1
0
mirror of https://codeberg.org/polarisfm/youtube-dl synced 2024-11-09 11:44:33 +01:00
youtube-dl/youtube_dl/postprocessor/metadatafromtitle.py
Jaime Marquínez Ferrándiz 88cf6fb368 [metadatafromtitle] Some improvements and cleanup
* Remove the 'songtitle' field, 'title' can be used instead.
* Remove newlines in the help text, for consistency with other options.
* Add 'from __future__ import unicode_literals'.
* Call '__init__' from the parent class.
* Add test for the format_to_regex method
2015-03-14 20:06:33 +01:00

48 lines
1.5 KiB
Python

from __future__ import unicode_literals
import re
from .common import PostProcessor
from ..utils import PostProcessingError
class MetadataFromTitlePPError(PostProcessingError):
pass
class MetadataFromTitlePP(PostProcessor):
def __init__(self, downloader, titleformat):
super(MetadataFromTitlePP, self).__init__(downloader)
self._titleformat = titleformat
self._titleregex = self.format_to_regex(titleformat)
def format_to_regex(self, fmt):
"""
Converts a string like
'%(title)s - %(artist)s'
to a regex like
'(?P<title>.+)\ \-\ (?P<artist>.+)'
"""
lastpos = 0
regex = ""
# replace %(..)s with regex group and escape other string parts
for match in re.finditer(r'%\((\w+)\)s', fmt):
regex += re.escape(fmt[lastpos:match.start()])
regex += r'(?P<' + match.group(1) + '>.+)'
lastpos = match.end()
if lastpos < len(fmt):
regex += re.escape(fmt[lastpos:len(fmt)])
return regex
def run(self, info):
title = info['title']
match = re.match(self._titleregex, title)
if match is None:
raise MetadataFromTitlePPError('Could not interpret title of video as "%s"' % self._titleformat)
for attribute, value in match.groupdict().items():
value = match.group(attribute)
info[attribute] = value
self._downloader.to_screen('[fromtitle] parsed ' + attribute + ': ' + value)
return True, info