CODEWITHSUNDEEP
pythonencodingffmpegmetadatayoutube-dl
mopsiok
mopsiok

Reputation: 565

Downloading youtube mp3 - metadata encoding issue (python, youtube-dl, ffmpeg)

I'm trying to download audio from youtube with youtube-dl.exe and ffmpeg.exe (Windows 7), but I am having some troubles with encoding. I have to parse metadata manually, because when I try to use

--metadata-from-title "%(artist) - %(title)" --extract-audio --audio-format mp3 https://www.youtube.com/watch?v=DaU94Ld3fuM

I get ERROR: Could not interpret title of video as "%(artist) - %(title)"

Anyway, I wrote some code to save metadata with ffmpeg:

def download(url, title_first=False):
    if (0 == subprocess.call('youtube-dl --extract-audio --audio-format mp3 %s' % url)):
        #saves file in current directory in format: VID_TITLE-VID_ID.mp3
        video_id = url[url.find('=')+1:] #video id from URL (after ?v=)
        for f in os.listdir('.'):
            if video_id in f:
                filename = f
                break
        os.rename(filename, video_id+'.mp3') #name without non-ascii chars (for tests)
        video_title = filename[: filename.find(video_id)-1]

        output = video_title + '.mp3'
        title, artist = '', ''
        try: #parsing the title
            x = video_title.find('-')
            artist = video_title[:x].strip()
            title = video_title[x+1:].strip()
            if (title_first): output = '%s - %s.mp3' % (title, artist)
        except:
            pass

        x = 'ffmpeg -i "%s" -metadata title="%s" -metadata artist="%s" -acodec copy -id3v2_version 3 -write_id3v1 1 "%s"' \
                        % (video_id+'.mp3', title, artist, output)
        print x
        subprocess.call(x)

The file is downloaded and then cropped to given start and duration times (the code above is a simplified version). Filename is fine, but when I open the file with AIMP3, it shows rubbish instead of non-ascii characters:

enter image description here

I've tried to re-encode the final command with iso-8859-2, utf-8 and mbcs:

x = x.decode('cp1250').encode('iso-8859-2')

But non-ascii chars are still not readable. Passing an unicode command returns UnicodeEncodeError...

Any idea how to solve this problem?

Upvotes: 0

Views: 5122

Answers (2)

dstftw
dstftw

Reputation: 3067

You are missing an s after each template field. Should be --metadata-from-title "%(artist)s - %(title)s". You should also pass --add-metadata to write the metadata to file. So that final command will look like this:

youtube-dl --metadata-from-title "%(artist)s - %(title)s" --extract-audio --audio-format mp3 --add-metadata https://www.youtube.com/watch?v=DaU94Ld3fuM

Upvotes: 2

nadermx
nadermx

Reputation: 2776

From this SO article I think you are having this problem

import re
from unicodedata import normalize

_punct_re = re.compile(r'[\t !"#$%&\'()*\-/<=>?@\[\\\]^_`{|},.:]+')

def slugify(text, delim=u'-'):
    """Generates an slightly worse ASCII-only slug."""
    result = []
    for word in _punct_re.split(text.lower()):
        word = normalize('NFKD', word).encode('ascii', 'ignore')
        if word:
            result.append(word)
    return unicode(delim.join(result))

Usage:

>>> slugify(u'My International Text: åäö')
u'my-international-text-aao'

Upvotes: 0

Related Questions