Fixed parsing in case of unknown entity.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 3e49491c7fc36ff5f1e8749fc86195f2a49983e8..fd8dc927c0d78d7adcd8394457774450ff9dd403 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -2,7 +2,7 @@
 """
    HTML Parsers wrapper
 
-   Written by BroytMann. Copyright (C) 1997-2008 PhiloSoft Design
+   Written by Broytman. Copyright (C) 1997-2008 PhiloSoft Design
 """
 
 import codecs
@@ -32,10 +32,14 @@ num_entity_re = re.compile("(&#[0-9]+;)")
 def recode_entities(title, charset):
    output = []
    for part in entity_re.split(title):
-      if part not in ("&", "<", ">", "&quote;") and \
+      if part not in ("&", "<", ">", """) and \
             entity_re.match(part):
-         part = unichr(name2codepoint.get(part[1:-1], part)).encode(charset)
-      output.append(part)
+         _part = name2codepoint.get(part[1:-1], None)
+         if _part is None:
+             _part = part
+         else:
+             _part = unichr(_part).encode(charset)
+      output.append(_part)
    title = ''.join(output)
 
    output = []
@@ -67,7 +71,7 @@ def parse_html(filename, charset=None, log=None):
       parser = None
       for c in charsets:
          try:
-            parser = p(filename, c)
+            parser = p(filename, c, log)
             break
          except UnicodeEncodeError:
             pass
@@ -77,7 +81,7 @@ def parse_html(filename, charset=None, log=None):
          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:
@@ -86,7 +90,7 @@ def parse_html(filename, charset=None, log=None):
    if parser.charset:
       parser.charset = parser.charset.lower().replace("windows-", "cp")
 
-   if parser.charset and (
+   if title and parser.charset and (
          (parser.charset <> universal_charset) or
          ((not charset) or (charset <> parser.charset))):
       try:
@@ -109,20 +113,28 @@ def parse_html(filename, charset=None, log=None):
    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,
-      log=lambda s: sys.stdout.write(s + '\n'))
+   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