blob: bb802df16274a3b96535a8de70a31d1d8f67ba31 [file] [log] [blame]
#!/usr/bin/env python
"""Universal feed parser
Handles RSS 0.9x, RSS 1.0, RSS 2.0, CDF, Atom 0.3, and Atom 1.0 feeds
Visit http://feedparser.org/ for the latest version
Visit http://feedparser.org/docs/ for the latest documentation
Required: Python 2.1 or later
Recommended: Python 2.3 or later
Recommended: CJKCodecs and iconv_codec <http://cjkpython.i18n.org/>
"""
__version__ = "4.1"# + "$Revision: 1.92 $"[11:15] + "-cvs"
__license__ = """Copyright (c) 2002-2006, Mark Pilgrim, All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 'AS IS'
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE."""
__author__ = "Mark Pilgrim <http://diveintomark.org/>"
__contributors__ = ["Jason Diamond <http://injektilo.org/>",
"John Beimler <http://john.beimler.org/>",
"Fazal Majid <http://www.majid.info/mylos/weblog/>",
"Aaron Swartz <http://aaronsw.com/>",
"Kevin Marks <http://epeus.blogspot.com/>"]
_debug = 0
# HTTP "User-Agent" header to send to servers when downloading feeds.
# If you are embedding feedparser in a larger application, you should
# change this to your application name and URL.
USER_AGENT = "UniversalFeedParser/%s +http://feedparser.org/" % __version__
# HTTP "Accept" header to send to servers when downloading feeds. If you don't
# want to send an Accept header, set this to None.
ACCEPT_HEADER = "application/atom+xml,application/rdf+xml,application/rss+xml,application/x-netcdf,application/xml;q=0.9,text/xml;q=0.2,*/*;q=0.1"
# List of preferred XML parsers, by SAX driver name. These will be tried first,
# but if they're not installed, Python will keep searching through its own list
# of pre-installed parsers until it finds one that supports everything we need.
PREFERRED_XML_PARSERS = ["drv_libxml2"]
# If you want feedparser to automatically run HTML markup through HTML Tidy, set
# this to 1. Requires mxTidy <http://www.egenix.com/files/python/mxTidy.html>
# or utidylib <http://utidylib.berlios.de/>.
TIDY_MARKUP = 0
# List of Python interfaces for HTML Tidy, in order of preference. Only useful
# if TIDY_MARKUP = 1
PREFERRED_TIDY_INTERFACES = ["uTidy", "mxTidy"]
# ---------- required modules (should come with any Python distribution) ----------
import sgmllib, re, sys, copy, urlparse, time, rfc822, types, cgi, urllib, urllib2
try:
from cStringIO import StringIO as _StringIO
except:
from StringIO import StringIO as _StringIO
# ---------- optional modules (feedparser will work without these, but with reduced functionality) ----------
# gzip is included with most Python distributions, but may not be available if you compiled your own
try:
import gzip
except:
gzip = None
try:
import zlib
except:
zlib = None
# If a real XML parser is available, feedparser will attempt to use it. feedparser has
# been tested with the built-in SAX parser, PyXML, and libxml2. On platforms where the
# Python distribution does not come with an XML parser (such as Mac OS X 10.2 and some
# versions of FreeBSD), feedparser will quietly fall back on regex-based parsing.
try:
import xml.sax
xml.sax.make_parser(PREFERRED_XML_PARSERS) # test for valid parsers
from xml.sax.saxutils import escape as _xmlescape
_XML_AVAILABLE = 1
except:
_XML_AVAILABLE = 0
def _xmlescape(data):
data = data.replace('&', '&amp;')
data = data.replace('>', '&gt;')
data = data.replace('<', '&lt;')
return data
# base64 support for Atom feeds that contain embedded binary data
try:
import base64, binascii
except:
base64 = binascii = None
# cjkcodecs and iconv_codec provide support for more character encodings.
# Both are available from http://cjkpython.i18n.org/
try:
import cjkcodecs.aliases
except:
pass
try:
import iconv_codec
except:
pass
# chardet library auto-detects character encodings
# Download from http://chardet.feedparser.org/
try:
import chardet
if _debug:
import chardet.constants
chardet.constants._debug = 1
except:
chardet = None
# ---------- don't touch these ----------
class ThingsNobodyCaresAboutButMe(Exception): pass
class CharacterEncodingOverride(ThingsNobodyCaresAboutButMe): pass
class CharacterEncodingUnknown(ThingsNobodyCaresAboutButMe): pass
class NonXMLContentType(ThingsNobodyCaresAboutButMe): pass
class UndeclaredNamespace(Exception): pass
sgmllib.tagfind = re.compile('[a-zA-Z][-_.:a-zA-Z0-9]*')
sgmllib.special = re.compile('<!')
sgmllib.charref = re.compile('&#(x?[0-9A-Fa-f]+)[^0-9A-Fa-f]')
SUPPORTED_VERSIONS = {'': 'unknown',
'rss090': 'RSS 0.90',
'rss091n': 'RSS 0.91 (Netscape)',
'rss091u': 'RSS 0.91 (Userland)',
'rss092': 'RSS 0.92',
'rss093': 'RSS 0.93',
'rss094': 'RSS 0.94',
'rss20': 'RSS 2.0',
'rss10': 'RSS 1.0',
'rss': 'RSS (unknown version)',
'atom01': 'Atom 0.1',
'atom02': 'Atom 0.2',
'atom03': 'Atom 0.3',
'atom10': 'Atom 1.0',
'atom': 'Atom (unknown version)',
'cdf': 'CDF',
'hotrss': 'Hot RSS'
}
try:
UserDict = dict
except NameError:
# Python 2.1 does not have dict
from UserDict import UserDict
def dict(aList):
rc = {}
for k, v in aList:
rc[k] = v
return rc
class FeedParserDict(UserDict):
keymap = {'channel': 'feed',
'items': 'entries',
'guid': 'id',
'date': 'updated',
'date_parsed': 'updated_parsed',
'description': ['subtitle', 'summary'],
'url': ['href'],
'modified': 'updated',
'modified_parsed': 'updated_parsed',
'issued': 'published',
'issued_parsed': 'published_parsed',
'copyright': 'rights',
'copyright_detail': 'rights_detail',
'tagline': 'subtitle',
'tagline_detail': 'subtitle_detail'}
def __getitem__(self, key):
if key == 'category':
return UserDict.__getitem__(self, 'tags')[0]['term']
if key == 'categories':
return [(tag['scheme'], tag['term']) for tag in UserDict.__getitem__(self, 'tags')]
realkey = self.keymap.get(key, key)
if type(realkey) == types.ListType:
for k in realkey:
if UserDict.has_key(self, k):
return UserDict.__getitem__(self, k)
if UserDict.has_key(self, key):
return UserDict.__getitem__(self, key)
return UserDict.__getitem__(self, realkey)
def __setitem__(self, key, value):
for k in self.keymap.keys():
if key == k:
key = self.keymap[k]
if type(key) == types.ListType:
key = key[0]
return UserDict.__setitem__(self, key, value)
def get(self, key, default=None):
if self.has_key(key):
return self[key]
else:
return default
def setdefault(self, key, value):
if not self.has_key(key):
self[key] = value
return self[key]
def has_key(self, key):
try:
return hasattr(self, key) or UserDict.has_key(self, key)
except AttributeError:
return False
def __getattr__(self, key):
try:
return self.__dict__[key]
except KeyError:
pass
try:
assert not key.startswith('_')
return self.__getitem__(key)
except:
raise AttributeError, "object has no attribute '%s'" % key
def __setattr__(self, key, value):
if key.startswith('_') or key == 'data':
self.__dict__[key] = value
else:
return self.__setitem__(key, value)
def __contains__(self, key):
return self.has_key(key)
def zopeCompatibilityHack():
global FeedParserDict
del FeedParserDict
def FeedParserDict(aDict=None):
rc = {}
if aDict:
rc.update(aDict)
return rc
_ebcdic_to_ascii_map = None
def _ebcdic_to_ascii(s):
global _ebcdic_to_ascii_map
if not _ebcdic_to_ascii_map:
emap = (
0,1,2,3,156,9,134,127,151,141,142,11,12,13,14,15,
16,17,18,19,157,133,8,135,24,25,146,143,28,29,30,31,
128,129,130,131,132,10,23,27,136,137,138,139,140,5,6,7,
144,145,22,147,148,149,150,4,152,153,154,155,20,21,158,26,
32,160,161,162,163,164,165,166,167,168,91,46,60,40,43,33,
38,169,170,171,172,173,174,175,176,177,93,36,42,41,59,94,
45,47,178,179,180,181,182,183,184,185,124,44,37,95,62,63,
186,187,188,189,190,191,192,193,194,96,58,35,64,39,61,34,
195,97,98,99,100,101,102,103,104,105,196,197,198,199,200,201,
202,106,107,108,109,110,111,112,113,114,203,204,205,206,207,208,
209,126,115,116,117,118,119,120,121,122,210,211,212,213,214,215,
216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,
123,65,66,67,68,69,70,71,72,73,232,233,234,235,236,237,
125,74,75,76,77,78,79,80,81,82,238,239,240,241,242,243,
92,159,83,84,85,86,87,88,89,90,244,245,246,247,248,249,
48,49,50,51,52,53,54,55,56,57,250,251,252,253,254,255
)
import string
_ebcdic_to_ascii_map = string.maketrans( \
''.join(map(chr, range(256))), ''.join(map(chr, emap)))
return s.translate(_ebcdic_to_ascii_map)
_urifixer = re.compile('^([A-Za-z][A-Za-z0-9+-.]*://)(/*)(.*?)')
def _urljoin(base, uri):
uri = _urifixer.sub(r'\1\3', uri)
return urlparse.urljoin(base, uri)
class _FeedParserMixin:
namespaces = {'': '',
'http://backend.userland.com/rss': '',
'http://blogs.law.harvard.edu/tech/rss': '',
'http://purl.org/rss/1.0/': '',
'http://my.netscape.com/rdf/simple/0.9/': '',
'http://example.com/newformat#': '',
'http://example.com/necho': '',
'http://purl.org/echo/': '',
'uri/of/echo/namespace#': '',
'http://purl.org/pie/': '',
'http://purl.org/atom/ns#': '',
'http://www.w3.org/2005/Atom': '',
'http://purl.org/rss/1.0/modules/rss091#': '',
'http://webns.net/mvcb/': 'admin',
'http://purl.org/rss/1.0/modules/aggregation/': 'ag',
'http://purl.org/rss/1.0/modules/annotate/': 'annotate',
'http://media.tangent.org/rss/1.0/': 'audio',
'http://backend.userland.com/blogChannelModule': 'blogChannel',
'http://web.resource.org/cc/': 'cc',
'http://backend.userland.com/creativeCommonsRssModule': 'creativeCommons',
'http://purl.org/rss/1.0/modules/company': 'co',
'http://purl.org/rss/1.0/modules/content/': 'content',
'http://my.theinfo.org/changed/1.0/rss/': 'cp',
'http://purl.org/dc/elements/1.1/': 'dc',
'http://purl.org/dc/terms/': 'dcterms',
'http://purl.org/rss/1.0/modules/email/': 'email',
'http://purl.org/rss/1.0/modules/event/': 'ev',
'http://rssnamespace.org/feedburner/ext/1.0': 'feedburner',
'http://freshmeat.net/rss/fm/': 'fm',
'http://xmlns.com/foaf/0.1/': 'foaf',
'http://www.w3.org/2003/01/geo/wgs84_pos#': 'geo',
'http://postneo.com/icbm/': 'icbm',
'http://purl.org/rss/1.0/modules/image/': 'image',
'http://www.itunes.com/DTDs/PodCast-1.0.dtd': 'itunes',
'http://example.com/DTDs/PodCast-1.0.dtd': 'itunes',
'http://purl.org/rss/1.0/modules/link/': 'l',
'http://search.yahoo.com/mrss': 'media',
'http://madskills.com/public/xml/rss/module/pingback/': 'pingback',
'http://prismstandard.org/namespaces/1.2/basic/': 'prism',
'http://www.w3.org/1999/02/22-rdf-syntax-ns#': 'rdf',
'http://www.w3.org/2000/01/rdf-schema#': 'rdfs',
'http://purl.org/rss/1.0/modules/reference/': 'ref',
'http://purl.org/rss/1.0/modules/richequiv/': 'reqv',
'http://purl.org/rss/1.0/modules/search/': 'search',
'http://purl.org/rss/1.0/modules/slash/': 'slash',
'http://schemas.xmlsoap.org/soap/envelope/': 'soap',
'http://purl.org/rss/1.0/modules/servicestatus/': 'ss',
'http://hacks.benhammersley.com/rss/streaming/': 'str',
'http://purl.org/rss/1.0/modules/subscription/': 'sub',
'http://purl.org/rss/1.0/modules/syndication/': 'sy',
'http://purl.org/rss/1.0/modules/taxonomy/': 'taxo',
'http://purl.org/rss/1.0/modules/threading/': 'thr',
'http://purl.org/rss/1.0/modules/textinput/': 'ti',
'http://madskills.com/public/xml/rss/module/trackback/':'trackback',
'http://wellformedweb.org/commentAPI/': 'wfw',
'http://purl.org/rss/1.0/modules/wiki/': 'wiki',
'http://www.w3.org/1999/xhtml': 'xhtml',
'http://www.w3.org/XML/1998/namespace': 'xml',
'http://schemas.pocketsoap.com/rss/myDescModule/': 'szf'
}
_matchnamespaces = {}
can_be_relative_uri = ['link', 'id', 'wfw_comment', 'wfw_commentrss', 'docs', 'url', 'href', 'comments', 'license', 'icon', 'logo']
can_contain_relative_uris = ['content', 'title', 'summary', 'info', 'tagline', 'subtitle', 'copyright', 'rights', 'description']
can_contain_dangerous_markup = ['content', 'title', 'summary', 'info', 'tagline', 'subtitle', 'copyright', 'rights', 'description']
html_types = ['text/html', 'application/xhtml+xml']
def __init__(self, baseuri=None, baselang=None, encoding='utf-8'):
if _debug: sys.stderr.write('initializing FeedParser\n')
if not self._matchnamespaces:
for k, v in self.namespaces.items():
self._matchnamespaces[k.lower()] = v
self.feeddata = FeedParserDict() # feed-level data
self.encoding = encoding # character encoding
self.entries = [] # list of entry-level data
self.version = '' # feed type/version, see SUPPORTED_VERSIONS
self.namespacesInUse = {} # dictionary of namespaces defined by the feed
# the following are used internally to track state;
# this is really out of control and should be refactored
self.infeed = 0
self.inentry = 0
self.incontent = 0
self.intextinput = 0
self.inimage = 0
self.inauthor = 0
self.incontributor = 0
self.inpublisher = 0
self.insource = 0
self.sourcedata = FeedParserDict()
self.contentparams = FeedParserDict()
self._summaryKey = None
self.namespacemap = {}
self.elementstack = []
self.basestack = []
self.langstack = []
self.baseuri = baseuri or ''
self.lang = baselang or None
if baselang:
self.feeddata['language'] = baselang
def unknown_starttag(self, tag, attrs):
if _debug: sys.stderr.write('start %s with %s\n' % (tag, attrs))
# normalize attrs
attrs = [(k.lower(), v) for k, v in attrs]
attrs = [(k, k in ('rel', 'type') and v.lower() or v) for k, v in attrs]
# track xml:base and xml:lang
attrsD = dict(attrs)
baseuri = attrsD.get('xml:base', attrsD.get('base')) or self.baseuri
self.baseuri = _urljoin(self.baseuri, baseuri)
lang = attrsD.get('xml:lang', attrsD.get('lang'))
if lang == '':
# xml:lang could be explicitly set to '', we need to capture that
lang = None
elif lang is None:
# if no xml:lang is specified, use parent lang
lang = self.lang
if lang:
if tag in ('feed', 'rss', 'rdf:RDF'):
self.feeddata['language'] = lang
self.lang = lang
self.basestack.append(self.baseuri)
self.langstack.append(lang)
# track namespaces
for prefix, uri in attrs:
if prefix.startswith('xmlns:'):
self.trackNamespace(prefix[6:], uri)
elif prefix == 'xmlns':
self.trackNamespace(None, uri)
# track inline content
if self.incontent and self.contentparams.has_key('type') and not self.contentparams.get('type', 'xml').endswith('xml'):
# element declared itself as escaped markup, but it isn't really
self.contentparams['type'] = 'application/xhtml+xml'
if self.incontent and self.contentparams.get('type') == 'application/xhtml+xml':
# Note: probably shouldn't simply recreate localname here, but
# our namespace handling isn't actually 100% correct in cases where
# the feed redefines the default namespace (which is actually
# the usual case for inline content, thanks Sam), so here we
# cheat and just reconstruct the element based on localname
# because that compensates for the bugs in our namespace handling.
# This will horribly munge inline content with non-empty qnames,
# but nobody actually does that, so I'm not fixing it.
tag = tag.split(':')[-1]
return self.handle_data('<%s%s>' % (tag, ''.join([' %s="%s"' % t for t in attrs])), escape=0)
# match namespaces
if tag.find(':') <> -1:
prefix, suffix = tag.split(':', 1)
else:
prefix, suffix = '', tag
prefix = self.namespacemap.get(prefix, prefix)
if prefix:
prefix = prefix + '_'
# special hack for better tracking of empty textinput/image elements in illformed feeds
if (not prefix) and tag not in ('title', 'link', 'description', 'name'):
self.intextinput = 0
if (not prefix) and tag not in ('title', 'link', 'description', 'url', 'href', 'width', 'height'):
self.inimage = 0
# call special handler (if defined) or default handler
methodname = '_start_' + prefix + suffix
try:
method = getattr(self, methodname)
return method(attrsD)
except AttributeError:
return self.push(prefix + suffix, 1)
def unknown_endtag(self, tag):
if _debug: sys.stderr.write('end %s\n' % tag)
# match namespaces
if tag.find(':') <> -1:
prefix, suffix = tag.split(':', 1)
else:
prefix, suffix = '', tag
prefix = self.namespacemap.get(prefix, prefix)
if prefix:
prefix = prefix + '_'
# call special handler (if defined) or default handler
methodname = '_end_' + prefix + suffix
try:
method = getattr(self, methodname)
method()
except AttributeError:
self.pop(prefix + suffix)
# track inline content
if self.incontent and self.contentparams.has_key('type') and not self.contentparams.get('type', 'xml').endswith('xml'):
# element declared itself as escaped markup, but it isn't really
self.contentparams['type'] = 'application/xhtml+xml'
if self.incontent and self.contentparams.get('type') == 'application/xhtml+xml':
tag = tag.split(':')[-1]
self.handle_data('</%s>' % tag, escape=0)
# track xml:base and xml:lang going out of scope
if self.basestack:
self.basestack.pop()
if self.basestack and self.basestack[-1]:
self.baseuri = self.basestack[-1]
if self.langstack:
self.langstack.pop()
if self.langstack: # and (self.langstack[-1] is not None):
self.lang = self.langstack[-1]
def handle_charref(self, ref):
# called for each character reference, e.g. for '&#160;', ref will be '160'
if not self.elementstack: return
ref = ref.lower()
if ref in ('34', '38', '39', '60', '62', 'x22', 'x26', 'x27', 'x3c', 'x3e'):
text = '&#%s;' % ref
else:
if ref[0] == 'x':
c = int(ref[1:], 16)
else:
c = int(ref)
text = unichr(c).encode('utf-8')
self.elementstack[-1][2].append(text)
def handle_entityref(self, ref):
# called for each entity reference, e.g. for '&copy;', ref will be 'copy'
if not self.elementstack: return
if _debug: sys.stderr.write('entering handle_entityref with %s\n' % ref)
if ref in ('lt', 'gt', 'quot', 'amp', 'apos'):
text = '&%s;' % ref
else:
# entity resolution graciously donated by Aaron Swartz
def name2cp(k):
import htmlentitydefs
if hasattr(htmlentitydefs, 'name2codepoint'): # requires Python 2.3
return htmlentitydefs.name2codepoint[k]
k = htmlentitydefs.entitydefs[k]
if k.startswith('&#') and k.endswith(';'):
return int(k[2:-1]) # not in latin-1
return ord(k)
try: name2cp(ref)
except KeyError: text = '&%s;' % ref
else: text = unichr(name2cp(ref)).encode('utf-8')
self.elementstack[-1][2].append(text)
def handle_data(self, text, escape=1):
# called for each block of plain text, i.e. outside of any tag and
# not containing any character or entity references
if not self.elementstack: return
if escape and self.contentparams.get('type') == 'application/xhtml+xml':
text = _xmlescape(text)
self.elementstack[-1][2].append(text)
def handle_comment(self, text):
# called for each comment, e.g. <!-- insert message here -->
pass
def handle_pi(self, text):
# called for each processing instruction, e.g. <?instruction>
pass
def handle_decl(self, text):
pass
def parse_declaration(self, i):
# override internal declaration handler to handle CDATA blocks
if _debug: sys.stderr.write('entering parse_declaration\n')
if self.rawdata[i:i+9] == '<![CDATA[':
k = self.rawdata.find(']]>', i)
if k == -1: k = len(self.rawdata)
self.handle_data(_xmlescape(self.rawdata[i+9:k]), 0)
return k+3
else:
k = self.rawdata.find('>', i)
return k+1
def mapContentType(self, contentType):
contentType = contentType.lower()
if contentType == 'text':
contentType = 'text/plain'
elif contentType == 'html':
contentType = 'text/html'
elif contentType == 'xhtml':
contentType = 'application/xhtml+xml'
return contentType
def trackNamespace(self, prefix, uri):
loweruri = uri.lower()
if (prefix, loweruri) == (None, 'http://my.netscape.com/rdf/simple/0.9/') and not self.version:
self.version = 'rss090'
if loweruri == 'http://purl.org/rss/1.0/' and not self.version:
self.version = 'rss10'
if loweruri == 'http://www.w3.org/2005/atom' and not self.version:
self.version = 'atom10'
if loweruri.find('backend.userland.com/rss') <> -1:
# match any backend.userland.com namespace
uri = 'http://backend.userland.com/rss'
loweruri = uri
if self._matchnamespaces.has_key(loweruri):
self.namespacemap[prefix] = self._matchnamespaces[loweruri]
self.namespacesInUse[self._matchnamespaces[loweruri]] = uri
else:
self.namespacesInUse[prefix or ''] = uri
def resolveURI(self, uri):
return _urljoin(self.baseuri or '', uri)
def decodeEntities(self, element, data):
return data
def push(self, element, expectingText):
self.elementstack.append([element, expectingText, []])
def pop(self, element, stripWhitespace=1):
if not self.elementstack: return
if self.elementstack[-1][0] != element: return
element, expectingText, pieces = self.elementstack.pop()
output = ''.join(pieces)
if stripWhitespace:
output = output.strip()
if not expectingText: return output
# decode base64 content
if base64 and self.contentparams.get('base64', 0):
try:
output = base64.decodestring(output)
except binascii.Error:
pass
except binascii.Incomplete:
pass
# resolve relative URIs
if (element in self.can_be_relative_uri) and output:
output = self.resolveURI(output)
# decode entities within embedded markup
if not self.contentparams.get('base64', 0):
output = self.decodeEntities(element, output)
# remove temporary cruft from contentparams
try:
del self.contentparams['mode']
except KeyError:
pass
try:
del self.contentparams['base64']
except KeyError:
pass
# resolve relative URIs within embedded markup
if self.mapContentType(self.contentparams.get('type', 'text/html')) in self.html_types:
if element in self.can_contain_relative_uris:
output = _resolveRelativeURIs(output, self.baseuri, self.encoding)
# sanitize embedded markup
if self.mapContentType(self.contentparams.get('type', 'text/html')) in self.html_types:
if element in self.can_contain_dangerous_markup:
output = _sanitizeHTML(output, self.encoding)
if self.encoding and type(output) != type(u''):
try:
output = unicode(output, self.encoding)
except:
pass
# categories/tags/keywords/whatever are handled in _end_category
if element == 'category':
return output
# store output in appropriate place(s)
if self.inentry and not self.insource:
if element == 'content':
self.entries[-1].setdefault(element, [])
contentparams = copy.deepcopy(self.contentparams)
contentparams['value'] = output
self.entries[-1][element].append(contentparams)
elif element == 'link':
self.entries[-1][element] = output
if output:
self.entries[-1]['links'][-1]['href'] = output
else:
if element == 'description':
element = 'summary'
self.entries[-1][element] = output
if self.incontent:
contentparams = copy.deepcopy(self.contentparams)
contentparams['value'] = output
self.entries[-1][element + '_detail'] = contentparams
elif (self.infeed or self.insource) and (not self.intextinput) and (not self.inimage):
context = self._getContext()
if element == 'description':
element = 'subtitle'
context[element] = output
if element == 'link':
context['links'][-1]['href'] = output
elif self.incontent:
contentparams = copy.deepcopy(self.contentparams)
contentparams['value'] = output
context[element + '_detail'] = contentparams
return output
def pushContent(self, tag, attrsD, defaultContentType, expectingText):
self.incontent += 1
self.contentparams = FeedParserDict({
'type': self.mapContentType(attrsD.get('type', defaultContentType)),
'language': self.lang,
'base': self.baseuri})
self.contentparams['base64'] = self._isBase64(attrsD, self.contentparams)
self.push(tag, expectingText)
def popContent(self, tag):
value = self.pop(tag)
self.incontent -= 1
self.contentparams.clear()
return value
def _mapToStandardPrefix(self, name):
colonpos = name.find(':')
if colonpos <> -1:
prefix = name[:colonpos]
suffix = name[colonpos+1:]
prefix = self.namespacemap.get(prefix, prefix)
name = prefix + ':' + suffix
return name
def _getAttribute(self, attrsD, name):
return attrsD.get(self._mapToStandardPrefix(name))
def _isBase64(self, attrsD, contentparams):
if attrsD.get('mode', '') == 'base64':
return 1
if self.contentparams['type'].startswith('text/'):
return 0
if self.contentparams['type'].endswith('+xml'):
return 0
if self.contentparams['type'].endswith('/xml'):
return 0
return 1
def _itsAnHrefDamnIt(self, attrsD):
href = attrsD.get('url', attrsD.get('uri', attrsD.get('href', None)))
if href:
try:
del attrsD['url']
except KeyError:
pass
try:
del attrsD['uri']
except KeyError:
pass
attrsD['href'] = href
return attrsD
def _save(self, key, value):
context = self._getContext()
context.setdefault(key, value)
def _start_rss(self, attrsD):
versionmap = {'0.91': 'rss091u',
'0.92': 'rss092',
'0.93': 'rss093',
'0.94': 'rss094'}
if not self.version:
attr_version = attrsD.get('version', '')
version = versionmap.get(attr_version)
if version:
self.version = version
elif attr_version.startswith('2.'):
self.version = 'rss20'
else:
self.version = 'rss'
def _start_dlhottitles(self, attrsD):
self.version = 'hotrss'
def _start_channel(self, attrsD):
self.infeed = 1
self._cdf_common(attrsD)
_start_feedinfo = _start_channel
def _cdf_common(self, attrsD):
if attrsD.has_key('lastmod'):
self._start_modified({})
self.elementstack[-1][-1] = attrsD['lastmod']
self._end_modified()
if attrsD.has_key('href'):
self._start_link({})
self.elementstack[-1][-1] = attrsD['href']
self._end_link()
def _start_feed(self, attrsD):
self.infeed = 1
versionmap = {'0.1': 'atom01',
'0.2': 'atom02',
'0.3': 'atom03'}
if not self.version:
attr_version = attrsD.get('version')
version = versionmap.get(attr_version)
if version:
self.version = version
else:
self.version = 'atom'
def _end_channel(self):
self.infeed = 0
_end_feed = _end_channel
def _start_image(self, attrsD):
self.inimage = 1
self.push('image', 0)
context = self._getContext()
context.setdefault('image', FeedParserDict())
def _end_image(self):
self.pop('image')
self.inimage = 0
def _start_textinput(self, attrsD):
self.intextinput = 1
self.push('textinput', 0)
context = self._getContext()
context.setdefault('textinput', FeedParserDict())
_start_textInput = _start_textinput
def _end_textinput(self):
self.pop('textinput')
self.intextinput = 0
_end_textInput = _end_textinput
def _start_author(self, attrsD):
self.inauthor = 1
self.push('author', 1)
_start_managingeditor = _start_author
_start_dc_author = _start_author
_start_dc_creator = _start_author
_start_itunes_author = _start_author
def _end_author(self):
self.pop('author')
self.inauthor = 0
self._sync_author_detail()
_end_managingeditor = _end_author
_end_dc_author = _end_author
_end_dc_creator = _end_author
_end_itunes_author = _end_author
def _start_itunes_owner(self, attrsD):
self.inpublisher = 1
self.push('publisher', 0)
def _end_itunes_owner(self):
self.pop('publisher')
self.inpublisher = 0
self._sync_author_detail('publisher')
def _start_contributor(self, attrsD):
self.incontributor = 1
context = self._getContext()
context.setdefault('contributors', [])
context['contributors'].append(FeedParserDict())
self.push('contributor', 0)
def _end_contributor(self):
self.pop('contributor')
self.incontributor = 0
def _start_dc_contributor(self, attrsD):
self.incontributor = 1
context = self._getContext()
context.setdefault('contributors', [])
context['contributors'].append(FeedParserDict())
self.push('name', 0)
def _end_dc_contributor(self):
self._end_name()
self.incontributor = 0
def _start_name(self, attrsD):
self.push('name', 0)
_start_itunes_name = _start_name
def _end_name(self):
value = self.pop('name')
if self.inpublisher:
self._save_author('name', value, 'publisher')
elif self.inauthor:
self._save_author('name', value)
elif self.incontributor:
self._save_contributor('name', value)
elif self.intextinput:
context = self._getContext()
context['textinput']['name'] = value
_end_itunes_name = _end_name
def _start_width(self, attrsD):
self.push('width', 0)
def _end_width(self):
value = self.pop('width')
try:
value = int(value)
except:
value = 0
if self.inimage:
context = self._getContext()
context['image']['width'] = value
def _start_height(self, attrsD):
self.push('height', 0)
def _end_height(self):
value = self.pop('height')
try:
value = int(value)
except:
value = 0
if self.inimage:
context = self._getContext()
context['image']['height'] = value
def _start_url(self, attrsD):
self.push('href', 1)
_start_homepage = _start_url
_start_uri = _start_url
def _end_url(self):
value = self.pop('href')
if self.inauthor:
self._save_author('href', value)
elif self.incontributor:
self._save_contributor('href', value)
elif self.inimage:
context = self._getContext()
context['image']['href'] = value
elif self.intextinput:
context = self._getContext()
context['textinput']['link'] = value
_end_homepage = _end_url
_end_uri = _end_url
def _start_email(self, attrsD):
self.push('email', 0)
_start_itunes_email = _start_email
def _end_email(self):
value = self.pop('email')
if self.inpublisher:
self._save_author('email', value, 'publisher')
elif self.inauthor:
self._save_author('email', value)
elif self.incontributor:
self._save_contributor('email', value)
_end_itunes_email = _end_email
def _getContext(self):
if self.insource:
context = self.sourcedata
elif self.inentry:
context = self.entries[-1]
else:
context = self.feeddata
return context
def _save_author(self, key, value, prefix='author'):
context = self._getContext()
context.setdefault(prefix + '_detail', FeedParserDict())
context[prefix + '_detail'][key] = value
self._sync_author_detail()
def _save_contributor(self, key, value):
context = self._getContext()
context.setdefault('contributors', [FeedParserDict()])
context['contributors'][-1][key] = value
def _sync_author_detail(self, key='author'):
context = self._getContext()
detail = context.get('%s_detail' % key)
if detail:
name = detail.get('name')
email = detail.get('email')
if name and email:
context[key] = '%s (%s)' % (name, email)
elif name:
context[key] = name
elif email:
context[key] = email
else:
author = context.get(key)
if not author: return
emailmatch = re.search(r'''(([a-zA-Z0-9\_\-\.\+]+)@((\[[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.)|(([a-zA-Z0-9\-]+\.)+))([a-zA-Z]{2,4}|[0-9]{1,3})(\]?))''', author)
if not emailmatch: return
email = emailmatch.group(0)
# probably a better way to do the following, but it passes all the tests
author = author.replace(email, '')
author = author.replace('()', '')
author = author.strip()
if author and (author[0] == '('):
author = author[1:]
if author and (author[-1] == ')'):
author = author[:-1]
author = author.strip()
context.setdefault('%s_detail' % key, FeedParserDict())
context['%s_detail' % key]['name'] = author
context['%s_detail' % key]['email'] = email
def _start_subtitle(self, attrsD):
self.pushContent('subtitle', attrsD, 'text/plain', 1)
_start_tagline = _start_subtitle
_start_itunes_subtitle = _start_subtitle
def _end_subtitle(self):
self.popContent('subtitle')
_end_tagline = _end_subtitle
_end_itunes_subtitle = _end_subtitle
def _start_rights(self, attrsD):
self.pushContent('rights', attrsD, 'text/plain', 1)
_start_dc_rights = _start_rights
_start_copyright = _start_rights
def _end_rights(self):
self.popContent('rights')
_end_dc_rights = _end_rights
_end_copyright = _end_rights
def _start_item(self, attrsD):
self.entries.append(FeedParserDict())
self.push('item', 0)
self.inentry = 1
self.guidislink = 0
id = self._getAttribute(attrsD, 'rdf:about')
if id:
context = self._getContext()
context['id'] = id
self._cdf_common(attrsD)
_start_entry = _start_item
_start_product = _start_item
def _end_item(self):
self.pop('item')
self.inentry = 0
_end_entry = _end_item
def _start_dc_language(self, attrsD):
self.push('language', 1)
_start_language = _start_dc_language
def _end_dc_language(self):
self.lang = self.pop('language')
_end_language = _end_dc_language
def _start_dc_publisher(self, attrsD):
self.push('publisher', 1)
_start_webmaster = _start_dc_publisher
def _end_dc_publisher(self):
self.pop('publisher')
self._sync_author_detail('publisher')
_end_webmaster = _end_dc_publisher
def _start_published(self, attrsD):
self.push('published', 1)
_start_dcterms_issued = _start_published
_start_issued = _start_published
def _end_published(self):
value = self.pop('published')
self._save('published_parsed', _parse_date(value))
_end_dcterms_issued = _end_published
_end_issued = _end_published
def _start_updated(self, attrsD):
self.push('updated', 1)
_start_modified = _start_updated
_start_dcterms_modified = _start_updated
_start_pubdate = _start_updated
_start_dc_date = _start_updated
def _end_updated(self):
value = self.pop('updated')
parsed_value = _parse_date(value)
self._save('updated_parsed', parsed_value)
_end_modified = _end_updated
_end_dcterms_modified = _end_updated
_end_pubdate = _end_updated
_end_dc_date = _end_updated
def _start_created(self, attrsD):
self.push('created', 1)
_start_dcterms_created = _start_created
def _end_created(self):
value = self.pop('created')
self._save('created_parsed', _parse_date(value))
_end_dcterms_created = _end_created
def _start_expirationdate(self, attrsD):
self.push('expired', 1)
def _end_expirationdate(self):
self._save('expired_parsed', _parse_date(self.pop('expired')))
def _start_cc_license(self, attrsD):
self.push('license', 1)
value = self._getAttribute(attrsD, 'rdf:resource')
if value:
self.elementstack[-1][2].append(value)
self.pop('license')
def _start_creativecommons_license(self, attrsD):
self.push('license', 1)
def _end_creativecommons_license(self):
self.pop('license')
def _addTag(self, term, scheme, label):
context = self._getContext()
tags = context.setdefault('tags', [])
if (not term) and (not scheme) and (not label): return
value = FeedParserDict({'term': term, 'scheme': scheme, 'label': label})
if value not in tags:
tags.append(FeedParserDict({'term': term, 'scheme': scheme, 'label': label}))
def _start_category(self, attrsD):
if _debug: sys.stderr.write('entering _start_category with %s\n' % repr(attrsD))
term = attrsD.get('term')
scheme = attrsD.get('scheme', attrsD.get('domain'))
label = attrsD.get('label')
self._addTag(term, scheme, label)
self.push('category', 1)
_start_dc_subject = _start_category
_start_keywords = _start_category
def _end_itunes_keywords(self):
for term in self.pop('itunes_keywords').split():
self._addTag(term, 'http://www.itunes.com/', None)
def _start_itunes_category(self, attrsD):
self._addTag(attrsD.get('text'), 'http://www.itunes.com/', None)
self.push('category', 1)
def _end_category(self):
value = self.pop('category')
if not value: return
context = self._getContext()
tags = context['tags']
if value and len(tags) and not tags[-1]['term']:
tags[-1]['term'] = value
else:
self._addTag(value, None, None)
_end_dc_subject = _end_category
_end_keywords = _end_category
_end_itunes_category = _end_category
def _start_cloud(self, attrsD):
self._getContext()['cloud'] = FeedParserDict(attrsD)
def _start_link(self, attrsD):
attrsD.setdefault('rel', 'alternate')
attrsD.setdefault('type', 'text/html')
attrsD = self._itsAnHrefDamnIt(attrsD)
if attrsD.has_key('href'):
attrsD['href'] = self.resolveURI(attrsD['href'])
expectingText = self.infeed or self.inentry or self.insource
context = self._getContext()
context.setdefault('links', [])
context['links'].append(FeedParserDict(attrsD))
if attrsD['rel'] == 'enclosure':
self._start_enclosure(attrsD)
if attrsD.has_key('href'):
expectingText = 0
if (attrsD.get('rel') == 'alternate') and (self.mapContentType(attrsD.get('type')) in self.html_types):
context['link'] = attrsD['href']
else:
self.push('link', expectingText)
_start_producturl = _start_link
def _end_link(self):
value = self.pop('link')
context = self._getContext()
if self.intextinput:
context['textinput']['link'] = value
if self.inimage:
context['image']['link'] = value
_end_producturl = _end_link
def _start_guid(self, attrsD):
self.guidislink = (attrsD.get('ispermalink', 'true') == 'true')
self.push('id', 1)
def _end_guid(self):
value = self.pop('id')
self._save('guidislink', self.guidislink and not self._getContext().has_key('link'))
if self.guidislink:
# guid acts as link, but only if 'ispermalink' is not present or is 'true',
# and only if the item doesn't already have a link element
self._save('link', value)
def _start_title(self, attrsD):
self.pushContent('title', attrsD, 'text/plain', self.infeed or self.inentry or self.insource)
_start_dc_title = _start_title
_start_media_title = _start_title
def _end_title(self):
value = self.popContent('title')
context = self._getContext()
if self.intextinput:
context['textinput']['title'] = value
elif self.inimage:
context['image']['title'] = value
_end_dc_title = _end_title
_end_media_title = _end_title
def _start_description(self, attrsD):
context = self._getContext()
if context.has_key('summary'):
self._summaryKey = 'content'
self._start_content(attrsD)
else:
self.pushContent('description', attrsD, 'text/html', self.infeed or self.inentry or self.insource)
def _start_abstract(self, attrsD):
self.pushContent('description', attrsD, 'text/plain', self.infeed or self.inentry or self.insource)
def _end_description(self):
if self._summaryKey == 'content':
self._end_content()
else:
value = self.popContent('description')
context = self._getContext()
if self.intextinput:
context['textinput']['description'] = value
elif self.inimage:
context['image']['description'] = value
self._summaryKey = None
_end_abstract = _end_description
def _start_info(self, attrsD):
self.pushContent('info', attrsD, 'text/plain', 1)
_start_feedburner_browserfriendly = _start_info
def _end_info(self):
self.popContent('info')
_end_feedburner_browserfriendly = _end_info
def _start_generator(self, attrsD):
if attrsD:
attrsD = self._itsAnHrefDamnIt(attrsD)
if attrsD.has_key('href'):
attrsD['href'] = self.resolveURI(attrsD['href'])
self._getContext()['generator_detail'] = FeedParserDict(attrsD)
self.push('generator', 1)
def _end_generator(self):
value = self.pop('generator')
context = self._getContext()
if context.has_key('generator_detail'):
context['generator_detail']['name'] = value
def _start_admin_generatoragent(self, attrsD):
self.push('generator', 1)
value = self._getAttribute(attrsD, 'rdf:resource')
if value:
self.elementstack[-1][2].append(value)
self.pop('generator')
self._getContext()['generator_detail'] = FeedParserDict({'href': value})
def _start_admin_errorreportsto(self, attrsD):
self.push('errorreportsto', 1)
value = self._getAttribute(attrsD, 'rdf:resource')
if value:
self.elementstack[-1][2].append(value)
self.pop('errorreportsto')
def _start_summary(self, attrsD):
context = self._getContext()
if context.has_key('summary'):
self._summaryKey = 'content'
self._start_content(attrsD)
else:
self._summaryKey = 'summary'
self.pushContent(self._summaryKey, attrsD, 'text/plain', 1)
_start_itunes_summary = _start_summary
def _end_summary(self):
if self._summaryKey == 'content':
self._end_content()
else:
self.popContent(self._summaryKey or 'summary')
self._summaryKey = None
_end_itunes_summary = _end_summary
def _start_enclosure(self, attrsD):
attrsD = self._itsAnHrefDamnIt(attrsD)
self._getContext().setdefault('enclosures', []).append(FeedParserDict(attrsD))
href = attrsD.get('href')
if href:
context = self._getContext()
if not context.get('id'):
context['id'] = href
def _start_source(self, attrsD):
self.insource = 1
def _end_source(self):
self.insource = 0
self._getContext()['source'] = copy.deepcopy(self.sourcedata)
self.sourcedata.clear()
def _start_content(self, attrsD):
self.pushContent('content', attrsD, 'text/plain', 1)
src = attrsD.get('src')
if src:
self.contentparams['src'] = src
self.push('content', 1)
def _start_prodlink(self, attrsD):
self.pushContent('content', attrsD, 'text/html', 1)
def _start_body(self, attrsD):
self.pushContent('content', attrsD, 'application/xhtml+xml', 1)
_start_xhtml_body = _start_body
def _start_content_encoded(self, attrsD):
self.pushContent('content', attrsD, 'text/html', 1)
_start_fullitem = _start_content_encoded
def _end_content(self):
copyToDescription = self.mapContentType(self.contentparams.get('type')) in (['text/plain'] + self.html_types)
value = self.popContent('content')
if copyToDescription:
self._save('description', value)
_end_body = _end_content
_end_xhtml_body = _end_content
_end_content_encoded = _end_content
_end_fullitem = _end_content
_end_prodlink = _end_content
def _start_itunes_image(self, attrsD):
self.push('itunes_image', 0)
self._getContext()['image'] = FeedParserDict({'href': attrsD.get('href')})
_start_itunes_link = _start_itunes_image
def _end_itunes_block(self):
value = self.pop('itunes_block', 0)
self._getContext()['itunes_block'] = (value == 'yes') and 1 or 0
def _end_itunes_explicit(self):
value = self.pop('itunes_explicit', 0)
self._getContext()['itunes_explicit'] = (value == 'yes') and 1 or 0
if _XML_AVAILABLE:
class _StrictFeedParser(_FeedParserMixin, xml.sax.handler.ContentHandler):
def __init__(self, baseuri, baselang, encoding):
if _debug: sys.stderr.write('trying StrictFeedParser\n')
xml.sax.handler.ContentHandler.__init__(self)
_FeedParserMixin.__init__(self, baseuri, baselang, encoding)
self.bozo = 0
self.exc = None
def startPrefixMapping(self, prefix, uri):
self.trackNamespace(prefix, uri)
def startElementNS(self, name, qname, attrs):
namespace, localname = name
lowernamespace = str(namespace or '').lower()
if lowernamespace.find('backend.userland.com/rss') <> -1:
# match any backend.userland.com namespace
namespace = 'http://backend.userland.com/rss'
lowernamespace = namespace
if qname and qname.find(':') > 0:
givenprefix = qname.split(':')[0]
else:
givenprefix = None
prefix = self._matchnamespaces.get(lowernamespace, givenprefix)
if givenprefix and (prefix == None or (prefix == '' and lowernamespace == '')) and not self.namespacesInUse.has_key(givenprefix):
raise UndeclaredNamespace, "'%s' is not associated with a namespace" % givenprefix
if prefix:
localname = prefix + ':' + localname
localname = str(localname).lower()
if _debug: sys.stderr.write('startElementNS: qname = %s, namespace = %s, givenprefix = %s, prefix = %s, attrs = %s, localname = %s\n' % (qname, namespace, givenprefix, prefix, attrs.items(), localname))
# qname implementation is horribly broken in Python 2.1 (it
# doesn't report any), and slightly broken in Python 2.2 (it
# doesn't report the xml: namespace). So we match up namespaces
# with a known list first, and then possibly override them with
# the qnames the SAX parser gives us (if indeed it gives us any
# at all). Thanks to MatejC for helping me test this and
# tirelessly telling me that it didn't work yet.
attrsD = {}
for (namespace, attrlocalname), attrvalue in attrs._attrs.items():
lowernamespace = (namespace or '').lower()
prefix = self._matchnamespaces.get(lowernamespace, '')
if prefix:
attrlocalname = prefix + ':' + attrlocalname
attrsD[str(attrlocalname).lower()] = attrvalue
for qname in attrs.getQNames():
attrsD[str(qname).lower()] = attrs.getValueByQName(qname)
self.unknown_starttag(localname, attrsD.items())
def characters(self, text):
self.handle_data(text)
def endElementNS(self, name, qname):
namespace, localname = name
lowernamespace = str(namespace or '').lower()
if qname and qname.find(':') > 0:
givenprefix = qname.split(':')[0]
else:
givenprefix = ''
prefix = self._matchnamespaces.get(lowernamespace, givenprefix)
if prefix:
localname = prefix + ':' + localname
localname = str(localname).lower()
self.unknown_endtag(localname)
def error(self, exc):
self.bozo = 1
self.exc = exc
def fatalError(self, exc):
self.error(exc)
raise exc
class _BaseHTMLProcessor(sgmllib.SGMLParser):
elements_no_end_tag = ['area', 'base', 'basefont', 'br', 'col', 'frame', 'hr',
'img', 'input', 'isindex', 'link', 'meta', 'param']
def __init__(self, encoding):
self.encoding = encoding
if _debug: sys.stderr.write('entering BaseHTMLProcessor, encoding=%s\n' % self.encoding)
sgmllib.SGMLParser.__init__(self)
def reset(self):
self.pieces = []
sgmllib.SGMLParser.reset(self)
def _shorttag_replace(self, match):
tag = match.group(1)
if tag in self.elements_no_end_tag:
return '<' + tag + ' />'
else:
return '<' + tag + '></' + tag + '>'
def feed(self, data):
data = re.compile(r'<!((?!DOCTYPE|--|\[))', re.IGNORECASE).sub(r'&lt;!\1', data)
#data = re.sub(r'<(\S+?)\s*?/>', self._shorttag_replace, data) # bug [ 1399464 ] Bad regexp for _shorttag_replace
data = re.sub(r'<([^<\s]+?)\s*/>', self._shorttag_replace, data)
data = data.replace('&#39;', "'")
data = data.replace('&#34;', '"')
if self.encoding and type(data) == type(u''):
data = data.encode(self.encoding)
sgmllib.SGMLParser.feed(self, data)
def normalize_attrs(self, attrs):
# utility method to be called by descendants
attrs = [(k.lower(), v) for k, v in attrs]
attrs = [(k, k in ('rel', 'type') and v.lower() or v) for k, v in attrs]
return attrs
def unknown_starttag(self, tag, attrs):
# called for each start tag
# attrs is a list of (attr, value) tuples
# e.g. for <pre class='screen'>, tag='pre', attrs=[('class', 'screen')]
if _debug: sys.stderr.write('_BaseHTMLProcessor, unknown_starttag, tag=%s\n' % tag)
uattrs = []
# thanks to Kevin Marks for this breathtaking hack to deal with (valid) high-bit attribute values in UTF-8 feeds
for key, value in attrs:
if type(value) != type(u''):
value = unicode(value, self.encoding)
uattrs.append((unicode(key, self.encoding), value))
strattrs = u''.join([u' %s="%s"' % (key, value) for key, value in uattrs]).encode(self.encoding)
if tag in self.elements_no_end_tag:
self.pieces.append('<%(tag)s%(strattrs)s />' % locals())
else:
self.pieces.append('<%(tag)s%(strattrs)s>' % locals())
def unknown_endtag(self, tag):
# called for each end tag, e.g. for </pre>, tag will be 'pre'
# Reconstruct the original end tag.
if tag not in self.elements_no_end_tag:
self.pieces.append("</%(tag)s>" % locals())
def handle_charref(self, ref):
# called for each character reference, e.g. for '&#160;', ref will be '160'
# Reconstruct the original character reference.
self.pieces.append('&#%(ref)s;' % locals())
def handle_entityref(self, ref):
# called for each entity reference, e.g. for '&copy;', ref will be 'copy'
# Reconstruct the original entity reference.
self.pieces.append('&%(ref)s;' % locals())
def handle_data(self, text):
# called for each block of plain text, i.e. outside of any tag and
# not containing any character or entity references
# Store the original text verbatim.
if _debug: sys.stderr.write('_BaseHTMLProcessor, handle_text, text=%s\n' % text)
self.pieces.append(text)
def handle_comment(self, text):
# called for each HTML comment, e.g. <!-- insert Javascript code here -->
# Reconstruct the original comment.
self.pieces.append('<!--%(text)s-->' % locals())
def handle_pi(self, text):
# called for each processing instruction, e.g. <?instruction>
# Reconstruct original processing instruction.
self.pieces.append('<?%(text)s>' % locals())
def handle_decl(self, text):
# called for the DOCTYPE, if present, e.g.
# <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
# "http://www.w3.org/TR/html4/loose.dtd">
# Reconstruct original DOCTYPE
self.pieces.append('<!%(text)s>' % locals())
_new_declname_match = re.compile(r'[a-zA-Z][-_.a-zA-Z0-9:]*\s*').match
def _scan_name(self, i, declstartpos):
rawdata = self.rawdata
n = len(rawdata)
if i == n:
return None, -1
m = self._new_declname_match(rawdata, i)
if m:
s = m.group()
name = s.strip()
if (i + len(s)) == n:
return None, -1 # end of buffer
return name.lower(), m.end()
else:
self.handle_data(rawdata)
# self.updatepos(declstartpos, i)
return None, -1
def output(self):
'''Return processed HTML as a single string'''
return ''.join([str(p) for p in self.pieces])
class _LooseFeedParser(_FeedParserMixin, _BaseHTMLProcessor):
def __init__(self, baseuri, baselang, encoding):
sgmllib.SGMLParser.__init__(self)
_FeedParserMixin.__init__(self, baseuri, baselang, encoding)
def decodeEntities(self, element, data):
data = data.replace('&#60;', '&lt;')
data = data.replace('&#x3c;', '&lt;')
data = data.replace('&#62;', '&gt;')
data = data.replace('&#x3e;', '&gt;')
data = data.replace('&#38;', '&amp;')
data = data.replace('&#x26;', '&amp;')
data = data.replace('&#34;', '&quot;')
data = data.replace('&#x22;', '&quot;')
data = data.replace('&#39;', '&apos;')
data = data.replace('&#x27;', '&apos;')
if self.contentparams.has_key('type') and not self.contentparams.get('type', 'xml').endswith('xml'):
data = data.replace('&lt;', '<')
data = data.replace('&gt;', '>')
data = data.replace('&amp;', '&')
data = data.replace('&quot;', '"')
data = data.replace('&apos;', "'")
return data
class _RelativeURIResolver(_BaseHTMLProcessor):
relative_uris = [('a', 'href'),
('applet', 'codebase'),
('area', 'href'),
('blockquote', 'cite'),
('body', 'background'),
('del', 'cite'),
('form', 'action'),
('frame', 'longdesc'),
('frame', 'src'),
('iframe', 'longdesc'),
('iframe', 'src'),
('head', 'profile'),
('img', 'longdesc'),
('img', 'src'),
('img', 'usemap'),
('input', 'src'),
('input', 'usemap'),
('ins', 'cite'),
('link', 'href'),
('object', 'classid'),
('object', 'codebase'),
('object', 'data'),
('object', 'usemap'),
('q', 'cite'),
('script', 'src')]
def __init__(self, baseuri, encoding):
_BaseHTMLProcessor.__init__(self, encoding)
self.baseuri = baseuri
def resolveURI(self, uri):
return _urljoin(self.baseuri, uri)
def unknown_starttag(self, tag, attrs):
attrs = self.normalize_attrs(attrs)
attrs = [(key, ((tag, key) in self.relative_uris) and self.resolveURI(value) or value) for key, value in attrs]
_BaseHTMLProcessor.unknown_starttag(self, tag, attrs)
def _resolveRelativeURIs(htmlSource, baseURI, encoding):
if _debug: sys.stderr.write('entering _resolveRelativeURIs\n')
p = _RelativeURIResolver(baseURI, encoding)
p.feed(htmlSource)
return p.output()
class _HTMLSanitizer(_BaseHTMLProcessor):
acceptable_elements = ['a', 'abbr', 'acronym', 'address', 'area', 'b', 'big',
'blockquote', 'br', 'button', 'caption', 'center', 'cite', 'code', 'col',
'colgroup', 'dd', 'del', 'dfn', 'dir', 'div', 'dl', 'dt', 'em', 'fieldset',
'font', 'form', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'hr', 'i', 'img', 'input',
'ins', 'kbd', 'label', 'legend', 'li', 'map', 'menu', 'ol', 'optgroup',
'option', 'p', 'pre', 'q', 's', 'samp', 'select', 'small', 'span', 'strike',
'strong', 'sub', 'sup', 'table', 'tbody', 'td', 'textarea', 'tfoot', 'th',
'thead', 'tr', 'tt', 'u', 'ul', 'var']
acceptable_attributes = ['abbr', 'accept', 'accept-charset', 'accesskey',
'action', 'align', 'alt', 'axis', 'border', 'cellpadding', 'cellspacing',
'char', 'charoff', 'charset', 'checked', 'cite', 'class', 'clear', 'cols',
'colspan', 'color', 'compact', 'coords', 'datetime', 'dir', 'disabled',
'enctype', 'for', 'frame', 'headers', 'height', 'href', 'hreflang', 'hspace',
'id', 'ismap', 'label', 'lang', 'longdesc', 'maxlength', 'media', 'method',
'multiple', 'name', 'nohref', 'noshade', 'nowrap', 'prompt', 'readonly',
'rel', 'rev', 'rows', 'rowspan', 'rules', 'scope', 'selected', 'shape', 'size',
'span', 'src', 'start', 'summary', 'tabindex', 'target', 'title', 'type',
'usemap', 'valign', 'value', 'vspace', 'width']
unacceptable_elements_with_end_tag = ['script', 'applet']
def reset(self):
_BaseHTMLProcessor.reset(self)
self.unacceptablestack = 0
def unknown_starttag(self, tag, attrs):
if not tag in self.acceptable_elements:
if tag in self.unacceptable_elements_with_end_tag:
self.unacceptablestack += 1
return
attrs = self.normalize_attrs(attrs)
attrs = [(key, value) for key, value in attrs if key in self.acceptable_attributes]
_BaseHTMLProcessor.unknown_starttag(self, tag, attrs)
def unknown_endtag(self, tag):
if not tag in self.acceptable_elements:
if tag in self.unacceptable_elements_with_end_tag:
self.unacceptablestack -= 1
return
_BaseHTMLProcessor.unknown_endtag(self, tag)
def handle_pi(self, text):
pass
def handle_decl(self, text):
pass
def handle_data(self, text):
if not self.unacceptablestack:
_BaseHTMLProcessor.handle_data(self, text)
def _sanitizeHTML(htmlSource, encoding):
p = _HTMLSanitizer(encoding)
p.feed(htmlSource)
data = p.output()
if TIDY_MARKUP:
# loop through list of preferred Tidy interfaces looking for one that's installed,
# then set up a common _tidy function to wrap the interface-specific API.
_tidy = None
for tidy_interface in PREFERRED_TIDY_INTERFACES:
try:
if tidy_interface == "uTidy":
from tidy import parseString as _utidy
def _tidy(data, **kwargs):
return str(_utidy(data, **kwargs))
break
elif tidy_interface == "mxTidy":
from mx.Tidy import Tidy as _mxtidy
def _tidy(data, **kwargs):
nerrors, nwarnings, data, errordata = _mxtidy.tidy(data, **kwargs)
return data
break
except:
pass
if _tidy:
utf8 = type(data) == type(u'')
if utf8:
data = data.encode('utf-8')
data = _tidy(data, output_xhtml=1, numeric_entities=1, wrap=0, char_encoding="utf8")
if utf8:
data = unicode(data, 'utf-8')
if data.count('<body'):
data = data.split('<body', 1)[1]
if data.count('>'):
data = data.split('>', 1)[1]
if data.count('</body'):
data = data.split('</body', 1)[0]
data = data.strip().replace('\r\n', '\n')
return data
class _FeedURLHandler(urllib2.HTTPDigestAuthHandler, urllib2.HTTPRedirectHandler, urllib2.HTTPDefaultErrorHandler):
def http_error_default(self, req, fp, code, msg, headers):
if ((code / 100) == 3) and (code != 304):
return self.http_error_302(req, fp, code, msg, headers)
infourl = urllib.addinfourl(fp, headers, req.get_full_url())
infourl.status = code
return infourl
def http_error_302(self, req, fp, code, msg, headers):
if headers.dict.has_key('location'):
infourl = urllib2.HTTPRedirectHandler.http_error_302(self, req, fp, code, msg, headers)
else:
infourl = urllib.addinfourl(fp, headers, req.get_full_url())
if not hasattr(infourl, 'status'):
infourl.status = code
return infourl
def http_error_301(self, req, fp, code, msg, headers):
if headers.dict.has_key('location'):
infourl = urllib2.HTTPRedirectHandler.http_error_301(self, req, fp, code, msg, headers)
else:
infourl = urllib.addinfourl(fp, headers, req.get_full_url())
if not hasattr(infourl, 'status'):
infourl.status = code
return infourl
http_error_300 = http_error_302
http_error_303 = http_error_302
http_error_307 = http_error_302
def http_error_401(self, req, fp, code, msg, headers):
# Check if
# - server requires digest auth, AND
# - we tried (unsuccessfully) with basic auth, AND
# - we're using Python 2.3.3 or later (digest auth is irreparably broken in earlier versions)
# If all conditions hold, parse authentication information
# out of the Authorization header we sent the first time
# (for the username and password) and the WWW-Authenticate
# header the server sent back (for the realm) and retry
# the request with the appropriate digest auth headers instead.
# This evil genius hack has been brought to you by Aaron Swartz.
host = urlparse.urlparse(req.get_full_url())[1]
try:
assert sys.version.split()[0] >= '2.3.3'
assert base64 != None
user, passw = base64.decodestring(req.headers['Authorization'].split(' ')[1]).split(':')
realm = re.findall('realm="([^"]*)"', headers['WWW-Authenticate'])[0]
self.add_password(realm, host, user, passw)
retry = self.http_error_auth_reqed('www-authenticate', host, req, headers)
self.reset_retry_count()
return retry
except:
return self.http_error_default(req, fp, code, msg, headers)
def _open_resource(url_file_stream_or_string, etag, modified, agent, referrer, handlers):
"""URL, filename, or string --> stream
This function lets you define parsers that take any input source
(URL, pathname to local or network file, or actual data as a string)
and deal with it in a uniform manner. Returned object is guaranteed
to have all the basic stdio read methods (read, readline, readlines).
Just .close() the object when you're done with it.
If the etag argument is supplied, it will be used as the value of an
If-None-Match request header.
If the modified argument is supplied, it must be a tuple of 9 integers
as returned by gmtime() in the standard Python time module. This MUST
be in GMT (Greenwich Mean Time). The formatted date/time will be used
as the value of an If-Modified-Since request header.
If the agent argument is supplied, it will be used as the value of a
User-Agent request header.
If the referrer argument is supplied, it will be used as the value of a
Referer[sic] request header.
If handlers is supplied, it is a list of handlers used to build a
urllib2 opener.
"""
if hasattr(url_file_stream_or_string, 'read'):
return url_file_stream_or_string
if url_file_stream_or_string == '-':
return sys.stdin
if urlparse.urlparse(url_file_stream_or_string)[0] in ('http', 'https', 'ftp'):
if not agent:
agent = USER_AGENT
# test for inline user:password for basic auth
auth = None
if base64:
urltype, rest = urllib.splittype(url_file_stream_or_string)
realhost, rest = urllib.splithost(rest)
if realhost:
user_passwd, realhost = urllib.splituser(realhost)
if user_passwd:
url_file_stream_or_string = '%s://%s%s' % (urltype, realhost, rest)
auth = base64.encodestring(user_passwd).strip()
# try to open with urllib2 (to use optional headers)
request = urllib2.Request(url_file_stream_or_string)
request.add_header('User-Agent', agent)
if etag:
request.add_header('If-None-Match', etag)
if modified:
# format into an RFC 1123-compliant timestamp. We can't use
# time.strftime() since the %a and %b directives can be affected
# by the current locale, but RFC 2616 states that dates must be
# in English.
short_weekdays = ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun']
months = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
request.add_header('If-Modified-Since', '%s, %02d %s %04d %02d:%02d:%02d GMT' % (short_weekdays[modified[6]], modified[2], months[modified[1] - 1], modified[0], modified[3], modified[4], modified[5]))
if referrer:
request.add_header('Referer', referrer)
if gzip and zlib:
request.add_header('Accept-encoding', 'gzip, deflate')
elif gzip:
request.add_header('Accept-encoding', 'gzip')
elif zlib:
request.add_header('Accept-encoding', 'deflate')
else:
request.add_header('Accept-encoding', '')
if auth:
request.add_header('Authorization', 'Basic %s' % auth)
if ACCEPT_HEADER:
request.add_header('Accept', ACCEPT_HEADER)
request.add_header('A-IM', 'feed') # RFC 3229 support
opener = apply(urllib2.build_opener, tuple([_FeedURLHandler()] + handlers))
opener.addheaders = [] # RMK - must clear so we only send our custom User-Agent
try:
return opener.open(request)
finally:
opener.close() # JohnD
# try to open with native open function (if url_file_stream_or_string is a filename)
try:
return open(url_file_stream_or_string)
except:
pass
# treat url_file_stream_or_string as string
return _StringIO(str(url_file_stream_or_string))
_date_handlers = []
def registerDateHandler(func):
'''Register a date handler function (takes string, returns 9-tuple date in GMT)'''
_date_handlers.insert(0, func)
# ISO-8601 date parsing routines written by Fazal Majid.
# The ISO 8601 standard is very convoluted and irregular - a full ISO 8601
# parser is beyond the scope of feedparser and would be a worthwhile addition
# to the Python library.
# A single regular expression cannot parse ISO 8601 date formats into groups
# as the standard is highly irregular (for instance is 030104 2003-01-04 or
# 0301-04-01), so we use templates instead.
# Please note the order in templates is significant because we need a
# greedy match.
_iso8601_tmpl = ['YYYY-?MM-?DD', 'YYYY-MM', 'YYYY-?OOO',
'YY-?MM-?DD', 'YY-?OOO', 'YYYY',
'-YY-?MM', '-OOO', '-YY',
'--MM-?DD', '--MM',
'---DD',
'CC', '']
_iso8601_re = [
tmpl.replace(
'YYYY', r'(?P<year>\d{4})').replace(
'YY', r'(?P<year>\d\d)').replace(
'MM', r'(?P<month>[01]\d)').replace(
'DD', r'(?P<day>[0123]\d)').replace(
'OOO', r'(?P<ordinal>[0123]\d\d)').replace(
'CC', r'(?P<century>\d\d$)')
+ r'(T?(?P<hour>\d{2}):(?P<minute>\d{2})'
+ r'(:(?P<second>\d{2}))?'
+ r'(?P<tz>[+-](?P<tzhour>\d{2})(:(?P<tzmin>\d{2}))?|Z)?)?'
for tmpl in _iso8601_tmpl]
del tmpl
_iso8601_matches = [re.compile(regex).match for regex in _iso8601_re]
del regex
def _parse_date_iso8601(dateString):
'''Parse a variety of ISO-8601-compatible formats like 20040105'''
m = None
for _iso8601_match in _iso8601_matches:
m = _iso8601_match(dateString)
if m: break
if not m: return
if m.span() == (0, 0): return
params = m.groupdict()
ordinal = params.get('ordinal', 0)
if ordinal:
ordinal = int(ordinal)
else:
ordinal = 0
year = params.get('year', '--')
if not year or year == '--':
year = time.gmtime()[0]
elif len(year) == 2:
# ISO 8601 assumes current century, i.e. 93 -> 2093, NOT 1993
year = 100 * int(time.gmtime()[0] / 100) + int(year)
else:
year = int(year)
month = params.get('month', '-')
if not month or month == '-':
# ordinals are NOT normalized by mktime, we simulate them
# by setting month=1, day=ordinal
if ordinal:
month = 1
else:
month = time.gmtime()[1]
month = int(month)
day = params.get('day', 0)
if not day:
# see above
if ordinal:
day = ordinal
elif params.get('century', 0) or \
params.get('year', 0) or params.get('month', 0):
day = 1
else:
day = time.gmtime()[2]
else:
day = int(day)
# special case of the century - is the first year of the 21st century
# 2000 or 2001 ? The debate goes on...
if 'century' in params.keys():
year = (int(params['century']) - 1) * 100 + 1
# in ISO 8601 most fields are optional
for field in ['hour', 'minute', 'second', 'tzhour', 'tzmin']:
if not params.get(field, None):
params[field] = 0
hour = int(params.get('hour', 0))
minute = int(params.get('minute', 0))
second = int(params.get('second', 0))
# weekday is normalized by mktime(), we can ignore it
weekday = 0
# daylight savings is complex, but not needed for feedparser's purposes
# as time zones, if specified, include mention of whether it is active
# (e.g. PST vs. PDT, CET). Using -1 is implementation-dependent and
# and most implementations have DST bugs
daylight_savings_flag = 0
tm = [year, month, day, hour, minute, second, weekday,
ordinal, daylight_savings_flag]
# ISO 8601 time zone adjustments
tz = params.get('tz')
if tz and tz != 'Z':
if tz[0] == '-':
tm[3] += int(params.get('tzhour', 0))
tm[4] += int(params.get('tzmin', 0))
elif tz[0] == '+':
tm[3] -= int(params.get('tzhour', 0))
tm[4] -= int(params.get('tzmin', 0))
else:
return None
# Python's time.mktime() is a wrapper around the ANSI C mktime(3c)
# which is guaranteed to normalize d/m/y/h/m/s.
# Many implementations have bugs, but we'll pretend they don't.
return time.localtime(time.mktime(tm))
registerDateHandler(_parse_date_iso8601)
# 8-bit date handling routines written by ytrewq1.
_korean_year = u'\ub144' # b3e2 in euc-kr
_korean_month = u'\uc6d4' # bff9 in euc-kr
_korean_day = u'\uc77c' # c0cf in euc-kr
_korean_am = u'\uc624\uc804' # bfc0 c0fc in euc-kr
_korean_pm = u'\uc624\ud6c4' # bfc0 c8c4 in euc-kr
_korean_onblog_date_re = \
re.compile('(\d{4})%s\s+(\d{2})%s\s+(\d{2})%s\s+(\d{2}):(\d{2}):(\d{2})' % \
(_korean_year, _korean_month, _korean_day))
_korean_nate_date_re = \
re.compile(u'(\d{4})-(\d{2})-(\d{2})\s+(%s|%s)\s+(\d{,2}):(\d{,2}):(\d{,2})' % \
(_korean_am, _korean_pm))
def _parse_date_onblog(dateString):
'''Parse a string according to the OnBlog 8-bit date format'''
m = _korean_onblog_date_re.match(dateString)
if not m: return
w3dtfdate = '%(year)s-%(month)s-%(day)sT%(hour)s:%(minute)s:%(second)s%(zonediff)s' % \
{'year': m.group(1), 'month': m.group(2), 'day': m.group(3),\
'hour': m.group(4), 'minute': m.group(5), 'second': m.group(6),\
'zonediff': '+09:00'}
if _debug: sys.stderr.write('OnBlog date parsed as: %s\n' % w3dtfdate)
return _parse_date_w3dtf(w3dtfdate)
registerDateHandler(_parse_date_onblog)
def _parse_date_nate(dateString):
'''Parse a string according to the Nate 8-bit date format'''
m = _korean_nate_date_re.match(dateString)
if not m: return
hour = int(m.group(5))
ampm = m.group(4)
if (ampm == _korean_pm):
hour += 12
hour = str(hour)
if len(hour) == 1:
hour = '0' + hour
w3dtfdate = '%(year)s-%(month)s-%(day)sT%(hour)s:%(minute)s:%(second)s%(zonediff)s' % \
{'year': m.group(1), 'month': m.group(2), 'day': m.group(3),\
'hour': hour, 'minute': m.group(6), 'second': m.group(7),\
'zonediff': '+09:00'}
if _debug: sys.stderr.write('Nate date parsed as: %s\n' % w3dtfdate)
return _parse_date_w3dtf(w3dtfdate)
registerDateHandler(_parse_date_nate)
_mssql_date_re = \
re.compile('(\d{4})-(\d{2})-(\d{2})\s+(\d{2}):(\d{2}):(\d{2})(\.\d+)?')
def _parse_date_mssql(dateString):
'''Parse a string according to the MS SQL date format'''
m = _mssql_date_re.match(dateString)
if not m: return
w3dtfdate = '%(year)s-%(month)s-%(day)sT%(hour)s:%(minute)s:%(second)s%(zonediff)s' % \
{'year': m.group(1), 'month': m.group(2), 'day': m.group(3),\
'hour': m.group(4), 'minute': m.group(5), 'second': m.group(6),\
'zonediff': '+09:00'}
if _debug: sys.stderr.write('MS SQL date parsed as: %s\n' % w3dtfdate)
return _parse_date_w3dtf(w3dtfdate)
registerDateHandler(_parse_date_mssql)
# Unicode strings for Greek date strings
_greek_months = \
{ \
u'\u0399\u03b1\u03bd': u'Jan', # c9e1ed in iso-8859-7
u'\u03a6\u03b5\u03b2': u'Feb', # d6e5e2 in iso-8859-7
u'\u039c\u03ac\u03ce': u'Mar', # ccdcfe in iso-8859-7
u'\u039c\u03b1\u03ce': u'Mar', # cce1fe in iso-8859-7
u'\u0391\u03c0\u03c1': u'Apr', # c1f0f1 in iso-8859-7
u'\u039c\u03ac\u03b9': u'May', # ccdce9 in iso-8859-7
u'\u039c\u03b1\u03ca': u'May', # cce1fa in iso-8859-7
u'\u039c\u03b1\u03b9': u'May', # cce1e9 in iso-8859-7
u'\u0399\u03bf\u03cd\u03bd': u'Jun', # c9effded in iso-8859-7
u'\u0399\u03bf\u03bd': u'Jun', # c9efed in iso-8859-7
u'\u0399\u03bf\u03cd\u03bb': u'Jul', # c9effdeb in iso-8859-7
u'\u0399\u03bf\u03bb': u'Jul', # c9f9eb in iso-8859-7
u'\u0391\u03cd\u03b3': u'Aug', # c1fde3 in iso-8859-7
u'\u0391\u03c5\u03b3': u'Aug', # c1f5e3 in iso-8859-7
u'\u03a3\u03b5\u03c0': u'Sep', # d3e5f0 in iso-8859-7
u'\u039f\u03ba\u03c4': u'Oct', # cfeaf4 in iso-8859-7
u'\u039d\u03bf\u03ad': u'Nov', # cdefdd in iso-8859-7
u'\u039d\u03bf\u03b5': u'Nov', # cdefe5 in iso-8859-7
u'\u0394\u03b5\u03ba': u'Dec', # c4e5ea in iso-8859-7
}
_greek_wdays = \
{ \
u'\u039a\u03c5\u03c1': u'Sun', # caf5f1 in iso-8859-7
u'\u0394\u03b5\u03c5': u'Mon', # c4e5f5 in iso-8859-7
u'\u03a4\u03c1\u03b9': u'Tue', # d4f1e9 in iso-8859-7
u'\u03a4\u03b5\u03c4': u'Wed', # d4e5f4 in iso-8859-7
u'\u03a0\u03b5\u03bc': u'Thu', # d0e5ec in iso-8859-7
u'\u03a0\u03b1\u03c1': u'Fri', # d0e1f1 in iso-8859-7
u'\u03a3\u03b1\u03b2': u'Sat', # d3e1e2 in iso-8859-7
}
_greek_date_format_re = \
re.compile(u'([^,]+),\s+(\d{2})\s+([^\s]+)\s+(\d{4})\s+(\d{2}):(\d{2}):(\d{2})\s+([^\s]+)')
def _parse_date_greek(dateString):
'''Parse a string according to a Greek 8-bit date format.'''
m = _greek_date_format_re.match(dateString)
if not m: return
try:
wday = _greek_wdays[m.group(1)]
month = _greek_months[m.group(3)]
except:
return
rfc822date = '%(wday)s, %(day)s %(month)s %(year)s %(hour)s:%(minute)s:%(second)s %(zonediff)s' % \
{'wday': wday, 'day': m.group(2), 'month': month, 'year': m.group(4),\
'hour': m.group(5), 'minute': m.group(6), 'second': m.group(7),\
'zonediff': m.group(8)}
if _debug: sys.stderr.write('Greek date parsed as: %s\n' % rfc822date)
return _parse_date_rfc822(rfc822date)
registerDateHandler(_parse_date_greek)
# Unicode strings for Hungarian date strings
_hungarian_months = \
{ \
u'janu\u00e1r': u'01', # e1 in iso-8859-2
u'febru\u00e1ri': u'02', # e1 in iso-8859-2
u'm\u00e1rcius': u'03', # e1 in iso-8859-2
u'\u00e1prilis': u'04', # e1 in iso-8859-2
u'm\u00e1ujus': u'05', # e1 in iso-8859-2
u'j\u00fanius': u'06', # fa in iso-8859-2
u'j\u00falius': u'07', # fa in iso-8859-2
u'augusztus': u'08',
u'szeptember': u'09',
u'okt\u00f3ber': u'10', # f3 in iso-8859-2
u'november': u'11',
u'december': u'12',
}
_hungarian_date_format_re = \
re.compile(u'(\d{4})-([^-]+)-(\d{,2})T(\d{,2}):(\d{2})((\+|-)(\d{,2}:\d{2}))')
def _parse_date_hungarian(dateString):
'''Parse a string according to a Hungarian 8-bit date format.'''
m = _hungarian_date_format_re.match(dateString)
if not m: return
try:
month = _hungarian_months[m.group(2)]
day = m.group(3)
if len(day) == 1:
day = '0' + day
hour = m.group(4)
if len(hour) == 1:
hour = '0' + hour
except:
return
w3dtfdate = '%(year)s-%(month)s-%(day)sT%(hour)s:%(minute)s%(zonediff)s' % \
{'year': m.group(1), 'month': month, 'day': day,\
'hour': hour, 'minute': m.group(5),\
'zonediff': m.group(6)}
if _debug: sys.stderr.write('Hungarian date parsed as: %s\n' % w3dtfdate)
return _parse_date_w3dtf(w3dtfdate)
registerDateHandler(_parse_date_hungarian)
# W3DTF-style date parsing adapted from PyXML xml.utils.iso8601, written by
# Drake and licensed under the Python license. Removed all range checking
# for month, day, hour, minute, and second, since mktime will normalize
# these later
def _parse_date_w3dtf(dateString):
def __extract_date(m):
year = int(m.group('year'))
if year < 100:
year = 100 * int(time.gmtime()[0] / 100) + int(year)
if year < 1000:
return 0, 0, 0
julian = m.group('julian')
if julian:
julian = int(julian)
month = julian / 30 + 1
day = julian % 30 + 1
jday = None
while jday != julian:
t = time.mktime((year, month, day, 0, 0, 0, 0, 0, 0))
jday = time.gmtime(t)[-2]
diff = abs(jday - julian)
if jday > julian:
if diff < day:
day = day - diff
else:
month = month - 1
day = 31
elif jday < julian:
if day + diff < 28:
day = day + diff
else:
month = month + 1
return year, month, day
month = m.group('month')
day = 1
if month is None:
month = 1
else:
month = int(month)
day = m.group('day')
if day:
day = int(day)
else:
day = 1
return year, month, day
def __extract_time(m):
if not m:
return 0, 0, 0
hours = m.group('hours')
if not hours:
return 0, 0, 0
hours = int(hours)
minutes = int(m.group('minutes'))
seconds = m.group('seconds')
if seconds:
seconds = int(seconds)
else:
seconds = 0
return hours, minutes, seconds
def __extract_tzd(m):
'''Return the Time Zone Designator as an offset in seconds from UTC.'''
if not m:
return 0
tzd = m.group('tzd')
if not tzd:
return 0
if tzd == 'Z':
return 0
hours = int(m.group('tzdhours'))
minutes = m.group('tzdminutes')
if minutes:
minutes = int(minutes)
else:
minutes = 0
offset = (hours*60 + minutes) * 60
if tzd[0] == '+':
return -offset
return offset
__date_re = ('(?P<year>\d\d\d\d)'
'(?:(?P<dsep>-|)'
'(?:(?P<julian>\d\d\d)'
'|(?P<month>\d\d)(?:(?P=dsep)(?P<day>\d\d))?))?')
__tzd_re = '(?P<tzd>[-+](?P<tzdhours>\d\d)(?::?(?P<tzdminutes>\d\d))|Z)'
__tzd_rx = re.compile(__tzd_re)
__time_re = ('(?P<hours>\d\d)(?P<tsep>:|)(?P<minutes>\d\d)'
'(?:(?P=tsep)(?P<seconds>\d\d(?:[.,]\d+)?))?'
+ __tzd_re)
__datetime_re = '%s(?:T%s)?' % (__date_re, __time_re)
__datetime_rx = re.compile(__datetime_re)
m = __datetime_rx.match(dateString)
if (m is None) or (m.group() != dateString): return
gmt = __extract_date(m) + __extract_time(m) + (0, 0, 0)
if gmt[0] == 0: return
return time.gmtime(time.mktime(gmt) + __extract_tzd(m) - time.timezone)
registerDateHandler(_parse_date_w3dtf)
def _parse_date_rfc822(dateString):
'''Parse an RFC822, RFC1123, RFC2822, or asctime-style date'''
data = dateString.split()
if data[0][-1] in (',', '.') or data[0].lower() in rfc822._daynames:
del data[0]
if len(data) == 4:
s = data[3]
i = s.find('+')
if i > 0:
data[3:] = [s[:i], s[i+1:]]
else:
data.append('')
dateString = " ".join(data)
if len(data) < 5:
dateString += ' 00:00:00 GMT'
tm = rfc822.parsedate_tz(dateString)
if tm:
return time.gmtime(rfc822.mktime_tz(tm))
# rfc822.py defines several time zones, but we define some extra ones.
# 'ET' is equivalent to 'EST', etc.
_additional_timezones = {'AT': -400, 'ET': -500, 'CT': -600, 'MT': -700, 'PT': -800}
rfc822._timezones.update(_additional_timezones)
registerDateHandler(_parse_date_rfc822)
def _parse_date(dateString):
'''Parses a variety of date formats into a 9-tuple in GMT'''
for handler in _date_handlers:
try:
date9tuple = handler(dateString)
if not date9tuple: continue
if len(date9tuple) != 9:
if _debug: sys.stderr.write('date handler function must return 9-tuple\n')
raise ValueError
map(int, date9tuple)
return date9tuple
except Exception, e:
if _debug: sys.stderr.write('%s raised %s\n' % (handler.__name__, repr(e)))
pass
return None
def _getCharacterEncoding(http_headers, xml_data):
'''Get the character encoding of the XML document
http_headers is a dictionary
xml_data is a raw string (not Unicode)
This is so much trickier than it sounds, it's not even funny.
According to RFC 3023 ('XML Media Types'), if the HTTP Content-Type
is application/xml, application/*+xml,
application/xml-external-parsed-entity, or application/xml-dtd,
the encoding given in the charset parameter of the HTTP Content-Type
takes precedence over the encoding given in the XML prefix within the
document, and defaults to 'utf-8' if neither are specified. But, if
the HTTP Content-Type is text/xml, text/*+xml, or
text/xml-external-parsed-entity, the encoding given in the XML prefix
within the document is ALWAYS IGNORED and only the encoding given in
the charset parameter of the HTTP Content-Type header should be
respected, and it defaults to 'us-ascii' if not specified.
Furthermore, discussion on the atom-syntax mailing list with the
author of RFC 3023 leads me to the conclusion that any document
served with a Content-Type of text/* and no charset parameter
must be treated as us-ascii. (We now do this.) And also that it
must always be flagged as non-well-formed. (We now do this too.)
If Content-Type is unspecified (input was local file or non-HTTP source)
or unrecognized (server just got it totally wrong), then go by the
encoding given in the XML prefix of the document and default to
'iso-8859-1' as per the HTTP specification (RFC 2616).
Then, assuming we didn't find a character encoding in the HTTP headers
(and the HTTP Content-type allowed us to look in the body), we need
to sniff the first few bytes of the XML data and try to determine
whether the encoding is ASCII-compatible. Section F of the XML
specification shows the way here:
http://www.w3.org/TR/REC-xml/#sec-guessing-no-ext-info
If the sniffed encoding is not ASCII-compatible, we need to make it
ASCII compatible so that we can sniff further into the XML declaration
to find the encoding attribute, which will tell us the true encoding.
Of course, none of this guarantees that we will be able to parse the
feed in the declared character encoding (assuming it was declared
correctly, which many are not). CJKCodecs and iconv_codec help a lot;
you should definitely install them if you can.
http://cjkpython.i18n.org/
'''
def _parseHTTPContentType(content_type):
'''takes HTTP Content-Type header and returns (content type, charset)
If no charset is specified, returns (content type, '')
If no content type is specified, returns ('', '')
Both return parameters are guaranteed to be lowercase strings
'''
content_type = content_type or ''
content_type, params = cgi.parse_header(content_type)
return content_type, params.get('charset', '').replace("'", '')
sniffed_xml_encoding = ''
xml_encoding = ''
true_encoding = ''
http_content_type, http_encoding = _parseHTTPContentType(http_headers.get('content-type'))
# Must sniff for non-ASCII-compatible character encodings before
# searching for XML declaration. This heuristic is defined in
# section F of the XML specification:
# http://www.w3.org/TR/REC-xml/#sec-guessing-no-ext-info
try:
if xml_data[:4] == '\x4c\x6f\xa7\x94':
# EBCDIC
xml_data = _ebcdic_to_ascii(xml_data)
elif xml_data[:4] == '\x00\x3c\x00\x3f':
# UTF-16BE
sniffed_xml_encoding = 'utf-16be'
xml_data = unicode(xml_data, 'utf-16be').encode('utf-8')
elif (len(xml_data) >= 4) and (xml_data[:2] == '\xfe\xff') and (xml_data[2:4] != '\x00\x00'):
# UTF-16BE with BOM
sniffed_xml_encoding = 'utf-16be'
xml_data = unicode(xml_data[2:], 'utf-16be').encode('utf-8')
elif xml_data[:4] == '\x3c\x00\x3f\x00':
# UTF-16LE
sniffed_xml_encoding = 'utf-16le'
xml_data = unicode(xml_data, 'utf-16le').encode('utf-8')
elif (len(xml_data) >= 4) and (xml_data[:2] == '\xff\xfe') and (xml_data[2:4] != '\x00\x00'):
# UTF-16LE with BOM
sniffed_xml_encoding = 'utf-16le'
xml_data = unicode(xml_data[2:], 'utf-16le').encode('utf-8')
elif xml_data[:4] == '\x00\x00\x00\x3c':
# UTF-32BE
sniffed_xml_encoding = 'utf-32be'
xml_data = unicode(xml_data, 'utf-32be').encode('utf-8')
elif xml_data[:4] == '\x3c\x00\x00\x00':
# UTF-32LE
sniffed_xml_encoding = 'utf-32le'
xml_data = unicode(xml_data, 'utf-32le').encode('utf-8')
elif xml_data[:4] == '\x00\x00\xfe\xff':
# UTF-32BE with BOM
sniffed_xml_encoding = 'utf-32be'
xml_data = unicode(xml_data[4:], 'utf-32be').encode('utf-8')
elif xml_data[:4] == '\xff\xfe\x00\x00':
# UTF-32LE with BOM
sniffed_xml_encoding = 'utf-32le'
xml_data = unicode(xml_data[4:], 'utf-32le').encode('utf-8')
elif xml_data[:3] == '\xef\xbb\xbf':
# UTF-8 with BOM
sniffed_xml_encoding = 'utf-8'
xml_data = unicode(xml_data[3:], 'utf-8').encode('utf-8')
else:
# ASCII-compatible
pass
xml_encoding_match = re.compile('^<\?.*encoding=[\'"](.*?)[\'"].*\?>').match(xml_data)
except:
xml_encoding_match = None
if xml_encoding_match:
xml_encoding = xml_encoding_match.groups()[0].lower()
if sniffed_xml_encoding and (xml_encoding in ('iso-10646-ucs-2', 'ucs-2', 'csunicode', 'iso-10646-ucs-4', 'ucs-4', 'csucs4', 'utf-16', 'utf-32', 'utf_16', 'utf_32', 'utf16', 'u16')):
xml_encoding = sniffed_xml_encoding
acceptable_content_type = 0
application_content_types = ('application/xml', 'application/xml-dtd', 'application/xml-external-parsed-entity')
text_content_types = ('text/xml', 'text/xml-external-parsed-entity')
if (http_content_type in application_content_types) or \
(http_content_type.startswith('application/') and http_content_type.endswith('+xml')):
acceptable_content_type = 1
true_encoding = http_encoding or xml_encoding or 'utf-8'
elif (http_content_type in text_content_types) or \
(http_content_type.startswith('text/')) and http_content_type.endswith('+xml'):
acceptable_content_type = 1
true_encoding = http_encoding or 'us-ascii'
elif http_content_type.startswith('text/'):
true_encoding = http_encoding or 'us-ascii'
elif http_headers and (not http_headers.has_key('content-type')):
true_encoding = xml_encoding or 'iso-8859-1'
else:
true_encoding = xml_encoding or 'utf-8'
return true_encoding, http_encoding, xml_encoding, sniffed_xml_encoding, acceptable_content_type
def _toUTF8(data, encoding):
'''Changes an XML data stream on the fly to specify a new encoding
data is a raw sequence of bytes (not Unicode) that is presumed to be in %encoding already
encoding is a string recognized by encodings.aliases
'''
if _debug: sys.stderr.write('entering _toUTF8, trying encoding %s\n' % encoding)
# strip Byte Order Mark (if present)
if (len(data) >= 4) and (data[:2] == '\xfe\xff') and (data[2:4] != '\x00\x00'):
if _debug:
sys.stderr.write('stripping BOM\n')
if encoding != 'utf-16be':
sys.stderr.write('trying utf-16be instead\n')
encoding = 'utf-16be'
data = data[2:]
elif (len(data) >= 4) and (data[:2] == '\xff\xfe') and (data[2:4] != '\x00\x00'):
if _debug:
sys.stderr.write('stripping BOM\n')
if encoding != 'utf-16le':
sys.stderr.write('trying utf-16le instead\n')
encoding = 'utf-16le'
data = data[2:]
elif data[:3] == '\xef\xbb\xbf':
if _debug:
sys.stderr.write('stripping BOM\n')
if encoding != 'utf-8':
sys.stderr.write('trying utf-8 instead\n')
encoding = 'utf-8'
data = data[3:]
elif data[:4] == '\x00\x00\xfe\xff':
if _debug:
sys.stderr.write('stripping BOM\n')
if encoding != 'utf-32be':
sys.stderr.write('trying utf-32be instead\n')
encoding = 'utf-32be'
data = data[4:]
elif data[:4] == '\xff\xfe\x00\x00':
if _debug:
sys.stderr.write('stripping BOM\n')
if encoding != 'utf-32le':
sys.stderr.write('trying utf-32le instead\n')
encoding = 'utf-32le'
data = data[4:]
newdata = unicode(data, encoding)
if _debug: sys.stderr.write('successfully converted %s data to unicode\n' % encoding)
declmatch = re.compile('^<\?xml[^>]*?>')
newdecl = '''<?xml version='1.0' encoding='utf-8'?>'''
if declmatch.search(newdata):
newdata = declmatch.sub(newdecl, newdata)
else:
newdata = newdecl + u'\n' + newdata
return newdata.encode('utf-8')
def _stripDoctype(data):
'''Strips DOCTYPE from XML document, returns (rss_version, stripped_data)
rss_version may be 'rss091n' or None
stripped_data is the same XML document, minus the DOCTYPE
'''
entity_pattern = re.compile(r'<!ENTITY([^>]*?)>', re.MULTILINE)
data = entity_pattern.sub('', data)
doctype_pattern = re.compile(r'<!DOCTYPE([^>]*?)>', re.MULTILINE)
doctype_results = doctype_pattern.findall(data)
doctype = doctype_results and doctype_results[0] or ''
if doctype.lower().count('netscape'):
version = 'rss091n'
else:
version = None
data = doctype_pattern.sub('', data)
return version, data
def parse(url_file_stream_or_string, etag=None, modified=None, agent=None, referrer=None, handlers=[]):
'''Parse a feed from a URL, file, stream, or string'''
result = FeedParserDict()
result['feed'] = FeedParserDict()
result['entries'] = []
if _XML_AVAILABLE:
result['bozo'] = 0
if type(handlers) == types.InstanceType:
handlers = [handlers]
try:
f = _open_resource(url_file_stream_or_string, etag, modified, agent, referrer, handlers)
data = f.read()
except Exception, e:
result['bozo'] = 1
result['bozo_exception'] = e
data = ''
f = None
# if feed is gzip-compressed, decompress it
if f and data and hasattr(f, 'headers'):
if gzip and f.headers.get('content-encoding', '') == 'gzip':
try:
data = gzip.GzipFile(fileobj=_StringIO(data)).read()
except Exception, e:
# Some feeds claim to be gzipped but they're not, so
# we get garbage. Ideally, we should re-request the
# feed without the 'Accept-encoding: gzip' header,
# but we don't.
result['bozo'] = 1
result['bozo_exception'] = e
data = ''
elif zlib and f.headers.get('content-encoding', '') == 'deflate':
try:
data = zlib.decompress(data, -zlib.MAX_WBITS)
except Exception, e:
result['bozo'] = 1
result['bozo_exception'] = e
data = ''
# save HTTP headers
if hasattr(f, 'info'):
info = f.info()
result['etag'] = info.getheader('ETag')
last_modified = info.getheader('Last-Modified')
if last_modified:
result['modified'] = _parse_date(last_modified)
if hasattr(f, 'url'):
result['href'] = f.url
result['status'] = 200
if hasattr(f, 'status'):
result['status'] = f.status
if hasattr(f, 'headers'):
result['headers'] = f.headers.dict
if hasattr(f, 'close'):
f.close()
# there are four encodings to keep track of:
# - http_encoding is the encoding declared in the Content-Type HTTP header
# - xml_encoding is the encoding declared in the <?xml declaration
# - sniffed_encoding is the encoding sniffed from the first 4 bytes of the XML data
# - result['encoding'] is the actual encoding, as per RFC 3023 and a variety of other conflicting specifications
http_headers = result.get('headers', {})
result['encoding'], http_encoding, xml_encoding, sniffed_xml_encoding, acceptable_content_type = \
_getCharacterEncoding(http_headers, data)
if http_headers and (not acceptable_content_type):
if http_headers.has_key('content-type'):
bozo_message = '%s is not an XML media type' % http_headers['content-type']
else:
bozo_message = 'no Content-type specified'
result['bozo'] = 1
result['bozo_exception'] = NonXMLContentType(bozo_message)
result['version'], data = _stripDoctype(data)
baseuri = http_headers.get('content-location', result.get('href'))
baselang = http_headers.get('content-language', None)
# if server sent 304, we're done
if result.get('status', 0) == 304:
result['version'] = ''
result['debug_message'] = 'The feed has not changed since you last checked, ' + \
'so the server sent no data. This is a feature, not a bug!'
return result
# if there was a problem downloading, we're done
if not data:
return result
# determine character encoding
use_strict_parser = 0
known_encoding = 0
tried_encodings = []
# try: HTTP encoding, declared XML encoding, encoding sniffed from BOM
for proposed_encoding in (result['encoding'], xml_encoding, sniffed_xml_encoding):
if not proposed_encoding: continue
if proposed_encoding in tried_encodings: continue
tried_encodings.append(proposed_encoding)
try:
data = _toUTF8(data, proposed_encoding)
known_encoding = use_strict_parser = 1
break
except:
pass
# if no luck and we have auto-detection library, try that
if (not known_encoding) and chardet:
try:
proposed_encoding = chardet.detect(data)['encoding']
if proposed_encoding and (proposed_encoding not in tried_encodings):
tried_encodings.append(proposed_encoding)
data = _toUTF8(data, proposed_encoding)
known_encoding = use_strict_parser = 1
except:
pass
# if still no luck and we haven't tried utf-8 yet, try that
if (not known_encoding) and ('utf-8' not in tried_encodings):
try:
proposed_encoding = 'utf-8'
tried_encodings.append(proposed_encoding)
data = _toUTF8(data, proposed_encoding)
known_encoding = use_strict_parser = 1
except:
pass
# if still no luck and we haven't tried windows-1252 yet, try that
if (not known_encoding) and ('windows-1252' not in tried_encodings):
try:
proposed_encoding = 'windows-1252'
tried_encodings.append(proposed_encoding)
data = _toUTF8(data, proposed_encoding)
known_encoding = use_strict_parser = 1
except:
pass
# if still no luck, give up
if not known_encoding:
result['bozo'] = 1
result['bozo_exception'] = CharacterEncodingUnknown( \
'document encoding unknown, I tried ' + \
'%s, %s, utf-8, and windows-1252 but nothing worked' % \
(result['encoding'], xml_encoding))
result['encoding'] = ''
elif proposed_encoding != result['encoding']:
result['bozo'] = 1
result['bozo_exception'] = CharacterEncodingOverride( \
'documented declared as %s,