Insert lxml-based parser at the beginning.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 80d4a9243bbbec31b96d98a4881182afefd05361..5e8061c793e307b0fad3d1163bf3b122884d185c 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -2,7 +2,7 @@
 """
    HTML Parsers wrapper
 
-   Written by BroytMann. Copyright (C) 1997-2008 PhiloSoft Design
+   Written by Broytman. Copyright (C) 1997-2010 PhiloSoft Design
 """
 
 import codecs
@@ -11,6 +11,14 @@ universal_charset = "utf-8"
 DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
 
 parsers = []
+
+try:
+   from parse_html_lxml import parse_html
+except ImportError:
+   pass
+else:
+    parsers.append(parse_html)
+
 try:
    import parse_html_beautifulsoup
    parse_html_beautifulsoup.DEFAULT_CHARSET = DEFAULT_CHARSET
@@ -19,8 +27,19 @@ except ImportError:
 else:
    parsers.append(parse_html_beautifulsoup.parse_html)
 
-from parse_html_htmlparser import parse_html
-parsers.append(parse_html)
+try:
+   from parse_html_htmlparser import parse_html
+except ImportError:
+   pass
+else:
+    parsers.append(parse_html)
+
+try:
+   import parse_html_html5
+except ImportError:
+   pass
+else:
+   parsers.append(parse_html_html5.parse_html)
 
 
 import re
@@ -32,9 +51,11 @@ 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 ("&", "<", ">", """) and \
             entity_re.match(part):
-         part = unichr(name2codepoint.get(part[1:-1], part)).encode(charset)
+         _part = name2codepoint.get(part[1:-1], None)
+         if _part is not None:
+             part = unichr(_part).encode(charset)
       output.append(part)
    title = ''.join(output)
 
@@ -51,6 +72,9 @@ def recode_entities(title, charset):
 
 
 def parse_html(filename, charset=None, log=None):
+   if not parsers:
+       return None
+
    if charset:
       try:
          codecs.lookup(charset) # In case of unknown charset...
@@ -60,8 +84,9 @@ def parse_html(filename, charset=None, log=None):
    charsets = [universal_charset, DEFAULT_CHARSET]
    if charset:
       charset = charset.lower().replace("windows-", "cp")
-      if charset not in charsets:
-         charsets.insert(0, charset)
+      if charset in charsets:
+         charsets.remove(charset)
+      charsets.insert(0, charset)
 
    for p in parsers:
       parser = None
@@ -76,8 +101,11 @@ def parse_html(filename, charset=None, log=None):
       else:
          if log: log("Parser %s.%s failed, trying next one." % (p.__module__, p.__name__))
 
+   if not parser:
+       return None
+
    converted_title = title = parser.title
-   if not parser.charset:
+   if title and (not parser.charset):
       try:
          unicode(title, "ascii")
       except UnicodeDecodeError:
@@ -86,7 +114,7 @@ def parse_html(filename, charset=None, log=None):
    if parser.charset:
       parser.charset = parser.charset.lower().replace("windows-", "cp")
 
-   if parser.charset and (
+   if title and parser.charset and (
          (parser.charset <> universal_charset) or
          ((not charset) or (charset <> parser.charset))):
       try:
@@ -109,20 +137,28 @@ def parse_html(filename, charset=None, log=None):
    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
+   if 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
-   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'))
+   l = len(sys.argv)
+   if l == 3:
+      filename = sys.argv[1]
+      charset = sys.argv[2]
+   elif l == 2:
+      filename = sys.argv[1]
+      charset = universal_charset
+   else:
+      sys.exit("Usage: %s filename [charset]" % sys.argv[0])
+
+   parser = parse_html(filename, charset, log=lambda s: sys.stdout.write(s + '\n'))
    print "   refresh:", parser.refresh
    print "   icon   :", parser.icon