HTMLParser.HTMLParser.__init__(self)
def error(self, message):
- print self.getpos()
+ print >> sys.stderr, self.getpos()
if self.error_count > 10 or self.started:
raise HTMLParser.HTMLParseError(message, self.getpos())
self.rawdata = '\n'.join(self.html.split('\n')[self.getpos()[0]:]) # skip one line
return html
-def sanitize_title(utitle):
- """Sanitizes a video title so it could be used as part of a filename."""
- utitle = unescapeHTML(utitle)
- return utitle.replace(unicode(os.sep), u'%')
-
-
def sanitize_open(filename, open_mode):
"""Try to open the given filename, and slightly tweak it if this fails.
if timetuple is not None:
timestamp = email.utils.mktime_tz(timetuple)
return timestamp
-
-def simplify_title(title):
- expr = re.compile(ur'[^\w\d_\-]+', flags=re.UNICODE)
- return expr.sub(u'_', title).strip(u'_')
+
+def sanitize_filename(s):
+ """Sanitizes a string so it could be used as part of a filename."""
+ def replace_insane(char):
+ if char in u' .\\/|?*<>:"' or ord(char) < 32:
+ return '_'
+ return char
+ return u''.join(map(replace_insane, s)).strip('_')
def orderedSet(iterable):
""" Remove all duplicates from the input iterable """
assert type(s) == type(u'')
- if sys.platform == 'win32' and sys.getwindowsversion().major >= 5:
+ if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
# Pass u'' directly to use Unicode APIs on Windows 2000 and up
# (Detecting Windows NT 4 is tricky because 'major >= 4' would
# match Windows 9x series as well. Besides, NT 4 is obsolete.)
self.expected = expected
+class Trouble(Exception):
+ """Trouble helper exception
+
+ This is an exception to be handled with
+ FileDownloader.trouble
+ """
+
class YoutubeDLHandler(urllib2.HTTPHandler):
"""Handler for HTTP requests and responses.