X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=Robots%2Fparse_html.py;h=680ad5573a741bb99402449a77d969202e49abde;hb=e7035a2a59f6a2cb55aa7ca852e2cff3b66b59d9;hp=3250a0de5980c43a00c6cce8ff5e91a9e5b5471c;hpb=daa40d9fc61a15fe51955d97b5711af5973a64d3;p=bookmarks_db.git

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 3250a0d..680ad55 100755
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -2,29 +2,49 @@
 """
    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 = []
 try:
-   from parse_html_beautifulsoup import parse_html as _parse_html
+   import parse_html_beautifulsoup
+   parse_html_beautifulsoup.DEFAULT_CHARSET = DEFAULT_CHARSET
 except ImportError:
-   from parse_html_htmlparser import parse_html as _parse_html
+   pass
+else:
+   parsers.append(parse_html_beautifulsoup.parse_html)
+
+from parse_html_htmlparser import parse_html
+parsers.append(parse_html)
 
 
 import re
-entity_re = re.compile("(&#[0-9]+;)")
+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 = unichr(int(part[2:-1])).encode(charset, "replace")
+      if part not in ("&", "<", ">", "&quote;", " ") and \
+            entity_re.match(part):
+         part = unichr(name2codepoint.get(part[1:-1], part)).encode(charset)
+      output.append(part)
+   title = ''.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
       output.append(part)
 
    return ''.join(output)
@@ -37,9 +57,23 @@ def parse_html(filename, charset=None, log=None):
       except (ValueError, LookupError):
          charset = None         # ...try charset from HTML
 
-   parser = _parse_html(filename, charset)
-   title = parser.title
-
+   for p in parsers:
+      charsets = [universal_charset, DEFAULT_CHARSET]
+      if charset not in charsets:
+         charsets.insert(0, charset)
+      parser = None
+      for c in charsets:
+         try:
+            parser = p(filename, c)
+            break
+         except UnicodeEncodeError:
+            pass
+      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")
@@ -49,37 +83,40 @@ 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("   guessed charset: %s" % parser.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):
-               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")
-            else:
-               title = unicode(title, parser.charset, "replace").encode(current_charset, "replace")
-         if log: log("   converted title: %s" % title)
+            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, 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)
-   print parser.charset
-   print parser.title
-   print parser.refresh
-   print parser.icon
+   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'))
+   print "   refresh:", parser.refresh
+   print "   icon   :", parser.icon