spradio-server-django/contrib/upload_s3/upload_s3.py

235 lines
6.6 KiB
Python
Raw Permalink Normal View History

'''
upload_s3.py
This is the helper script that uploads songs from an exported playlist into
an Amazon S3 instance (or other implementations, like DigialOcean Spaces).
'''
import argparse
import json
2019-06-12 19:53:06 +00:00
import logging
import os
import sys
import threading
2019-06-12 19:53:06 +00:00
import traceback
from unicodedata import normalize
from decouple import config
import boto3
# If these four are not defined, then boto3 will look for defaults in the
# ~/.aws configurations
S3_REGION = config('S3_REGION', default=None)
S3_ENDPOINT = config('S3_ENDPOINT', default=None)
S3_ACCESS_KEY = config('S3_ACCESS_KEY', default=None)
S3_SECRET_KEY = config('S3_SECRET_KEY', default=None)
# This has to be defined regardless.
S3_BUCKET = config('S3_BUCKET')
# Radio name for metadata
RADIO_NAME = config('RADIO_NAME', default='Save Point Radio')
2019-06-12 19:53:06 +00:00
logging.basicConfig(
handlers=[logging.FileHandler('./s3_uploads.log', encoding='utf8')],
level=logging.INFO,
format=('[%(asctime)s] [%(levelname)s]'
' [%(name)s.%(funcName)s] === %(message)s'),
datefmt='%Y-%m-%dT%H:%M:%S'
)
LOGGER = logging.getLogger('upload_s3')
class Progress(object):
'''
A callback class for the Amazon S3 upload to detect how far along in an
upload we are.
'''
def __init__(self, filepath):
self._filepath = filepath
self._filename = os.path.basename(filepath)
self._size = float(os.path.getsize(filepath))
self._seen_so_far = 0
self._lock = threading.Lock()
def __call__(self, bytes_amount):
with self._lock:
self._seen_so_far += bytes_amount
percentage = (self._seen_so_far / self._size) * 100
sys.stdout.write(
"\r%s %s / %s (%.2f%%)" % (
self._filename, self._seen_so_far, self._size,
percentage
)
)
sys.stdout.flush()
def asciify(text):
'''
Converts a unicode string to pure ascii.
'''
normal = normalize('NFKC', text)
return normal.encode('ascii', 'backslashreplace').decode('ascii')
def get_fullname(artist):
'''
String representing the artist's full name including an alias, if
available.
'''
if artist['alias']:
if artist['first_name'] or artist['last_name']:
return '{} "{}" {}'.format(artist['first_name'],
artist['alias'],
artist['last_name'])
return artist['alias']
return '{} {}'.format(artist['first_name'], artist['last_name'])
def beautify_artists(artists):
'''
Turns a list of one or more artists into a proper English listing.
'''
fullnames = [get_fullname(artist) for artist in artists]
output = ', '
if len(fullnames) == 2:
output = ' & '
return output.join(fullnames)
def import_playlist(playlist_file):
'''
Imports a playlist from a JSON file, uploads the files to an S3[-like]
instance, and exports a new JSON file with the updated paths.
'''
if not os.path.isfile(playlist_file):
raise FileNotFoundError
with open(playlist_file, 'r', encoding='utf8') as pfile:
playlist = json.load(pfile)
session = boto3.session.Session()
client = session.client(
's3',
region_name=S3_REGION,
endpoint_url=S3_ENDPOINT,
aws_access_key_id=S3_ACCESS_KEY,
aws_secret_access_key=S3_SECRET_KEY
)
2019-06-12 19:53:06 +00:00
totals = {'success': 0, 'fail': 0}
for song in playlist['songs']:
old_path = song['store']['path']
if song['type'] == 'S':
prefix = 'songs'
metadata = {
'album': asciify(song['album']),
'artists': asciify(beautify_artists(song['artists'])),
'game': asciify(song['game']),
'title': asciify(song['title']),
'length': str(song['store']['length']),
'original-path': asciify(old_path)
}
else:
prefix = 'jingles'
metadata = {
'artists': asciify(RADIO_NAME),
'title': asciify(song['title']),
'length': str(song['store']['length']),
'original-path': asciify(old_path)
}
file_hash = song['store']['filehash']
ext = os.path.splitext(old_path)[1]
new_path = '{}/{}{}'.format(prefix, file_hash, ext)
2019-06-12 19:53:06 +00:00
LOGGER.info('Begin upload of: %s', old_path)
try:
client.upload_file(
old_path,
S3_BUCKET,
new_path,
ExtraArgs={
'Metadata': metadata,
'ContentType': song['store']['mime']
},
Callback=Progress(old_path)
)
except Exception:
LOGGER.error(
'Upload failed for: %s -- %s',
old_path,
traceback.print_exc()
)
totals['fail'] += 1
else:
song['store']['path'] = 's3://{}/{}'.format(S3_BUCKET, new_path)
LOGGER.info(
'Successful upload of: %s to %s',
old_path,
song['store']['path']
)
totals['success'] += 1
sys.stdout.write("\r\n")
sys.stdout.flush()
2019-06-12 19:53:06 +00:00
result_message = 'Uploads complete -- {} successful, {} failures'.format(
totals['success'],
totals['fail']
)
print(result_message)
LOGGER.info(result_message)
return playlist
def main():
'''Main loop of the program'''
description = 'Uploads song files to an Amazon S3 (or similar) instance.'
parser = argparse.ArgumentParser(description=description)
subparsers = parser.add_subparsers(dest='command')
parser_playlist = subparsers.add_parser(
'playlist',
help='Import playlist song data.'
)
parser_playlist.add_argument(
'filepath',
help='Path to the playlist file.',
nargs=1
)
if len(sys.argv) == 1:
sys.stderr.write('Error: please specify a command\n\n')
parser.print_help(sys.stderr)
sys.exit(1)
results = None
args = parser.parse_args()
if args.command == 'playlist':
results = import_playlist(args.filepath[0])
if results:
2019-06-12 19:53:06 +00:00
LOGGER.info('Exporting new playlist file to \'playlist_s3.json\'')
with open('playlist_s3.json', 'w', encoding='utf8') as file:
json.dump(
results,
file,
ensure_ascii=False,
sort_keys=True,
indent=4
)
2019-06-12 19:53:06 +00:00
LOGGER.info('Program finished. Exiting.')
if __name__ == '__main__':
main()