Switched to utf-8.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index fc9514d005e5c02dce4fe4717c7e1732060d7cd6..dbca770543d7b6cc7b4e4c367413e53d505b1b5d 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -2,13 +2,15 @@
 """
    HTML Parsers wrapper
 
-   Written by BroytMann. Copyright (C) 1997-2007 PhiloSoft Design
+   Written by BroytMann. Copyright (C) 1997-2008 PhiloSoft Design
 """
 
 import codecs
 
 from m_lib.defenc import default_encoding
 current_charset = default_encoding.replace("windows-", "cp")
+
+universal_charset = "utf-8"
 DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
 
 parsers = []
@@ -24,16 +26,33 @@ parsers.append(parse_html)
 
 
 import re
-entity_re = re.compile("(&#[0-9]+;)")
+from htmlentitydefs import entitydefs
+
+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")
+   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)
+   output2 = []
+   for part in entity_re.split(''.join(output)):
+      if entity_re.match(part):
+         part = entitydefs.get(part[1:-1], part)
+         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)
+
+   return ''.join(output2)
 
 
 def parse_html(filename, charset=None, log=None):
@@ -48,9 +67,9 @@ def parse_html(filename, charset=None, log=None):
       if parser:
          break
       else:
-         if log: log("Parser %s failed, trying next one." % p)
+         if log: log("Parser %s.%s failed, trying next one." % (p.__module__, p.__name__))
 
-   title = parser.title
+   converted_title = title = parser.title
    if not parser.charset:
       try:
          unicode(title, "ascii")
@@ -60,37 +79,44 @@ def parse_html(filename, charset=None, log=None):
    if parser.charset:
       parser.charset = parser.charset.replace("windows-", "cp").lower()
 
-   if parser.charset and (parser.charset <> current_charset):
+   if parser.charset and (parser.charset <> universal_charset):
       try:
          if parser.meta_charset:
             if log: log("   META charset   : %s" % parser.charset)
          else:
             if log: log("   HTTP charset   : %s" % parser.charset)
-         if log: log("   current charset: %s" % current_charset)
+         if log: log("   current charset: %s" % universal_charset)
          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(universal_charset)
          except UnicodeError:
             if parser.meta_charset and parser.charset.endswith("1252") and \
-                  not DEFAULT_CHARSET.endswith("1252") and (DEFAULT_CHARSET <> current_charset):
+                  not DEFAULT_CHARSET.endswith("1252") and (DEFAULT_CHARSET <> universal_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(universal_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(universal_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)
+         if log: log("   unknown charset: `%s' or `%s'" % (parser.charset, universal_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__':
    import sys
-   parser = parse_html(sys.argv[1], current_charset)
+   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