Added _read_mp4box_stem_titles function

.github/workflows/test_unittests.yml

@@ -30,6 +30,7 @@ jobs:
         env:
           FFMPEG_INSTALL: ${{ matrix.pytorch-version }}
         run: |
+          sudo apt-get update
           sudo apt-get -y install gpac
           conda install -c conda-forge ffmpeg==${{ matrix.ffmpeg-version }}
           python -m pip install -e .['tests']
@@ -40,4 +41,4 @@ jobs:
         run: conda list
       - name: Run tests
         run: |
-          py.test tests -v
+          py.test tests -v

.gitignore

@@ -7,6 +7,8 @@ envffmpeg/
 .vscode/
 .circleci/
 
+# Pycharm
+.idea/
 
 #####=== OSX ===#####
 .DS_Store

stempeg/read.py

@@ -4,6 +4,11 @@
 
 
 """
+import codecs
+import json
+import os
+import subprocess
+
 from stempeg.write import FilesWriter
 import numpy as np
 import warnings
@@ -13,6 +18,8 @@
 import atexit
 from functools import partial
 import datetime as dt
+from .cmds import mp4box_exists, find_cmd
+
 
 class Reader(object):
     """Base class for reader
@@ -287,7 +294,7 @@ def read_stems(
         ]
     stem_durations = np.array([t.shape[0] for t in stems])
     if not (stem_durations == stem_durations[0]).all():
-        warnings.warning("Stems differ in length and were shortend")
+        warnings.warn("Stems differ in length and were shortend")
         min_length = np.min(stem_durations)
         stems = [t[:min_length, :] for t in stems]
 
@@ -321,6 +328,11 @@ def __init__(self, filename):
             stream for stream in self.info['streams']
             if stream['codec_type'] == 'audio'
         ]
+        # Try to get the stem titles using MP4Box if possible, otherwise fall back to numbered stems
+        stem_titles = _read_mp4box_stem_titles(filename)
+        if stem_titles is not None:
+            for i, stream in enumerate(self.audio_streams):
+                stream['tags']['handler_name'] = stem_titles[i]
 
     @property
     def nb_audio_streams(self):
@@ -385,3 +397,38 @@ def channels(self, idx):
     def __repr__(self):
         """Print stream information"""
         return pprint.pformat(self.audio_streams)
+
+
+def _read_mp4box_stem_titles(filename):
+    """Reads a mp4 stem titles file using MP4Box
+    Mainly taken from https://github.com/axeldelafosse/stemgen/blob/master/ni-stem/_internal.py
+    """
+    stem_titles = None
+    if mp4box_exists():
+
+        mp4box = find_cmd("MP4Box")
+
+        try:
+            callArgs = [mp4box]
+            callArgs.extend(["-dump-udta", "0:stem", filename, '-quiet'])
+            subprocess.check_call(callArgs)
+
+        except subprocess.CalledProcessError as e:
+            return None
+
+        try:
+            root, ext = os.path.splitext(filename)
+            udtaFile = root + "_stem.udta"
+            fileObj = codecs.open(udtaFile, encoding="utf-8")
+            # fileObj.seek(8)   # Not sure why in the original code this is needed?
+            _metadata = json.load(fileObj)
+            os.remove(udtaFile)
+
+            # add the mixture stem first since its index is 0 as per rest of the project
+            stem_titles = ['Mixture'] + [d['name'] for d in _metadata['stems']]
+
+        except FileNotFoundError as e:
+            return None
+
+    return stem_titles
+

stempeg/write.py

@@ -758,7 +758,7 @@ def write_stems(
     """
     # check if ffmpeg installed
     if int(stempeg.ffmpeg_version()[0]) < 3:
-        warnings.warning(
+        warnings.warn(
             "Writing stems with FFMPEG version < 3 is unsupported",
             UserWarning
         )

tests/test_read.py

@@ -104,3 +104,9 @@ def test_info():
     fp = stempeg.example_stem_path()
     info = stempeg.Info(fp)
     S, rate = stempeg.read_stems(fp, info=info)
+
+
+def test_info_stem_titles():
+    fp = stempeg.example_stem_path()
+    info = stempeg.Info(fp)
+    assert info.title_streams == ['Mixture', 'Drums', 'Bass', 'Other', 'Vox']