youtubero/youtubero.py

#!/usr/bin/env python3

#Youtubero - simpatico script che dato un link di youtube scarica i file rinominandoli a modino.

#esempi presi da:
#https://www.programcreek.com/python/example/98358/youtube_dl.YoutubeDL
#https://github.com/ytdl-org/youtube-dl/blob/master/README.md


from __future__ import unicode_literals
import youtube_dl
import shutil
import sys
import re
import os
import validators
import glob
import json


def download_video(url, user = ""):
    print ('--- Inizio ---')
    if not os.path.exists("media"):
        os.makedirs("media")
    if not os.path.exists("temp"):
        os.makedirs("temp")

    ydl_opts = {
        'format': 'bestaudio[ext=m4a]',       
        'outtmpl': 'temp/%(id)s.m4a',        
        'noplaylist': True,
    }

    url = url.strip()
    print (url)
    print (user)
    
    if not validators.url(url):
       print ('--- URL malformato ---')
       return ("Errore: url non valido")


    with youtube_dl.YoutubeDL(ydl_opts) as ydl:
        try:
            meta = ydl.extract_info(url, download = False)
        except youtube_dl.DownloadError as detail:
            print ('--- Errore video non disponibile ---')
            print(str(detail))
            return (str(detail))

    id = meta.get('id').strip()
    title = normalizetext(meta.get('title'))
    alt_title = normalizetext(meta.get('alt_title'))
    creator = normalizetext(meta.get('creator'))
    track = normalizetext(meta.get('track'))
    artist = normalizetext(meta.get('artist'))
    album = normalizetext(meta.get('album'))
        
    print ('id          : %s' %(id))
    print ('title       : %s' %(title))
    print ('artist      : %s' %(artist))
    print ('track       : %s' %(track))

    #scrivo  il json
    with open(os.path.join("temp", id + ".json"), 'w') as outfile:
        json.dump(meta, outfile, indent=4)

    #ho letto le info, ora controllo se il file esiste altrimenti lo scarico
    #miglioria: controllare se upload_date e' uguale a quella del json gia' esistente
    filetemp = os.path.join("temp", id + ".m4a")
    if not glob.glob(filetemp):
        print ('--- Scarico ---')
        ydl.download([url]) #non ho capito perche' ma senza [] fa un carattere per volta
    
    #se il file esiste gia' in Media salto (potrebbe esserci, anche rinominato)
    if glob.glob(os.path.join("media/" + user, '*[[]' + id + '[]]*')):
        print ('--- Skippo ---')
        return ("Errore: %s [%s] già presente" %(title, id))

    print ('--- Converto ---')                
    #qui compone il nome del file
    if not artist is None:
        fileout = artist + " - " + track + " [" + id + "].mp3"
        id3title = track
    else:
        fileout = "_" + title + " [" + id + "].mp3"
        id3title = title
        
    print (fileout)

    if not os.path.exists("media/" + user):
        os.makedirs("media/" + user)

    fileout = os.path.join("media/" + user, fileout)
    
    if os.path.isfile(filetemp):
        #copia il file nella cartella media
        #implementare conversione in mp3 e metadata
        #shutil.copy(filetemp, fileout)
        
        cmd = "ffmpeg -i %s -y -vn -ar 44100 -ac 2 -b:a 192k -acodec libmp3lame \
            -metadata title=\"%s\" -metadata artist=\"%s\" -metadata album=\"%s\" \
            \"%s\"" \
            %(filetemp, id3title, artist, album, fileout)
        os.system(cmd)


    print ('--- Fine ---')
    print ("")

    return ("Scaricato %s [%s]" %(title, id))





def normalizetext(s):
    if s is None:
        return None
    else:
        s = re.sub(r'[\\|/|:|*|?|"|<|>|\|]',r'',s)
        s = " ".join(s.split())
    return s
    
    
    
    

if __name__ == '__main__':
    url = sys.argv[1]
    download_video(url)