X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=Robots%2Fparse_html.py;h=80d4a9243bbbec31b96d98a4881182afefd05361;hb=5f637b5a5fe27098985975928632b9fea5ea3c62;hp=dbca770543d7b6cc7b4e4c367413e53d505b1b5d;hpb=3054de4d1a2dc190c06943d021fdb836464e6b19;p=bookmarks_db.git

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index dbca770..80d4a92 100755
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -7,31 +7,37 @@
 
 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
+   import parse_html_beautifulsoup
+   parse_html_beautifulsoup.DEFAULT_CHARSET = DEFAULT_CHARSET
 except ImportError:
    pass
 else:
-   parsers.append(parse_html)
+   parsers.append(parse_html_beautifulsoup.parse_html)
 
 from parse_html_htmlparser import parse_html
 parsers.append(parse_html)
 
 
 import re
-from htmlentitydefs import entitydefs
+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 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):
@@ -41,18 +47,7 @@ def recode_entities(title, charset):
             pass # Leave the entity as is
       output.append(part)
 
-   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)
+   return ''.join(output)
 
 
 def parse_html(filename, charset=None, log=None):
@@ -62,8 +57,20 @@ def parse_html(filename, charset=None, log=None):
       except (ValueError, LookupError):
          charset = None         # ...try charset from HTML
 
+   charsets = [universal_charset, DEFAULT_CHARSET]
+   if charset:
+      charset = charset.lower().replace("windows-", "cp")
+      if charset not in charsets:
+         charsets.insert(0, charset)
+
    for p in parsers:
-      parser = p(filename, charset)
+      parser = None
+      for c in charsets:
+         try:
+            parser = p(filename, c, log)
+            break
+         except UnicodeEncodeError:
+            pass
       if parser:
          break
       else:
@@ -77,30 +84,28 @@ def parse_html(filename, charset=None, log=None):
          parser.charset = DEFAULT_CHARSET
 
    if parser.charset:
-      parser.charset = parser.charset.replace("windows-", "cp").lower()
+      parser.charset = parser.charset.lower().replace("windows-", "cp")
 
-   if parser.charset and (parser.charset <> universal_charset):
+   if parser.charset and (
+         (parser.charset <> universal_charset) or
+         ((not charset) or (charset <> parser.charset))):
       try:
          if parser.meta_charset:
             if log: log("   META charset   : %s" % parser.charset)
-         else:
-            if log: log("   HTTP charset   : %s" % parser.charset)
+         elif (not charset) or (charset <> parser.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:
-            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 <> universal_charset):
+         if parser.charset <> universal_charset:
+            try:
+               converted_title = unicode(title, parser.charset).encode(universal_charset)
+            except UnicodeError:
+               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: log("   incorrect conversion from cp1252, converting from %s" % DEFAULT_CHARSET)
-               converted_title = unicode(save_title, DEFAULT_CHARSET, "replace").encode(universal_charset, "replace")
-            else:
-               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, universal_charset))
+         if log: log("   unknown charset: '%s'" % parser.charset)
    else:
       if log: log("   title          : %s" % title)
 
@@ -114,9 +119,10 @@ def parse_html(filename, charset=None, log=None):
 
 if __name__ == '__main__':
    import sys
-   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
+   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