sya

commit d78c56c9406479fd020a424117b650352ee2eade
parent c3bb455d0a00ece5dd6b7dcb788a3ccef5d2e342
Author: gearsix <gearsix@tuta.io>
Date:   Tue,  4 May 2021 17:22:38 +0100

session 2: need to start making these commits more incrementally...

- finished get_audio arg handling
- added '-o', '--output' argument
- improved arg handling through functions
- added check to see if timestamps are missing
- added TODO for find_times where timestamps are missing
- start split_tracks

Diffstat:
M
sya.py
 | 
91
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-----------------

1 file changed, 72 insertions(+), 19 deletions(-)
diff --git a/sya.py b/sya.py
@@ -2,19 +2,25 @@
 
 import argparse
 import subprocess
+import re
+import os
+
+Timestamp = re.compile(':?\d\d')
 
 class TracklistItem:
     def __init__(self, timestamp, title):
         self.timestamp = timestamp
         self.title = title
 
-def parseargs():
+def parse_args():
     parser = argparse.ArgumentParser(
         description='download & split audio tracks long youtube videos')
     # arguments
     parser.add_argument('tracklist', metavar='TRACKLIST',
         help='tracklist to split audio by')
     # options
+    parser.add_argument('-o', '--output', metavar='PATH', type=str, nargs='?', default='out',
+        help='specify the directory to write output files to (default: ./out)')
     parser.add_argument('-f', '--format', type=str, nargs='?', default='mp3',
         help='specify the --audio-format argument to pass to youtube-dl (default: mp3)')
     parser.add_argument('-q', '--quality', type=str, nargs='?', default='320K',
@@ -25,11 +31,11 @@ def parseargs():
     parser.add_argument('--ffmpeg', metavar='PATH', type=str, nargs='?',
         default='ffmpeg', dest='ffmpeg',
         help='path of the "ffmpeg" binary to use')
-    parser.add_argument('-k', '--keep-full', action='store_true',
-        help='don\'t remove the full audio file after splitting')
+    parser.add_argument('-k', '--keep', action='store_true',
+        help='keep any files removed during processing (full video/audio file)')
     return parser.parse_args()
 
-def checkbin(*binaries):
+def check_bin(*binaries):
     for b in binaries:
         try:
             subprocess.call([b], stderr=subprocess.DEVNULL, stdout=subprocess.DEVNULL)
@@ -37,24 +43,71 @@ def checkbin(*binaries):
             print(b, 'failed to execute, check it exists in your $PATH.\n'
             'Otherwise you can point to the binary using the relevant optional argument.')
 
-def getaudio(youtubedl, url, format, quality):
-    subprocess.call([youtubedl, url, '-x', '--audio-format', format, '--audio-quality', quality, '-k'])
+def get_audio(youtubedl, url, format='mp3', quality='320K', keep=True, ffmpeg='ffmpeg'):
+    cmd = [youtubedl, url, '--extract-audio', '--audio-format', format,
+        '--audio-quality', quality, '-o', 'audio.%(ext)s', '--prefer-ffmpeg']
+    if keep == True:
+        cmd.append('-k')
+    if ffmpeg != 'ffmpeg':
+        cmd.append('--ffmpeg-location ', ffmpeg)
+    subprocess.call(cmd)
+    return './audio.mp3'
 
-def readtracks(tracklist_path):
-    tracklist_file = open(tracklist_path, mode = 'r')
-    tracklist_lines = tracklist.readlines()
+def load_tracklist(path):
+    tracklist = []
+    tracklist_file = open(path, mode = 'r')
+    for t in tracklist_file.readlines():
+        tracklist.append(t.strip('\n'))
     tracklist_file.close()
+    return tracklist
 
-    tracklist = []
-    for line_count, line in enumerate(tracklist_lines):
-        if line_count == 0:
-            getaudio(args.youtubedl, line, args.format, args.quality)
-            continue
+def parse_tracks(tracklist):
+    tracks = []
+    for lcount, line in enumerate(tracklist):
         sline = line.split(' ', maxsplit=1)
-        tracklist.append(TracklistItem(sline[0], sline[1]))
-    return tracklist
+        if Timestamp.match(sline[0]):
+            tracks.append(TracklistItem(sline[0], sline[1]))
+        else: # assume there's no timestamp, whole line = track title
+            tracks.append(TracklistItem(None, line))
+    return tracks
+
+def missing_times(tracks):
+    missing = []
+    for i, t in enumerate(tracks):
+        if t.timestamp == None:
+            missing.append(i)
+    return missing
+
+def find_times(tracks, missing):
+    j = 0
+    for i, t in enumerate(tracks):
+        if i == missing[j]:
+            # TODO
+            j += 1
+    return tracks
+
+def split_tracks(ffmpeg, audio_fpath, tracks, outpath):
+    ret = subprocess.run(['ffmpeg', '-v', 'quiet', '-stats', '-i', 'audio.mp3',
+        '-f', 'null', '-'], stdout=subprocess.PIPE)
+    print(ret)
+    for i, t in enumerate(tracks):
+        end = 'END'
+        if i < len(tracks)-1:
+            end = tracks[i+1].timestamp
+        subprocess.call(['ffmpeg', '-nostdin', '-y', '-loglevel', 'error',
+            '-i', audio_fpath, '-ss', t.timestamp, '-to', end, '-acodec', 'copy',
+            '{}/{} - {}.mp3'.format(outpath, str(i).zfill(2), t.title)])
+    return
 
 if __name__ == '__main__':
-    args = parseargs()
-    checkbin(args.youtubedl, args.ffmpeg)
-    tracklist = readtracks(args.tracklist)
+    args = parse_args()
+    check_bin(args.youtubedl, args.ffmpeg)
+    tracklist = load_tracklist(args.tracklist)
+    audio_fpath = get_audio(args.youtubedl, tracklist[0],
+            args.format, args.quality, args.keep, args.ffmpeg)
+    tracks = parse_tracks(tracklist[1:])
+    missing = missing_times(tracks)
+    if len(missing) > 0:
+        tracks = find_times(tracks, missing)
+    os.makedirs(args.output, exist_ok=True)
+    split_tracks(args.ffmpeg, audio_fpath, tracks, args.output)