def test_remove_windows_drive(self):
self.assertEqual(
('album1', None),
album_artist_from_path(r'C:\album1\foo.mp3', None, None))
self.assertEqual(
('album1', None),
album_artist_from_path(r'\\myserver\myshare\album1\foo.mp3', None, None))
def test_strip_disc_dir(self):
self.assertEqual(
('albumy', 'artistx'),
album_artist_from_path(r'/artistx/albumy/CD 1/file.flac', '', ''))
self.assertEqual(
('albumy', 'artistx'),
album_artist_from_path(r'/artistx/albumy/the DVD 23 B/file.flac', '', ''))
self.assertEqual(
('albumy', 'artistx'),
album_artist_from_path(r'/artistx/albumy/disc23/file.flac', '', ''))
self.assertNotEqual(
('albumy', 'artistx'),
album_artist_from_path(r'/artistx/albumy/disc/file.flac', '', ''))
def cluster(files):
"""Group the provided files into clusters, based on album tag in metadata.
Args:
files: List of File objects.
Yields:
FileCluster objects
"""
config = get_config()
various_artists = config.setting['va_name']
cluster_list = defaultdict(FileCluster)
for file in files:
artist = file.metadata["albumartist"] or file.metadata["artist"]
album = file.metadata["album"]
# Improve clustering from directory structure if no existing tags
# Only used for grouping and to provide cluster title / artist - not added to file tags.
album, artist = album_artist_from_path(file.filename, album,
artist)
token = tokenize(album)
if token:
cluster_list[token].add(album, artist or various_artists, file)
yield from cluster_list.values()
def cluster(files, threshold):
win_compat = config.setting["windows_compatibility"] or IS_WIN
artist_dict = ClusterDict()
album_dict = ClusterDict()
tracks = []
for file in files:
artist = file.metadata["albumartist"] or file.metadata["artist"]
album = file.metadata["album"]
# Improve clustering from directory structure if no existing tags
# Only used for grouping and to provide cluster title / artist - not added to file tags.
if win_compat:
filename = ntpath.splitdrive(file.filename)[1]
else:
filename = file.filename
album, artist = album_artist_from_path(filename, album, artist)
# For each track, record the index of the artist and album within the clusters
tracks.append((artist_dict.add(artist), album_dict.add(album)))
artist_cluster_engine = ClusterEngine(artist_dict)
artist_cluster_engine.cluster(threshold)
album_cluster_engine = ClusterEngine(album_dict)
album_cluster_engine.cluster(threshold)
# Arrange tracks into albums
albums = {}
for i, track in enumerate(tracks):
cluster = album_cluster_engine.get_cluster_from_id(track[1])
if cluster is not None:
albums.setdefault(cluster, []).append(i)
# Now determine the most prominent names in the cluster and build the
# final cluster list
for album_id, album in albums.items():
album_name = album_cluster_engine.get_cluster_title(album_id)
artist_max = 0
artist_id = None
artist_hist = {}
for track_id in album:
cluster = artist_cluster_engine.get_cluster_from_id(
tracks[track_id][0])
if cluster is not None:
cnt = artist_hist.get(cluster, 0) + 1
if cnt > artist_max:
artist_max = cnt
artist_id = cluster
artist_hist[cluster] = cnt
if artist_id is None:
artist_name = "Various Artists"
else:
artist_name = artist_cluster_engine.get_cluster_title(
artist_id)
yield album_name, artist_name, (files[i] for i in album)
def cluster(files, threshold):
artistDict = ClusterDict()
albumDict = ClusterDict()
tracks = []
for file in files:
artist = file.metadata["albumartist"] or file.metadata["artist"]
album = file.metadata["album"]
# Improve clustering from directory structure if no existing tags
# Only used for grouping and to provide cluster title / artist - not added to file tags.
filename = file.filename
if config.setting["windows_compatibility"] or sys.platform == "win32":
filename = ntpath.splitdrive(filename)[1]
album, artist = album_artist_from_path(filename, album, artist)
# For each track, record the index of the artist and album within the clusters
tracks.append((artistDict.add(artist),
albumDict.add(album)))
artist_cluster_engine = ClusterEngine(artistDict)
artist_cluster_engine.cluster(threshold)
album_cluster_engine = ClusterEngine(albumDict)
album_cluster_engine.cluster(threshold)
# Arrange tracks into albums
albums = {}
for i in xrange(len(tracks)):
cluster = album_cluster_engine.getClusterFromId(tracks[i][1])
if cluster is not None:
albums.setdefault(cluster, []).append(i)
# Now determine the most prominent names in the cluster and build the
# final cluster list
for album_id, album in albums.items():
album_name = album_cluster_engine.getClusterTitle(album_id)
artist_max = 0
artist_id = None
artist_hist = {}
for track_id in album:
cluster = artist_cluster_engine.getClusterFromId(
tracks[track_id][0])
if cluster is not None:
cnt = artist_hist.get(cluster, 0) + 1
if cnt > artist_max:
artist_max = cnt
artist_id = cluster
artist_hist[cluster] = cnt
if artist_id is None:
artist_name = u"Various Artists"
else:
artist_name = artist_cluster_engine.getClusterTitle(artist_id)
yield album_name, artist_name, (files[i] for i in album)
def cluster(files, threshold, tagger=None):
config = get_config()
win_compat = config.setting["windows_compatibility"] or IS_WIN
artist_dict = ClusterDict()
album_dict = ClusterDict()
tracks = []
num_files = len(files)
# 10 evenly spaced indexes of files being clustered, used as checkpoints for every 10% progress
status_update_steps = ProgressCheckpoints(num_files, 10)
for i, file in process_events_iter(enumerate(files)):
artist = file.metadata["albumartist"] or file.metadata["artist"]
album = file.metadata["album"]
# Improve clustering from directory structure if no existing tags
# Only used for grouping and to provide cluster title / artist - not added to file tags.
if win_compat:
filename = ntpath.splitdrive(file.filename)[1]
else:
filename = file.filename
album, artist = album_artist_from_path(filename, album, artist)
# For each track, record the index of the artist and album within the clusters
tracks.append((artist_dict.add(artist), album_dict.add(album)))
if tagger and status_update_steps.is_checkpoint(i):
statusmsg = N_(
"Clustering - step %(step)d/3: %(cluster_type)s (%(update)d%%)"
)
mparams = {
'step':
ClusterType.METADATA.value,
'cluster_type':
_(ClusterEngine.cluster_type_label(ClusterType.METADATA)),
'update':
status_update_steps.progress(i),
}
tagger.window.set_statusbar_message(statusmsg, mparams)
artist_cluster_engine = ClusterEngine(artist_dict, ClusterType.ARTIST)
artist_cluster_engine.cluster(threshold, tagger)
album_cluster_engine = ClusterEngine(album_dict, ClusterType.ALBUM)
album_cluster_engine.cluster(threshold, tagger)
# Arrange tracks into albums
albums = {}
for i, track in enumerate(tracks):
cluster = album_cluster_engine.get_cluster_from_id(track[1])
if cluster is not None:
albums.setdefault(cluster, []).append(i)
# Now determine the most prominent names in the cluster and build the
# final cluster list
for album_id, album in albums.items():
album_name = album_cluster_engine.get_cluster_title(album_id)
artist_max = 0
artist_id = None
artist_hist = {}
for track_id in album:
cluster = artist_cluster_engine.get_cluster_from_id(
tracks[track_id][0])
if cluster is not None:
cnt = artist_hist.get(cluster, 0) + 1
if cnt > artist_max:
artist_max = cnt
artist_id = cluster
artist_hist[cluster] = cnt
if artist_id is None:
artist_name = "Various Artists"
else:
artist_name = artist_cluster_engine.get_cluster_title(
artist_id)
yield album_name, artist_name, (files[i] for i in album)
def test_path_no_dirs(self):
for name in ('', 'x', '/', '\\', '///'):
self.assertEqual(('', 'artist'), album_artist_from_path(name, '', 'artist'))
def cluster(files, threshold):
artistDict = ClusterDict()
albumDict = ClusterDict()
tracks = []
for file in files:
artist = file.metadata["albumartist"] or file.metadata["artist"]
album = file.metadata["album"]
# Improve clustering from directory structure if no existing tags
# Only used for grouping and to provide cluster title / artist - not added to file tags.
filename = file.filename
if config.setting[
"windows_compatibility"] or sys.platform == "win32":
filename = ntpath.splitdrive(filename)[1]
album, artist = album_artist_from_path(filename, album, artist)
# For each track, record the index of the artist and album within the clusters
tracks.append((artistDict.add(artist), albumDict.add(album)))
artist_cluster_engine = ClusterEngine(artistDict)
artist_cluster_engine.cluster(threshold)
album_cluster_engine = ClusterEngine(albumDict)
album_cluster_engine.cluster(threshold)
# Arrange tracks into albums
albums = {}
for i in xrange(len(tracks)):
cluster = album_cluster_engine.getClusterFromId(tracks[i][1])
if cluster is not None:
albums.setdefault(cluster, []).append(i)
# Now determine the most prominent names in the cluster and build the
# final cluster list
for album_id, album in albums.items():
album_name = album_cluster_engine.getClusterTitle(album_id)
artist_max = 0
artist_id = None
artist_hist = {}
main_artist = None
do_all = False
do_cluster = True
to_remove = []
for track_id in album:
artist = tracks[track_id][0]
if main_artist is None:
main_artist = artist
cluster = artist_cluster_engine.getClusterFromId(
tracks[track_id][0])
# if it isn't the first track the user hasn't chosen an action to do for all
if artist is not main_artist:
if not do_all:
do_cluster, do_all = Cluster.cluster_warning(
files, track_id, album)
if not do_cluster:
to_remove.append(track_id)
if do_cluster and cluster is not None:
cnt = artist_hist.get(cluster, 0) + 1
if cnt > artist_max:
artist_max = cnt
artist_id = cluster
artist_hist[cluster] = cnt
for id in to_remove:
album.remove(id)
if artist_id is None:
artist_name = u"Various Artists"
else:
artist_name = artist_cluster_engine.getClusterTitle(artist_id)
yield album_name, artist_name, (files[i] for i in album)
