X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=Robots%2Fparse_html.py;h=a83585df318afb1602a916a1a17768485a099138;hb=8ce74f839238093d7278aa041ff55dbcb7abd3a0;hp=c7acb6426e7dfbf386702ffb18fe5f12b3bed1ce;hpb=240c6b89a3825d71e040702dc9f9f948a71da107;p=bookmarks_db.git diff --git a/Robots/parse_html.py b/Robots/parse_html.py index c7acb64..a83585d 100755 --- 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 +11,7 @@ universal_charset = "utf-8" DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic parsers = [] + try: import parse_html_beautifulsoup parse_html_beautifulsoup.DEFAULT_CHARSET = DEFAULT_CHARSET @@ -19,12 +20,30 @@ 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_lxml import parse_html +except ImportError: + pass +else: + 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 -from htmlentitydefs import entitydefs +from htmlentitydefs import name2codepoint entity_re = re.compile("(&\w+;)") num_entity_re = re.compile("(&#[0-9]+;)") @@ -32,82 +51,122 @@ 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 = entitydefs.get(part[1:-1], part) + if part not in ("&", "<", ">", """) and \ + entity_re.match(part): + _part = name2codepoint.get(part[1:-1], None) + if _part is not None: + part = unichr(_part).encode(charset) output.append(part) + title = ''.join(output) - output2 = [] - for part in num_entity_re.split(''.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 - output2.append(part) + output.append(part) - return ''.join(output2) + return ''.join(output) def parse_html(filename, charset=None, log=None): + if not parsers: + return None + if charset: try: codecs.lookup(charset) # In case of unknown charset... except (ValueError, LookupError): charset = None # ...try charset from HTML + charsets = [universal_charset, DEFAULT_CHARSET] + if charset: + charset = charset.lower().replace("windows-", "cp") + if charset in charsets: + charsets.remove(charset) + 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: 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: 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 title and 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) - 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) - 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 + + icon = parser.icon + if isinstance(icon, unicode): + try: + parser.icon = icon.encode('ascii') + except UnicodeEncodeError: + if parser.charset: + parser.icon = icon.encode(parser.charset) 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) - print parser.charset - print parser.title - print parser.title.decode(universal_charset).encode(current_charset, "replace") - print parser.refresh - print parser.icon + + 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