current_charset = default_encoding.replace("windows-", "cp")
DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
+parsers = []
try:
- from parse_html_beautifulsoup import parse_html as _parse_html
+ from parse_html_beautifulsoup import parse_html
except ImportError:
- from parse_html_htmlparser import parse_html as _parse_html
+ pass
+else:
+ parsers.append(parse_html)
+
+from parse_html_htmlparser import parse_html
+parsers.append(parse_html)
import re
output = []
for part in entity_re.split(title):
if entity_re.match(part):
- part = unichr(int(part[2:-1])).encode(charset, "replace")
+ 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
- parser = _parse_html(filename, charset)
- title = parser.title
+ for p in parsers:
+ parser = p(filename, charset)
+ if parser:
+ break
+ else:
+ if log: log("Parser %s.%s failed, trying next one." % (p.__module__, p.__name__))
+ converted_title = title = parser.title
if not parser.charset:
try:
unicode(title, "ascii")
if log: log(" title : %s" % title)
save_title = title
try:
- title = unicode(title, parser.charset).encode(current_charset)
+ converted_title = unicode(title, parser.charset).encode(current_charset)
except UnicodeError:
if parser.meta_charset and parser.charset.endswith("1252") and \
not DEFAULT_CHARSET.endswith("1252") and (DEFAULT_CHARSET <> current_charset):
parser.charset = DEFAULT_CHARSET
if log: log(" incorrect conversion from cp1252, converting from %s" % DEFAULT_CHARSET)
- title = unicode(save_title, DEFAULT_CHARSET, "replace").encode(current_charset, "replace")
+ converted_title = unicode(save_title, DEFAULT_CHARSET, "replace").encode(current_charset, "replace")
else:
- title = unicode(title, parser.charset, "replace").encode(current_charset, "replace")
- if log: log(" converted title: %s" % title)
+ converted_title = unicode(title, parser.charset, "replace").encode(current_charset, "replace")
+ 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)
+ else:
+ if log: log(" title : %s" % title)
+
+ final_title = recode_entities(converted_title, current_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