ElementTidy often segfaults.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 80c7aa89d28efed06db8a598088303f44d2b338b..cbb45d612f9b33699b6f2d0bbe397b11a3562fe1 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -2,7 +2,7 @@
 """
    HTML Parsers wrapper
 
-   Written by Broytman. Copyright (C) 1997-2008 PhiloSoft Design
+   Written by Broytman. Copyright (C) 1997-2011 PhiloSoft Design
 """
 
 import codecs
@@ -11,6 +11,7 @@ universal_charset = "utf-8"
 DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
 
 parsers = []
+
 try:
    import parse_html_beautifulsoup
    parse_html_beautifulsoup.DEFAULT_CHARSET = DEFAULT_CHARSET
@@ -19,9 +20,34 @@ except ImportError:
 else:
    parsers.append(parse_html_beautifulsoup.parse_html)
 
-from parse_html_htmlparser import parse_html
-parsers.append(parse_html)
+try:
+   from parse_html_lxml import parse_html
+except ImportError:
+   pass
+else:
+    parsers.append(parse_html)
+
+try:
+   from parse_html_htmlparser import parse_html
+except ImportError:
+   pass
+else:
+    parsers.append(parse_html)
+
+try:
+   import parse_html_html5
+except ImportError:
+   pass
+else:
+   parsers.append(parse_html_html5.parse_html)
 
+# ElementTidy often segfaults
+#try:
+#   import parse_html_etreetidy
+#except ImportError:
+#   pass
+#else:
+#   parsers.append(parse_html_etreetidy.parse_html)
 
 import re
 from htmlentitydefs import name2codepoint
@@ -34,7 +60,9 @@ def recode_entities(title, charset):
    for part in entity_re.split(title):
       if part not in ("&", "<", ">", """) and \
             entity_re.match(part):
-         part = unichr(name2codepoint.get(part[1:-1], part)).encode(charset)
+         _part = name2codepoint.get(part[1:-1], None)
+         if _part is not None:
+             part = unichr(_part).encode(charset)
       output.append(part)
    title = ''.join(output)
 
@@ -51,6 +79,9 @@ def recode_entities(title, charset):
 
 
 def parse_html(filename, charset=None, log=None):
+   if not parsers:
+       return None
+
    if charset:
       try:
          codecs.lookup(charset) # In case of unknown charset...
@@ -60,8 +91,9 @@ def parse_html(filename, charset=None, log=None):
    charsets = [universal_charset, DEFAULT_CHARSET]
    if charset:
       charset = charset.lower().replace("windows-", "cp")
-      if charset not in charsets:
-         charsets.insert(0, charset)
+      if charset in charsets:
+         charsets.remove(charset)
+      charsets.insert(0, charset)
 
    for p in parsers:
       parser = None
@@ -76,6 +108,9 @@ def parse_html(filename, charset=None, log=None):
       else:
          if log: log("Parser %s.%s failed, trying next one." % (p.__module__, p.__name__))
 
+   if not parser:
+       return None
+
    converted_title = title = parser.title
    if title and (not parser.charset):
       try:
@@ -115,6 +150,14 @@ def parse_html(filename, charset=None, log=None):
       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
+
+   icon = parser.icon
+   if isinstance(icon, unicode):
+       try:
+           parser.icon = icon.encode('ascii')
+       except UnicodeEncodeError:
+           if parser.charset:
+               parser.icon = icon.encode(parser.charset)
    return parser