def retry_queue(self):
self.logger.info('Retrying scrobble cache.')
a = True
lastfm = LastFm(self.config)
for key in self.cache:
# do submissions retry
try:
a = lastfm.scrobble(self.cache[key][0], self.cache[key][1],
self.cache[key][3])
self.cache[key][2] += 1
except:
pass
# if it was a failed submission
if not a:
# remove this record from retry cache, if we're at the retry limit
if self.cache[key][2] >= MAX_CACHE_AGE:
self.logger.info(u'MAX_CACHE_AGE for {key} : {artist} - {track}'.format(
key, self.cache[key][0], self.key[1]))
self.remove(key)
else:
# successful send to last.fm, remove from cache
self.remove(key)
def retry_queue(self):
self.logger.info('Retrying scrobble cache.')
a = True
lastfm = LastFm(self.config)
for key in self.cache:
# do submissions retry
try:
a = lastfm.scrobble(self.cache[key][0], self.cache[key][1],
self.cache[key][3])
self.cache[key][2] += 1
except:
pass
# if it was a failed submission
if not a:
# remove this record from retry cache, if we're at the retry limit
if self.cache[key][2] >= MAX_CACHE_AGE:
self.logger.info(
u'MAX_CACHE_AGE for {key} : {artist} - {track}'.format(
key, self.cache[key][0], self.key[1]))
self.remove(key)
else:
# successful send to last.fm, remove from cache
self.remove(key)
def main_loop():
last_fm = LastFm(keys.last_fm_username, keys.last_fm_password, keys.last_fm_api_key, keys.last_fm_shared_secret)
vk = Vk(keys.vk_access_token)
default_status = None
old_track_status = ""
while True:
track = last_fm.get_last_fm_now_playing(keys.last_fm_username)
if track:
print('Now playing track:{}'.format(track.name))
new_track_status = 'Слушает {} - {} через Яндекс Музыку'.format(track.artist, track.name)
if default_status is None:
default_status = vk.get_status()
print('Default status is none, setting default: {}'.format(default_status))
vk.set_status(new_track_status)
print('Setting new status on vk: {}'.format(new_track_status))
old_track_status = new_track_status
print('Setting old status: {}'.format(old_track_status))
else:
print('Default status: {}'.format(default_status))
if old_track_status != new_track_status:
print('Old track status differs from new, setting to: {}'.format(new_track_status))
old_track_status = new_track_status
vk.set_status(new_track_status)
else:
print('Old track status and new are the same: {}'.format(new_track_status))
else:
print('No now playing track')
if default_status:
print('Default status exist, setting it to vk: {}'.format(default_status))
vk.set_status(default_status)
default_status = None
else:
print('Default status is None')
time.sleep(10)
def lastfm_reg():
token = request.args.get('token')
if token in ('', None):
flash('Missing LastFM auth token')
return redirect(url_for('user_profile'))
lfm = LastFm(UserManager.get(session.get('userid'))[1], app.logger)
status, error = lfm.link_account(token)
flash(error if not status else 'Successfully linked LastFM account')
return redirect(url_for('user_profile'))
def lastfm_reg():
token = request.args.get('token')
if token in ('', None):
flash('Missing LastFM auth token')
return redirect(url_for('user_profile'))
lfm = LastFm(UserManager.get(session.get('userid'))[1], app.logger)
status, error = lfm.link_account(token)
flash(error if not status else 'Successfully linked LastFM account')
return redirect(url_for('user_profile'))
def lastfm_reg():
token = request.args.get("token")
if token in ("", None):
flash("Missing LastFM auth token")
return redirect(url_for("user_profile"))
lfm = LastFm(UserManager.get(session.get("userid"))[1], app.logger)
status, error = lfm.link_account(token)
flash(error if not status else "Successfully linked LastFM account")
return redirect(url_for("user_profile"))
def df_purge_no_tag(merged_df: pd.DataFrame, lf: lastfm.LastFm):
''' Remove tracks which are not matched to any tag. '''
tids_with_tag = lf.get_tids()
tids_with_tag_df = pd.DataFrame(data={'track_id': tids_with_tag})
return pd.merge(merged_df, tids_with_tag_df, on='track_id', how='inner')
def clean_database():
# check if different .txt path has been provided
if args.supp_txt_path:
lastfm_utils.set_txt_path(args.supp_txt_path)
# check if user provided a .csv folder or a .db file (if using .csv, load .csv into LastFm2Pandas; otherwise, load .db into LastFm)
if os.path.isdir(args.input):
try:
tags = pd.read_csv(os.path.join(args.input, 'lastfm_tags.csv'))
except FileNotFoundError:
raise FileNotFoundError('Please make sure {} contains a file "lastfm_tags.csv".'.format(args.input))
try:
tids = pd.read_csv(os.path.join(args.input, 'lastfm_tids.csv'))
except FileNotFoundError:
raise FileNotFoundError('Please make sure {} contains a file "lastfm_tids.csv".'.format(args.input))
try:
tid_tag = pd.read_csv(os.path.join(args.input, 'lastfm_tid_tag.csv'))
except FileNotFoundError:
raise FileNotFoundError('Please make sure {} contains a file "lastfm_tid_tag.csv".'.format(args.input))
lastfm = LastFm2Pandas.load_from(tags=tags, tids=tids, tid_tag=tid_tag)
else:
lastfm = LastFm(args.input)
df = lastfm_utils.generate_final_df(lastfm)
df.reset_index(drop=True, inplace=True) # sanity check
df.index += 1
assert all(df.columns == ['tag', 'merge_tags']) # sanity check
# generate tables which will go into output database
tags = df['tag'].str.lower()
print('Matching all tags to the "clean" few ones...', end=' ', flush=True)
tag_tag = create_tag_tag_table(lastfm, df)
print('done')
print('Matching all tids to tags...', end=' ', flush=True)
tid_tag = create_tid_tag_table(lastfm, tag_tag, args.val)
print('done')
print('Purging tids...', end=' ', flush=True)
tids = tid_tag['tid'].drop_duplicates()
tids.index = tids.values
tids = tids.map(lastfm.tid_num_to_tid).reindex(pd.RangeIndex(1, len(lastfm.get_tid_nums())+1))
print('done')
return lastfm.LastFm2Pandas.load_from(tags=tags, tids=tids, tid_tag=tid_tag) # wrap into LastFm2Pandas class
def scrobble():
status, res = get_entity(request, Track)
if not status:
return res
t, submission = map(request.args.get, [ 'time', 'submission' ])
if t:
try:
t = int(t) / 1000
except:
return request.error_formatter(0, 'Invalid time value')
else:
t = int(time.time())
lfm = LastFm(request.user, app.logger)
if submission in (None, '', True, 'true', 'True', 1, '1'):
lfm.scrobble(res, t)
else:
lfm.now_playing(res)
return request.formatter({})
def __parse(self):
inst = _mysql.connect(host=self.db['host'],
db=self.db['database'],
user=self.db['user'],
passwd=self.db['password'])
inst.query("SELECT * FROM `%s` LIMIT 0, 1000" % (self.TABLE))
r = inst.store_result()
for row in r.fetch_row(maxrows=0, how=1):
gu = GrooveUser(row['grooveshark'], row['sync'])
lu = LastFm(row['lastfm'], row['session'], self.api)
for track in gu.tracks:
lu.luv_track(track['artist'], track['title'])
del gu, lu
# update last sync time
last_sync = time.strftime("%a, %d %b %Y %H:%M:%S +0000",
time.gmtime())
update_query = "UPDATE `%s` SET sync='%s' \
WHERE lastfm='%s' AND grooveshark='%s'" % \
(self.TABLE,
last_sync,
row['lastfm'],
row['grooveshark'])
inst.query(update_query)
def scrobble():
status, res = get_entity(request, Track)
if not status:
return res
t, submission = map(request.args.get, ['time', 'submission'])
if t:
try:
t = int(t) / 1000
except:
return request.error_formatter(0, 'Invalid time value')
else:
t = int(time.time())
lfm = LastFm(request.user, app.logger)
if submission in (None, '', True, 'true', 'True', 1, '1'):
lfm.scrobble(res, t)
else:
lfm.now_playing(res)
return request.formatter({})
def lastfm_unreg():
lfm = LastFm(UserManager.get(session.get('userid'))[1], app.logger)
lfm.unlink_account()
flash('Unliked LastFM account')
return redirect(url_for('user_profile'))
def lastfm_unreg():
lfm = LastFm(UserManager.get(session.get('userid'))[1], app.logger)
lfm.unlink_account()
flash('Unliked LastFM account')
return redirect(url_for('user_profile'))
def save_example_to_tfrecord(df,
output_path,
audio_format,
root_dir,
tag_path,
sample_rate=16000,
num_mels=96,
multitag=False,
verbose=False):
''' Creates and saves a TFRecord file.
Parameters
----------
df: DataFrame
A pandas DataFrame containing the following columns: "track_id", "mp3_path", "npz_path".
output_path: str
The path or filename to save TFRecord file as.
If not a path, the current folder will be used with output_path as filename.
audio_format: {'waveform', 'log-mel-spectrogram'}
If 'log-mel-spectrogram', audio will be converted to that format; otherwise, it will default to raw waveform.
root_dir: str
The root directory to where the .npz files (or the .mp3 files) are stored.
tag_path: str
The path to the lastfm_clean.db database.
sample_rate: int
The sample rate to use when serializing the audio.
num_mels: int
The number of mels in the mel-spectrogram.
multitag: list
If True, encode multiple tags at the same time (provide as list of filenames; feature names will be 'tags-0', 'tags-1' etc.)
verbose: bool
If True, print progress.
'''
with tf.io.TFRecordWriter(output_path) as writer:
if not multitag:
fm = LastFm(tag_path)
n_tags = len(fm.get_tag_nums())
else:
fm = [LastFm(os.path.join(tag_path, path)) for path in multitag]
n_tags = [len(fm.get_tag_nums()) for fm in fm]
assert all(
x == n_tags[0] for x in
n_tags), 'all databases need to have the same number of tags'
n_tags = n_tags[0] # cast back to int
# initialize
exceptions = []
df.reset_index(drop=True, inplace=True)
if verbose:
progbar = Progbar(
len(df)) # create an instance of the progress bar
for i, cols in df.iterrows():
if verbose:
progbar.add(1) # update progress bar
# unpack cols
tid, path = cols
# encode tags
if not multitag:
encoded_tags = get_encoded_tags(fm, tid, n_tags)
else:
encoded_tags = np.array(
[get_encoded_tags(fm, tid, n_tags) for fm in fm]
) # convert to ndarray to ensure consistency with one-dimensional case
# skip tracks which dont have any "clean" tags
if encoded_tags.size == 0:
if verbose:
print("{} has no tags. Skipping...".format(tid))
continue
path = os.path.join(root_dir, path)
if set(df.columns) == {'track_id', 'npz_path'}:
# get the unsampled array from the .npz file
unsampled_audio = np.load(path)
else:
# get the unsampled array from the original .mp3 file
try:
array, sr = librosa.core.load(path, sr=None)
except:
exceptions.append({
'path': path,
'tid': tid,
'encoded_tags': encoded_tags
})
continue
unsampled_audio = {'array': array, 'sr': sr}
# resample audio array into 'sample_rate' and convert into 'audio_format'
processed_array = process_array(unsampled_audio['array'],
audio_format,
sr_in=unsampled_audio['sr'],
sr_out=sample_rate,
num_mels=num_mels)
# load the tf.Example
example = get_example(processed_array, tid, encoded_tags)
# save the tf.Example into a .tfrecord file
writer.write(example.SerializeToString())
# print exceptions
if set(df.columns) == {'track_id', 'npz_path'}:
return
else:
if exceptions:
print('Could not process the following tracks:')
for i, exception in enumerate(exceptions):
print(" {:3d}. {} {}".format(
i, exception["tid"] + exception["path"]))
return
def parse_config_json(config_path, lastfm):
''' Parse a JSON configuration file into a handy Namespace.
Parameters
-----------
config_path: str
The path to the .json file, or the directory where it is saved.
lastfm: str, LastFm, LastFm2Pandas
Instance of the tags database. If a string is passed, try to instantiate the tags database from the (string as a) path.
Returns
-------
config: argparse.Namespace
'''
if not isinstance(lastfm, (LastFm, LastFm2Pandas)):
lastfm = LastFm(os.path.expanduser(lastfm))
# if config_path is a folder, assume the folder contains a config.json
if os.path.isdir(os.path.expanduser(config_path)):
config_path = os.path.join(os.path.abspath(os.path.expanduser(config_path)), 'config.json')
else:
config_path = os.path.expanduser(config_path)
# load json
with open(config_path, 'r') as f:
config_dict = json.loads(f.read())
# create config namespace
config = argparse.Namespace(**config_dict['model'], **config_dict['model-training'], **config_dict['tfrecords'])
config.path = os.path.abspath(config_path)
# update config (optimizer will be instantiated with tf.get_optimizer using {"class_name": config.optimizer_name, "config": config.optimizer})
config.optimizer_name = config.optimizer.pop('name')
# read tags from popularity dataframe
top = config_dict['tags']['top']
if (top is not None) and (top != config.n_tags):
top_tags = lastfm.popularity()['tag'][:top].tolist()
tags = set(top_tags)
else:
tags = None
# update tags according to 'with' (to be added) and 'without' (to be discarded)
if tags is not None:
if config_dict['tags']['with']:
tags.update(config_dict['tags']['with'])
if config_dict['tags']['without']:
tags.difference_update(config_dict['tags']['without'])
tags = list(tags)
else:
raise ValueError("parameter 'with' is inconsistent to parameter 'top'")
# write final tags
config.tags = np.sort(lastfm.tag_to_tag_num(tags)) if tags is not None else None # sorting is necessary to aviod unexpected behaviour
# write final tags to merge together
config.tags_to_merge = lastfm.tag_to_tag_num(config_dict['tags']['merge']) if config_dict['tags']['merge'] is not None else None
# count number of classes
config.n_output_neurons = len(tags) if tags is not None else config.n_tags
return config
def monitor_log(config):
logger = logging.getLogger(__name__)
st_mtime = False
last_played = None
try:
f = open(config.get('plex-scrobble', 'mediaserver_log_location'))
except IOError:
logger.error('Unable to read log-file {0}. Shutting down.'.format(
config.get('plex-scrobble', 'mediaserver_log_location')))
return
f.seek(0, 2)
while True:
time.sleep(.03)
# reset our file handle in the event the log file was not written to
# within the last 60 seconds. This is a very crude attempt to support
# the log file i/o rotation detection cross-platform.
if int(time.time()) - int(os.fstat(f.fileno()).st_mtime) >= 60:
if int(os.fstat(f.fileno()).st_mtime) == st_mtime: continue
logger.debug(
'Possible log file rotation, resetting file handle (st_mtime={mtime})'
.format(mtime=time.ctime(os.fstat(f.fileno()).st_mtime)))
f.close()
try:
f = open(
config.get('plex-scrobble', 'mediaserver_log_location'))
except IOError:
logger.error(
'Unable to read log-file {0}. Shutting down.'.format(
config.get('plex-scrobble',
'mediaserver_log_location')))
return
f.seek(0, 2)
st_mtime = int(os.fstat(f.fileno()).st_mtime)
line = f.readline()
# read all new lines starting at the end. We attempt to match
# based on a regex value. If we have a match, extract the media file
# id and send it off to last.fm for scrobble.
if line:
played = parse_line(line)
if not played: continue
# when playing via a client, log lines are duplicated (seen via iOS)
# this skips dupes. Note: will also miss songs that have been repeated
if played == last_played:
logger.warn(
'Dupe detection : {0}, not submitting'.format(last_played))
continue
metadata = fetch_metadata(played, config)
if not metadata: continue
# submit to last.fm
lastfm = LastFm(config)
a = lastfm.scrobble(metadata['artist'], metadata['track'],
metadata['album'])
# scrobble was not successful , add to our retry queue
if not a:
cache = ScrobbleCache(config)
cache.add(metadata['artist'], metadata['track'],
metadata['album'])
cache.close
last_played = played
def monitor_log(config):
logger = logging.getLogger(__name__)
st_mtime = False
last_played = None
try:
f = open(config.get('plex-scrobble', 'mediaserver_log_location'))
except IOError:
logger.error('Unable to read log-file {0}. Shutting down.'.format(config.get(
'plex-scrobble', 'mediaserver_log_location')))
return
f.seek(0, 2)
account_name = config.get('plex-scrobble', 'account_name')
while True:
time.sleep(.03)
# reset our file handle in the event the log file was not written to
# within the last 60 seconds. This is a very crude attempt to support
# the log file i/o rotation detection cross-platform.
if int(time.time()) - int(os.fstat(f.fileno()).st_mtime) >= 60:
if int(os.fstat(f.fileno()).st_mtime) == st_mtime: continue
logger.debug('Possible log file rotation, resetting file handle (st_mtime={mtime})'.format(
mtime=time.ctime(os.fstat(f.fileno()).st_mtime) ))
f.close()
try:
f = open(config.get('plex-scrobble', 'mediaserver_log_location'))
except IOError:
logger.error('Unable to read log-file {0}. Shutting down.'.format(config.get(
'plex-scrobble', 'mediaserver_log_location')))
return
f.seek(0, 2)
st_mtime = int(os.fstat(f.fileno()).st_mtime)
line = f.readline()
# read all new lines starting at the end. We attempt to match
# based on a regex value. If we have a match, extract the media file
# id and send it off to last.fm for scrobble.
if line:
played = parse_line(line, account_name)
if not played: continue
# when playing via a client, log lines are duplicated (seen via iOS)
# this skips dupes. Note: will also miss songs that have been repeated
if played == last_played:
logger.warn('Dupe detection : {0}, not submitting'.format(last_played))
continue
metadata = fetch_metadata(played, config)
if not metadata: continue
# submit to last.fm
lastfm = LastFm(config)
a = lastfm.scrobble(metadata['artist'], metadata['track'],
metadata['album'])
# scrobble was not successful , add to our retry queue
if not a:
cache = ScrobbleCache(config)
cache.add(metadata['artist'], metadata['track'], metadata['album'])
cache.close
last_played = played
def lastfm(self):
if self._lastfm is None:
self._lastfm = LastFm()
return self._lastfm
