forked from luna/jorts
ast_nodes: add ReturnType, Identifier, Import
handling of ReturnType follows the hello.jt rule, that it's implicitly void. - add Reader.peek, a repr, Reader.expect_val, Reader.ignore - add _fn_ret_type for ReturnType generation - fix some things when reading named functions - add read_import - add handler for unexpected reserved word - add some ast cleanups on a read_loop - add ast skip when token is None
This commit is contained in:
parent
b7d5646bad
commit
92a365d908
2 changed files with 130 additions and 14 deletions
|
@ -6,8 +6,24 @@ class TypedVar:
|
||||||
name: str
|
name: str
|
||||||
|
|
||||||
|
|
||||||
|
@dataclass
|
||||||
|
class ReturnType:
|
||||||
|
type_: str
|
||||||
|
|
||||||
|
|
||||||
@dataclass
|
@dataclass
|
||||||
class Function:
|
class Function:
|
||||||
name: str
|
name: str
|
||||||
arguments: str
|
arguments: str
|
||||||
|
ret_type: ReturnType
|
||||||
block: list
|
block: list
|
||||||
|
|
||||||
|
|
||||||
|
@dataclass
|
||||||
|
class Identifier:
|
||||||
|
name: str
|
||||||
|
|
||||||
|
|
||||||
|
@dataclass
|
||||||
|
class Import:
|
||||||
|
module: str
|
||||||
|
|
|
@ -1,7 +1,9 @@
|
||||||
from typing import Optional, Any, List
|
from typing import Optional, Any, List
|
||||||
|
|
||||||
from jortsc.parser.lexer import Token, TokenType
|
from jortsc.parser.lexer import Token, TokenType
|
||||||
from jortsc.parser.ast_nodes import Function, TypedVar
|
from jortsc.parser.ast_nodes import (
|
||||||
|
Function, TypedVar, Identifier, Import, ReturnType
|
||||||
|
)
|
||||||
|
|
||||||
|
|
||||||
class ParseError(Exception):
|
class ParseError(Exception):
|
||||||
|
@ -15,13 +17,20 @@ class Reader:
|
||||||
self.tokens = tokens
|
self.tokens = tokens
|
||||||
self.cur = 0
|
self.cur = 0
|
||||||
|
|
||||||
def next(self) -> Optional[Token]:
|
def __repr__(self):
|
||||||
"""Fetch the current token then skip to the next one."""
|
return f'<Reader cur={self.cur} cur_tok={self.peek()}>'
|
||||||
|
|
||||||
|
def peek(self) -> Optional[Token]:
|
||||||
|
"""Peek at the current token."""
|
||||||
try:
|
try:
|
||||||
token = self.tokens[self.cur]
|
token = self.tokens[self.cur]
|
||||||
|
return token
|
||||||
except IndexError:
|
except IndexError:
|
||||||
return None
|
return None
|
||||||
|
|
||||||
|
def next(self) -> Optional[Token]:
|
||||||
|
"""Fetch the current token then skip to the next one."""
|
||||||
|
token = self.peek()
|
||||||
self.cur += 1
|
self.cur += 1
|
||||||
return token
|
return token
|
||||||
|
|
||||||
|
@ -30,10 +39,23 @@ class Reader:
|
||||||
token = self.next()
|
token = self.next()
|
||||||
|
|
||||||
if token.type_ != token_type:
|
if token.type_ != token_type:
|
||||||
raise ParseError(f'Expected {token_type}, got {token.type_}')
|
raise ParseError(f'Expected {token_type}, got '
|
||||||
|
f'{token.type_} {token.value!r}')
|
||||||
|
|
||||||
return token
|
return token
|
||||||
|
|
||||||
|
def expect_val(self, value: str) -> Token:
|
||||||
|
"""Check the next token to see if it matches against a given value,
|
||||||
|
instead of a type."""
|
||||||
|
token = self.next()
|
||||||
|
|
||||||
|
if token.value != value:
|
||||||
|
raise ParseError(f'Expected {value!r}, got '
|
||||||
|
f'{token.type_} {token.value!r}')
|
||||||
|
|
||||||
|
return token
|
||||||
|
|
||||||
|
|
||||||
def next_safe(self) -> Token:
|
def next_safe(self) -> Token:
|
||||||
"""'Safe' version of next().
|
"""'Safe' version of next().
|
||||||
|
|
||||||
|
@ -46,6 +68,14 @@ class Reader:
|
||||||
|
|
||||||
return token
|
return token
|
||||||
|
|
||||||
|
def ignore(self, token_type: TokenType):
|
||||||
|
"""Only increase self.cur if token_type is the upcoming token."""
|
||||||
|
try:
|
||||||
|
assert self.tokens[self.cur].type_ == token_type
|
||||||
|
self.cur += 1
|
||||||
|
except AssertionError:
|
||||||
|
pass
|
||||||
|
|
||||||
|
|
||||||
def _fn_read_args(reader: Reader, cur: List = None) -> List:
|
def _fn_read_args(reader: Reader, cur: List = None) -> List:
|
||||||
"""Recursively read the arguments of the function."""
|
"""Recursively read the arguments of the function."""
|
||||||
|
@ -68,6 +98,18 @@ def _fn_read_args(reader: Reader, cur: List = None) -> List:
|
||||||
return _fn_read_args(reader, cur)
|
return _fn_read_args(reader, cur)
|
||||||
|
|
||||||
|
|
||||||
|
def _fn_ret_type(reader: Reader) -> ReturnType:
|
||||||
|
"""Fetch the return type of a function. Defaults to void."""
|
||||||
|
try:
|
||||||
|
reader.expect_val('->')
|
||||||
|
except ParseError:
|
||||||
|
return ReturnType('void')
|
||||||
|
|
||||||
|
reader.ignore(TokenType.whitespace)
|
||||||
|
token = reader.expect(TokenType.identifier)
|
||||||
|
return ReturnType(token.value)
|
||||||
|
|
||||||
|
|
||||||
def read_function(reader: Reader):
|
def read_function(reader: Reader):
|
||||||
"""Read a function block."""
|
"""Read a function block."""
|
||||||
reader.expect(TokenType.whitespace)
|
reader.expect(TokenType.whitespace)
|
||||||
|
@ -81,19 +123,49 @@ def read_function(reader: Reader):
|
||||||
|
|
||||||
if token.type_ == TokenType.identifier:
|
if token.type_ == TokenType.identifier:
|
||||||
fn_name = token.value
|
fn_name = token.value
|
||||||
|
|
||||||
|
reader.expect(TokenType.whitespace)
|
||||||
|
reader.expect_val('(')
|
||||||
|
|
||||||
fn_args = _fn_read_args(reader)
|
fn_args = _fn_read_args(reader)
|
||||||
|
|
||||||
|
reader.expect(TokenType.whitespace)
|
||||||
|
fn_ret_type = _fn_ret_type(reader)
|
||||||
|
|
||||||
|
# only skip whitespace if we see it
|
||||||
|
reader.ignore(TokenType.whitespace)
|
||||||
block = read_start(reader)
|
block = read_start(reader)
|
||||||
elif token.value == '(':
|
elif token.value == '(':
|
||||||
fn_args = _fn_read_args(reader)
|
fn_args = _fn_read_args(reader)
|
||||||
|
fn_ret_type = _fn_ret_type(reader)
|
||||||
block = read_start(reader)
|
block = read_start(reader)
|
||||||
|
|
||||||
return Function(fn_name, fn_args, block)
|
print('final function', fn_name, fn_args, fn_ret_type, block)
|
||||||
|
|
||||||
|
return Function(fn_name, fn_args, fn_ret_type, block)
|
||||||
|
|
||||||
|
|
||||||
|
def read_import(reader):
|
||||||
|
"""Read an import"""
|
||||||
|
reader.expect(TokenType.whitespace)
|
||||||
|
module = reader.next_safe()
|
||||||
|
return Import(module.value)
|
||||||
|
|
||||||
|
|
||||||
|
HANDLERS = {
|
||||||
|
'fn': read_function,
|
||||||
|
'import': read_import,
|
||||||
|
}
|
||||||
|
|
||||||
|
|
||||||
def read_reserved(token: Token, reader: Reader):
|
def read_reserved(token: Token, reader: Reader):
|
||||||
"""Read reserved statements."""
|
"""Read reserved statements."""
|
||||||
if token.value == 'fn':
|
try:
|
||||||
return read_function(reader)
|
handler = HANDLERS[token.value]
|
||||||
|
except KeyError:
|
||||||
|
raise ParseError(f'Unexpected reserved word {token.value!r}')
|
||||||
|
|
||||||
|
return handler(reader)
|
||||||
|
|
||||||
|
|
||||||
def read_start(reader: Reader):
|
def read_start(reader: Reader):
|
||||||
|
@ -101,29 +173,57 @@ def read_start(reader: Reader):
|
||||||
print('reader', reader)
|
print('reader', reader)
|
||||||
|
|
||||||
token = reader.next()
|
token = reader.next()
|
||||||
|
|
||||||
|
if token is None:
|
||||||
|
return []
|
||||||
|
|
||||||
ast = []
|
ast = []
|
||||||
res = []
|
res = []
|
||||||
|
|
||||||
print('cur', token)
|
print('cur', token)
|
||||||
|
|
||||||
# TODO: statement handling
|
|
||||||
if token.type_ == TokenType.reserved:
|
|
||||||
res = read_reserved(token, reader)
|
|
||||||
|
|
||||||
# handle blocks
|
# handle blocks
|
||||||
elif token.value == '{':
|
if token.value == '{':
|
||||||
while True:
|
while True:
|
||||||
token = reader.peek()
|
token = reader.next()
|
||||||
|
|
||||||
if token.value == '}':
|
if token.value == '}':
|
||||||
break
|
break
|
||||||
|
|
||||||
res.extend(read_start(reader))
|
res.extend(read_start(reader))
|
||||||
|
|
||||||
|
# import, fn, etc
|
||||||
|
elif token.type_ == TokenType.reserved:
|
||||||
|
res = read_reserved(token, reader)
|
||||||
|
|
||||||
|
elif token.type_ == TokenType.comment:
|
||||||
|
return []
|
||||||
|
|
||||||
|
elif token.type_ == TokenType.identifier:
|
||||||
|
res = [Identifier(token.value)]
|
||||||
|
|
||||||
ast.append(res)
|
ast.append(res)
|
||||||
return ast
|
return ast
|
||||||
|
|
||||||
|
def read_loop(reader: Reader):
|
||||||
|
final_ast = []
|
||||||
|
|
||||||
|
while True:
|
||||||
|
ast = read_start(reader)
|
||||||
|
|
||||||
|
if not ast:
|
||||||
|
break
|
||||||
|
|
||||||
|
inner = ast[0]
|
||||||
|
|
||||||
|
if not inner:
|
||||||
|
break
|
||||||
|
|
||||||
|
final_ast.append(ast)
|
||||||
|
|
||||||
|
return final_ast
|
||||||
|
|
||||||
|
|
||||||
def syntatic(tokens: List[Token]):
|
def syntatic(tokens: List[Token]):
|
||||||
"""Create an AST out of the tokens."""
|
"""Create an AST out of the tokens."""
|
||||||
return read_start(Reader(tokens))
|
return read_loop(Reader(tokens))
|
||||||
|
|
Loading…
Reference in a new issue