Add metadata from title parser

(Closes #5125)

youtube_dl/__init__.py

@@ -213,6 +213,11 @@ def _real_main(argv=None):
     # PostProcessors
     postprocessors = []
     # Add the metadata pp first, the other pps will copy it
+    if opts.metafromtitle:
+        postprocessors.append({
+            'key': 'MetadataFromTitle',
+            'titleformat': opts.metafromtitle
+        })
     if opts.addmetadata:
         postprocessors.append({'key': 'FFmpegMetadata'})
     if opts.extractaudio:

youtube_dl/options.py

@@ -735,6 +735,15 @@ def parseOpts(overrideArguments=None):
         '--add-metadata',
         action='store_true', dest='addmetadata', default=False,
         help='write metadata to the video file')
+    postproc.add_option(
+        '--metadata-from-title',
+        metavar='FORMAT', dest='metafromtitle',
+        help='parse additional metadata like song title / artist from the video title. \n'
+             'The format syntax is the same as --output, '
+             'the parsed parameters replace existing values.\n'
+             'Additional templates: %(songtitle), %(album), %(artist). \n'
+             'Example: --metadata-from-title "%(artist)s - %(title)s" matches a title like '
+             '"Coldplay - Paradise"')
     postproc.add_option(
         '--xattrs',
         action='store_true', dest='xattrs', default=False,

youtube_dl/postprocessor/__init__.py

@@ -15,6 +15,7 @@ from .ffmpeg import (
 )
 from .xattrpp import XAttrMetadataPP
 from .execafterdownload import ExecAfterDownloadPP
+from .metadatafromtitle import MetadataFromTitlePP
 
 
 def get_postprocessor(key):
@@ -34,5 +35,6 @@ __all__ = [
     'FFmpegPostProcessor',
     'FFmpegSubtitlesConvertorPP',
     'FFmpegVideoConvertorPP',
+    'MetadataFromTitlePP',
     'XAttrMetadataPP',
 ]

youtube_dl/postprocessor/ffmpeg.py

@@ -541,11 +541,15 @@ class FFmpegEmbedSubtitlePP(FFmpegPostProcessor):
 class FFmpegMetadataPP(FFmpegPostProcessor):
     def run(self, info):
         metadata = {}
-        if info.get('title') is not None:
+        if info.get('songtitle') is not None:
+            metadata['title'] = info['songtitle']
+        elif info.get('title') is not None:
             metadata['title'] = info['title']
         if info.get('upload_date') is not None:
             metadata['date'] = info['upload_date']
-        if info.get('uploader') is not None:
+        if info.get('artist') is not None:
+            metadata['artist'] = info['artist']
+        elif info.get('uploader') is not None:
             metadata['artist'] = info['uploader']
         elif info.get('uploader_id') is not None:
             metadata['artist'] = info['uploader_id']
@@ -554,6 +558,8 @@ class FFmpegMetadataPP(FFmpegPostProcessor):
             metadata['comment'] = info['description']
         if info.get('webpage_url') is not None:
             metadata['purl'] = info['webpage_url']
+        if info.get('album') is not None:
+            metadata['album'] = info['album']
 
         if not metadata:
             self._downloader.to_screen('[ffmpeg] There isn\'t any metadata to add')

youtube_dl/postprocessor/metadatafromtitle.py

@@ -0,0 +1,48 @@
+# -*- coding: utf-8 -*-
+
+import re
+
+from .common import PostProcessor
+from ..utils import PostProcessingError
+
+
+class MetadataFromTitlePPError(PostProcessingError):
+    pass
+
+
+class MetadataFromTitlePP(PostProcessor):
+    def __init__(self, downloader, titleformat):
+        self._titleformat = titleformat
+        self._titleregex = self.fmtToRegex(titleformat)
+
+    def fmtToRegex(self, fmt):
+        """
+        Converts a string like
+           '%(title)s - %(artist)s'
+        to a regex like
+           '(?P<title>.+)\ \-\ (?P<artist>.+)'
+        and a list of the named groups [title, artist]
+        """
+        lastpos = 0
+        regex = ""
+        groups = []
+        # 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