#! /usr/bin/env python
"""
- HTML Parser
+ HTML Parsers wrapper
- Written by BroytMann. Copyright (C) 1997-2007 PhiloSoft Design
+ Written by BroytMann. Copyright (C) 1997-2008 PhiloSoft Design
"""
-from m_lib.defenc import default_encoding
-current_charset = default_encoding.replace("windows-", "cp")
-DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
-
-from HTMLParser import HTMLParseError
-from m_lib.net.www.html import HTMLParser as _HTMLParser
-
-
-class HTMLHeadDone(Exception): pass
-
-
-class HTMLParser(_HTMLParser):
- def __init__(self, charset=None):
- _HTMLParser.__init__(self)
- self.charset = charset
- self.meta_charset = 0
- self.title = ''
- self.refresh = ''
- self.icon = None
-
- def end_head(self):
- raise HTMLHeadDone()
-
-
- def do_meta(self, attrs):
- http_equiv = ""
- content = ""
-
- for attrname, value in attrs:
- if value:
- value = value.strip()
- if attrname == 'http-equiv':
- http_equiv = value.lower()
- elif attrname == 'content':
- content = value
-
- if (not self.charset) and (http_equiv == "content-type"):
- try:
- # extract charset from "text/html; foo; charset=UTF-8; bar;"
- self.charset = content.lower().split('charset=')[1].split(';')[0]
- self.meta_charset = 1 # Remember that the charset was retrieved from
- # META tag, not from the Content-Type header
- except IndexError:
- pass
-
- if http_equiv == "refresh":
- self.refresh = content
-
-
- def start_title(self, attrs):
- self.accumulator = ''
-
- def end_title(self):
- if not self.title: # use only the first title
- self.title = self.accumulator
+import codecs
+universal_charset = "utf-8"
+DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
- def do_link(self, attrs):
- has_icon = False
- href = None
-
- for attrname, value in attrs:
- if value:
- value = value.strip().lower()
- if (attrname == 'rel') and (value in ('icon', 'shortcut icon')):
- has_icon = True
- elif attrname == 'href':
- href = value
+parsers = []
+try:
+ import parse_html_beautifulsoup
+ parse_html_beautifulsoup.DEFAULT_CHARSET = DEFAULT_CHARSET
+except ImportError:
+ pass
+else:
+ parsers.append(parse_html_beautifulsoup.parse_html)
- if has_icon:
- self.icon = href
- else:
- self.icon = None
+from parse_html_htmlparser import parse_html
+parsers.append(parse_html)
import re
-entity_re = re.compile("(&#[0-9]+;)")
+from htmlentitydefs import name2codepoint
+
+entity_re = re.compile("(&\w+;)")
+num_entity_re = re.compile("(&#[0-9]+;)")
def recode_entities(title, charset):
output = []
for part in entity_re.split(title):
- if entity_re.match(part):
- part = unichr(int(part[2:-1])).encode(charset, "replace")
+ if part not in ("&", "<", ">", ""e;") and \
+ entity_re.match(part):
+ part = unichr(name2codepoint.get(part[1:-1], part)).encode(charset)
+ output.append(part)
+ title = ''.join(output)
+
+ output = []
+ for part in num_entity_re.split(title):
+ if num_entity_re.match(part):
+ try:
+ part = unichr(int(part[2:-1])).encode(charset)
+ except UnicodeEncodeError:
+ pass # Leave the entity as is
output.append(part)
return ''.join(output)
except (ValueError, LookupError):
charset = None # ...try charset from HTML
- infile = open(filename, 'r')
- parser = HTMLParser(charset)
+ charsets = [universal_charset, DEFAULT_CHARSET]
+ if charset:
+ charset = charset.lower().replace("windows-", "cp")
+ if charset not in charsets:
+ charsets.insert(0, charset)
- for line in infile:
- try:
- parser.feed(line)
- except (HTMLParseError, HTMLHeadDone):
+ for p in parsers:
+ parser = None
+ for c in charsets:
+ try:
+ parser = p(filename, c, log)
+ break
+ except UnicodeEncodeError:
+ pass
+ if parser:
break
+ else:
+ if log: log("Parser %s.%s failed, trying next one." % (p.__module__, p.__name__))
- infile.close()
-
- try:
- parser.close()
- except (HTMLParseError, HTMLHeadDone):
- pass
-
- title = parser.title
-
+ converted_title = title = parser.title
if not parser.charset:
try:
unicode(title, "ascii")
parser.charset = DEFAULT_CHARSET
if parser.charset:
- parser.charset = parser.charset.replace("windows-", "cp").lower()
+ parser.charset = parser.charset.lower().replace("windows-", "cp")
- if parser.charset and (parser.charset <> current_charset):
+ if parser.charset and (
+ (parser.charset <> universal_charset) or
+ ((not charset) or (charset <> parser.charset))):
try:
if parser.meta_charset:
if log: log(" META charset : %s" % parser.charset)
- else:
- if log: log(" charset : %s" % parser.charset)
+ elif (not charset) or (charset <> parser.charset):
+ if log: log(" guessed charset: %s" % parser.charset)
+ if log: log(" current charset: %s" % universal_charset)
if log: log(" title : %s" % title)
- title = unicode(title, parser.charset, "replace").encode(current_charset, "replace")
- if log: log(" current charset: %s" % current_charset)
- if log: log(" converted title: %s" % title)
+ if parser.charset <> universal_charset:
+ try:
+ converted_title = unicode(title, parser.charset).encode(universal_charset)
+ except UnicodeError:
+ if log: log(" incorrect conversion from %s, converting from %s" % (parser.charset, DEFAULT_CHARSET))
+ converted_title = unicode(title, DEFAULT_CHARSET, "replace").encode(universal_charset, "replace")
+ parser.charset = DEFAULT_CHARSET
+ if log and (converted_title <> title): log(" converted title: %s" % converted_title)
except LookupError:
- if log: log(" unknown charset: `%s' or `%s'" % (parser.charset, current_charset))
-
- parser.title = recode_entities(title, current_charset)
+ if log: log(" unknown charset: '%s'" % parser.charset)
+ else:
+ if log: log(" title : %s" % title)
+
+ final_title = recode_entities(converted_title, universal_charset)
+ parts = [s.strip() for s in final_title.replace('\r', '').split('\n')]
+ final_title = ' '.join([s for s in parts if s])
+ if log and (final_title <> converted_title): log(" final title : %s" % final_title)
+ parser.title = final_title
return parser
if __name__ == '__main__':
- parser = parse_html(sys.argv[1])
- print parser.charset
- print parser.title
- print parser.refresh
- print parser.icon
+ import sys
+ from m_lib.defenc import default_encoding
+ current_charset = default_encoding.replace("windows-", "cp")
+
+ parser = parse_html(sys.argv[1], universal_charset,
+ log=lambda s: sys.stdout.write(s + '\n'))
+ print " refresh:", parser.refresh
+ print " icon :", parser.icon