Try a list of charsets, including the universal (utf-8) and the default ones.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index bc507679b034f20f06124171dba3b1e8ec5e2945..6fe1df954236f855e2f08ae1ca048b4f275a704a 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -12,11 +12,12 @@ 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)
@@ -55,7 +56,15 @@ def parse_html(filename, charset=None, log=None):
          charset = None         # ...try charset from HTML
 
    for p in parsers:
-      parser = p(filename, charset)
+      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)
+         except UnicodeEncodeError:
+            pass
       if parser:
          break
       else:
@@ -79,17 +88,12 @@ def parse_html(filename, charset=None, log=None):
             if log: log("   HTTP 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):
-               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: 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))