Always log guessed charset even if it's utf-8.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 680ad5573a741bb99402449a77d969202e49abde..3e49491c7fc36ff5f1e8749fc86195f2a49983e8 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -32,7 +32,7 @@ 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 ("&", "<", ">", "&quote;") and \
             entity_re.match(part):
          part = unichr(name2codepoint.get(part[1:-1], part)).encode(charset)
       output.append(part)
@@ -57,10 +57,13 @@ def parse_html(filename, charset=None, log=None):
       except (ValueError, LookupError):
          charset = None         # ...try charset from HTML
 
-   for p in parsers:
-      charsets = [universal_charset, DEFAULT_CHARSET]
+   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 = None
       for c in charsets:
          try:
@@ -81,25 +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:
+         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)
-         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 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 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)