return static_cast(-1) - get_number(); - } - - case 0x39: // Negative integer -1-n (two-byte uint16_t follows) - { - return static_cast(-1) - get_number(); - } - - case 0x3a: // Negative integer -1-n (four-byte uint32_t follows) - { - return static_cast(-1) - get_number(); - } - - case 0x3b: // Negative integer -1-n (eight-byte uint64_t follows) - { - return static_cast(-1) - static_cast(get_number()); - } - - // UTF-8 string (0x00..0x17 bytes follow) - case 0x60: - case 0x61: - case 0x62: - case 0x63: - case 0x64: - case 0x65: - case 0x66: - case 0x67: - case 0x68: - case 0x69: - case 0x6a: - case 0x6b: - case 0x6c: - case 0x6d: - case 0x6e: - case 0x6f: - case 0x70: - case 0x71: - case 0x72: - case 0x73: - case 0x74: - case 0x75: - case 0x76: - case 0x77: - case 0x78: // UTF-8 string (one-byte uint8_t for n follows) - case 0x79: // UTF-8 string (two-byte uint16_t for n follow) - case 0x7a: // UTF-8 string (four-byte uint32_t for n follow) - case 0x7b: // UTF-8 string (eight-byte uint64_t for n follow) - case 0x7f: // UTF-8 string (indefinite length) - { - return get_cbor_string(); - } - - // array (0x00..0x17 data items follow) - case 0x80: - case 0x81: - case 0x82: - case 0x83: - case 0x84: - case 0x85: - case 0x86: - case 0x87: - case 0x88: - case 0x89: - case 0x8a: - case 0x8b: - case 0x8c: - case 0x8d: - case 0x8e: - case 0x8f: - case 0x90: - case 0x91: - case 0x92: - case 0x93: - case 0x94: - case 0x95: - case 0x96: - case 0x97: - { - basic_json result = value_t::array; - const auto len = static_cast(current & 0x1f); - for (size_t i = 0; i < len; ++i) - { - result.push_back(parse_cbor()); - } - return result; - } - - case 0x98: // array (one-byte uint8_t for n follows) - { - basic_json result = value_t::array; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - result.push_back(parse_cbor()); - } - return result; - } - - case 0x99: // array (two-byte uint16_t for n follow) - { - basic_json result = value_t::array; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - result.push_back(parse_cbor()); - } - return result; - } - - case 0x9a: // array (four-byte uint32_t for n follow) - { - basic_json result = value_t::array; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - result.push_back(parse_cbor()); - } - return result; - } - - case 0x9b: // array (eight-byte uint64_t for n follow) - { - basic_json result = value_t::array; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - result.push_back(parse_cbor()); - } - return result; - } - - case 0x9f: // array (indefinite length) - { - basic_json result = value_t::array; - while (get() != 0xff) - { - result.push_back(parse_cbor(false)); - } - return result; - } - - // map (0x00..0x17 pairs of data items follow) - case 0xa0: - case 0xa1: - case 0xa2: - case 0xa3: - case 0xa4: - case 0xa5: - case 0xa6: - case 0xa7: - case 0xa8: - case 0xa9: - case 0xaa: - case 0xab: - case 0xac: - case 0xad: - case 0xae: - case 0xaf: - case 0xb0: - case 0xb1: - case 0xb2: - case 0xb3: - case 0xb4: - case 0xb5: - case 0xb6: - case 0xb7: - { - basic_json result = value_t::object; - const auto len = static_cast(current & 0x1f); - for (size_t i = 0; i < len; ++i) - { - get(); - auto key = get_cbor_string(); - result[key] = parse_cbor(); - } - return result; - } - - case 0xb8: // map (one-byte uint8_t for n follows) - { - basic_json result = value_t::object; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - get(); - auto key = get_cbor_string(); - result[key] = parse_cbor(); - } - return result; - } - - case 0xb9: // map (two-byte uint16_t for n follow) - { - basic_json result = value_t::object; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - get(); - auto key = get_cbor_string(); - result[key] = parse_cbor(); - } - return result; - } - - case 0xba: // map (four-byte uint32_t for n follow) - { - basic_json result = value_t::object; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - get(); - auto key = get_cbor_string(); - result[key] = parse_cbor(); - } - return result; - } - - case 0xbb: // map (eight-byte uint64_t for n follow) - { - basic_json result = value_t::object; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - get(); - auto key = get_cbor_string(); - result[key] = parse_cbor(); - } - return result; - } - - case 0xbf: // map (indefinite length) - { - basic_json result = value_t::object; - while (get() != 0xff) - { - auto key = get_cbor_string(); - result[key] = parse_cbor(); - } - return result; - } - - case 0xf4: // false - { - return false; - } - - case 0xf5: // true - { - return true; - } - - case 0xf6: // null - { - return value_t::null; - } - - case 0xf9: // Half-Precision Float (two-byte IEEE 754) - { - const int byte1 = get(); - check_eof(); - const int byte2 = get(); - check_eof(); - - // code from RFC 7049, Appendix D, Figure 3: - // As half-precision floating-point numbers were only added - // to IEEE 754 in 2008, today's programming platforms often - // still only have limited support for them. It is very - // easy to include at least decoding support for them even - // without such support. An example of a small decoder for - // half-precision floating-point numbers in the C language - // is shown in Fig. 3. - const int half = (byte1 << 8) + byte2; - const int exp = (half >> 10) & 0x1f; - const int mant = half & 0x3ff; - double val; - if (exp == 0) - { - val = std::ldexp(mant, -24); - } - else if (exp != 31) - { - val = std::ldexp(mant + 1024, exp - 25); - } - else - { - val = mant == 0 - ? std::numeric_limits::infinity() - : std::numeric_limits::quiet_NaN(); - } - return (half & 0x8000) != 0 ? -val : val; - } - - case 0xfa: // Single-Precision Float (four-byte IEEE 754) - { - return get_number(); - } - - case 0xfb: // Double-Precision Float (eight-byte IEEE 754) - { - return get_number(); - } - - default: // anything else (0xFF is handled inside the other types) - { - std::stringstream ss; - ss << std::setw(2) << std::setfill('0') << std::hex << current; - JSON_THROW(parse_error::create(112, chars_read, "error reading CBOR; last byte: 0x" + ss.str())); - } - } - } - - /*! - @brief create a JSON value from MessagePack input - - @return JSON value created from MessagePack input - - @throw parse_error.110 if input ended unexpectedly - @throw parse_error.112 if unsupported byte was read - */ - basic_json parse_msgpack() - { - switch (get()) - { - // EOF - case std::char_traits::eof(): - { - JSON_THROW(parse_error::create(110, chars_read, "unexpected end of input")); - } - - // positive fixint - case 0x00: - case 0x01: - case 0x02: - case 0x03: - case 0x04: - case 0x05: - case 0x06: - case 0x07: - case 0x08: - case 0x09: - case 0x0a: - case 0x0b: - case 0x0c: - case 0x0d: - case 0x0e: - case 0x0f: - case 0x10: - case 0x11: - case 0x12: - case 0x13: - case 0x14: - case 0x15: - case 0x16: - case 0x17: - case 0x18: - case 0x19: - case 0x1a: - case 0x1b: - case 0x1c: - case 0x1d: - case 0x1e: - case 0x1f: - case 0x20: - case 0x21: - case 0x22: - case 0x23: - case 0x24: - case 0x25: - case 0x26: - case 0x27: - case 0x28: - case 0x29: - case 0x2a: - case 0x2b: - case 0x2c: - case 0x2d: - case 0x2e: - case 0x2f: - case 0x30: - case 0x31: - case 0x32: - case 0x33: - case 0x34: - case 0x35: - case 0x36: - case 0x37: - case 0x38: - case 0x39: - case 0x3a: - case 0x3b: - case 0x3c: - case 0x3d: - case 0x3e: - case 0x3f: - case 0x40: - case 0x41: - case 0x42: - case 0x43: - case 0x44: - case 0x45: - case 0x46: - case 0x47: - case 0x48: - case 0x49: - case 0x4a: - case 0x4b: - case 0x4c: - case 0x4d: - case 0x4e: - case 0x4f: - case 0x50: - case 0x51: - case 0x52: - case 0x53: - case 0x54: - case 0x55: - case 0x56: - case 0x57: - case 0x58: - case 0x59: - case 0x5a: - case 0x5b: - case 0x5c: - case 0x5d: - case 0x5e: - case 0x5f: - case 0x60: - case 0x61: - case 0x62: - case 0x63: - case 0x64: - case 0x65: - case 0x66: - case 0x67: - case 0x68: - case 0x69: - case 0x6a: - case 0x6b: - case 0x6c: - case 0x6d: - case 0x6e: - case 0x6f: - case 0x70: - case 0x71: - case 0x72: - case 0x73: - case 0x74: - case 0x75: - case 0x76: - case 0x77: - case 0x78: - case 0x79: - case 0x7a: - case 0x7b: - case 0x7c: - case 0x7d: - case 0x7e: - case 0x7f: - { - return static_cast(current); - } - - // fixmap - case 0x80: - case 0x81: - case 0x82: - case 0x83: - case 0x84: - case 0x85: - case 0x86: - case 0x87: - case 0x88: - case 0x89: - case 0x8a: - case 0x8b: - case 0x8c: - case 0x8d: - case 0x8e: - case 0x8f: - { - basic_json result = value_t::object; - const auto len = static_cast(current & 0x0f); - for (size_t i = 0; i < len; ++i) - { - get(); - auto key = get_msgpack_string(); - result[key] = parse_msgpack(); - } - return result; - } - - // fixarray - case 0x90: - case 0x91: - case 0x92: - case 0x93: - case 0x94: - case 0x95: - case 0x96: - case 0x97: - case 0x98: - case 0x99: - case 0x9a: - case 0x9b: - case 0x9c: - case 0x9d: - case 0x9e: - case 0x9f: - { - basic_json result = value_t::array; - const auto len = static_cast(current & 0x0f); - for (size_t i = 0; i < len; ++i) - { - result.push_back(parse_msgpack()); - } - return result; - } - - // fixstr - case 0xa0: - case 0xa1: - case 0xa2: - case 0xa3: - case 0xa4: - case 0xa5: - case 0xa6: - case 0xa7: - case 0xa8: - case 0xa9: - case 0xaa: - case 0xab: - case 0xac: - case 0xad: - case 0xae: - case 0xaf: - case 0xb0: - case 0xb1: - case 0xb2: - case 0xb3: - case 0xb4: - case 0xb5: - case 0xb6: - case 0xb7: - case 0xb8: - case 0xb9: - case 0xba: - case 0xbb: - case 0xbc: - case 0xbd: - case 0xbe: - case 0xbf: - { - return get_msgpack_string(); - } - - case 0xc0: // nil - { - return value_t::null; - } - - case 0xc2: // false - { - return false; - } - - case 0xc3: // true - { - return true; - } - - case 0xca: // float 32 - { - return get_number(); - } - - case 0xcb: // float 64 - { - return get_number(); - } - - case 0xcc: // uint 8 - { - return get_number(); - } - - case 0xcd: // uint 16 - { - return get_number(); - } - - case 0xce: // uint 32 - { - return get_number(); - } - - case 0xcf: // uint 64 - { - return get_number(); - } - - case 0xd0: // int 8 - { - return get_number(); - } - - case 0xd1: // int 16 - { - return get_number(); - } - - case 0xd2: // int 32 - { - return get_number(); - } - - case 0xd3: // int 64 - { - return get_number(); - } - - case 0xd9: // str 8 - case 0xda: // str 16 - case 0xdb: // str 32 - { - return get_msgpack_string(); - } - - case 0xdc: // array 16 - { - basic_json result = value_t::array; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - result.push_back(parse_msgpack()); - } - return result; - } - - case 0xdd: // array 32 - { - basic_json result = value_t::array; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - result.push_back(parse_msgpack()); - } - return result; - } - - case 0xde: // map 16 - { - basic_json result = value_t::object; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - get(); - auto key = get_msgpack_string(); - result[key] = parse_msgpack(); - } - return result; - } - - case 0xdf: // map 32 - { - basic_json result = value_t::object; - const auto len = static_cast(get_number()); - for (size_t i = 0; i < len; ++i) - { - get(); - auto key = get_msgpack_string(); - result[key] = parse_msgpack(); - } - return result; - } - - // positive fixint - case 0xe0: - case 0xe1: - case 0xe2: - case 0xe3: - case 0xe4: - case 0xe5: - case 0xe6: - case 0xe7: - case 0xe8: - case 0xe9: - case 0xea: - case 0xeb: - case 0xec: - case 0xed: - case 0xee: - case 0xef: - case 0xf0: - case 0xf1: - case 0xf2: - case 0xf3: - case 0xf4: - case 0xf5: - case 0xf6: - case 0xf7: - case 0xf8: - case 0xf9: - case 0xfa: - case 0xfb: - case 0xfc: - case 0xfd: - case 0xfe: - case 0xff: - { - return static_cast(current); - } - - default: // anything else - { - std::stringstream ss; - ss << std::setw(2) << std::setfill('0') << std::hex << current; - JSON_THROW(parse_error::create(112, chars_read, "error reading MessagePack; last byte: 0x" + ss.str())); - } - } - } - - /*! - @brief determine system byte order - - @return true iff system's byte order is little endian - - @note from http://stackoverflow.com/a/1001328/266378 - */ - static bool little_endianess() noexcept - { - int num = 1; - return (*reinterpret_cast(&num) == 1); - } - - private: - /*! - @brief get next character from the input - - This function provides the interface to the used input adapter. It does - not throw in case the input reached EOF, but returns - `std::char_traits::eof()` in that case. - - @return character read from the input - */ - int get() - { - ++chars_read; - return (current = ia->get_character()); - } - - /* - @brief read a number from the input - - @tparam T the type of the number - - @return number of type @a T - - @note This function needs to respect the system's endianess, because - bytes in CBOR and MessagePack are stored in network order (big - endian) and therefore need reordering on little endian systems. - - @throw parse_error.110 if input has less than `sizeof(T)` bytes - */ - template - T get_number() - { - // step 1: read input into array with system's byte order - std::array vec; - for (size_t i = 0; i < sizeof(T); ++i) - { - get(); - check_eof(); - - // reverse byte order prior to conversion if necessary - if (is_little_endian) - { - vec[sizeof(T) - i - 1] = static_cast(current); - } - else - { - vec[i] = static_cast(current); - } - } - - // step 2: convert array into number of type T and return - T result; - std::memcpy(&result, vec.data(), sizeof(T)); - return result; - } - - /*! - @brief create a string by reading characters from the input - - @param[in] len number of bytes to read - - @return string created by reading @a len bytes - - @throw parse_error.110 if input has less than @a len bytes - */ - std::string get_string(const size_t len) - { - std::string result; - for (size_t i = 0; i < len; ++i) - { - get(); - check_eof(); - result.append(1, static_cast(current)); - } - return result; - } - - /*! - @brief reads a CBOR string - - This function first reads starting bytes to determine the expected - string length and then copies this number of bytes into a string. - Additionally, CBOR's strings with indefinite lengths are supported. - - @return string - - @throw parse_error.110 if input ended - @throw parse_error.113 if an unexpexted byte is read - */ - std::string get_cbor_string() - { - check_eof(); - - switch (current) - { - // UTF-8 string (0x00..0x17 bytes follow) - case 0x60: - case 0x61: - case 0x62: - case 0x63: - case 0x64: - case 0x65: - case 0x66: - case 0x67: - case 0x68: - case 0x69: - case 0x6a: - case 0x6b: - case 0x6c: - case 0x6d: - case 0x6e: - case 0x6f: - case 0x70: - case 0x71: - case 0x72: - case 0x73: - case 0x74: - case 0x75: - case 0x76: - case 0x77: - { - const auto len = static_cast(current & 0x1f); - return get_string(len); - } - - case 0x78: // UTF-8 string (one-byte uint8_t for n follows) - { - const auto len = static_cast(get_number()); - return get_string(len); - } - - case 0x79: // UTF-8 string (two-byte uint16_t for n follow) - { - const auto len = static_cast(get_number()); - return get_string(len); - } - - case 0x7a: // UTF-8 string (four-byte uint32_t for n follow) - { - const auto len = static_cast(get_number()); - return get_string(len); - } - - case 0x7b: // UTF-8 string (eight-byte uint64_t for n follow) - { - const auto len = static_cast(get_number()); - return get_string(len); - } - - case 0x7f: // UTF-8 string (indefinite length) - { - std::string result; - while (get() != 0xff) - { - check_eof(); - result.append(1, static_cast(current)); - } - return result; - } - - default: - { - std::stringstream ss; - ss << std::setw(2) << std::setfill('0') << std::hex << current; - JSON_THROW(parse_error::create(113, chars_read, "expected a CBOR string; last byte: 0x" + ss.str())); - } - } - } - - /*! - @brief reads a MessagePack string - - This function first reads starting bytes to determine the expected - string length and then copies this number of bytes into a string. - - @return string - - @throw parse_error.110 if input ended - @throw parse_error.113 if an unexpexted byte is read - */ - std::string get_msgpack_string() - { - check_eof(); - - switch (current) - { - // fixstr - case 0xa0: - case 0xa1: - case 0xa2: - case 0xa3: - case 0xa4: - case 0xa5: - case 0xa6: - case 0xa7: - case 0xa8: - case 0xa9: - case 0xaa: - case 0xab: - case 0xac: - case 0xad: - case 0xae: - case 0xaf: - case 0xb0: - case 0xb1: - case 0xb2: - case 0xb3: - case 0xb4: - case 0xb5: - case 0xb6: - case 0xb7: - case 0xb8: - case 0xb9: - case 0xba: - case 0xbb: - case 0xbc: - case 0xbd: - case 0xbe: - case 0xbf: - { - const auto len = static_cast(current & 0x1f); - return get_string(len); - } - - case 0xd9: // str 8 - { - const auto len = static_cast(get_number()); - return get_string(len); - } - - case 0xda: // str 16 - { - const auto len = static_cast(get_number()); - return get_string(len); - } - - case 0xdb: // str 32 - { - const auto len = static_cast(get_number()); - return get_string(len); - } - - default: - { - std::stringstream ss; - ss << std::setw(2) << std::setfill('0') << std::hex << current; - JSON_THROW(parse_error::create(113, chars_read, "expected a MessagePack string; last byte: 0x" + ss.str())); - } - } - } - - /*! - @brief check if input ended - @throw parse_error.110 if input ended - */ - void check_eof() const - { - if (JSON_UNLIKELY(current == std::char_traits::eof())) - { - JSON_THROW(parse_error::create(110, chars_read, "unexpected end of input")); - } - } - - private: - /// input adapter - input_adapter_t ia = nullptr; - - /// the current character - int current = std::char_traits::eof(); - - /// the number of characters read - size_t chars_read = 0; - - /// whether we can assume little endianess - const bool is_little_endian = true; - }; - - /*! - @brief serialization to CBOR and MessagePack values - */ - class binary_writer - { - public: - /*! - @brief create a binary writer - - @param[in] adapter output adapter to write to - */ - explicit binary_writer(output_adapter_t adapter) - : is_little_endian(binary_reader::little_endianess()), oa(adapter) - { - assert(oa); - } - - /*! - @brief[in] j JSON value to serialize - */ - void write_cbor(const basic_json& j) - { - switch (j.type()) - { - case value_t::null: - { - oa->write_character(0xf6); - break; - } - - case value_t::boolean: - { - oa->write_character(j.m_value.boolean ? 0xf5 : 0xf4); - break; - } - - case value_t::number_integer: - { - if (j.m_value.number_integer >= 0) - { - // CBOR does not differentiate between positive signed - // integers and unsigned integers. Therefore, we used the - // code from the value_t::number_unsigned case here. - if (j.m_value.number_integer <= 0x17) - { - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_integer <= (std::numeric_limits::max)()) - { - oa->write_character(0x18); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_integer <= (std::numeric_limits::max)()) - { - oa->write_character(0x19); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_integer <= (std::numeric_limits::max)()) - { - oa->write_character(0x1a); - write_number(static_cast(j.m_value.number_integer)); - } - else - { - oa->write_character(0x1b); - write_number(static_cast(j.m_value.number_integer)); - } - } - else - { - // The conversions below encode the sign in the first - // byte, and the value is converted to a positive number. - const auto positive_number = -1 - j.m_value.number_integer; - if (j.m_value.number_integer >= -24) - { - write_number(static_cast(0x20 + positive_number)); - } - else if (positive_number <= (std::numeric_limits::max)()) - { - oa->write_character(0x38); - write_number(static_cast(positive_number)); - } - else if (positive_number <= (std::numeric_limits::max)()) - { - oa->write_character(0x39); - write_number(static_cast(positive_number)); - } - else if (positive_number <= (std::numeric_limits::max)()) - { - oa->write_character(0x3a); - write_number(static_cast(positive_number)); - } - else - { - oa->write_character(0x3b); - write_number(static_cast(positive_number)); - } - } - break; - } - - case value_t::number_unsigned: - { - if (j.m_value.number_unsigned <= 0x17) - { - write_number(static_cast(j.m_value.number_unsigned)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - oa->write_character(0x18); - write_number(static_cast(j.m_value.number_unsigned)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - oa->write_character(0x19); - write_number(static_cast(j.m_value.number_unsigned)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - oa->write_character(0x1a); - write_number(static_cast(j.m_value.number_unsigned)); - } - else - { - oa->write_character(0x1b); - write_number(static_cast(j.m_value.number_unsigned)); - } - break; - } - - case value_t::number_float: - { - // Double-Precision Float - oa->write_character(0xfb); - write_number(j.m_value.number_float); - break; - } - - case value_t::string: - { - // step 1: write control byte and the string length - const auto N = j.m_value.string->size(); - if (N <= 0x17) - { - write_number(static_cast(0x60 + N)); - } - else if (N <= 0xff) - { - oa->write_character(0x78); - write_number(static_cast(N)); - } - else if (N <= 0xffff) - { - oa->write_character(0x79); - write_number(static_cast(N)); - } - else if (N <= 0xffffffff) - { - oa->write_character(0x7a); - write_number(static_cast(N)); - } - // LCOV_EXCL_START - else if (N <= 0xffffffffffffffff) - { - oa->write_character(0x7b); - write_number(static_cast(N)); - } - // LCOV_EXCL_STOP - - // step 2: write the string - oa->write_characters(reinterpret_cast(j.m_value.string->c_str()), - j.m_value.string->size()); - break; - } - - case value_t::array: - { - // step 1: write control byte and the array size - const auto N = j.m_value.array->size(); - if (N <= 0x17) - { - write_number(static_cast(0x80 + N)); - } - else if (N <= 0xff) - { - oa->write_character(0x98); - write_number(static_cast(N)); - } - else if (N <= 0xffff) - { - oa->write_character(0x99); - write_number(static_cast(N)); - } - else if (N <= 0xffffffff) - { - oa->write_character(0x9a); - write_number(static_cast(N)); - } - // LCOV_EXCL_START - else if (N <= 0xffffffffffffffff) - { - oa->write_character(0x9b); - write_number(static_cast(N)); - } - // LCOV_EXCL_STOP - - // step 2: write each element - for (const auto& el : *j.m_value.array) - { - write_cbor(el); - } - break; - } - - case value_t::object: - { - // step 1: write control byte and the object size - const auto N = j.m_value.object->size(); - if (N <= 0x17) - { - write_number(static_cast(0xa0 + N)); - } - else if (N <= 0xff) - { - oa->write_character(0xb8); - write_number(static_cast(N)); - } - else if (N <= 0xffff) - { - oa->write_character(0xb9); - write_number(static_cast(N)); - } - else if (N <= 0xffffffff) - { - oa->write_character(0xba); - write_number(static_cast(N)); - } - // LCOV_EXCL_START - else if (N <= 0xffffffffffffffff) - { - oa->write_character(0xbb); - write_number(static_cast(N)); - } - // LCOV_EXCL_STOP - - // step 2: write each element - for (const auto& el : *j.m_value.object) - { - write_cbor(el.first); - write_cbor(el.second); - } - break; - } - - default: - { - break; - } - } - } - - /*! - @brief[in] j JSON value to serialize - */ - void write_msgpack(const basic_json& j) - { - switch (j.type()) - { - case value_t::null: - { - // nil - oa->write_character(0xc0); - break; - } - - case value_t::boolean: - { - // true and false - oa->write_character(j.m_value.boolean ? 0xc3 : 0xc2); - break; - } - - case value_t::number_integer: - { - if (j.m_value.number_integer >= 0) - { - // MessagePack does not differentiate between positive - // signed integers and unsigned integers. Therefore, we - // used the code from the value_t::number_unsigned case - // here. - if (j.m_value.number_unsigned < 128) - { - // positive fixnum - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - // uint 8 - oa->write_character(0xcc); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - // uint 16 - oa->write_character(0xcd); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - // uint 32 - oa->write_character(0xce); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - // uint 64 - oa->write_character(0xcf); - write_number(static_cast(j.m_value.number_integer)); - } - } - else - { - if (j.m_value.number_integer >= -32) - { - // negative fixnum - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_integer >= (std::numeric_limits::min)() and j.m_value.number_integer <= (std::numeric_limits::max)()) - { - // int 8 - oa->write_character(0xd0); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_integer >= (std::numeric_limits::min)() and j.m_value.number_integer <= (std::numeric_limits::max)()) - { - // int 16 - oa->write_character(0xd1); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_integer >= (std::numeric_limits::min)() and j.m_value.number_integer <= (std::numeric_limits::max)()) - { - // int 32 - oa->write_character(0xd2); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_integer >= (std::numeric_limits::min)() and j.m_value.number_integer <= (std::numeric_limits::max)()) - { - // int 64 - oa->write_character(0xd3); - write_number(static_cast(j.m_value.number_integer)); - } - } - break; - } - - case value_t::number_unsigned: - { - if (j.m_value.number_unsigned < 128) - { - // positive fixnum - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - // uint 8 - oa->write_character(0xcc); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - // uint 16 - oa->write_character(0xcd); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - // uint 32 - oa->write_character(0xce); - write_number(static_cast(j.m_value.number_integer)); - } - else if (j.m_value.number_unsigned <= (std::numeric_limits::max)()) - { - // uint 64 - oa->write_character(0xcf); - write_number(static_cast(j.m_value.number_integer)); - } - break; - } - - case value_t::number_float: - { - // float 64 - oa->write_character(0xcb); - write_number(j.m_value.number_float); - break; - } - - case value_t::string: - { - // step 1: write control byte and the string length - const auto N = j.m_value.string->size(); - if (N <= 31) - { - // fixstr - write_number(static_cast(0xa0 | N)); - } - else if (N <= 255) - { - // str 8 - oa->write_character(0xd9); - write_number(static_cast(N)); - } - else if (N <= 65535) - { - // str 16 - oa->write_character(0xda); - write_number(static_cast(N)); - } - else if (N <= 4294967295) - { - // str 32 - oa->write_character(0xdb); - write_number(static_cast(N)); - } - - // step 2: write the string - oa->write_characters(reinterpret_cast(j.m_value.string->c_str()), - j.m_value.string->size()); - break; - } - - case value_t::array: - { - // step 1: write control byte and the array size - const auto N = j.m_value.array->size(); - if (N <= 15) - { - // fixarray - write_number(static_cast(0x90 | N)); - } - else if (N <= 0xffff) - { - // array 16 - oa->write_character(0xdc); - write_number(static_cast(N)); - } - else if (N <= 0xffffffff) - { - // array 32 - oa->write_character(0xdd); - write_number(static_cast(N)); - } - - // step 2: write each element - for (const auto& el : *j.m_value.array) - { - write_msgpack(el); - } - break; - } - - case value_t::object: - { - // step 1: write control byte and the object size - const auto N = j.m_value.object->size(); - if (N <= 15) - { - // fixmap - write_number(static_cast(0x80 | (N & 0xf))); - } - else if (N <= 65535) - { - // map 16 - oa->write_character(0xde); - write_number(static_cast(N)); - } - else if (N <= 4294967295) - { - // map 32 - oa->write_character(0xdf); - write_number(static_cast(N)); - } - - // step 2: write each element - for (const auto& el : *j.m_value.object) - { - write_msgpack(el.first); - write_msgpack(el.second); - } - break; - } - - default: - { - break; - } - } - } - - private: - /* - @brief write a number to output input - - @param[in] n number of type @a T - @tparam T the type of the number - - @note This function needs to respect the system's endianess, because - bytes in CBOR and MessagePack are stored in network order (big - endian) and therefore need reordering on little endian systems. - */ - template - void write_number(T n) - { - // step 1: write number to array of length T - std::array vec; - std::memcpy(vec.data(), &n, sizeof(T)); - - // step 2: write array to output (with possible reordering) - for (size_t i = 0; i < sizeof(T); ++i) - { - // reverse byte order prior to conversion if necessary - if (is_little_endian) - { - oa->write_character(vec[sizeof(T) - i - 1]); - } - else - { - oa->write_character(vec[i]); - } - } - } - - private: - /// whether we can assume little endianess - const bool is_little_endian = true; - - /// the output - output_adapter_t oa = nullptr; - }; - public: /*! @brief create a CBOR serialization of a given JSON value @@ -10647,59 +16986,63 @@ class basic_json JSON value type | value/range | CBOR type | first byte --------------- | ------------------------------------------ | ---------------------------------- | --------------- - null | `null` | Null | 0xf6 - boolean | `true` | True | 0xf5 - boolean | `false` | False | 0xf4 - number_integer | -9223372036854775808..-2147483649 | Negative integer (8 bytes follow) | 0x3b - number_integer | -2147483648..-32769 | Negative integer (4 bytes follow) | 0x3a + null | `null` | Null | 0xF6 + boolean | `true` | True | 0xF5 + boolean | `false` | False | 0xF4 + number_integer | -9223372036854775808..-2147483649 | Negative integer (8 bytes follow) | 0x3B + number_integer | -2147483648..-32769 | Negative integer (4 bytes follow) | 0x3A number_integer | -32768..-129 | Negative integer (2 bytes follow) | 0x39 number_integer | -128..-25 | Negative integer (1 byte follow) | 0x38 number_integer | -24..-1 | Negative integer | 0x20..0x37 number_integer | 0..23 | Integer | 0x00..0x17 number_integer | 24..255 | Unsigned integer (1 byte follow) | 0x18 number_integer | 256..65535 | Unsigned integer (2 bytes follow) | 0x19 - number_integer | 65536..4294967295 | Unsigned integer (4 bytes follow) | 0x1a - number_integer | 4294967296..18446744073709551615 | Unsigned integer (8 bytes follow) | 0x1b + number_integer | 65536..4294967295 | Unsigned integer (4 bytes follow) | 0x1A + number_integer | 4294967296..18446744073709551615 | Unsigned integer (8 bytes follow) | 0x1B number_unsigned | 0..23 | Integer | 0x00..0x17 number_unsigned | 24..255 | Unsigned integer (1 byte follow) | 0x18 number_unsigned | 256..65535 | Unsigned integer (2 bytes follow) | 0x19 - number_unsigned | 65536..4294967295 | Unsigned integer (4 bytes follow) | 0x1a - number_unsigned | 4294967296..18446744073709551615 | Unsigned integer (8 bytes follow) | 0x1b - number_float | *any value* | Double-Precision Float | 0xfb + number_unsigned | 65536..4294967295 | Unsigned integer (4 bytes follow) | 0x1A + number_unsigned | 4294967296..18446744073709551615 | Unsigned integer (8 bytes follow) | 0x1B + number_float | *any value* | Double-Precision Float | 0xFB string | *length*: 0..23 | UTF-8 string | 0x60..0x77 string | *length*: 23..255 | UTF-8 string (1 byte follow) | 0x78 string | *length*: 256..65535 | UTF-8 string (2 bytes follow) | 0x79 - string | *length*: 65536..4294967295 | UTF-8 string (4 bytes follow) | 0x7a - string | *length*: 4294967296..18446744073709551615 | UTF-8 string (8 bytes follow) | 0x7b + string | *length*: 65536..4294967295 | UTF-8 string (4 bytes follow) | 0x7A + string | *length*: 4294967296..18446744073709551615 | UTF-8 string (8 bytes follow) | 0x7B array | *size*: 0..23 | array | 0x80..0x97 array | *size*: 23..255 | array (1 byte follow) | 0x98 array | *size*: 256..65535 | array (2 bytes follow) | 0x99 - array | *size*: 65536..4294967295 | array (4 bytes follow) | 0x9a - array | *size*: 4294967296..18446744073709551615 | array (8 bytes follow) | 0x9b - object | *size*: 0..23 | map | 0xa0..0xb7 - object | *size*: 23..255 | map (1 byte follow) | 0xb8 - object | *size*: 256..65535 | map (2 bytes follow) | 0xb9 - object | *size*: 65536..4294967295 | map (4 bytes follow) | 0xba - object | *size*: 4294967296..18446744073709551615 | map (8 bytes follow) | 0xbb + array | *size*: 65536..4294967295 | array (4 bytes follow) | 0x9A + array | *size*: 4294967296..18446744073709551615 | array (8 bytes follow) | 0x9B + object | *size*: 0..23 | map | 0xA0..0xB7 + object | *size*: 23..255 | map (1 byte follow) | 0xB8 + object | *size*: 256..65535 | map (2 bytes follow) | 0xB9 + object | *size*: 65536..4294967295 | map (4 bytes follow) | 0xBA + object | *size*: 4294967296..18446744073709551615 | map (8 bytes follow) | 0xBB @note The mapping is **complete** in the sense that any JSON value type can be converted to a CBOR value. + @note If NaN or Infinity are stored inside a JSON number, they are + serialized properly. This behavior differs from the @ref dump() + function which serializes NaN or Infinity to `null`. + @note The following CBOR types are not used in the conversion: - - byte strings (0x40..0x5f) - - UTF-8 strings terminated by "break" (0x7f) - - arrays terminated by "break" (0x9f) - - maps terminated by "break" (0xbf) - - date/time (0xc0..0xc1) - - bignum (0xc2..0xc3) - - decimal fraction (0xc4) - - bigfloat (0xc5) - - tagged items (0xc6..0xd4, 0xd8..0xdb) - - expected conversions (0xd5..0xd7) - - simple values (0xe0..0xf3, 0xf8) - - undefined (0xf7) - - half and single-precision floats (0xf9-0xfa) - - break (0xff) + - byte strings (0x40..0x5F) + - UTF-8 strings terminated by "break" (0x7F) + - arrays terminated by "break" (0x9F) + - maps terminated by "break" (0xBF) + - date/time (0xC0..0xC1) + - bignum (0xC2..0xC3) + - decimal fraction (0xC4) + - bigfloat (0xC5) + - tagged items (0xC6..0xD4, 0xD8..0xDB) + - expected conversions (0xD5..0xD7) + - simple values (0xE0..0xF3, 0xF8) + - undefined (0xF7) + - half and single-precision floats (0xF9-0xFA) + - break (0xFF) @param[in] j JSON value to serialize @return MessagePack serialization as byte vector @@ -10710,20 +17053,31 @@ class basic_json vector in CBOR format.,to_cbor} @sa http://cbor.io - @sa @ref from_cbor(const std::vector&, const size_t) for the + @sa @ref from_cbor(detail::input_adapter, const bool strict) for the analogous deserialization - @sa @ref to_msgpack(const basic_json& for the related MessagePack format + @sa @ref to_msgpack(const basic_json&) for the related MessagePack format + @sa @ref to_ubjson(const basic_json&, const bool, const bool) for the + related UBJSON format @since version 2.0.9 */ static std::vector to_cbor(const basic_json& j) { std::vector result; - binary_writer bw(output_adapter::create(result)); - bw.write_cbor(j); + to_cbor(j, result); return result; } + static void to_cbor(const basic_json& j, detail::output_adapter o) + { + binary_writer(o).write_cbor(j); + } + + static void to_cbor(const basic_json& j, detail::output_adapter o) + { + binary_writer(o).write_cbor(j); + } + /*! @brief create a MessagePack serialization of a given JSON value @@ -10736,35 +17090,35 @@ class basic_json JSON value type | value/range | MessagePack type | first byte --------------- | --------------------------------- | ---------------- | ---------- - null | `null` | nil | 0xc0 - boolean | `true` | true | 0xc3 - boolean | `false` | false | 0xc2 - number_integer | -9223372036854775808..-2147483649 | int64 | 0xd3 - number_integer | -2147483648..-32769 | int32 | 0xd2 - number_integer | -32768..-129 | int16 | 0xd1 - number_integer | -128..-33 | int8 | 0xd0 - number_integer | -32..-1 | negative fixint | 0xe0..0xff - number_integer | 0..127 | positive fixint | 0x00..0x7f - number_integer | 128..255 | uint 8 | 0xcc - number_integer | 256..65535 | uint 16 | 0xcd - number_integer | 65536..4294967295 | uint 32 | 0xce - number_integer | 4294967296..18446744073709551615 | uint 64 | 0xcf - number_unsigned | 0..127 | positive fixint | 0x00..0x7f - number_unsigned | 128..255 | uint 8 | 0xcc - number_unsigned | 256..65535 | uint 16 | 0xcd - number_unsigned | 65536..4294967295 | uint 32 | 0xce - number_unsigned | 4294967296..18446744073709551615 | uint 64 | 0xcf - number_float | *any value* | float 64 | 0xcb - string | *length*: 0..31 | fixstr | 0xa0..0xbf - string | *length*: 32..255 | str 8 | 0xd9 - string | *length*: 256..65535 | str 16 | 0xda - string | *length*: 65536..4294967295 | str 32 | 0xdb - array | *size*: 0..15 | fixarray | 0x90..0x9f - array | *size*: 16..65535 | array 16 | 0xdc - array | *size*: 65536..4294967295 | array 32 | 0xdd - object | *size*: 0..15 | fix map | 0x80..0x8f - object | *size*: 16..65535 | map 16 | 0xde - object | *size*: 65536..4294967295 | map 32 | 0xdf + null | `null` | nil | 0xC0 + boolean | `true` | true | 0xC3 + boolean | `false` | false | 0xC2 + number_integer | -9223372036854775808..-2147483649 | int64 | 0xD3 + number_integer | -2147483648..-32769 | int32 | 0xD2 + number_integer | -32768..-129 | int16 | 0xD1 + number_integer | -128..-33 | int8 | 0xD0 + number_integer | -32..-1 | negative fixint | 0xE0..0xFF + number_integer | 0..127 | positive fixint | 0x00..0x7F + number_integer | 128..255 | uint 8 | 0xCC + number_integer | 256..65535 | uint 16 | 0xCD + number_integer | 65536..4294967295 | uint 32 | 0xCE + number_integer | 4294967296..18446744073709551615 | uint 64 | 0xCF + number_unsigned | 0..127 | positive fixint | 0x00..0x7F + number_unsigned | 128..255 | uint 8 | 0xCC + number_unsigned | 256..65535 | uint 16 | 0xCD + number_unsigned | 65536..4294967295 | uint 32 | 0xCE + number_unsigned | 4294967296..18446744073709551615 | uint 64 | 0xCF + number_float | *any value* | float 64 | 0xCB + string | *length*: 0..31 | fixstr | 0xA0..0xBF + string | *length*: 32..255 | str 8 | 0xD9 + string | *length*: 256..65535 | str 16 | 0xDA + string | *length*: 65536..4294967295 | str 32 | 0xDB + array | *size*: 0..15 | fixarray | 0x90..0x9F + array | *size*: 16..65535 | array 16 | 0xDC + array | *size*: 65536..4294967295 | array 32 | 0xDD + object | *size*: 0..15 | fix map | 0x80..0x8F + object | *size*: 16..65535 | map 16 | 0xDE + object | *size*: 65536..4294967295 | map 32 | 0xDF @note The mapping is **complete** in the sense that any JSON value type can be converted to a MessagePack value. @@ -10775,14 +17129,18 @@ class basic_json - objects with more than 4294967295 elements @note The following MessagePack types are not used in the conversion: - - bin 8 - bin 32 (0xc4..0xc6) - - ext 8 - ext 32 (0xc7..0xc9) - - float 32 (0xca) - - fixext 1 - fixext 16 (0xd4..0xd8) + - bin 8 - bin 32 (0xC4..0xC6) + - ext 8 - ext 32 (0xC7..0xC9) + - float 32 (0xCA) + - fixext 1 - fixext 16 (0xD4..0xD8) @note Any MessagePack output created @ref to_msgpack can be successfully parsed by @ref from_msgpack. + @note If NaN or Infinity are stored inside a JSON number, they are + serialized properly. This behavior differs from the @ref dump() + function which serializes NaN or Infinity to `null`. + @param[in] j JSON value to serialize @return MessagePack serialization as byte vector @@ -10795,22 +17153,134 @@ class basic_json @sa @ref from_msgpack(const std::vector&, const size_t) for the analogous deserialization @sa @ref to_cbor(const basic_json& for the related CBOR format + @sa @ref to_ubjson(const basic_json&, const bool, const bool) for the + related UBJSON format @since version 2.0.9 */ static std::vector to_msgpack(const basic_json& j) { std::vector result; - binary_writer bw(output_adapter::create(result)); - bw.write_msgpack(j); + to_msgpack(j, result); return result; } - /*! - @brief create a JSON value from a byte vector in CBOR format + static void to_msgpack(const basic_json& j, detail::output_adapter o) + { + binary_writer(o).write_msgpack(j); + } - Deserializes a given byte vector @a v to a JSON value using the CBOR - (Concise Binary Object Representation) serialization format. + static void to_msgpack(const basic_json& j, detail::output_adapter o) + { + binary_writer(o).write_msgpack(j); + } + + /*! + @brief create a UBJSON serialization of a given JSON value + + Serializes a given JSON value @a j to a byte vector using the UBJSON + (Universal Binary JSON) serialization format. UBJSON aims to be more compact + than JSON itself, yet more efficient to parse. + + The library uses the following mapping from JSON values types to + UBJSON types according to the UBJSON specification: + + JSON value type | value/range | UBJSON type | marker + --------------- | --------------------------------- | ----------- | ------ + null | `null` | null | `Z` + boolean | `true` | true | `T` + boolean | `false` | false | `F` + number_integer | -9223372036854775808..-2147483649 | int64 | `L` + number_integer | -2147483648..-32769 | int32 | `l` + number_integer | -32768..-129 | int16 | `I` + number_integer | -128..127 | int8 | `i` + number_integer | 128..255 | uint8 | `U` + number_integer | 256..32767 | int16 | `I` + number_integer | 32768..2147483647 | int32 | `l` + number_integer | 2147483648..9223372036854775807 | int64 | `L` + number_unsigned | 0..127 | int8 | `i` + number_unsigned | 128..255 | uint8 | `U` + number_unsigned | 256..32767 | int16 | `I` + number_unsigned | 32768..2147483647 | int32 | `l` + number_unsigned | 2147483648..9223372036854775807 | int64 | `L` + number_float | *any value* | float64 | `D` + string | *with shortest length indicator* | string | `S` + array | *see notes on optimized format* | array | `[` + object | *see notes on optimized format* | map | `{` + + @note The mapping is **complete** in the sense that any JSON value type + can be converted to a UBJSON value. + + @note The following values can **not** be converted to a UBJSON value: + - strings with more than 9223372036854775807 bytes (theoretical) + - unsigned integer numbers above 9223372036854775807 + + @note The following markers are not used in the conversion: + - `Z`: no-op values are not created. + - `C`: single-byte strings are serialized with `S` markers. + + @note Any UBJSON output created @ref to_ubjson can be successfully parsed + by @ref from_ubjson. + + @note If NaN or Infinity are stored inside a JSON number, they are + serialized properly. This behavior differs from the @ref dump() + function which serializes NaN or Infinity to `null`. + + @note The optimized formats for containers are supported: Parameter + @a use_size adds size information to the beginning of a container and + removes the closing marker. Parameter @a use_type further checks + whether all elements of a container have the same type and adds the + type marker to the beginning of the container. The @a use_type + parameter must only be used together with @a use_size = true. Note + that @a use_size = true alone may result in larger representations - + the benefit of this parameter is that the receiving side is + immediately informed on the number of elements of the container. + + @param[in] j JSON value to serialize + @param[in] use_size whether to add size annotations to container types + @param[in] use_type whether to add type annotations to container types + (must be combined with @a use_size = true) + @return UBJSON serialization as byte vector + + @complexity Linear in the size of the JSON value @a j. + + @liveexample{The example shows the serialization of a JSON value to a byte + vector in UBJSON format.,to_ubjson} + + @sa http://ubjson.org + @sa @ref from_ubjson(detail::input_adapter, const bool strict) for the + analogous deserialization + @sa @ref to_cbor(const basic_json& for the related CBOR format + @sa @ref to_msgpack(const basic_json&) for the related MessagePack format + + @since version 3.1.0 + */ + static std::vector to_ubjson(const basic_json& j, + const bool use_size = false, + const bool use_type = false) + { + std::vector result; + to_ubjson(j, result, use_size, use_type); + return result; + } + + static void to_ubjson(const basic_json& j, detail::output_adapter o, + const bool use_size = false, const bool use_type = false) + { + binary_writer(o).write_ubjson(j, use_size, use_type); + } + + static void to_ubjson(const basic_json& j, detail::output_adapter o, + const bool use_size = false, const bool use_type = false) + { + binary_writer(o).write_ubjson(j, use_size, use_type); + } + + /*! + @brief create a JSON value from an input in CBOR format + + Deserializes a given input @a i to a JSON value using the CBOR (Concise + Binary Object Representation) serialization format. The library maps CBOR types to JSON value types as follows: @@ -10819,51 +17289,51 @@ class basic_json Integer | number_unsigned | 0x00..0x17 Unsigned integer | number_unsigned | 0x18 Unsigned integer | number_unsigned | 0x19 - Unsigned integer | number_unsigned | 0x1a - Unsigned integer | number_unsigned | 0x1b + Unsigned integer | number_unsigned | 0x1A + Unsigned integer | number_unsigned | 0x1B Negative integer | number_integer | 0x20..0x37 Negative integer | number_integer | 0x38 Negative integer | number_integer | 0x39 - Negative integer | number_integer | 0x3a - Negative integer | number_integer | 0x3b + Negative integer | number_integer | 0x3A + Negative integer | number_integer | 0x3B Negative integer | number_integer | 0x40..0x57 UTF-8 string | string | 0x60..0x77 UTF-8 string | string | 0x78 UTF-8 string | string | 0x79 - UTF-8 string | string | 0x7a - UTF-8 string | string | 0x7b - UTF-8 string | string | 0x7f + UTF-8 string | string | 0x7A + UTF-8 string | string | 0x7B + UTF-8 string | string | 0x7F array | array | 0x80..0x97 array | array | 0x98 array | array | 0x99 - array | array | 0x9a - array | array | 0x9b - array | array | 0x9f - map | object | 0xa0..0xb7 - map | object | 0xb8 - map | object | 0xb9 - map | object | 0xba - map | object | 0xbb - map | object | 0xbf - False | `false` | 0xf4 - True | `true` | 0xf5 - Nill | `null` | 0xf6 - Half-Precision Float | number_float | 0xf9 - Single-Precision Float | number_float | 0xfa - Double-Precision Float | number_float | 0xfb + array | array | 0x9A + array | array | 0x9B + array | array | 0x9F + map | object | 0xA0..0xB7 + map | object | 0xB8 + map | object | 0xB9 + map | object | 0xBA + map | object | 0xBB + map | object | 0xBF + False | `false` | 0xF4 + True | `true` | 0xF5 + Nill | `null` | 0xF6 + Half-Precision Float | number_float | 0xF9 + Single-Precision Float | number_float | 0xFA + Double-Precision Float | number_float | 0xFB @warning The mapping is **incomplete** in the sense that not all CBOR types can be converted to a JSON value. The following CBOR types are not supported and will yield parse errors (parse_error.112): - - byte strings (0x40..0x5f) - - date/time (0xc0..0xc1) - - bignum (0xc2..0xc3) - - decimal fraction (0xc4) - - bigfloat (0xc5) - - tagged items (0xc6..0xd4, 0xd8..0xdb) - - expected conversions (0xd5..0xd7) - - simple values (0xe0..0xf3, 0xf8) - - undefined (0xf7) + - byte strings (0x40..0x5F) + - date/time (0xC0..0xC1) + - bignum (0xC2..0xC3) + - decimal fraction (0xC4) + - bigfloat (0xC5) + - tagged items (0xC6..0xD4, 0xD8..0xDB) + - expected conversions (0xD5..0xD7) + - simple values (0xE0..0xF3, 0xF8) + - undefined (0xF7) @warning CBOR allows map keys of any type, whereas JSON only allows strings as keys in object values. Therefore, CBOR maps with keys @@ -10872,2788 +17342,249 @@ class basic_json @note Any CBOR output created @ref to_cbor can be successfully parsed by @ref from_cbor. - @param[in] v a byte vector in CBOR format - @param[in] start_index the index to start reading from @a v (0 by default) + @param[in] i an input in CBOR format convertible to an input adapter + @param[in] strict whether to expect the input to be consumed until EOF + (true by default) + @param[in] allow_exceptions whether to throw exceptions in case of a + parse error (optional, true by default) + @return deserialized JSON value - @throw parse_error.110 if the given vector ends prematurely + @throw parse_error.110 if the given input ends prematurely or the end of + file was not reached when @a strict was set to true @throw parse_error.112 if unsupported features from CBOR were - used in the given vector @a v or if the input is not valid CBOR + used in the given input @a v or if the input is not valid CBOR @throw parse_error.113 if a string was expected as map key, but not found - @complexity Linear in the size of the byte vector @a v. + @complexity Linear in the size of the input @a i. @liveexample{The example shows the deserialization of a byte vector in CBOR format to a JSON value.,from_cbor} @sa http://cbor.io @sa @ref to_cbor(const basic_json&) for the analogous serialization - @sa @ref from_msgpack(const std::vector&, const size_t) for the + @sa @ref from_msgpack(detail::input_adapter, const bool, const bool) for the related MessagePack format + @sa @ref from_ubjson(detail::input_adapter, const bool, const bool) for the + related UBJSON format - @since version 2.0.9, parameter @a start_index since 2.1.1 + @since version 2.0.9; parameter @a start_index since 2.1.1; changed to + consume input adapters, removed start_index parameter, and added + @a strict parameter since 3.0.0; added @allow_exceptions parameter + since 3.2.0 */ - static basic_json from_cbor(const std::vector& v, - const size_t start_index = 0) + static basic_json from_cbor(detail::input_adapter&& i, + const bool strict = true, + const bool allow_exceptions = true) { - binary_reader br(input_adapter::create(v.begin() + static_cast(start_index), v.end())); - return br.parse_cbor(); + basic_json result; + detail::json_sax_dom_parser sdp(result, allow_exceptions); + const bool res = binary_reader(detail::input_adapter(i)).sax_parse(input_format_t::cbor, &sdp, strict); + return res ? result : basic_json(value_t::discarded); } + /*! + @copydoc from_cbor(detail::input_adapter, const bool, const bool) + */ + template::value, int> = 0> + static basic_json from_cbor(A1 && a1, A2 && a2, + const bool strict = true, + const bool allow_exceptions = true) + { + basic_json result; + detail::json_sax_dom_parser sdp(result, allow_exceptions); + const bool res = binary_reader(detail::input_adapter(std::forward(a1), std::forward(a2))).sax_parse(input_format_t::cbor, &sdp, strict); + return res ? result : basic_json(value_t::discarded); + } /*! - @brief create a JSON value from a byte vector in MessagePack format + @brief create a JSON value from an input in MessagePack format - Deserializes a given byte vector @a v to a JSON value using the MessagePack + Deserializes a given input @a i to a JSON value using the MessagePack serialization format. The library maps MessagePack types to JSON value types as follows: MessagePack type | JSON value type | first byte ---------------- | --------------- | ---------- - positive fixint | number_unsigned | 0x00..0x7f - fixmap | object | 0x80..0x8f - fixarray | array | 0x90..0x9f - fixstr | string | 0xa0..0xbf - nil | `null` | 0xc0 - false | `false` | 0xc2 - true | `true` | 0xc3 - float 32 | number_float | 0xca - float 64 | number_float | 0xcb - uint 8 | number_unsigned | 0xcc - uint 16 | number_unsigned | 0xcd - uint 32 | number_unsigned | 0xce - uint 64 | number_unsigned | 0xcf - int 8 | number_integer | 0xd0 - int 16 | number_integer | 0xd1 - int 32 | number_integer | 0xd2 - int 64 | number_integer | 0xd3 - str 8 | string | 0xd9 - str 16 | string | 0xda - str 32 | string | 0xdb - array 16 | array | 0xdc - array 32 | array | 0xdd - map 16 | object | 0xde - map 32 | object | 0xdf - negative fixint | number_integer | 0xe0-0xff + positive fixint | number_unsigned | 0x00..0x7F + fixmap | object | 0x80..0x8F + fixarray | array | 0x90..0x9F + fixstr | string | 0xA0..0xBF + nil | `null` | 0xC0 + false | `false` | 0xC2 + true | `true` | 0xC3 + float 32 | number_float | 0xCA + float 64 | number_float | 0xCB + uint 8 | number_unsigned | 0xCC + uint 16 | number_unsigned | 0xCD + uint 32 | number_unsigned | 0xCE + uint 64 | number_unsigned | 0xCF + int 8 | number_integer | 0xD0 + int 16 | number_integer | 0xD1 + int 32 | number_integer | 0xD2 + int 64 | number_integer | 0xD3 + str 8 | string | 0xD9 + str 16 | string | 0xDA + str 32 | string | 0xDB + array 16 | array | 0xDC + array 32 | array | 0xDD + map 16 | object | 0xDE + map 32 | object | 0xDF + negative fixint | number_integer | 0xE0-0xFF @warning The mapping is **incomplete** in the sense that not all MessagePack types can be converted to a JSON value. The following MessagePack types are not supported and will yield parse errors: - - bin 8 - bin 32 (0xc4..0xc6) - - ext 8 - ext 32 (0xc7..0xc9) - - fixext 1 - fixext 16 (0xd4..0xd8) + - bin 8 - bin 32 (0xC4..0xC6) + - ext 8 - ext 32 (0xC7..0xC9) + - fixext 1 - fixext 16 (0xD4..0xD8) @note Any MessagePack output created @ref to_msgpack can be successfully parsed by @ref from_msgpack. - @param[in] v a byte vector in MessagePack format - @param[in] start_index the index to start reading from @a v (0 by default) + @param[in] i an input in MessagePack format convertible to an input + adapter + @param[in] strict whether to expect the input to be consumed until EOF + (true by default) + @param[in] allow_exceptions whether to throw exceptions in case of a + parse error (optional, true by default) + @return deserialized JSON value - @throw parse_error.110 if the given vector ends prematurely + @throw parse_error.110 if the given input ends prematurely or the end of + file was not reached when @a strict was set to true @throw parse_error.112 if unsupported features from MessagePack were - used in the given vector @a v or if the input is not valid MessagePack + used in the given input @a i or if the input is not valid MessagePack @throw parse_error.113 if a string was expected as map key, but not found - @complexity Linear in the size of the byte vector @a v. + @complexity Linear in the size of the input @a i. @liveexample{The example shows the deserialization of a byte vector in MessagePack format to a JSON value.,from_msgpack} @sa http://msgpack.org @sa @ref to_msgpack(const basic_json&) for the analogous serialization - @sa @ref from_cbor(const std::vector&, const size_t) for the + @sa @ref from_cbor(detail::input_adapter, const bool, const bool) for the related CBOR format + @sa @ref from_ubjson(detail::input_adapter, const bool, const bool) for + the related UBJSON format - @since version 2.0.9, parameter @a start_index since 2.1.1 + @since version 2.0.9; parameter @a start_index since 2.1.1; changed to + consume input adapters, removed start_index parameter, and added + @a strict parameter since 3.0.0; added @allow_exceptions parameter + since 3.2.0 */ - static basic_json from_msgpack(const std::vector& v, - const size_t start_index = 0) + static basic_json from_msgpack(detail::input_adapter&& i, + const bool strict = true, + const bool allow_exceptions = true) { - binary_reader br(input_adapter::create(v.begin() + static_cast(start_index), v.end())); - return br.parse_msgpack(); + basic_json result; + detail::json_sax_dom_parser sdp(result, allow_exceptions); + const bool res = binary_reader(detail::input_adapter(i)).sax_parse(input_format_t::msgpack, &sdp, strict); + return res ? result : basic_json(value_t::discarded); + } + + /*! + @copydoc from_msgpack(detail::input_adapter, const bool, const bool) + */ + template::value, int> = 0> + static basic_json from_msgpack(A1 && a1, A2 && a2, + const bool strict = true, + const bool allow_exceptions = true) + { + basic_json result; + detail::json_sax_dom_parser sdp(result, allow_exceptions); + const bool res = binary_reader(detail::input_adapter(std::forward(a1), std::forward(a2))).sax_parse(input_format_t::msgpack, &sdp, strict); + return res ? result : basic_json(value_t::discarded); + } + + /*! + @brief create a JSON value from an input in UBJSON format + + Deserializes a given input @a i to a JSON value using the UBJSON (Universal + Binary JSON) serialization format. + + The library maps UBJSON types to JSON value types as follows: + + UBJSON type | JSON value type | marker + ----------- | --------------------------------------- | ------ + no-op | *no value, next value is read* | `N` + null | `null` | `Z` + false | `false` | `F` + true | `true` | `T` + float32 | number_float | `d` + float64 | number_float | `D` + uint8 | number_unsigned | `U` + int8 | number_integer | `i` + int16 | number_integer | `I` + int32 | number_integer | `l` + int64 | number_integer | `L` + string | string | `S` + char | string | `C` + array | array (optimized values are supported) | `[` + object | object (optimized values are supported) | `{` + + @note The mapping is **complete** in the sense that any UBJSON value can + be converted to a JSON value. + + @param[in] i an input in UBJSON format convertible to an input adapter + @param[in] strict whether to expect the input to be consumed until EOF + (true by default) + @param[in] allow_exceptions whether to throw exceptions in case of a + parse error (optional, true by default) + + @return deserialized JSON value + + @throw parse_error.110 if the given input ends prematurely or the end of + file was not reached when @a strict was set to true + @throw parse_error.112 if a parse error occurs + @throw parse_error.113 if a string could not be parsed successfully + + @complexity Linear in the size of the input @a i. + + @liveexample{The example shows the deserialization of a byte vector in + UBJSON format to a JSON value.,from_ubjson} + + @sa http://ubjson.org + @sa @ref to_ubjson(const basic_json&, const bool, const bool) for the + analogous serialization + @sa @ref from_cbor(detail::input_adapter, const bool, const bool) for the + related CBOR format + @sa @ref from_msgpack(detail::input_adapter, const bool, const bool) for + the related MessagePack format + + @since version 3.1.0; added @allow_exceptions parameter since 3.2.0 + */ + static basic_json from_ubjson(detail::input_adapter&& i, + const bool strict = true, + const bool allow_exceptions = true) + { + basic_json result; + detail::json_sax_dom_parser sdp(result, allow_exceptions); + const bool res = binary_reader(detail::input_adapter(i)).sax_parse(input_format_t::ubjson, &sdp, strict); + return res ? result : basic_json(value_t::discarded); + } + + /*! + @copydoc from_ubjson(detail::input_adapter, const bool, const bool) + */ + template::value, int> = 0> + static basic_json from_ubjson(A1 && a1, A2 && a2, + const bool strict = true, + const bool allow_exceptions = true) + { + basic_json result; + detail::json_sax_dom_parser sdp(result, allow_exceptions); + const bool res = binary_reader(detail::input_adapter(std::forward(a1), std::forward(a2))).sax_parse(input_format_t::ubjson, &sdp, strict); + return res ? result : basic_json(value_t::discarded); } /// @} - ////////////////////// - // lexer and parser // - ////////////////////// - - private: - /*! - @brief lexical analysis - - This class organizes the lexical analysis during JSON deserialization. - */ - class lexer - { - public: - /// token types for the parser - enum class token_type - { - uninitialized, ///< indicating the scanner is uninitialized - literal_true, ///< the `true` literal - literal_false, ///< the `false` literal - literal_null, ///< the `null` literal - value_string, ///< a string -- use get_string() for actual value - value_unsigned, ///< an unsigned integer -- use get_number_unsigned() for actual value - value_integer, ///< a signed integer -- use get_number_integer() for actual value - value_float, ///< an floating point number -- use get_number_float() for actual value - begin_array, ///< the character for array begin `[` - begin_object, ///< the character for object begin `{` - end_array, ///< the character for array end `]` - end_object, ///< the character for object end `}` - name_separator, ///< the name separator `:` - value_separator, ///< the value separator `,` - parse_error, ///< indicating a parse error - end_of_input ///< indicating the end of the input buffer - }; - - /// return name of values of type token_type (only used for errors) - static const char* token_type_name(const token_type t) noexcept - { - switch (t) - { - case token_type::uninitialized: - return ""; - case token_type::literal_true: - return "true literal"; - case token_type::literal_false: - return "false literal"; - case token_type::literal_null: - return "null literal"; - case token_type::value_string: - return "string literal"; - case lexer::token_type::value_unsigned: - case lexer::token_type::value_integer: - case lexer::token_type::value_float: - return "number literal"; - case token_type::begin_array: - return "'['"; - case token_type::begin_object: - return "'{'"; - case token_type::end_array: - return "']'"; - case token_type::end_object: - return "'}'"; - case token_type::name_separator: - return "':'"; - case token_type::value_separator: - return "','"; - case token_type::parse_error: - return ""; - case token_type::end_of_input: - return "end of input"; - default: - { - // catch non-enum values - return "unknown token"; // LCOV_EXCL_LINE - } - } - } - - explicit lexer(input_adapter_t adapter) - : ia(adapter), decimal_point_char(get_decimal_point()) - {} - - private: - ///////////////////// - // locales - ///////////////////// - - /// return the locale-dependent decimal point - static char get_decimal_point() noexcept - { - const auto loc = localeconv(); - assert(loc != nullptr); - return (loc->decimal_point == nullptr) ? '.' : loc->decimal_point[0]; - } - - ///////////////////// - // scan functions - ///////////////////// - - /*! - @brief get codepoint from 4 hex characters following `\u` - - @return codepoint or -1 in case of an error (e.g. EOF or non-hex - character) - */ - int get_codepoint() - { - // this function only makes sense after reading `\u` - assert(current == 'u'); - int codepoint = 0; - - // byte 1: \uXxxx - switch (get()) - { - case '0': - break; - case '1': - codepoint += 0x1000; - break; - case '2': - codepoint += 0x2000; - break; - case '3': - codepoint += 0x3000; - break; - case '4': - codepoint += 0x4000; - break; - case '5': - codepoint += 0x5000; - break; - case '6': - codepoint += 0x6000; - break; - case '7': - codepoint += 0x7000; - break; - case '8': - codepoint += 0x8000; - break; - case '9': - codepoint += 0x9000; - break; - case 'A': - case 'a': - codepoint += 0xa000; - break; - case 'B': - case 'b': - codepoint += 0xb000; - break; - case 'C': - case 'c': - codepoint += 0xc000; - break; - case 'D': - case 'd': - codepoint += 0xd000; - break; - case 'E': - case 'e': - codepoint += 0xe000; - break; - case 'F': - case 'f': - codepoint += 0xf000; - break; - default: - return -1; - } - - // byte 2: \uxXxx - switch (get()) - { - case '0': - break; - case '1': - codepoint += 0x0100; - break; - case '2': - codepoint += 0x0200; - break; - case '3': - codepoint += 0x0300; - break; - case '4': - codepoint += 0x0400; - break; - case '5': - codepoint += 0x0500; - break; - case '6': - codepoint += 0x0600; - break; - case '7': - codepoint += 0x0700; - break; - case '8': - codepoint += 0x0800; - break; - case '9': - codepoint += 0x0900; - break; - case 'A': - case 'a': - codepoint += 0x0a00; - break; - case 'B': - case 'b': - codepoint += 0x0b00; - break; - case 'C': - case 'c': - codepoint += 0x0c00; - break; - case 'D': - case 'd': - codepoint += 0x0d00; - break; - case 'E': - case 'e': - codepoint += 0x0e00; - break; - case 'F': - case 'f': - codepoint += 0x0f00; - break; - default: - return -1; - } - - // byte 3: \uxxXx - switch (get()) - { - case '0': - break; - case '1': - codepoint += 0x0010; - break; - case '2': - codepoint += 0x0020; - break; - case '3': - codepoint += 0x0030; - break; - case '4': - codepoint += 0x0040; - break; - case '5': - codepoint += 0x0050; - break; - case '6': - codepoint += 0x0060; - break; - case '7': - codepoint += 0x0070; - break; - case '8': - codepoint += 0x0080; - break; - case '9': - codepoint += 0x0090; - break; - case 'A': - case 'a': - codepoint += 0x00a0; - break; - case 'B': - case 'b': - codepoint += 0x00b0; - break; - case 'C': - case 'c': - codepoint += 0x00c0; - break; - case 'D': - case 'd': - codepoint += 0x00d0; - break; - case 'E': - case 'e': - codepoint += 0x00e0; - break; - case 'F': - case 'f': - codepoint += 0x00f0; - break; - default: - return -1; - } - - // byte 4: \uxxxX - switch (get()) - { - case '0': - break; - case '1': - codepoint += 0x0001; - break; - case '2': - codepoint += 0x0002; - break; - case '3': - codepoint += 0x0003; - break; - case '4': - codepoint += 0x0004; - break; - case '5': - codepoint += 0x0005; - break; - case '6': - codepoint += 0x0006; - break; - case '7': - codepoint += 0x0007; - break; - case '8': - codepoint += 0x0008; - break; - case '9': - codepoint += 0x0009; - break; - case 'A': - case 'a': - codepoint += 0x000a; - break; - case 'B': - case 'b': - codepoint += 0x000b; - break; - case 'C': - case 'c': - codepoint += 0x000c; - break; - case 'D': - case 'd': - codepoint += 0x000d; - break; - case 'E': - case 'e': - codepoint += 0x000e; - break; - case 'F': - case 'f': - codepoint += 0x000f; - break; - default: - return -1; - } - - return codepoint; - } - - /*! - @brief create diagnostic representation of a codepoint - @return string "U+XXXX" for codepoint XXXX - */ - static std::string codepoint_to_string(int codepoint) - { - std::stringstream ss; - ss << "U+" << std::setw(4) << std::uppercase << std::setfill('0') << std::hex << codepoint; - return ss.str(); - } - - /*! - @brief scan a string literal - - This function scans a string according to Sect. 7 of RFC 7159. While - scanning, bytes are escaped and copied into buffer yytext. Then the - function returns successfully, yytext is null-terminated and yylen - contains the number of bytes in the string. - - @return token_type::value_string if string could be successfully - scanned, token_type::parse_error otherwise - - @note In case of errors, variable error_message contains a textual - description. - */ - token_type scan_string() - { - // reset yytext (ignore opening quote) - reset(); - - // we entered the function by reading an open quote - assert(current == '\"'); - - while (true) - { - // get next character - get(); - - switch (current) - { - // end of file while parsing string - case std::char_traits::eof(): - { - error_message = "invalid string: missing closing quote"; - return token_type::parse_error; - } - - // closing quote - case '\"': - { - // terminate yytext - add('\0'); - --yylen; - return token_type::value_string; - } - - // escapes - case '\\': - { - switch (get()) - { - // quotation mark - case '\"': - add('\"'); - break; - // reverse solidus - case '\\': - add('\\'); - break; - // solidus - case '/': - add('/'); - break; - // backspace - case 'b': - add('\b'); - break; - // form feed - case 'f': - add('\f'); - break; - // line feed - case 'n': - add('\n'); - break; - // carriage return - case 'r': - add('\r'); - break; - // tab - case 't': - add('\t'); - break; - - // unicode escapes - case 'u': - { - int codepoint; - int codepoint1 = get_codepoint(); - - if (JSON_UNLIKELY(codepoint1 == -1)) - { - error_message = "invalid string: '\\u' must be followed by 4 hex digits"; - return token_type::parse_error; - } - - // check if code point is a high surrogate - if (0xD800 <= codepoint1 and codepoint1 <= 0xDBFF) - { - // expect next \uxxxx entry - if (JSON_LIKELY(get() == '\\' and get() == 'u')) - { - const int codepoint2 = get_codepoint(); - - if (JSON_UNLIKELY(codepoint2 == -1)) - { - error_message = "invalid string: '\\u' must be followed by 4 hex digits"; - return token_type::parse_error; - } - - // check if codepoint2 is a low surrogate - if (JSON_LIKELY(0xDC00 <= codepoint2 and codepoint2 <= 0xDFFF)) - { - codepoint = - // high surrogate occupies the most significant 22 bits - (codepoint1 << 10) - // low surrogate occupies the least significant 15 bits - + codepoint2 - // there is still the 0xD800, 0xDC00 and 0x10000 noise - // in the result so we have to subtract with: - // (0xD800 << 10) + DC00 - 0x10000 = 0x35FDC00 - - 0x35FDC00; - } - else - { - error_message = "invalid string: surrogate " + codepoint_to_string(codepoint1) + " must be followed by U+DC00..U+DFFF instead of " + codepoint_to_string(codepoint2); - return token_type::parse_error; - } - } - else - { - error_message = "invalid string: surrogate " + codepoint_to_string(codepoint1) + " must be followed by U+DC00..U+DFFF"; - return token_type::parse_error; - } - } - else - { - if (JSON_UNLIKELY(0xDC00 <= codepoint1 and codepoint1 <= 0xDFFF)) - { - error_message = "invalid string: surrogate " + codepoint_to_string(codepoint1) + " must follow U+D800..U+DBFF"; - return token_type::parse_error; - } - - // only work with first code point - codepoint = codepoint1; - } - - // result of the above calculation yields a proper codepoint - assert(0x00 <= codepoint and codepoint <= 0x10FFFF); - - // translate code point to bytes - if (codepoint < 0x80) - { - // 1-byte characters: 0xxxxxxx (ASCII) - add(codepoint); - } - else if (codepoint <= 0x7ff) - { - // 2-byte characters: 110xxxxx 10xxxxxx - add(0xC0 | (codepoint >> 6)); - add(0x80 | (codepoint & 0x3F)); - } - else if (codepoint <= 0xffff) - { - // 3-byte characters: 1110xxxx 10xxxxxx 10xxxxxx - add(0xE0 | (codepoint >> 12)); - add(0x80 | ((codepoint >> 6) & 0x3F)); - add(0x80 | (codepoint & 0x3F)); - } - else - { - // 4-byte characters: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx - add(0xF0 | (codepoint >> 18)); - add(0x80 | ((codepoint >> 12) & 0x3F)); - add(0x80 | ((codepoint >> 6) & 0x3F)); - add(0x80 | (codepoint & 0x3F)); - } - - break; - } - - // other characters after escape - default: - error_message = "invalid string: forbidden character after backslash"; - return token_type::parse_error; - } - - break; - } - - // invalid control characters - case 0x00: - case 0x01: - case 0x02: - case 0x03: - case 0x04: - case 0x05: - case 0x06: - case 0x07: - case 0x08: - case 0x09: - case 0x0a: - case 0x0b: - case 0x0c: - case 0x0d: - case 0x0e: - case 0x0f: - case 0x10: - case 0x11: - case 0x12: - case 0x13: - case 0x14: - case 0x15: - case 0x16: - case 0x17: - case 0x18: - case 0x19: - case 0x1a: - case 0x1b: - case 0x1c: - case 0x1d: - case 0x1e: - case 0x1f: - { - error_message = "invalid string: control character " + codepoint_to_string(current) + " must be escaped"; - return token_type::parse_error; - } - - // U+0020..U+007F (except U+0022 (quote) and U+005C (backspace)) - case 0x20: - case 0x21: - case 0x23: - case 0x24: - case 0x25: - case 0x26: - case 0x27: - case 0x28: - case 0x29: - case 0x2a: - case 0x2b: - case 0x2c: - case 0x2d: - case 0x2e: - case 0x2f: - case 0x30: - case 0x31: - case 0x32: - case 0x33: - case 0x34: - case 0x35: - case 0x36: - case 0x37: - case 0x38: - case 0x39: - case 0x3a: - case 0x3b: - case 0x3c: - case 0x3d: - case 0x3e: - case 0x3f: - case 0x40: - case 0x41: - case 0x42: - case 0x43: - case 0x44: - case 0x45: - case 0x46: - case 0x47: - case 0x48: - case 0x49: - case 0x4a: - case 0x4b: - case 0x4c: - case 0x4d: - case 0x4e: - case 0x4f: - case 0x50: - case 0x51: - case 0x52: - case 0x53: - case 0x54: - case 0x55: - case 0x56: - case 0x57: - case 0x58: - case 0x59: - case 0x5a: - case 0x5b: - case 0x5d: - case 0x5e: - case 0x5f: - case 0x60: - case 0x61: - case 0x62: - case 0x63: - case 0x64: - case 0x65: - case 0x66: - case 0x67: - case 0x68: - case 0x69: - case 0x6a: - case 0x6b: - case 0x6c: - case 0x6d: - case 0x6e: - case 0x6f: - case 0x70: - case 0x71: - case 0x72: - case 0x73: - case 0x74: - case 0x75: - case 0x76: - case 0x77: - case 0x78: - case 0x79: - case 0x7a: - case 0x7b: - case 0x7c: - case 0x7d: - case 0x7e: - case 0x7f: - { - add(current); - break; - } - - // U+0080..U+07FF: bytes C2..DF 80..BF - case 0xc2: - case 0xc3: - case 0xc4: - case 0xc5: - case 0xc6: - case 0xc7: - case 0xc8: - case 0xc9: - case 0xca: - case 0xcb: - case 0xcc: - case 0xcd: - case 0xce: - case 0xcf: - case 0xd0: - case 0xd1: - case 0xd2: - case 0xd3: - case 0xd4: - case 0xd5: - case 0xd6: - case 0xd7: - case 0xd8: - case 0xd9: - case 0xda: - case 0xdb: - case 0xdc: - case 0xdd: - case 0xde: - case 0xdf: - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - continue; - } - - error_message = "invalid string: ill-formed UTF-8 byte"; - return token_type::parse_error; - } - - // U+0800..U+0FFF: bytes E0 A0..BF 80..BF - case 0xe0: - { - add(current); - get(); - if (JSON_LIKELY(0xa0 <= current and current <= 0xbf)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - continue; - } - } - - error_message = "invalid string: ill-formed UTF-8 byte"; - return token_type::parse_error; - } - - // U+1000..U+CFFF: bytes E1..EC 80..BF 80..BF - // U+E000..U+FFFF: bytes EE..EF 80..BF 80..BF - case 0xe1: - case 0xe2: - case 0xe3: - case 0xe4: - case 0xe5: - case 0xe6: - case 0xe7: - case 0xe8: - case 0xe9: - case 0xea: - case 0xeb: - case 0xec: - case 0xee: - case 0xef: - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - continue; - } - } - - error_message = "invalid string: ill-formed UTF-8 byte"; - return token_type::parse_error; - } - - // U+D000..U+D7FF: bytes ED 80..9F 80..BF - case 0xed: - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0x9f)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - continue; - } - } - - error_message = "invalid string: ill-formed UTF-8 byte"; - return token_type::parse_error; - } - - // U+10000..U+3FFFF F0 90..BF 80..BF 80..BF - case 0xf0: - { - add(current); - get(); - if (JSON_LIKELY(0x90 <= current and current <= 0xbf)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - continue; - } - } - } - - error_message = "invalid string: ill-formed UTF-8 byte"; - return token_type::parse_error; - } - - // U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF - case 0xf1: - case 0xf2: - case 0xf3: - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - continue; - } - } - } - - error_message = "invalid string: ill-formed UTF-8 byte"; - return token_type::parse_error; - } - - // U+100000..U+10FFFF F4 80..8F 80..BF 80..BF - case 0xf4: - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0x8f)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - get(); - if (JSON_LIKELY(0x80 <= current and current <= 0xbf)) - { - add(current); - continue; - } - } - } - - error_message = "invalid string: ill-formed UTF-8 byte"; - return token_type::parse_error; - } - - // remaining bytes (80..C1 and F5..FF) are ill-formed - default: - { - error_message = "invalid string: ill-formed UTF-8 byte"; - return token_type::parse_error; - } - } - } - } - - static void strtof(float& f, const char* str, char** endptr) noexcept - { - f = std::strtof(str, endptr); - } - - static void strtof(double& f, const char* str, char** endptr) noexcept - { - f = std::strtod(str, endptr); - } - - static void strtof(long double& f, const char* str, char** endptr) noexcept - { - f = std::strtold(str, endptr); - } - - /*! - @brief scan a number literal - - This function scans a string according to Sect. 6 of RFC 7159. - - The function is realized with a deterministic finite state machine - derived from the grammar described in RFC 7159. Starting in state - "init", the input is read and used to determined the next state. Only - state "done" accepts the number. State "error" is a trap state to model - errors. In the table below, "anything" means any character but the ones - listed before. - - state | 0 | 1-9 | e E | + | - | . | anything - ---------|----------|----------|----------|---------|---------|----------|----------- - init | zero | any1 | [error] | [error] | minus | [error] | [error] - minus | zero | any1 | [error] | [error] | [error] | [error] | [error] - zero | done | done | exponent | done | done | decimal1 | done - any1 | any1 | any1 | exponent | done | done | decimal1 | done - decimal1 | decimal2 | [error] | [error] | [error] | [error] | [error] | [error] - decimal2 | decimal2 | decimal2 | exponent | done | done | done | done - exponent | any2 | any2 | [error] | sign | sign | [error] | [error] - sign | any2 | any2 | [error] | [error] | [error] | [error] | [error] - any2 | any2 | any2 | done | done | done | done | done - - The state machine is realized with one label per state (prefixed with - "scan_number_") and `goto` statements between them. The state machine - contains cycles, but any cycle can be left when EOF is read. Therefore, - the function is guaranteed to terminate. - - During scanning, the read bytes are stored in yytext. This string is - then converted to a signed integer, an unsigned integer, or a - floating-point number. - - @return token_type::value_unsigned, token_type::value_integer, or - token_type::value_float if number could be successfully scanned, - token_type::parse_error otherwise - - @note The scanner is independent of the current locale. Internally, the - locale's decimal point is used instead of `.` to work with the - locale-dependent converters. - */ - token_type scan_number() - { - // reset yytext to store the number's bytes - reset(); - - // the type of the parsed number; initially set to unsigned; will be - // changed if minus sign, decimal point or exponent is read - token_type number_type = token_type::value_unsigned; - - // state (init): we just found out we need to scan a number - switch (current) - { - case '-': - { - add(current); - goto scan_number_minus; - } - - case '0': - { - add(current); - goto scan_number_zero; - } - - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - { - add(current); - goto scan_number_any1; - } - - default: - { - // all other characters are rejected outside scan_number() - assert(false); // LCOV_EXCL_LINE - } - } - -scan_number_minus: - // state: we just parsed a leading minus sign - number_type = token_type::value_integer; - switch (get()) - { - case '0': - { - add(current); - goto scan_number_zero; - } - - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - { - add(current); - goto scan_number_any1; - } - - default: - { - error_message = "invalid number; expected digit after '-'"; - return token_type::parse_error; - } - } - -scan_number_zero: - // state: we just parse a zero (maybe with a leading minus sign) - switch (get()) - { - case '.': - { - add(decimal_point_char); - goto scan_number_decimal1; - } - - case 'e': - case 'E': - { - add(current); - goto scan_number_exponent; - } - - default: - { - goto scan_number_done; - } - } - -scan_number_any1: - // state: we just parsed a number 0-9 (maybe with a leading minus sign) - switch (get()) - { - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - { - add(current); - goto scan_number_any1; - } - - case '.': - { - add(decimal_point_char); - goto scan_number_decimal1; - } - - case 'e': - case 'E': - { - add(current); - goto scan_number_exponent; - } - - default: - { - goto scan_number_done; - } - } - -scan_number_decimal1: - // state: we just parsed a decimal point - number_type = token_type::value_float; - switch (get()) - { - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - { - add(current); - goto scan_number_decimal2; - } - - default: - { - error_message = "invalid number; expected digit after '.'"; - return token_type::parse_error; - } - } - -scan_number_decimal2: - // we just parsed at least one number after a decimal point - switch (get()) - { - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - { - add(current); - goto scan_number_decimal2; - } - - case 'e': - case 'E': - { - add(current); - goto scan_number_exponent; - } - - default: - { - goto scan_number_done; - } - } - -scan_number_exponent: - // we just parsed an exponent - number_type = token_type::value_float; - switch (get()) - { - case '+': - case '-': - { - add(current); - goto scan_number_sign; - } - - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - { - add(current); - goto scan_number_any2; - } - - default: - { - error_message = "invalid number; expected '+', '-', or digit after exponent"; - return token_type::parse_error; - } - } - -scan_number_sign: - // we just parsed an exponent sign - switch (get()) - { - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - { - add(current); - goto scan_number_any2; - } - - default: - { - error_message = "invalid number; expected digit after exponent sign"; - return token_type::parse_error; - } - } - -scan_number_any2: - // we just parsed a number after the exponent or exponent sign - switch (get()) - { - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - { - add(current); - goto scan_number_any2; - } - - default: - { - goto scan_number_done; - } - } - -scan_number_done: - // unget the character after the number (we only read it to know - // that we are done scanning a number) - --chars_read; - next_unget = true; - - // terminate token - add('\0'); - --yylen; - - // try to parse integers first and fall back to floats - if (number_type == token_type::value_unsigned) - { - char* endptr = nullptr; - errno = 0; - const auto x = std::strtoull(yytext.data(), &endptr, 10); - - // we checked the number format before - assert(endptr == yytext.data() + yylen); - - if (errno == 0) - { - value_unsigned = static_cast(x); - if (value_unsigned == x) - { - return token_type::value_unsigned; - } - } - } - else if (number_type == token_type::value_integer) - { - char* endptr = nullptr; - errno = 0; - const auto x = std::strtoll(yytext.data(), &endptr, 10); - - // we checked the number format before - assert(endptr == yytext.data() + yylen); - - if (errno == 0) - { - value_integer = static_cast(x); - if (value_integer == x) - { - return token_type::value_integer; - } - } - } - - // this code is reached if we parse a floating-point number or if - // an integer conversion above failed - strtof(value_float, yytext.data(), nullptr); - return token_type::value_float; - } - - token_type scan_true() - { - assert(current == 't'); - if (JSON_LIKELY((get() == 'r' and get() == 'u' and get() == 'e'))) - { - return token_type::literal_true; - } - - error_message = "invalid literal; expected 'true'"; - return token_type::parse_error; - } - - token_type scan_false() - { - assert(current == 'f'); - if (JSON_LIKELY((get() == 'a' and get() == 'l' and get() == 's' and get() == 'e'))) - { - return token_type::literal_false; - } - - error_message = "invalid literal; expected 'false'"; - return token_type::parse_error; - } - - token_type scan_null() - { - assert(current == 'n'); - if (JSON_LIKELY((get() == 'u' and get() == 'l' and get() == 'l'))) - { - return token_type::literal_null; - } - - error_message = "invalid literal; expected 'null'"; - return token_type::parse_error; - } - - ///////////////////// - // input management - ///////////////////// - - /// reset yytext - void reset() noexcept - { - yylen = 0; - start_pos = chars_read - 1; - } - - /// get a character from the input - int get() - { - ++chars_read; - return next_unget - ? (next_unget = false, current) - : (current = ia->get_character()); - } - - /// add a character to yytext - void add(int c) - { - // resize yytext if necessary; this condition is deemed unlikely, - // because we start with a 1024-byte buffer - if (JSON_UNLIKELY((yylen + 1 > yytext.capacity()))) - { - yytext.resize(2 * yytext.capacity(), '\0'); - } - yytext[yylen++] = static_cast(c); - } - - public: - ///////////////////// - // value getters - ///////////////////// - - /// return integer value - constexpr number_integer_t get_number_integer() const noexcept - { - return value_integer; - } - - /// return unsigned integer value - constexpr number_unsigned_t get_number_unsigned() const noexcept - { - return value_unsigned; - } - - /// return floating-point value - constexpr number_float_t get_number_float() const noexcept - { - return value_float; - } - - /// return string value - const std::string get_string() - { - // yytext cannot be returned as char*, because it may contain a - // null byte (parsed as "\u0000") - return std::string(yytext.data(), yylen); - } - - ///////////////////// - // diagnostics - ///////////////////// - - /// return position of last read token - constexpr size_t get_position() const noexcept - { - return chars_read; - } - - /// return the last read token (for errors only) - std::string get_token_string() const - { - // get the raw byte sequence of the last token - std::string s = ia->read(start_pos, chars_read - start_pos); - - // escape control characters - std::string result; - for (auto c : s) - { - if (c == '\0' or c == std::char_traits::eof()) - { - // ignore EOF - continue; - } - else if ('\x00' <= c and c <= '\x1f') - { - // escape control characters - result += "<" + codepoint_to_string(c) + ">"; - } - else - { - // add character as is - result.append(1, c); - } - } - - return result; - } - - /// return syntax error message - const std::string& get_error_message() const noexcept - { - return error_message; - } - - ///////////////////// - // actual scanner - ///////////////////// - - token_type scan() - { - // read next character and ignore whitespace - do - { - get(); - } - while (current == ' ' or current == '\t' or current == '\n' or current == '\r'); - - switch (current) - { - // structural characters - case '[': - return token_type::begin_array; - case ']': - return token_type::end_array; - case '{': - return token_type::begin_object; - case '}': - return token_type::end_object; - case ':': - return token_type::name_separator; - case ',': - return token_type::value_separator; - - // literals - case 't': - return scan_true(); - case 'f': - return scan_false(); - case 'n': - return scan_null(); - - // string - case '\"': - return scan_string(); - - // number - case '-': - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - return scan_number(); - - // end of input (the null byte is needed when parsing from - // string literals) - case '\0': - case std::char_traits::eof(): - return token_type::end_of_input; - - // error - default: - error_message = "invalid literal"; - return token_type::parse_error; - } - } - - private: - /// input adapter - input_adapter_t ia = nullptr; - - /// the current character - int current = std::char_traits::eof(); - - /// whether get() should return the last character again - bool next_unget = false; - - /// the number of characters read - size_t chars_read = 0; - /// the start position of the current token - size_t start_pos = 0; - - /// buffer for variable-length tokens (numbers, strings) - std::vector yytext = std::vector(1024, '\0'); - /// current index in yytext - size_t yylen = 0; - - /// a description of occurred lexer errors - std::string error_message = ""; - - // number values - number_integer_t value_integer = 0; - number_unsigned_t value_unsigned = 0; - number_float_t value_float = 0; - - /// the decimal point - const char decimal_point_char = '.'; - }; - - /*! - @brief syntax analysis - - This class implements a recursive decent parser. - */ - class parser - { - public: - /// a parser reading from an input adapter - explicit parser(input_adapter_t adapter, - const parser_callback_t cb = nullptr) - : callback(cb), m_lexer(adapter) - {} - - /*! - @brief public parser interface - - @param[in] strict whether to expect the last token to be EOF - @return parsed JSON value - - @throw parse_error.101 in case of an unexpected token - @throw parse_error.102 if to_unicode fails or surrogate error - @throw parse_error.103 if to_unicode fails - */ - basic_json parse(const bool strict = true) - { - // read first token - get_token(); - - basic_json result = parse_internal(true); - result.assert_invariant(); - - if (strict) - { - get_token(); - expect(lexer::token_type::end_of_input); - } - - // return parser result and replace it with null in case the - // top-level value was discarded by the callback function - return result.is_discarded() ? basic_json() : std::move(result); - } - - /*! - @brief public accept interface - - @param[in] strict whether to expect the last token to be EOF - @return whether the input is a proper JSON text - */ - bool accept(const bool strict = true) - { - // read first token - get_token(); - - if (not accept_internal()) - { - return false; - } - - if (strict and get_token() != lexer::token_type::end_of_input) - { - return false; - } - - return true; - } - - private: - /*! - @brief the actual parser - @throw parse_error.101 in case of an unexpected token - @throw parse_error.102 if to_unicode fails or surrogate error - @throw parse_error.103 if to_unicode fails - */ - basic_json parse_internal(bool keep) - { - auto result = basic_json(value_t::discarded); - - switch (last_token) - { - case lexer::token_type::begin_object: - { - if (keep and (not callback - or ((keep = callback(depth++, parse_event_t::object_start, result)) != 0))) - { - // explicitly set result to object to cope with {} - result.m_type = value_t::object; - result.m_value = value_t::object; - } - - // read next token - get_token(); - - // closing } -> we are done - if (last_token == lexer::token_type::end_object) - { - if (keep and callback and not callback(--depth, parse_event_t::object_end, result)) - { - result = basic_json(value_t::discarded); - } - return result; - } - - // parse values - while (true) - { - // store key - expect(lexer::token_type::value_string); - const auto key = m_lexer.get_string(); - - bool keep_tag = false; - if (keep) - { - if (callback) - { - basic_json k(key); - keep_tag = callback(depth, parse_event_t::key, k); - } - else - { - keep_tag = true; - } - } - - // parse separator (:) - get_token(); - expect(lexer::token_type::name_separator); - - // parse and add value - get_token(); - auto value = parse_internal(keep); - if (keep and keep_tag and not value.is_discarded()) - { - result[key] = std::move(value); - } - - // comma -> next value - get_token(); - if (last_token == lexer::token_type::value_separator) - { - get_token(); - continue; - } - - // closing } - expect(lexer::token_type::end_object); - break; - } - - if (keep and callback and not callback(--depth, parse_event_t::object_end, result)) - { - result = basic_json(value_t::discarded); - } - - return result; - } - - case lexer::token_type::begin_array: - { - if (keep and (not callback - or ((keep = callback(depth++, parse_event_t::array_start, result)) != 0))) - { - // explicitly set result to object to cope with [] - result.m_type = value_t::array; - result.m_value = value_t::array; - } - - // read next token - get_token(); - - // closing ] -> we are done - if (last_token == lexer::token_type::end_array) - { - if (callback and not callback(--depth, parse_event_t::array_end, result)) - { - result = basic_json(value_t::discarded); - } - return result; - } - - // parse values - while (true) - { - // parse value - auto value = parse_internal(keep); - if (keep and not value.is_discarded()) - { - result.push_back(std::move(value)); - } - - // comma -> next value - get_token(); - if (last_token == lexer::token_type::value_separator) - { - get_token(); - continue; - } - - // closing ] - expect(lexer::token_type::end_array); - break; - } - - if (keep and callback and not callback(--depth, parse_event_t::array_end, result)) - { - result = basic_json(value_t::discarded); - } - - return result; - } - - case lexer::token_type::literal_null: - { - result.m_type = value_t::null; - break; - } - - case lexer::token_type::value_string: - { - result = basic_json(m_lexer.get_string()); - break; - } - - case lexer::token_type::literal_true: - { - result.m_type = value_t::boolean; - result.m_value = true; - break; - } - - case lexer::token_type::literal_false: - { - result.m_type = value_t::boolean; - result.m_value = false; - break; - } - - case lexer::token_type::value_unsigned: - { - result.m_type = value_t::number_unsigned; - result.m_value = m_lexer.get_number_unsigned(); - break; - } - - case lexer::token_type::value_integer: - { - result.m_type = value_t::number_integer; - result.m_value = m_lexer.get_number_integer(); - break; - } - - case lexer::token_type::value_float: - { - result.m_type = value_t::number_float; - result.m_value = m_lexer.get_number_float(); - - // throw in case of infinity or NAN - if (JSON_UNLIKELY(not std::isfinite(result.m_value.number_float))) - { - JSON_THROW(out_of_range::create(406, "number overflow parsing '" + m_lexer.get_token_string() + "'")); - } - - break; - } - - default: - { - // the last token was unexpected - unexpect(last_token); - } - } - - if (keep and callback and not callback(depth, parse_event_t::value, result)) - { - result = basic_json(value_t::discarded); - } - return result; - } - - /*! - @brief the acutal acceptor - - @invariant 1. The last token is not yet processed. Therefore, the - caller of this function must make sure a token has - been read. - 2. When this function returns, the last token is processed. - That is, the last read character was already considered. - - This invariant makes sure that no token needs to be "unput". - */ - bool accept_internal() - { - switch (last_token) - { - case lexer::token_type::begin_object: - { - // read next token - get_token(); - - // closing } -> we are done - if (last_token == lexer::token_type::end_object) - { - return true; - } - - // parse values - while (true) - { - // parse key - if (last_token != lexer::token_type::value_string) - { - return false; - } - - // parse separator (:) - get_token(); - if (last_token != lexer::token_type::name_separator) - { - return false; - } - - // parse value - get_token(); - if (not accept_internal()) - { - return false; - } - - // comma -> next value - get_token(); - if (last_token == lexer::token_type::value_separator) - { - get_token(); - continue; - } - - // closing } - if (last_token != lexer::token_type::end_object) - { - return false; - } - - return true; - } - } - - case lexer::token_type::begin_array: - { - // read next token - get_token(); - - // closing ] -> we are done - if (last_token == lexer::token_type::end_array) - { - return true; - } - - // parse values - while (true) - { - // parse value - if (not accept_internal()) - { - return false; - } - - // comma -> next value - get_token(); - if (last_token == lexer::token_type::value_separator) - { - get_token(); - continue; - } - - // closing ] - if (last_token != lexer::token_type::end_array) - { - return false; - } - - return true; - } - } - - case lexer::token_type::literal_false: - case lexer::token_type::literal_null: - case lexer::token_type::literal_true: - case lexer::token_type::value_float: - case lexer::token_type::value_integer: - case lexer::token_type::value_string: - case lexer::token_type::value_unsigned: - { - return true; - } - - default: - { - // the last token was unexpected - return false; - } - } - } - - /// get next token from lexer - typename lexer::token_type get_token() - { - last_token = m_lexer.scan(); - return last_token; - } - - /*! - @throw parse_error.101 if expected token did not occur - */ - void expect(typename lexer::token_type t) const - { - if (JSON_UNLIKELY(t != last_token)) - { - std::string error_msg = "syntax error - "; - if (last_token == lexer::token_type::parse_error) - { - error_msg += m_lexer.get_error_message() + "; last read: '" + m_lexer.get_token_string() + "'"; - } - else - { - error_msg += "unexpected " + std::string(lexer::token_type_name(last_token)); - } - - error_msg += "; expected " + std::string(lexer::token_type_name(t)); - JSON_THROW(parse_error::create(101, m_lexer.get_position(), error_msg)); - } - } - - /*! - @throw parse_error.101 if unexpected token occurred - */ - void unexpect(typename lexer::token_type t) const - { - if (JSON_UNLIKELY(t == last_token)) - { - std::string error_msg = "syntax error - "; - if (last_token == lexer::token_type::parse_error) - { - error_msg += m_lexer.get_error_message() + "; last read '" + m_lexer.get_token_string() + "'"; - } - else - { - error_msg += "unexpected " + std::string(lexer::token_type_name(last_token)); - } - - JSON_THROW(parse_error::create(101, m_lexer.get_position(), error_msg)); - } - } - - private: - /// current level of recursion - int depth = 0; - /// callback function - const parser_callback_t callback = nullptr; - /// the type of the last read token - typename lexer::token_type last_token = lexer::token_type::uninitialized; - /// the lexer - lexer m_lexer; - }; - - public: - /*! - @brief JSON Pointer - - A JSON pointer defines a string syntax for identifying a specific value - within a JSON document. It can be used with functions `at` and - `operator[]`. Furthermore, JSON pointers are the base for JSON patches. - - @sa [RFC 6901](https://tools.ietf.org/html/rfc6901) - - @since version 2.0.0 - */ - class json_pointer - { - /// allow basic_json to access private members - friend class basic_json; - - public: - /*! - @brief create JSON pointer - - Create a JSON pointer according to the syntax described in - [Section 3 of RFC6901](https://tools.ietf.org/html/rfc6901#section-3). - - @param[in] s string representing the JSON pointer; if omitted, the - empty string is assumed which references the whole JSON - value - - @throw parse_error.107 if the given JSON pointer @a s is nonempty and - does not begin with a slash (`/`); see example below - - @throw parse_error.108 if a tilde (`~`) in the given JSON pointer @a s - is not followed by `0` (representing `~`) or `1` (representing `/`); - see example below - - @liveexample{The example shows the construction several valid JSON - pointers as well as the exceptional behavior.,json_pointer} - - @since version 2.0.0 - */ - explicit json_pointer(const std::string& s = "") - : reference_tokens(split(s)) - {} - - /*! - @brief return a string representation of the JSON pointer - - @invariant For each JSON pointer `ptr`, it holds: - @code {.cpp} - ptr == json_pointer(ptr.to_string()); - @endcode - - @return a string representation of the JSON pointer - - @liveexample{The example shows the result of `to_string`., - json_pointer__to_string} - - @since version 2.0.0 - */ - std::string to_string() const noexcept - { - return std::accumulate(reference_tokens.begin(), - reference_tokens.end(), std::string{}, - [](const std::string & a, const std::string & b) - { - return a + "/" + escape(b); - }); - } - - /// @copydoc to_string() - operator std::string() const - { - return to_string(); - } - - private: - /*! - @brief remove and return last reference pointer - @throw out_of_range.405 if JSON pointer has no parent - */ - std::string pop_back() - { - if (is_root()) - { - JSON_THROW(out_of_range::create(405, "JSON pointer has no parent")); - } - - auto last = reference_tokens.back(); - reference_tokens.pop_back(); - return last; - } - - /// return whether pointer points to the root document - bool is_root() const - { - return reference_tokens.empty(); - } - - json_pointer top() const - { - if (is_root()) - { - JSON_THROW(out_of_range::create(405, "JSON pointer has no parent")); - } - - json_pointer result = *this; - result.reference_tokens = {reference_tokens[0]}; - return result; - } - - /*! - @brief create and return a reference to the pointed to value - - @complexity Linear in the number of reference tokens. - - @throw parse_error.109 if array index is not a number - @throw type_error.313 if value cannot be unflattened - */ - reference get_and_create(reference j) const - { - pointer result = &j; - - // in case no reference tokens exist, return a reference to the - // JSON value j which will be overwritten by a primitive value - for (const auto& reference_token : reference_tokens) - { - switch (result->m_type) - { - case value_t::null: - { - if (reference_token == "0") - { - // start a new array if reference token is 0 - result = &result->operator[](0); - } - else - { - // start a new object otherwise - result = &result->operator[](reference_token); - } - break; - } - - case value_t::object: - { - // create an entry in the object - result = &result->operator[](reference_token); - break; - } - - case value_t::array: - { - // create an entry in the array - JSON_TRY - { - result = &result->operator[](static_cast(std::stoi(reference_token))); - } - JSON_CATCH (std::invalid_argument&) - { - JSON_THROW(parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); - } - break; - } - - /* - The following code is only reached if there exists a - reference token _and_ the current value is primitive. In - this case, we have an error situation, because primitive - values may only occur as single value; that is, with an - empty list of reference tokens. - */ - default: - { - JSON_THROW(type_error::create(313, "invalid value to unflatten")); - } - } - } - - return *result; - } - - /*! - @brief return a reference to the pointed to value - - @note This version does not throw if a value is not present, but tries - to create nested values instead. For instance, calling this function - with pointer `"/this/that"` on a null value is equivalent to calling - `operator[]("this").operator[]("that")` on that value, effectively - changing the null value to an object. - - @param[in] ptr a JSON value - - @return reference to the JSON value pointed to by the JSON pointer - - @complexity Linear in the length of the JSON pointer. - - @throw parse_error.106 if an array index begins with '0' - @throw parse_error.109 if an array index was not a number - @throw out_of_range.404 if the JSON pointer can not be resolved - */ - reference get_unchecked(pointer ptr) const - { - for (const auto& reference_token : reference_tokens) - { - // convert null values to arrays or objects before continuing - if (ptr->m_type == value_t::null) - { - // check if reference token is a number - const bool nums = std::all_of(reference_token.begin(), - reference_token.end(), - [](const char x) - { - return (x >= '0' and x <= '9'); - }); - - // change value to array for numbers or "-" or to object - // otherwise - if (nums or reference_token == "-") - { - *ptr = value_t::array; - } - else - { - *ptr = value_t::object; - } - } - - switch (ptr->m_type) - { - case value_t::object: - { - // use unchecked object access - ptr = &ptr->operator[](reference_token); - break; - } - - case value_t::array: - { - // error condition (cf. RFC 6901, Sect. 4) - if (reference_token.size() > 1 and reference_token[0] == '0') - { - JSON_THROW(parse_error::create(106, 0, "array index '" + reference_token + "' must not begin with '0'")); - } - - if (reference_token == "-") - { - // explicitly treat "-" as index beyond the end - ptr = &ptr->operator[](ptr->m_value.array->size()); - } - else - { - // convert array index to number; unchecked access - JSON_TRY - { - ptr = &ptr->operator[](static_cast(std::stoi(reference_token))); - } - JSON_CATCH (std::invalid_argument&) - { - JSON_THROW(parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); - } - } - break; - } - - default: - { - JSON_THROW(out_of_range::create(404, "unresolved reference token '" + reference_token + "'")); - } - } - } - - return *ptr; - } - - /*! - @throw parse_error.106 if an array index begins with '0' - @throw parse_error.109 if an array index was not a number - @throw out_of_range.402 if the array index '-' is used - @throw out_of_range.404 if the JSON pointer can not be resolved - */ - reference get_checked(pointer ptr) const - { - for (const auto& reference_token : reference_tokens) - { - switch (ptr->m_type) - { - case value_t::object: - { - // note: at performs range check - ptr = &ptr->at(reference_token); - break; - } - - case value_t::array: - { - if (reference_token == "-") - { - // "-" always fails the range check - JSON_THROW(out_of_range::create(402, "array index '-' (" + - std::to_string(ptr->m_value.array->size()) + - ") is out of range")); - } - - // error condition (cf. RFC 6901, Sect. 4) - if (reference_token.size() > 1 and reference_token[0] == '0') - { - JSON_THROW(parse_error::create(106, 0, "array index '" + reference_token + "' must not begin with '0'")); - } - - // note: at performs range check - JSON_TRY - { - ptr = &ptr->at(static_cast(std::stoi(reference_token))); - } - JSON_CATCH (std::invalid_argument&) - { - JSON_THROW(parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); - } - break; - } - - default: - { - JSON_THROW(out_of_range::create(404, "unresolved reference token '" + reference_token + "'")); - } - } - } - - return *ptr; - } - - /*! - @brief return a const reference to the pointed to value - - @param[in] ptr a JSON value - - @return const reference to the JSON value pointed to by the JSON - pointer - - @throw parse_error.106 if an array index begins with '0' - @throw parse_error.109 if an array index was not a number - @throw out_of_range.402 if the array index '-' is used - @throw out_of_range.404 if the JSON pointer can not be resolved - */ - const_reference get_unchecked(const_pointer ptr) const - { - for (const auto& reference_token : reference_tokens) - { - switch (ptr->m_type) - { - case value_t::object: - { - // use unchecked object access - ptr = &ptr->operator[](reference_token); - break; - } - - case value_t::array: - { - if (reference_token == "-") - { - // "-" cannot be used for const access - JSON_THROW(out_of_range::create(402, "array index '-' (" + - std::to_string(ptr->m_value.array->size()) + - ") is out of range")); - } - - // error condition (cf. RFC 6901, Sect. 4) - if (reference_token.size() > 1 and reference_token[0] == '0') - { - JSON_THROW(parse_error::create(106, 0, "array index '" + reference_token + "' must not begin with '0'")); - } - - // use unchecked array access - JSON_TRY - { - ptr = &ptr->operator[](static_cast(std::stoi(reference_token))); - } - JSON_CATCH (std::invalid_argument&) - { - JSON_THROW(parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); - } - break; - } - - default: - { - JSON_THROW(out_of_range::create(404, "unresolved reference token '" + reference_token + "'")); - } - } - } - - return *ptr; - } - - /*! - @throw parse_error.106 if an array index begins with '0' - @throw parse_error.109 if an array index was not a number - @throw out_of_range.402 if the array index '-' is used - @throw out_of_range.404 if the JSON pointer can not be resolved - */ - const_reference get_checked(const_pointer ptr) const - { - for (const auto& reference_token : reference_tokens) - { - switch (ptr->m_type) - { - case value_t::object: - { - // note: at performs range check - ptr = &ptr->at(reference_token); - break; - } - - case value_t::array: - { - if (reference_token == "-") - { - // "-" always fails the range check - JSON_THROW(out_of_range::create(402, "array index '-' (" + - std::to_string(ptr->m_value.array->size()) + - ") is out of range")); - } - - // error condition (cf. RFC 6901, Sect. 4) - if (reference_token.size() > 1 and reference_token[0] == '0') - { - JSON_THROW(parse_error::create(106, 0, "array index '" + reference_token + "' must not begin with '0'")); - } - - // note: at performs range check - JSON_TRY - { - ptr = &ptr->at(static_cast(std::stoi(reference_token))); - } - JSON_CATCH (std::invalid_argument&) - { - JSON_THROW(parse_error::create(109, 0, "array index '" + reference_token + "' is not a number")); - } - break; - } - - default: - { - JSON_THROW(out_of_range::create(404, "unresolved reference token '" + reference_token + "'")); - } - } - } - - return *ptr; - } - - /*! - @brief split the string input to reference tokens - - @note This function is only called by the json_pointer constructor. - All exceptions below are documented there. - - @throw parse_error.107 if the pointer is not empty or begins with '/' - @throw parse_error.108 if character '~' is not followed by '0' or '1' - */ - static std::vector split(const std::string& reference_string) - { - std::vector result; - - // special case: empty reference string -> no reference tokens - if (reference_string.empty()) - { - return result; - } - - // check if nonempty reference string begins with slash - if (reference_string[0] != '/') - { - JSON_THROW(parse_error::create(107, 1, "JSON pointer must be empty or begin with '/' - was: '" + reference_string + "'")); - } - - // extract the reference tokens: - // - slash: position of the last read slash (or end of string) - // - start: position after the previous slash - for ( - // search for the first slash after the first character - size_t slash = reference_string.find_first_of('/', 1), - // set the beginning of the first reference token - start = 1; - // we can stop if start == string::npos+1 = 0 - start != 0; - // set the beginning of the next reference token - // (will eventually be 0 if slash == std::string::npos) - start = slash + 1, - // find next slash - slash = reference_string.find_first_of('/', start)) - { - // use the text between the beginning of the reference token - // (start) and the last slash (slash). - auto reference_token = reference_string.substr(start, slash - start); - - // check reference tokens are properly escaped - for (size_t pos = reference_token.find_first_of('~'); - pos != std::string::npos; - pos = reference_token.find_first_of('~', pos + 1)) - { - assert(reference_token[pos] == '~'); - - // ~ must be followed by 0 or 1 - if (pos == reference_token.size() - 1 or - (reference_token[pos + 1] != '0' and - reference_token[pos + 1] != '1')) - { - JSON_THROW(parse_error::create(108, 0, "escape character '~' must be followed with '0' or '1'")); - } - } - - // finally, store the reference token - unescape(reference_token); - result.push_back(reference_token); - } - - return result; - } - - /*! - @brief replace all occurrences of a substring by another string - - @param[in,out] s the string to manipulate; changed so that all - occurrences of @a f are replaced with @a t - @param[in] f the substring to replace with @a t - @param[in] t the string to replace @a f - - @pre The search string @a f must not be empty. **This precondition is - enforced with an assertion.** - - @since version 2.0.0 - */ - static void replace_substring(std::string& s, - const std::string& f, - const std::string& t) - { - assert(not f.empty()); - - for ( - size_t pos = s.find(f); // find first occurrence of f - pos != std::string::npos; // make sure f was found - s.replace(pos, f.size(), t), // replace with t - pos = s.find(f, pos + t.size()) // find next occurrence of f - ); - } - - /// escape tilde and slash - static std::string escape(std::string s) - { - // escape "~"" to "~0" and "/" to "~1" - replace_substring(s, "~", "~0"); - replace_substring(s, "/", "~1"); - return s; - } - - /// unescape tilde and slash - static void unescape(std::string& s) - { - // first transform any occurrence of the sequence '~1' to '/' - replace_substring(s, "~1", "/"); - // then transform any occurrence of the sequence '~0' to '~' - replace_substring(s, "~0", "~"); - } - - /*! - @param[in] reference_string the reference string to the current value - @param[in] value the value to consider - @param[in,out] result the result object to insert values to - - @note Empty objects or arrays are flattened to `null`. - */ - static void flatten(const std::string& reference_string, - const basic_json& value, - basic_json& result) - { - switch (value.m_type) - { - case value_t::array: - { - if (value.m_value.array->empty()) - { - // flatten empty array as null - result[reference_string] = nullptr; - } - else - { - // iterate array and use index as reference string - for (size_t i = 0; i < value.m_value.array->size(); ++i) - { - flatten(reference_string + "/" + std::to_string(i), - value.m_value.array->operator[](i), result); - } - } - break; - } - - case value_t::object: - { - if (value.m_value.object->empty()) - { - // flatten empty object as null - result[reference_string] = nullptr; - } - else - { - // iterate object and use keys as reference string - for (const auto& element : *value.m_value.object) - { - flatten(reference_string + "/" + escape(element.first), - element.second, result); - } - } - break; - } - - default: - { - // add primitive value with its reference string - result[reference_string] = value; - break; - } - } - } - - /*! - @param[in] value flattened JSON - - @return unflattened JSON - - @throw parse_error.109 if array index is not a number - @throw type_error.314 if value is not an object - @throw type_error.315 if object values are not primitive - @throw type_error.313 if value cannot be unflattened - */ - static basic_json unflatten(const basic_json& value) - { - if (not value.is_object()) - { - JSON_THROW(type_error::create(314, "only objects can be unflattened")); - } - - basic_json result; - - // iterate the JSON object values - for (const auto& element : *value.m_value.object) - { - if (not element.second.is_primitive()) - { - JSON_THROW(type_error::create(315, "values in object must be primitive")); - } - - // assign value to reference pointed to by JSON pointer; Note - // that if the JSON pointer is "" (i.e., points to the whole - // value), function get_and_create returns a reference to - // result itself. An assignment will then create a primitive - // value. - json_pointer(element.first).get_and_create(result) = element.second; - } - - return result; - } - - friend bool operator==(json_pointer const& lhs, - json_pointer const& rhs) noexcept - { - return lhs.reference_tokens == rhs.reference_tokens; - } - - friend bool operator!=(json_pointer const& lhs, - json_pointer const& rhs) noexcept - { - return !(lhs == rhs); - } - - /// the reference tokens - std::vector reference_tokens {}; - }; - ////////////////////////// // JSON Pointer support // ////////////////////////// @@ -13750,6 +17681,9 @@ scan_number_done: pointer @a ptr. As `at` provides checked access (and no elements are implicitly inserted), the index '-' is always invalid. See example below. + @throw out_of_range.403 if the JSON pointer describes a key of an object + which cannot be found. See example below. + @throw out_of_range.404 if the JSON pointer @a ptr can not be resolved. See example below. @@ -13790,6 +17724,9 @@ scan_number_done: pointer @a ptr. As `at` provides checked access (and no elements are implicitly inserted), the index '-' is always invalid. See example below. + @throw out_of_range.403 if the JSON pointer describes a key of an object + which cannot be found. See example below. + @throw out_of_range.404 if the JSON pointer @a ptr can not be resolved. See example below. @@ -13857,7 +17794,7 @@ scan_number_done: @complexity Linear in the size the JSON value. @throw type_error.314 if value is not an object - @throw type_error.315 if object values are not primitve + @throw type_error.315 if object values are not primitive @liveexample{The following code shows how a flattened JSON object is unflattened into the original nested JSON object.,unflatten} @@ -14005,8 +17942,8 @@ scan_number_done: } else { - const auto idx = std::stoi(last_path); - if (static_cast(idx) > parent.size()) + const auto idx = json_pointer::array_index(last_path); + if (JSON_UNLIKELY(static_cast(idx) > parent.size())) { // avoid undefined behavior JSON_THROW(out_of_range::create(401, "array index " + std::to_string(idx) + " is out of range")); @@ -14041,7 +17978,7 @@ scan_number_done: { // perform range check auto it = parent.find(last_path); - if (it != parent.end()) + if (JSON_LIKELY(it != parent.end())) { parent.erase(it); } @@ -14053,12 +17990,12 @@ scan_number_done: else if (parent.is_array()) { // note erase performs range check - parent.erase(static_cast(std::stoi(last_path))); + parent.erase(static_cast(json_pointer::array_index(last_path))); } }; // type check: top level value must be an array - if (not json_patch.is_array()) + if (JSON_UNLIKELY(not json_patch.is_array())) { JSON_THROW(parse_error::create(104, 0, "JSON patch must be an array of objects")); } @@ -14069,7 +18006,7 @@ scan_number_done: // wrapper to get a value for an operation const auto get_value = [&val](const std::string & op, const std::string & member, - bool string_type) -> basic_json& + bool string_type) -> basic_json & { // find value auto it = val.m_value.object->find(member); @@ -14078,13 +18015,13 @@ scan_number_done: const auto error_msg = (op == "op") ? "operation" : "operation '" + op + "'"; // check if desired value is present - if (it == val.m_value.object->end()) + if (JSON_UNLIKELY(it == val.m_value.object->end())) { JSON_THROW(parse_error::create(105, 0, error_msg + " must have member '" + member + "'")); } // check if result is of type string - if (string_type and not it->second.is_string()) + if (JSON_UNLIKELY(string_type and not it->second.is_string())) { JSON_THROW(parse_error::create(105, 0, error_msg + " must have string member '" + member + "'")); } @@ -14094,7 +18031,7 @@ scan_number_done: }; // type check: every element of the array must be an object - if (not val.is_object()) + if (JSON_UNLIKELY(not val.is_object())) { JSON_THROW(parse_error::create(104, 0, "JSON patch must be an array of objects")); } @@ -14144,11 +18081,16 @@ scan_number_done: case patch_operations::copy: { - const std::string from_path = get_value("copy", "from", true);; + const std::string from_path = get_value("copy", "from", true); const json_pointer from_ptr(from_path); // the "from" location must exist - use at() - result[ptr] = result.at(from_ptr); + basic_json v = result.at(from_ptr); + + // The copy is functionally identical to an "add" + // operation at the target location using the value + // specified in the "from" member. + operation_add(ptr, v); break; } @@ -14167,7 +18109,7 @@ scan_number_done: } // throw an exception if test fails - if (not success) + if (JSON_UNLIKELY(not success)) { JSON_THROW(other_error::create(501, "unsuccessful: " + val.dump())); } @@ -14214,13 +18156,13 @@ scan_number_done: diff for two JSON values.,diff} @sa @ref patch -- apply a JSON patch + @sa @ref merge_patch -- apply a JSON Merge Patch @sa [RFC 6902 (JSON Patch)](https://tools.ietf.org/html/rfc6902) @since version 2.0.0 */ - static basic_json diff(const basic_json& source, - const basic_json& target, + static basic_json diff(const basic_json& source, const basic_json& target, const std::string& path = "") { // the patch @@ -14237,9 +18179,7 @@ scan_number_done: // different types: replace value result.push_back( { - {"op", "replace"}, - {"path", path}, - {"value", target} + {"op", "replace"}, {"path", path}, {"value", target} }); } else @@ -14249,7 +18189,7 @@ scan_number_done: case value_t::array: { // first pass: traverse common elements - size_t i = 0; + std::size_t i = 0; while (i < source.size() and i < target.size()) { // recursive call to compare array values at index i @@ -14293,7 +18233,7 @@ scan_number_done: case value_t::object: { // first pass: traverse this object's elements - for (auto it = source.begin(); it != source.end(); ++it) + for (auto it = source.cbegin(); it != source.cend(); ++it) { // escape the key name to be used in a JSON patch const auto key = json_pointer::escape(it.key()); @@ -14309,14 +18249,13 @@ scan_number_done: // found a key that is not in o -> remove it result.push_back(object( { - {"op", "remove"}, - {"path", path + "/" + key} + {"op", "remove"}, {"path", path + "/" + key} })); } } // second pass: traverse other object's elements - for (auto it = target.begin(); it != target.end(); ++it) + for (auto it = target.cbegin(); it != target.cend(); ++it) { if (source.find(it.key()) == source.end()) { @@ -14324,8 +18263,7 @@ scan_number_done: const auto key = json_pointer::escape(it.key()); result.push_back( { - {"op", "add"}, - {"path", path + "/" + key}, + {"op", "add"}, {"path", path + "/" + key}, {"value", it.value()} }); } @@ -14339,9 +18277,7 @@ scan_number_done: // both primitive type: replace value result.push_back( { - {"op", "replace"}, - {"path", path}, - {"value", target} + {"op", "replace"}, {"path", path}, {"value", target} }); break; } @@ -14352,24 +18288,86 @@ scan_number_done: } /// @} + + //////////////////////////////// + // JSON Merge Patch functions // + //////////////////////////////// + + /// @name JSON Merge Patch functions + /// @{ + + /*! + @brief applies a JSON Merge Patch + + The merge patch format is primarily intended for use with the HTTP PATCH + method as a means of describing a set of modifications to a target + resource's content. This function applies a merge patch to the current + JSON value. + + The function implements the following algorithm from Section 2 of + [RFC 7396 (JSON Merge Patch)](https://tools.ietf.org/html/rfc7396): + + ``` + define MergePatch(Target, Patch): + if Patch is an Object: + if Target is not an Object: + Target = {} // Ignore the contents and set it to an empty Object + for each Name/Value pair in Patch: + if Value is null: + if Name exists in Target: + remove the Name/Value pair from Target + else: + Target[Name] = MergePatch(Target[Name], Value) + return Target + else: + return Patch + ``` + + Thereby, `Target` is the current object; that is, the patch is applied to + the current value. + + @param[in] patch the patch to apply + + @complexity Linear in the lengths of @a patch. + + @liveexample{The following code shows how a JSON Merge Patch is applied to + a JSON document.,merge_patch} + + @sa @ref patch -- apply a JSON patch + @sa [RFC 7396 (JSON Merge Patch)](https://tools.ietf.org/html/rfc7396) + + @since version 3.0.0 + */ + void merge_patch(const basic_json& patch) + { + if (patch.is_object()) + { + if (not is_object()) + { + *this = object(); + } + for (auto it = patch.begin(); it != patch.end(); ++it) + { + if (it.value().is_null()) + { + erase(it.key()); + } + else + { + operator[](it.key()).merge_patch(it.value()); + } + } + } + else + { + *this = patch; + } + } + + /// @} }; - -///////////// -// presets // -///////////// - -/*! -@brief default JSON class - -This type is the default specialization of the @ref basic_json class which -uses the standard template types. - -@since version 1.0.0 -*/ -using json = basic_json<>; } // namespace nlohmann - /////////////////////// // nonmember support // /////////////////////// @@ -14410,8 +18408,10 @@ struct hash }; /// specialization for std::less -template <> -struct less<::nlohmann::detail::value_t> +/// @note: do not remove the space after '<', +/// see https://github.com/nlohmann/json/pull/679 +template<> +struct less< ::nlohmann::detail::value_t> { /*! @brief compare two value_t enum values @@ -14462,6 +18462,9 @@ inline nlohmann::json::json_pointer operator "" _json_pointer(const char* s, std return nlohmann::json::json_pointer(std::string(s, n)); } +// #include + + // restore GCC/clang diagnostic settings #if defined(__clang__) || defined(__GNUC__) || defined(__GNUG__) #pragma GCC diagnostic pop @@ -14477,5 +18480,11 @@ inline nlohmann::json::json_pointer operator "" _json_pointer(const char* s, std #undef JSON_LIKELY #undef JSON_UNLIKELY #undef JSON_DEPRECATED +#undef JSON_HAS_CPP_14 +#undef JSON_HAS_CPP_17 +#undef NLOHMANN_BASIC_JSON_TPL_DECLARATION +#undef NLOHMANN_BASIC_JSON_TPL +#undef NLOHMANN_JSON_HAS_HELPER + #endif diff --git a/src/page.h b/src/page.h index 9e65bbb..0f0f05f 100644 --- a/src/page.h +++ b/src/page.h @@ -114,6 +114,93 @@ namespace std } + +/** + * visitor to produc json representations of + * vallues stored in mstch::node + */ +class mstch_node_to_json: public boost::static_visitor { +public: + +// json operator()(const int& value) const { +// return json {value}; +// } +// +// json operator()(const double& value) const { +// return json {value}; +// } +// +// json operator()(const uint64_t& value) const { +// return json {value}; +// } +// +// json operator()(const int64_t& value) const { +// return json {value}; +// } +// +// json operator()(const uint32_t& value) const { +// return json {value}; +// } +// +// json operator()(const bool& value) const { +// return json {value}; +// } + + // enabled for numeric types + template + std::enable_if_t::value, nlohmann::json> + operator()(T const& value) const { + return nlohmann::json {value}; + } + + nlohmann::json operator()(std::string const& value) const { + return nlohmann::json {value}; + } + + nlohmann::json operator()(mstch::map const& n_map) const + { + nlohmann::json j; + + for (auto const& kv: n_map) + j[kv.first] = boost::apply_visitor(mstch_node_to_json(), kv.second); + + return j; + } + + nlohmann::json operator()(mstch::array const& n_array) const + { + nlohmann::json j; + + for (auto const& v: n_array) + j.push_back(boost::apply_visitor(mstch_node_to_json(), v)); + + return j; + + } + + // catch other types that are non-numeric and not listed above + template + std::enable_if_t::value, nlohmann::json> + operator()(const T&) const { + return nlohmann::json {}; + } + +}; + +namespace mstch +{ +namespace internal +{ + // add conversion from mstch::map to nlohmann::json + void + to_json(nlohmann::json& j, mstch::map const &m) + { + for (auto const& kv: m) + j[kv.first] = boost::apply_visitor(mstch_node_to_json(), kv.second); + } +} +} + namespace xmreg { @@ -125,6 +212,8 @@ using namespace std; using epee::string_tools::pod_to_hex; using epee::string_tools::hex_to_pod; + + /** * @brief The tx_details struct * @@ -273,80 +362,6 @@ struct tx_details }; -class my_render_node: public boost::static_visitor { -public: - -// my_render_node(json& _j) : j {_j} -// {} - - json operator()(const int& value) const { - //return std::to_string(value); - return json {value}; - } - - json operator()(const double& value) const { - return json {value}; - } - - json operator()(const uint64_t& value) const { - return json {value}; - } - - json operator()(const int64_t& value) const { - return json {value}; - } - - json operator()(const uint32_t& value) const { - return json {value}; - } - - json operator()(const bool& value) const { - return json {value}; - } - - json operator()(const string& value) const { - return json {value}; - } - - json operator()(const mstch::map& n_map) const - { - json j; - - for (auto const& kv: n_map) - { - j[kv.first] = boost::apply_visitor(my_render_node(), kv.second); - } - - return j; - - } - - json operator()(const mstch::array& n_array) const - { - json j; - - for (auto const& v: n_array) - j.push_back(boost::apply_visitor(my_render_node(), v)); - - return j; - - } - - json operator()(const mstch::lambda& value) const { - return json {"lambda"}; - } - - template - json operator()(const T&) const { - return json {}; - } - -//private: -// json &j; -}; - - - class page { @@ -4320,24 +4335,27 @@ public: return j_response; } + // get detailed tx information mstch::map tx_context = construct_tx_context(tx, 1 /*full detailed */); - std::string view{"{{#bold}}{{yay}} :){{/bold}}"}; + // remove some page specific and html stuff + tx_context.erase("timescales"); + tx_context.erase("tx_json"); + tx_context.erase("tx_json_raw"); + tx_context.erase("enable_mixins_details"); + tx_context.erase("with_ring_signatures"); + tx_context.erase("show_part_of_inputs"); + tx_context.erase("show_more_details_link"); + tx_context.erase("max_no_of_inputs_to_show"); + tx_context.erase("inputs_xmr_sum_not_zero"); + tx_context.erase("have_raw_tx"); + tx_context.erase("have_any_unknown_amount"); + tx_context.erase("has_error"); + tx_context.erase("error_msg"); + tx_context.erase("server_time"); + tx_context.erase("construction_time"); - json j; - //% boost::apply_visitor(my_render_node(j), kv.second); - - for (auto const& kv: tx_context) - { - //j_data[kv.first] = boost::apply_visitor(render_node2(), kv.second); - - //string a = boost::apply_visitor(my_render_node(), kv.second); - j[kv.first] = boost::apply_visitor(my_render_node(), kv.second); - - //cout << kv.first << " = " << a << endl; - } - - cout << j.dump() << '\n'; + j_data = tx_context; j_response["status"] = "success"; diff --git a/src/tools.h b/src/tools.h index 0321360..1063d20 100644 --- a/src/tools.h +++ b/src/tools.h @@ -370,6 +370,7 @@ calc_median(It it_begin, It it_end) void pause_execution(uint64_t no_seconds, const string& text = "now"); + } #endif //XMREG01_TOOLS_H \ No newline at end of file