"""Small, fast HTTP client library for Python. Features persistent connections, cache, and Google App Engine Standard Environment support. """ from __future__ import print_function __author__ = "Joe Gregorio (joe@bitworking.org)" __copyright__ = "Copyright 2006, Joe Gregorio" __contributors__ = [ "Thomas Broyer (t.broyer@ltgt.net)", "James Antill", "Xavier Verges Farrero", "Jonathan Feinberg", "Blair Zajac", "Sam Ruby", "Louis Nyffenegger", "Alex Yu", ] __license__ = "MIT" __version__ = '0.17.0' import base64 import calendar import copy import email import email.FeedParser import email.Message import email.Utils import errno import gzip import httplib import os import random import re import StringIO import sys import time import urllib import urlparse import zlib try: from hashlib import sha1 as _sha, md5 as _md5 except ImportError: # prior to Python 2.5, these were separate modules import sha import md5 _sha = sha.new _md5 = md5.new import hmac from gettext import gettext as _ import socket try: import socks except ImportError: try: import socks except (ImportError, AttributeError): socks = None # Build the appropriate socket wrapper for ssl ssl = None ssl_SSLError = None ssl_CertificateError = None try: import ssl # python 2.6 except ImportError: pass if ssl is not None: ssl_SSLError = getattr(ssl, "SSLError", None) ssl_CertificateError = getattr(ssl, "CertificateError", None) def _ssl_wrap_socket( sock, key_file, cert_file, disable_validation, ca_certs, ssl_version, hostname, key_password ): if disable_validation: cert_reqs = ssl.CERT_NONE else: cert_reqs = ssl.CERT_REQUIRED if ssl_version is None: ssl_version = ssl.PROTOCOL_SSLv23 if hasattr(ssl, "SSLContext"): # Python 2.7.9 context = ssl.SSLContext(ssl_version) context.verify_mode = cert_reqs context.check_hostname = cert_reqs != ssl.CERT_NONE if cert_file: if key_password: context.load_cert_chain(cert_file, key_file, key_password) else: context.load_cert_chain(cert_file, key_file) if ca_certs: context.load_verify_locations(ca_certs) return context.wrap_socket(sock, server_hostname=hostname) else: if key_password: raise NotSupportedOnThisPlatform("Certificate with password is not supported.") return ssl.wrap_socket( sock, keyfile=key_file, certfile=cert_file, cert_reqs=cert_reqs, ca_certs=ca_certs, ssl_version=ssl_version, ) def _ssl_wrap_socket_unsupported( sock, key_file, cert_file, disable_validation, ca_certs, ssl_version, hostname, key_password ): if not disable_validation: raise CertificateValidationUnsupported( "SSL certificate validation is not supported without " "the ssl module installed. To avoid this error, install " "the ssl module, or explicity disable validation." ) if key_password: raise NotSupportedOnThisPlatform("Certificate with password is not supported.") ssl_sock = socket.ssl(sock, key_file, cert_file) return httplib.FakeSocket(sock, ssl_sock) if ssl is None: _ssl_wrap_socket = _ssl_wrap_socket_unsupported if sys.version_info >= (2, 3): from iri2uri import iri2uri else: def iri2uri(uri): return uri def has_timeout(timeout): # python 2.6 if hasattr(socket, "_GLOBAL_DEFAULT_TIMEOUT"): return timeout is not None and timeout is not socket._GLOBAL_DEFAULT_TIMEOUT return timeout is not None __all__ = [ "Http", "Response", "ProxyInfo", "HttpLib2Error", "RedirectMissingLocation", "RedirectLimit", "FailedToDecompressContent", "UnimplementedDigestAuthOptionError", "UnimplementedHmacDigestAuthOptionError", "debuglevel", "ProxiesUnavailableError", ] # The httplib debug level, set to a non-zero value to get debug output debuglevel = 0 # A request will be tried 'RETRIES' times if it fails at the socket/connection level. RETRIES = 2 # Python 2.3 support if sys.version_info < (2, 4): def sorted(seq): seq.sort() return seq # Python 2.3 support def HTTPResponse__getheaders(self): """Return list of (header, value) tuples.""" if self.msg is None: raise httplib.ResponseNotReady() return self.msg.items() if not hasattr(httplib.HTTPResponse, "getheaders"): httplib.HTTPResponse.getheaders = HTTPResponse__getheaders # All exceptions raised here derive from HttpLib2Error class HttpLib2Error(Exception): pass # Some exceptions can be caught and optionally # be turned back into responses. class HttpLib2ErrorWithResponse(HttpLib2Error): def __init__(self, desc, response, content): self.response = response self.content = content HttpLib2Error.__init__(self, desc) class RedirectMissingLocation(HttpLib2ErrorWithResponse): pass class RedirectLimit(HttpLib2ErrorWithResponse): pass class FailedToDecompressContent(HttpLib2ErrorWithResponse): pass class UnimplementedDigestAuthOptionError(HttpLib2ErrorWithResponse): pass class UnimplementedHmacDigestAuthOptionError(HttpLib2ErrorWithResponse): pass class MalformedHeader(HttpLib2Error): pass class RelativeURIError(HttpLib2Error): pass class ServerNotFoundError(HttpLib2Error): pass class ProxiesUnavailableError(HttpLib2Error): pass class CertificateValidationUnsupported(HttpLib2Error): pass class SSLHandshakeError(HttpLib2Error): pass class NotSupportedOnThisPlatform(HttpLib2Error): pass class CertificateHostnameMismatch(SSLHandshakeError): def __init__(self, desc, host, cert): HttpLib2Error.__init__(self, desc) self.host = host self.cert = cert class NotRunningAppEngineEnvironment(HttpLib2Error): pass # Open Items: # ----------- # Proxy support # Are we removing the cached content too soon on PUT (only delete on 200 Maybe?) # Pluggable cache storage (supports storing the cache in # flat files by default. We need a plug-in architecture # that can support Berkeley DB and Squid) # == Known Issues == # Does not handle a resource that uses conneg and Last-Modified but no ETag as a cache validator. # Does not handle Cache-Control: max-stale # Does not use Age: headers when calculating cache freshness. # The number of redirections to follow before giving up. # Note that only GET redirects are automatically followed. # Will also honor 301 requests by saving that info and never # requesting that URI again. DEFAULT_MAX_REDIRECTS = 5 import certs CA_CERTS = certs.where() # Which headers are hop-by-hop headers by default HOP_BY_HOP = [ "connection", "keep-alive", "proxy-authenticate", "proxy-authorization", "te", "trailers", "transfer-encoding", "upgrade", ] # https://tools.ietf.org/html/rfc7231#section-8.1.3 SAFE_METHODS = ("GET", "HEAD") # TODO add "OPTIONS", "TRACE" # To change, assign to `Http().redirect_codes` REDIRECT_CODES = frozenset((300, 301, 302, 303, 307, 308)) def _get_end2end_headers(response): hopbyhop = list(HOP_BY_HOP) hopbyhop.extend([x.strip() for x in response.get("connection", "").split(",")]) return [header for header in response.keys() if header not in hopbyhop] URI = re.compile(r"^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))?") def parse_uri(uri): """Parses a URI using the regex given in Appendix B of RFC 3986. (scheme, authority, path, query, fragment) = parse_uri(uri) """ groups = URI.match(uri).groups() return (groups[1], groups[3], groups[4], groups[6], groups[8]) def urlnorm(uri): (scheme, authority, path, query, fragment) = parse_uri(uri) if not scheme or not authority: raise RelativeURIError("Only absolute URIs are allowed. uri = %s" % uri) authority = authority.lower() scheme = scheme.lower() if not path: path = "/" # Could do syntax based normalization of the URI before # computing the digest. See Section 6.2.2 of Std 66. request_uri = query and "?".join([path, query]) or path scheme = scheme.lower() defrag_uri = scheme + "://" + authority + request_uri return scheme, authority, request_uri, defrag_uri # Cache filename construction (original borrowed from Venus http://intertwingly.net/code/venus/) re_url_scheme = re.compile(r"^\w+://") re_unsafe = re.compile(r"[^\w\-_.()=!]+") def safename(filename): """Return a filename suitable for the cache. Strips dangerous and common characters to create a filename we can use to store the cache in. """ if isinstance(filename, str): filename_bytes = filename filename = filename.decode("utf-8") else: filename_bytes = filename.encode("utf-8") filemd5 = _md5(filename_bytes).hexdigest() filename = re_url_scheme.sub("", filename) filename = re_unsafe.sub("", filename) # limit length of filename (vital for Windows) # https://github.com/httplib2/httplib2/pull/74 # C:\Users\ \AppData\Local\Temp\ , # 9 chars + max 104 chars + 20 chars + x + 1 + 32 = max 259 chars # Thus max safe filename x = 93 chars. Let it be 90 to make a round sum: filename = filename[:90] return ",".join((filename, filemd5)) NORMALIZE_SPACE = re.compile(r"(?:\r\n)?[ \t]+") def _normalize_headers(headers): return dict( [ (key.lower(), NORMALIZE_SPACE.sub(value, " ").strip()) for (key, value) in headers.iteritems() ] ) def _parse_cache_control(headers): retval = {} if "cache-control" in headers: parts = headers["cache-control"].split(",") parts_with_args = [ tuple([x.strip().lower() for x in part.split("=", 1)]) for part in parts if -1 != part.find("=") ] parts_wo_args = [ (name.strip().lower(), 1) for name in parts if -1 == name.find("=") ] retval = dict(parts_with_args + parts_wo_args) return retval # Whether to use a strict mode to parse WWW-Authenticate headers # Might lead to bad results in case of ill-formed header value, # so disabled by default, falling back to relaxed parsing. # Set to true to turn on, usefull for testing servers. USE_WWW_AUTH_STRICT_PARSING = 0 # In regex below: # [^\0-\x1f\x7f-\xff()<>@,;:\\\"/[\]?={} \t]+ matches a "token" as defined by HTTP # "(?:[^\0-\x08\x0A-\x1f\x7f-\xff\\\"]|\\[\0-\x7f])*?" matches a "quoted-string" as defined by HTTP, when LWS have already been replaced by a single space # Actually, as an auth-param value can be either a token or a quoted-string, they are combined in a single pattern which matches both: # \"?((?<=\")(?:[^\0-\x1f\x7f-\xff\\\"]|\\[\0-\x7f])*?(?=\")|(?@,;:\\\"/[\]?={} \t]+(?!\"))\"? WWW_AUTH_STRICT = re.compile( r"^(?:\s*(?:,\s*)?([^\0-\x1f\x7f-\xff()<>@,;:\\\"/[\]?={} \t]+)\s*=\s*\"?((?<=\")(?:[^\0-\x08\x0A-\x1f\x7f-\xff\\\"]|\\[\0-\x7f])*?(?=\")|(?@,;:\\\"/[\]?={} \t]+(?!\"))\"?)(.*)$" ) WWW_AUTH_RELAXED = re.compile( r"^(?:\s*(?:,\s*)?([^ \t\r\n=]+)\s*=\s*\"?((?<=\")(?:[^\\\"]|\\.)*?(?=\")|(? current_age: retval = "FRESH" return retval def _decompressContent(response, new_content): content = new_content try: encoding = response.get("content-encoding", None) if encoding in ["gzip", "deflate"]: if encoding == "gzip": content = gzip.GzipFile(fileobj=StringIO.StringIO(new_content)).read() if encoding == "deflate": content = zlib.decompress(content, -zlib.MAX_WBITS) response["content-length"] = str(len(content)) # Record the historical presence of the encoding in a way the won't interfere. response["-content-encoding"] = response["content-encoding"] del response["content-encoding"] except (IOError, zlib.error): content = "" raise FailedToDecompressContent( _("Content purported to be compressed with %s but failed to decompress.") % response.get("content-encoding"), response, content, ) return content def _updateCache(request_headers, response_headers, content, cache, cachekey): if cachekey: cc = _parse_cache_control(request_headers) cc_response = _parse_cache_control(response_headers) if "no-store" in cc or "no-store" in cc_response: cache.delete(cachekey) else: info = email.Message.Message() for key, value in response_headers.iteritems(): if key not in ["status", "content-encoding", "transfer-encoding"]: info[key] = value # Add annotations to the cache to indicate what headers # are variant for this request. vary = response_headers.get("vary", None) if vary: vary_headers = vary.lower().replace(" ", "").split(",") for header in vary_headers: key = "-varied-%s" % header try: info[key] = request_headers[header] except KeyError: pass status = response_headers.status if status == 304: status = 200 status_header = "status: %d\r\n" % status header_str = info.as_string() header_str = re.sub("\r(?!\n)|(? 0: service = "cl" # No point in guessing Base or Spreadsheet # elif request_uri.find("spreadsheets") > 0: # service = "wise" auth = dict( Email=credentials[0], Passwd=credentials[1], service=service, source=headers["user-agent"], ) resp, content = self.http.request( "https://www.google.com/accounts/ClientLogin", method="POST", body=urlencode(auth), headers={"Content-Type": "application/x-www-form-urlencoded"}, ) lines = content.split("\n") d = dict([tuple(line.split("=", 1)) for line in lines if line]) if resp.status == 403: self.Auth = "" else: self.Auth = d["Auth"] def request(self, method, request_uri, headers, content): """Modify the request headers to add the appropriate Authorization header.""" headers["authorization"] = "GoogleLogin Auth=" + self.Auth AUTH_SCHEME_CLASSES = { "basic": BasicAuthentication, "wsse": WsseAuthentication, "digest": DigestAuthentication, "hmacdigest": HmacDigestAuthentication, "googlelogin": GoogleLoginAuthentication, } AUTH_SCHEME_ORDER = ["hmacdigest", "googlelogin", "digest", "wsse", "basic"] class FileCache(object): """Uses a local directory as a store for cached files. Not really safe to use if multiple threads or processes are going to be running on the same cache. """ def __init__( self, cache, safe=safename ): # use safe=lambda x: md5.new(x).hexdigest() for the old behavior self.cache = cache self.safe = safe if not os.path.exists(cache): os.makedirs(self.cache) def get(self, key): retval = None cacheFullPath = os.path.join(self.cache, self.safe(key)) try: f = file(cacheFullPath, "rb") retval = f.read() f.close() except IOError: pass return retval def set(self, key, value): cacheFullPath = os.path.join(self.cache, self.safe(key)) f = file(cacheFullPath, "wb") f.write(value) f.close() def delete(self, key): cacheFullPath = os.path.join(self.cache, self.safe(key)) if os.path.exists(cacheFullPath): os.remove(cacheFullPath) class Credentials(object): def __init__(self): self.credentials = [] def add(self, name, password, domain=""): self.credentials.append((domain.lower(), name, password)) def clear(self): self.credentials = [] def iter(self, domain): for (cdomain, name, password) in self.credentials: if cdomain == "" or domain == cdomain: yield (name, password) class KeyCerts(Credentials): """Identical to Credentials except that name/password are mapped to key/cert.""" def add(self, key, cert, domain, password): self.credentials.append((domain.lower(), key, cert, password)) def iter(self, domain): for (cdomain, key, cert, password) in self.credentials: if cdomain == "" or domain == cdomain: yield (key, cert, password) class AllHosts(object): pass class ProxyInfo(object): """Collect information required to use a proxy.""" bypass_hosts = () def __init__( self, proxy_type, proxy_host, proxy_port, proxy_rdns=True, proxy_user=None, proxy_pass=None, proxy_headers=None, ): """Args: proxy_type: The type of proxy server. This must be set to one of socks.PROXY_TYPE_XXX constants. For example: p = ProxyInfo(proxy_type=socks.PROXY_TYPE_HTTP, proxy_host='localhost', proxy_port=8000) proxy_host: The hostname or IP address of the proxy server. proxy_port: The port that the proxy server is running on. proxy_rdns: If True (default), DNS queries will not be performed locally, and instead, handed to the proxy to resolve. This is useful if the network does not allow resolution of non-local names. In httplib2 0.9 and earlier, this defaulted to False. proxy_user: The username used to authenticate with the proxy server. proxy_pass: The password used to authenticate with the proxy server. proxy_headers: Additional or modified headers for the proxy connect request. """ self.proxy_type = proxy_type self.proxy_host = proxy_host self.proxy_port = proxy_port self.proxy_rdns = proxy_rdns self.proxy_user = proxy_user self.proxy_pass = proxy_pass self.proxy_headers = proxy_headers def astuple(self): return ( self.proxy_type, self.proxy_host, self.proxy_port, self.proxy_rdns, self.proxy_user, self.proxy_pass, self.proxy_headers, ) def isgood(self): return (self.proxy_host != None) and (self.proxy_port != None) def applies_to(self, hostname): return not self.bypass_host(hostname) def bypass_host(self, hostname): """Has this host been excluded from the proxy config""" if self.bypass_hosts is AllHosts: return True hostname = "." + hostname.lstrip(".") for skip_name in self.bypass_hosts: # *.suffix if skip_name.startswith(".") and hostname.endswith(skip_name): return True # exact match if hostname == "." + skip_name: return True return False def __repr__(self): return ( "" ).format(p=self) def proxy_info_from_environment(method="http"): """Read proxy info from the environment variables. """ if method not in ["http", "https"]: return env_var = method + "_proxy" url = os.environ.get(env_var, os.environ.get(env_var.upper())) if not url: return return proxy_info_from_url(url, method, None) def proxy_info_from_url(url, method="http", noproxy=None): """Construct a ProxyInfo from a URL (such as http_proxy env var) """ url = urlparse.urlparse(url) username = None password = None port = None if "@" in url[1]: ident, host_port = url[1].split("@", 1) if ":" in ident: username, password = ident.split(":", 1) else: password = ident else: host_port = url[1] if ":" in host_port: host, port = host_port.split(":", 1) else: host = host_port if port: port = int(port) else: port = dict(https=443, http=80)[method] proxy_type = 3 # socks.PROXY_TYPE_HTTP pi = ProxyInfo( proxy_type=proxy_type, proxy_host=host, proxy_port=port, proxy_user=username or None, proxy_pass=password or None, proxy_headers=None, ) bypass_hosts = [] # If not given an explicit noproxy value, respect values in env vars. if noproxy is None: noproxy = os.environ.get("no_proxy", os.environ.get("NO_PROXY", "")) # Special case: A single '*' character means all hosts should be bypassed. if noproxy == "*": bypass_hosts = AllHosts elif noproxy.strip(): bypass_hosts = noproxy.split(",") bypass_hosts = filter(bool, bypass_hosts) # To exclude empty string. pi.bypass_hosts = bypass_hosts return pi class HTTPConnectionWithTimeout(httplib.HTTPConnection): """HTTPConnection subclass that supports timeouts All timeouts are in seconds. If None is passed for timeout then Python's default timeout for sockets will be used. See for example the docs of socket.setdefaulttimeout(): http://docs.python.org/library/socket.html#socket.setdefaulttimeout """ def __init__(self, host, port=None, strict=None, timeout=None, proxy_info=None): httplib.HTTPConnection.__init__(self, host, port, strict) self.timeout = timeout self.proxy_info = proxy_info def connect(self): """Connect to the host and port specified in __init__.""" # Mostly verbatim from httplib.py. if self.proxy_info and socks is None: raise ProxiesUnavailableError( "Proxy support missing but proxy use was requested!" ) if self.proxy_info and self.proxy_info.isgood(): use_proxy = True proxy_type, proxy_host, proxy_port, proxy_rdns, proxy_user, proxy_pass, proxy_headers = ( self.proxy_info.astuple() ) host = proxy_host port = proxy_port else: use_proxy = False host = self.host port = self.port socket_err = None for res in socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM): af, socktype, proto, canonname, sa = res try: if use_proxy: self.sock = socks.socksocket(af, socktype, proto) self.sock.setproxy( proxy_type, proxy_host, proxy_port, proxy_rdns, proxy_user, proxy_pass, proxy_headers, ) else: self.sock = socket.socket(af, socktype, proto) self.sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1) # Different from httplib: support timeouts. if has_timeout(self.timeout): self.sock.settimeout(self.timeout) # End of difference from httplib. if self.debuglevel > 0: print("connect: (%s, %s) ************" % (self.host, self.port)) if use_proxy: print( "proxy: %s ************" % str( ( proxy_host, proxy_port, proxy_rdns, proxy_user, proxy_pass, proxy_headers, ) ) ) if use_proxy: self.sock.connect((self.host, self.port) + sa[2:]) else: self.sock.connect(sa) except socket.error as e: socket_err = e if self.debuglevel > 0: print("connect fail: (%s, %s)" % (self.host, self.port)) if use_proxy: print( "proxy: %s" % str( ( proxy_host, proxy_port, proxy_rdns, proxy_user, proxy_pass, proxy_headers, ) ) ) if self.sock: self.sock.close() self.sock = None continue break if not self.sock: raise socket_err or socket.error("getaddrinfo returns an empty list") class HTTPSConnectionWithTimeout(httplib.HTTPSConnection): """This class allows communication via SSL. All timeouts are in seconds. If None is passed for timeout then Python's default timeout for sockets will be used. See for example the docs of socket.setdefaulttimeout(): http://docs.python.org/library/socket.html#socket.setdefaulttimeout """ def __init__( self, host, port=None, key_file=None, cert_file=None, strict=None, timeout=None, proxy_info=None, ca_certs=None, disable_ssl_certificate_validation=False, ssl_version=None, key_password=None, ): if key_password: httplib.HTTPSConnection.__init__(self, host, port=port, strict=strict) self._context.load_cert_chain(cert_file, key_file, key_password) self.key_file = key_file self.cert_file = cert_file self.key_password = key_password else: httplib.HTTPSConnection.__init__( self, host, port=port, key_file=key_file, cert_file=cert_file, strict=strict ) self.key_password = None self.timeout = timeout self.proxy_info = proxy_info if ca_certs is None: ca_certs = CA_CERTS self.ca_certs = ca_certs self.disable_ssl_certificate_validation = disable_ssl_certificate_validation self.ssl_version = ssl_version # The following two methods were adapted from https_wrapper.py, released # with the Google Appengine SDK at # http://googleappengine.googlecode.com/svn-history/r136/trunk/python/google/appengine/tools/https_wrapper.py # under the following license: # # Copyright 2007 Google Inc. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # def _GetValidHostsForCert(self, cert): """Returns a list of valid host globs for an SSL certificate. Args: cert: A dictionary representing an SSL certificate. Returns: list: A list of valid host globs. """ if "subjectAltName" in cert: return [x[1] for x in cert["subjectAltName"] if x[0].lower() == "dns"] else: return [x[0][1] for x in cert["subject"] if x[0][0].lower() == "commonname"] def _ValidateCertificateHostname(self, cert, hostname): """Validates that a given hostname is valid for an SSL certificate. Args: cert: A dictionary representing an SSL certificate. hostname: The hostname to test. Returns: bool: Whether or not the hostname is valid for this certificate. """ hosts = self._GetValidHostsForCert(cert) for host in hosts: host_re = host.replace(".", "\.").replace("*", "[^.]*") if re.search("^%s$" % (host_re,), hostname, re.I): return True return False def connect(self): "Connect to a host on a given (SSL) port." if self.proxy_info and self.proxy_info.isgood(): use_proxy = True proxy_type, proxy_host, proxy_port, proxy_rdns, proxy_user, proxy_pass, proxy_headers = ( self.proxy_info.astuple() ) host = proxy_host port = proxy_port else: use_proxy = False host = self.host port = self.port socket_err = None address_info = socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM) for family, socktype, proto, canonname, sockaddr in address_info: try: if use_proxy: sock = socks.socksocket(family, socktype, proto) sock.setproxy( proxy_type, proxy_host, proxy_port, proxy_rdns, proxy_user, proxy_pass, proxy_headers, ) else: sock = socket.socket(family, socktype, proto) sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1) if has_timeout(self.timeout): sock.settimeout(self.timeout) if use_proxy: sock.connect((self.host, self.port) + sockaddr[:2]) else: sock.connect(sockaddr) self.sock = _ssl_wrap_socket( sock, self.key_file, self.cert_file, self.disable_ssl_certificate_validation, self.ca_certs, self.ssl_version, self.host, self.key_password, ) if self.debuglevel > 0: print("connect: (%s, %s)" % (self.host, self.port)) if use_proxy: print( "proxy: %s" % str( ( proxy_host, proxy_port, proxy_rdns, proxy_user, proxy_pass, proxy_headers, ) ) ) if not self.disable_ssl_certificate_validation: cert = self.sock.getpeercert() hostname = self.host.split(":", 0)[0] if not self._ValidateCertificateHostname(cert, hostname): raise CertificateHostnameMismatch( "Server presented certificate that does not match " "host %s: %s" % (hostname, cert), hostname, cert, ) except ( ssl_SSLError, ssl_CertificateError, CertificateHostnameMismatch, ) as e: if sock: sock.close() if self.sock: self.sock.close() self.sock = None # Unfortunately the ssl module doesn't seem to provide any way # to get at more detailed error information, in particular # whether the error is due to certificate validation or # something else (such as SSL protocol mismatch). if getattr(e, "errno", None) == ssl.SSL_ERROR_SSL: raise SSLHandshakeError(e) else: raise except (socket.timeout, socket.gaierror): raise except socket.error as e: socket_err = e if self.debuglevel > 0: print("connect fail: (%s, %s)" % (self.host, self.port)) if use_proxy: print( "proxy: %s" % str( ( proxy_host, proxy_port, proxy_rdns, proxy_user, proxy_pass, proxy_headers, ) ) ) if self.sock: self.sock.close() self.sock = None continue break if not self.sock: raise socket_err or socket.error("getaddrinfo returns an empty list") SCHEME_TO_CONNECTION = { "http": HTTPConnectionWithTimeout, "https": HTTPSConnectionWithTimeout, } def _new_fixed_fetch(validate_certificate): def fixed_fetch( url, payload=None, method="GET", headers={}, allow_truncated=False, follow_redirects=True, deadline=None, ): return fetch( url, payload=payload, method=method, headers=headers, allow_truncated=allow_truncated, follow_redirects=follow_redirects, deadline=deadline, validate_certificate=validate_certificate, ) return fixed_fetch class AppEngineHttpConnection(httplib.HTTPConnection): """Use httplib on App Engine, but compensate for its weirdness. The parameters key_file, cert_file, proxy_info, ca_certs, disable_ssl_certificate_validation, and ssl_version are all dropped on the ground. """ def __init__( self, host, port=None, key_file=None, cert_file=None, strict=None, timeout=None, proxy_info=None, ca_certs=None, disable_ssl_certificate_validation=False, ssl_version=None, ): httplib.HTTPConnection.__init__( self, host, port=port, strict=strict, timeout=timeout ) class AppEngineHttpsConnection(httplib.HTTPSConnection): """Same as AppEngineHttpConnection, but for HTTPS URIs. The parameters proxy_info, ca_certs, disable_ssl_certificate_validation, and ssl_version are all dropped on the ground. """ def __init__( self, host, port=None, key_file=None, cert_file=None, strict=None, timeout=None, proxy_info=None, ca_certs=None, disable_ssl_certificate_validation=False, ssl_version=None, key_password=None, ): if key_password: raise NotSupportedOnThisPlatform("Certificate with password is not supported.") httplib.HTTPSConnection.__init__( self, host, port=port, key_file=key_file, cert_file=cert_file, strict=strict, timeout=timeout, ) self._fetch = _new_fixed_fetch(not disable_ssl_certificate_validation) # Use a different connection object for Google App Engine Standard Environment. def is_gae_instance(): server_software = os.environ.get('SERVER_SOFTWARE', '') if (server_software.startswith('Google App Engine/') or server_software.startswith('Development/') or server_software.startswith('testutil/')): return True return False try: if not is_gae_instance(): raise NotRunningAppEngineEnvironment() from google.appengine.api import apiproxy_stub_map if apiproxy_stub_map.apiproxy.GetStub("urlfetch") is None: raise ImportError from google.appengine.api.urlfetch import fetch # Update the connection classes to use the Googel App Engine specific ones. SCHEME_TO_CONNECTION = { "http": AppEngineHttpConnection, "https": AppEngineHttpsConnection, } except (ImportError, NotRunningAppEngineEnvironment): pass class Http(object): """An HTTP client that handles: - all methods - caching - ETags - compression, - HTTPS - Basic - Digest - WSSE and more. """ def __init__( self, cache=None, timeout=None, proxy_info=proxy_info_from_environment, ca_certs=None, disable_ssl_certificate_validation=False, ssl_version=None, ): """If 'cache' is a string then it is used as a directory name for a disk cache. Otherwise it must be an object that supports the same interface as FileCache. All timeouts are in seconds. If None is passed for timeout then Python's default timeout for sockets will be used. See for example the docs of socket.setdefaulttimeout(): http://docs.python.org/library/socket.html#socket.setdefaulttimeout `proxy_info` may be: - a callable that takes the http scheme ('http' or 'https') and returns a ProxyInfo instance per request. By default, uses proxy_nfo_from_environment. - a ProxyInfo instance (static proxy config). - None (proxy disabled). ca_certs is the path of a file containing root CA certificates for SSL server certificate validation. By default, a CA cert file bundled with httplib2 is used. If disable_ssl_certificate_validation is true, SSL cert validation will not be performed. By default, ssl.PROTOCOL_SSLv23 will be used for the ssl version. """ self.proxy_info = proxy_info self.ca_certs = ca_certs self.disable_ssl_certificate_validation = disable_ssl_certificate_validation self.ssl_version = ssl_version # Map domain name to an httplib connection self.connections = {} # The location of the cache, for now a directory # where cached responses are held. if cache and isinstance(cache, basestring): self.cache = FileCache(cache) else: self.cache = cache # Name/password self.credentials = Credentials() # Key/cert self.certificates = KeyCerts() # authorization objects self.authorizations = [] # If set to False then no redirects are followed, even safe ones. self.follow_redirects = True self.redirect_codes = REDIRECT_CODES # Which HTTP methods do we apply optimistic concurrency to, i.e. # which methods get an "if-match:" etag header added to them. self.optimistic_concurrency_methods = ["PUT", "PATCH"] self.safe_methods = list(SAFE_METHODS) # If 'follow_redirects' is True, and this is set to True then # all redirecs are followed, including unsafe ones. self.follow_all_redirects = False self.ignore_etag = False self.force_exception_to_status_code = False self.timeout = timeout # Keep Authorization: headers on a redirect. self.forward_authorization_headers = False def close(self): """Close persistent connections, clear sensitive data. Not thread-safe, requires external synchronization against concurrent requests. """ existing, self.connections = self.connections, {} for _, c in existing.iteritems(): c.close() self.certificates.clear() self.clear_credentials() def __getstate__(self): state_dict = copy.copy(self.__dict__) # In case request is augmented by some foreign object such as # credentials which handle auth if "request" in state_dict: del state_dict["request"] if "connections" in state_dict: del state_dict["connections"] return state_dict def __setstate__(self, state): self.__dict__.update(state) self.connections = {} def _auth_from_challenge(self, host, request_uri, headers, response, content): """A generator that creates Authorization objects that can be applied to requests. """ challenges = _parse_www_authenticate(response, "www-authenticate") for cred in self.credentials.iter(host): for scheme in AUTH_SCHEME_ORDER: if scheme in challenges: yield AUTH_SCHEME_CLASSES[scheme]( cred, host, request_uri, headers, response, content, self ) def add_credentials(self, name, password, domain=""): """Add a name and password that will be used any time a request requires authentication.""" self.credentials.add(name, password, domain) def add_certificate(self, key, cert, domain, password=None): """Add a key and cert that will be used any time a request requires authentication.""" self.certificates.add(key, cert, domain, password) def clear_credentials(self): """Remove all the names and passwords that are used for authentication""" self.credentials.clear() self.authorizations = [] def _conn_request(self, conn, request_uri, method, body, headers): i = 0 seen_bad_status_line = False while i < RETRIES: i += 1 try: if hasattr(conn, "sock") and conn.sock is None: conn.connect() conn.request(method, request_uri, body, headers) except socket.timeout: raise except socket.gaierror: conn.close() raise ServerNotFoundError("Unable to find the server at %s" % conn.host) except ssl_SSLError: conn.close() raise except socket.error as e: err = 0 if hasattr(e, "args"): err = getattr(e, "args")[0] else: err = e.errno if err == errno.ECONNREFUSED: # Connection refused raise if err in (errno.ENETUNREACH, errno.EADDRNOTAVAIL) and i < RETRIES: continue # retry on potentially transient socket errors except httplib.HTTPException: # Just because the server closed the connection doesn't apparently mean # that the server didn't send a response. if hasattr(conn, "sock") and conn.sock is None: if i < RETRIES - 1: conn.close() conn.connect() continue else: conn.close() raise if i < RETRIES - 1: conn.close() conn.connect() continue try: response = conn.getresponse() except httplib.BadStatusLine: # If we get a BadStatusLine on the first try then that means # the connection just went stale, so retry regardless of the # number of RETRIES set. if not seen_bad_status_line and i == 1: i = 0 seen_bad_status_line = True conn.close() conn.connect() continue else: conn.close() raise except (socket.error, httplib.HTTPException): if i < RETRIES - 1: conn.close() conn.connect() continue else: conn.close() raise else: content = "" if method == "HEAD": conn.close() else: content = response.read() response = Response(response) if method != "HEAD": content = _decompressContent(response, content) break return (response, content) def _request( self, conn, host, absolute_uri, request_uri, method, body, headers, redirections, cachekey, ): """Do the actual request using the connection object and also follow one level of redirects if necessary""" auths = [ (auth.depth(request_uri), auth) for auth in self.authorizations if auth.inscope(host, request_uri) ] auth = auths and sorted(auths)[0][1] or None if auth: auth.request(method, request_uri, headers, body) (response, content) = self._conn_request( conn, request_uri, method, body, headers ) if auth: if auth.response(response, body): auth.request(method, request_uri, headers, body) (response, content) = self._conn_request( conn, request_uri, method, body, headers ) response._stale_digest = 1 if response.status == 401: for authorization in self._auth_from_challenge( host, request_uri, headers, response, content ): authorization.request(method, request_uri, headers, body) (response, content) = self._conn_request( conn, request_uri, method, body, headers ) if response.status != 401: self.authorizations.append(authorization) authorization.response(response, body) break if ( self.follow_all_redirects or method in self.safe_methods or response.status in (303, 308) ): if self.follow_redirects and response.status in self.redirect_codes: # Pick out the location header and basically start from the beginning # remembering first to strip the ETag header and decrement our 'depth' if redirections: if "location" not in response and response.status != 300: raise RedirectMissingLocation( _( "Redirected but the response is missing a Location: header." ), response, content, ) # Fix-up relative redirects (which violate an RFC 2616 MUST) if "location" in response: location = response["location"] (scheme, authority, path, query, fragment) = parse_uri(location) if authority == None: response["location"] = urlparse.urljoin( absolute_uri, location ) if response.status == 308 or (response.status == 301 and method in self.safe_methods): response["-x-permanent-redirect-url"] = response["location"] if "content-location" not in response: response["content-location"] = absolute_uri _updateCache(headers, response, content, self.cache, cachekey) if "if-none-match" in headers: del headers["if-none-match"] if "if-modified-since" in headers: del headers["if-modified-since"] if ( "authorization" in headers and not self.forward_authorization_headers ): del headers["authorization"] if "location" in response: location = response["location"] old_response = copy.deepcopy(response) if "content-location" not in old_response: old_response["content-location"] = absolute_uri redirect_method = method if response.status in [302, 303]: redirect_method = "GET" body = None (response, content) = self.request( location, method=redirect_method, body=body, headers=headers, redirections=redirections - 1, ) response.previous = old_response else: raise RedirectLimit( "Redirected more times than rediection_limit allows.", response, content, ) elif response.status in [200, 203] and method in self.safe_methods: # Don't cache 206's since we aren't going to handle byte range requests if "content-location" not in response: response["content-location"] = absolute_uri _updateCache(headers, response, content, self.cache, cachekey) return (response, content) def _normalize_headers(self, headers): return _normalize_headers(headers) # Need to catch and rebrand some exceptions # Then need to optionally turn all exceptions into status codes # including all socket.* and httplib.* exceptions. def request( self, uri, method="GET", body=None, headers=None, redirections=DEFAULT_MAX_REDIRECTS, connection_type=None, ): """ Performs a single HTTP request. The 'uri' is the URI of the HTTP resource and can begin with either 'http' or 'https'. The value of 'uri' must be an absolute URI. The 'method' is the HTTP method to perform, such as GET, POST, DELETE, etc. There is no restriction on the methods allowed. The 'body' is the entity body to be sent with the request. It is a string object. Any extra headers that are to be sent with the request should be provided in the 'headers' dictionary. The maximum number of redirect to follow before raising an exception is 'redirections. The default is 5. The return value is a tuple of (response, content), the first being and instance of the 'Response' class, the second being a string that contains the response entity body. """ conn_key = '' try: if headers is None: headers = {} else: headers = self._normalize_headers(headers) if "user-agent" not in headers: headers["user-agent"] = "Python-httplib2/%s (gzip)" % __version__ uri = iri2uri(uri) (scheme, authority, request_uri, defrag_uri) = urlnorm(uri) proxy_info = self._get_proxy_info(scheme, authority) conn_key = scheme + ":" + authority conn = self.connections.get(conn_key) if conn is None: if not connection_type: connection_type = SCHEME_TO_CONNECTION[scheme] certs = list(self.certificates.iter(authority)) if scheme == "https": if certs: conn = self.connections[conn_key] = connection_type( authority, key_file=certs[0][0], cert_file=certs[0][1], timeout=self.timeout, proxy_info=proxy_info, ca_certs=self.ca_certs, disable_ssl_certificate_validation=self.disable_ssl_certificate_validation, ssl_version=self.ssl_version, key_password=certs[0][2], ) else: conn = self.connections[conn_key] = connection_type( authority, timeout=self.timeout, proxy_info=proxy_info, ca_certs=self.ca_certs, disable_ssl_certificate_validation=self.disable_ssl_certificate_validation, ssl_version=self.ssl_version, ) else: conn = self.connections[conn_key] = connection_type( authority, timeout=self.timeout, proxy_info=proxy_info ) conn.set_debuglevel(debuglevel) if "range" not in headers and "accept-encoding" not in headers: headers["accept-encoding"] = "gzip, deflate" info = email.Message.Message() cachekey = None cached_value = None if self.cache: cachekey = defrag_uri.encode("utf-8") cached_value = self.cache.get(cachekey) if cached_value: # info = email.message_from_string(cached_value) # # Need to replace the line above with the kludge below # to fix the non-existent bug not fixed in this # bug report: http://mail.python.org/pipermail/python-bugs-list/2005-September/030289.html try: info, content = cached_value.split("\r\n\r\n", 1) feedparser = email.FeedParser.FeedParser() feedparser.feed(info) info = feedparser.close() feedparser._parse = None except (IndexError, ValueError): self.cache.delete(cachekey) cachekey = None cached_value = None if ( method in self.optimistic_concurrency_methods and self.cache and "etag" in info and not self.ignore_etag and "if-match" not in headers ): # http://www.w3.org/1999/04/Editing/ headers["if-match"] = info["etag"] # https://tools.ietf.org/html/rfc7234 # A cache MUST invalidate the effective Request URI as well as [...] Location and Content-Location # when a non-error status code is received in response to an unsafe request method. if self.cache and cachekey and method not in self.safe_methods: self.cache.delete(cachekey) # Check the vary header in the cache to see if this request # matches what varies in the cache. if method in self.safe_methods and "vary" in info: vary = info["vary"] vary_headers = vary.lower().replace(" ", "").split(",") for header in vary_headers: key = "-varied-%s" % header value = info[key] if headers.get(header, None) != value: cached_value = None break if ( self.cache and cached_value and (method in self.safe_methods or info["status"] == "308") and "range" not in headers ): redirect_method = method if info["status"] not in ("307", "308"): redirect_method = "GET" if "-x-permanent-redirect-url" in info: # Should cached permanent redirects be counted in our redirection count? For now, yes. if redirections <= 0: raise RedirectLimit( "Redirected more times than rediection_limit allows.", {}, "", ) (response, new_content) = self.request( info["-x-permanent-redirect-url"], method=redirect_method, headers=headers, redirections=redirections - 1, ) response.previous = Response(info) response.previous.fromcache = True else: # Determine our course of action: # Is the cached entry fresh or stale? # Has the client requested a non-cached response? # # There seems to be three possible answers: # 1. [FRESH] Return the cache entry w/o doing a GET # 2. [STALE] Do the GET (but add in cache validators if available) # 3. [TRANSPARENT] Do a GET w/o any cache validators (Cache-Control: no-cache) on the request entry_disposition = _entry_disposition(info, headers) if entry_disposition == "FRESH": if not cached_value: info["status"] = "504" content = "" response = Response(info) if cached_value: response.fromcache = True return (response, content) if entry_disposition == "STALE": if ( "etag" in info and not self.ignore_etag and not "if-none-match" in headers ): headers["if-none-match"] = info["etag"] if "last-modified" in info and not "last-modified" in headers: headers["if-modified-since"] = info["last-modified"] elif entry_disposition == "TRANSPARENT": pass (response, new_content) = self._request( conn, authority, uri, request_uri, method, body, headers, redirections, cachekey, ) if response.status == 304 and method == "GET": # Rewrite the cache entry with the new end-to-end headers # Take all headers that are in response # and overwrite their values in info. # unless they are hop-by-hop, or are listed in the connection header. for key in _get_end2end_headers(response): info[key] = response[key] merged_response = Response(info) if hasattr(response, "_stale_digest"): merged_response._stale_digest = response._stale_digest _updateCache( headers, merged_response, content, self.cache, cachekey ) response = merged_response response.status = 200 response.fromcache = True elif response.status == 200: content = new_content else: self.cache.delete(cachekey) content = new_content else: cc = _parse_cache_control(headers) if "only-if-cached" in cc: info["status"] = "504" response = Response(info) content = "" else: (response, content) = self._request( conn, authority, uri, request_uri, method, body, headers, redirections, cachekey, ) except Exception as e: is_timeout = isinstance(e, socket.timeout) if is_timeout: conn = self.connections.pop(conn_key, None) if conn: conn.close() if self.force_exception_to_status_code: if isinstance(e, HttpLib2ErrorWithResponse): response = e.response content = e.content response.status = 500 response.reason = str(e) elif is_timeout: content = "Request Timeout" response = Response( { "content-type": "text/plain", "status": "408", "content-length": len(content), } ) response.reason = "Request Timeout" else: content = str(e) response = Response( { "content-type": "text/plain", "status": "400", "content-length": len(content), } ) response.reason = "Bad Request" else: raise return (response, content) def _get_proxy_info(self, scheme, authority): """Return a ProxyInfo instance (or None) based on the scheme and authority. """ hostname, port = urllib.splitport(authority) proxy_info = self.proxy_info if callable(proxy_info): proxy_info = proxy_info(scheme) if hasattr(proxy_info, "applies_to") and not proxy_info.applies_to(hostname): proxy_info = None return proxy_info class Response(dict): """An object more like email.Message than httplib.HTTPResponse.""" """Is this response from our local cache""" fromcache = False """HTTP protocol version used by server. 10 for HTTP/1.0, 11 for HTTP/1.1. """ version = 11 "Status code returned by server. " status = 200 """Reason phrase returned by server.""" reason = "Ok" previous = None def __init__(self, info): # info is either an email.Message or # an httplib.HTTPResponse object. if isinstance(info, httplib.HTTPResponse): for key, value in info.getheaders(): self[key.lower()] = value self.status = info.status self["status"] = str(self.status) self.reason = info.reason self.version = info.version elif isinstance(info, email.Message.Message): for key, value in info.items(): self[key.lower()] = value self.status = int(self["status"]) else: for key, value in info.iteritems(): self[key.lower()] = value self.status = int(self.get("status", self.status)) self.reason = self.get("reason", self.reason) def __getattr__(self, name): if name == "dict": return self else: raise AttributeError(name)