"""
HTML Parsers wrapper
- Written by BroytMann. Copyright (C) 1997-2008 PhiloSoft Design
+ Written by Broytman. Copyright (C) 1997-2010 PhiloSoft Design
"""
import codecs
DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
parsers = []
+
+try:
+ import parse_html_etreetidy
+except ImportError:
+ pass
+else:
+ parsers.append(parse_html_etreetidy.parse_html)
+
try:
import parse_html_beautifulsoup
parse_html_beautifulsoup.DEFAULT_CHARSET = DEFAULT_CHARSET
import re
-from htmlentitydefs import entitydefs
+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 = entitydefs.get(part[1:-1], part)
+ if part not in ("&", "<", ">", """) and \
+ entity_re.match(part):
+ _part = name2codepoint.get(part[1:-1], None)
+ if _part is not None:
+ part = unichr(_part).encode(charset)
output.append(part)
+ title = ''.join(output)
- output2 = []
- for part in num_entity_re.split(''.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
- output2.append(part)
+ output.append(part)
- return ''.join(output2)
+ return ''.join(output)
def parse_html(filename, charset=None, log=None):
except (ValueError, LookupError):
charset = None # ...try charset from HTML
- for p in parsers:
- charsets = [universal_charset, DEFAULT_CHARSET]
+ charsets = [universal_charset, DEFAULT_CHARSET]
+ if charset:
+ charset = charset.lower().replace("windows-", "cp")
if charset not in charsets:
charsets.insert(0, charset)
+
+ for p in parsers:
parser = None
for c in charsets:
try:
- parser = p(filename, c)
+ parser = p(filename, c, log)
+ break
except UnicodeEncodeError:
pass
if parser:
if log: log("Parser %s.%s failed, trying next one." % (p.__module__, p.__name__))
converted_title = title = parser.title
- if not parser.charset:
+ if title and (not parser.charset):
try:
unicode(title, "ascii")
except UnicodeDecodeError:
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 <> universal_charset):
+ if title and 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(" HTTP 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)
- 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 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, universal_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
+ if 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__':
import sys
- from m_lib.defenc import default_encoding
- current_charset = default_encoding.replace("windows-", "cp")
-
- parser = parse_html(sys.argv[1], universal_charset)
- print parser.charset
- print parser.title
- print parser.title.decode(universal_charset).encode(current_charset, "replace")
- print parser.refresh
- print parser.icon
+
+ l = len(sys.argv)
+ if l == 3:
+ filename = sys.argv[1]
+ charset = sys.argv[2]
+ elif l == 2:
+ filename = sys.argv[1]
+ charset = universal_charset
+ else:
+ sys.exit("Usage: %s filename [charset]" % sys.argv[0])
+
+ parser = parse_html(filename, charset, log=lambda s: sys.stdout.write(s + '\n'))
+ print " refresh:", parser.refresh
+ print " icon :", parser.icon