added yt-dlp for youtube music download

This commit is contained in:
Alexander Karpov 2023-08-11 22:38:38 +03:00
parent 08b1479401
commit b7773181ae
8 changed files with 653 additions and 56 deletions

View File

@ -28,8 +28,10 @@ def __init__(self, app):
async def __call__(self, scope, receive, send): async def __call__(self, scope, receive, send):
scope["user"] = await get_user(dict(scope["headers"])) scope["user"] = await get_user(dict(scope["headers"]))
try:
return await self.app(scope, receive, send) return await self.app(scope, receive, send)
except ValueError:
return
class BaseConsumer(AsyncJsonWebsocketConsumer): class BaseConsumer(AsyncJsonWebsocketConsumer):

View File

@ -6,11 +6,12 @@
def login() -> spotipy.Spotify: def login() -> spotipy.Spotify:
if not settings.SPOTIFY_ID or not settings.SPOTIFY_SECRET: if not settings.MUSIC_SPOTIFY_ID or not settings.MUSIC_SPOTIFY_SECRET:
raise ConnectionError("No spotify credentials provided") raise ConnectionError("No spotify credentials provided")
return spotipy.Spotify( return spotipy.Spotify(
auth_manager=SpotifyClientCredentials( auth_manager=SpotifyClientCredentials(
client_id=settings.SPOTIFY_ID, client_secret=settings.SPOTIFY_SECRET client_id=settings.MUSIC_SPOTIFY_ID,
client_secret=settings.MUSIC_SPOTIFY_SECRET,
) )
) )

View File

@ -16,9 +16,9 @@
def login() -> Client: def login() -> Client:
if not settings.YANDEX_TOKEN: if not settings.MUSIC_YANDEX_TOKEN:
raise ConnectionError("No yandex credentials provided") raise ConnectionError("No yandex credentials provided")
return Client(settings.YANDEX_TOKEN).init() return Client(settings.MUSIC_YANDEX_TOKEN).init()
def search_ym(name: str): def search_ym(name: str):

View File

@ -1,51 +1,145 @@
import datetime
import os import os
import re
from random import randint from random import randint
import requests import requests
import yt_dlp
from django.conf import settings from django.conf import settings
from PIL import Image from PIL import Image
from pydub import AudioSegment from pydub import AudioSegment
from pytube import Search, YouTube from pytube import Search, YouTube
from yt_dlp import YoutubeDL
from akarpov.music.models import Song, SongInQue from akarpov.music.models import Song
from akarpov.music.services.db import load_track from akarpov.music.services.db import load_track
from akarpov.music.services.spotify import get_track_info from akarpov.music.services.spotify import get_track_info
final_filename = None
ydl_opts = {
"format": "m4a/bestaudio/best",
"postprocessors": [
{ # Extract audio using ffmpeg
"key": "FFmpegExtractAudio",
"preferredcodec": "m4a",
}
],
"outtmpl": f"{settings.MEDIA_ROOT}/%(uploader)s_%(title)s.%(ext)s",
}
def download_file(url):
with yt_dlp.YoutubeDL(ydl_opts) as ydl:
info = ydl.extract_info(url)
return info["requested_downloads"][0]["_filename"]
def parse_description(description: str) -> list:
# Read the description file
# Split into time and chapter name
list_of_chapters = []
# only increment chapter number on a chapter line
# chapter lines start with timecode
line_counter = 1
for line in description.split("\n"):
result = re.search(r"\(?(\d?[:]?\d+[:]\d+)\)?", line)
try:
time_count = datetime.datetime.strptime(result.group(1), "%H:%M:%S")
except Exception:
try:
time_count = datetime.datetime.strptime(result.group(1), "%M:%S")
except Exception:
continue
chap_name = line.replace(result.group(0), "").rstrip(" :\n")
chap_pos = (
time_count.timestamp() - datetime.datetime(1900, 1, 1, 0, 0).timestamp()
) * 1000
list_of_chapters.append((str(line_counter).zfill(2), chap_pos, chap_name))
line_counter += 1
return list_of_chapters
def download_from_youtube_link(link: str) -> Song: def download_from_youtube_link(link: str) -> Song:
que = SongInQue.objects.create() song = None
try:
yt = YouTube(link)
if yt.length > 900: with YoutubeDL(ydl_opts) as ydl:
# TODO: add long video splitting info_dict = ydl.extract_info(link, download=False)
raise ValueError("Track is too long") title = info_dict.get("title", None)
description = info_dict.get("description", None)
chapters = parse_description(description)
orig_path = download_file(link)
if not len(yt.streams): # convert to mp3
raise ValueError("There is no such song") print(f"[processing] {title} converting to mp3")
path = orig_path.replace(orig_path.split(".")[-1], "mp3")
AudioSegment.from_file(orig_path).export(path)
os.remove(orig_path)
print(f"[processing] {title} converting to mp3: done")
info = get_track_info(yt.title) # split in chapters
que.name = info["title"] if len(chapters) > 1:
que.save() sound = AudioSegment.from_mp3(path)
if sng := Song.objects.filter( for i in range(len(chapters)):
name=info["title"], album__name=info["album_name"] if i != len(chapters) - 1:
): print(
que.delete() f"[processing] loading {chapters[i][2]} from {chapters[i][1] // 1000} to",
return sng.first() f"{chapters[i + 1][1] // 1000}",
)
st = chapters[i][1]
end = chapters[i + 1][1]
audio = sound[st:end]
else:
print(
f"[processing] loading {chapters[i][2]} from {chapters[i][1] // 1000}"
)
st = chapters[i][1]
audio = sound[st:]
chapter_path = path.split(".")[0] + chapters[i][2] + ".mp3"
info = get_track_info(chapters[i][2])
audio.export(chapter_path, format="mp3")
r = requests.get(info["album_image"])
img_pth = str(
settings.MEDIA_ROOT
+ f"/{info['album_image'].split('/')[-1]}_{str(randint(100, 999))}"
)
with open(img_pth, "wb") as f:
f.write(r.content)
audio = yt.streams.filter(only_audio=True).order_by("abr").desc().first() im = Image.open(img_pth)
orig_path = audio.download(output_path=settings.MEDIA_ROOT) im.save(str(f"{img_pth}.png"))
# convert to mp3 os.remove(img_pth)
path = orig_path.replace(orig_path.split(".")[-1], "mp3") if "genre" in info:
AudioSegment.from_file(orig_path).export(path) song = load_track(
os.remove(orig_path) chapter_path,
f"{img_pth}.png",
info["artists"],
info["album_name"],
chapters[i][2],
genre=info["genre"],
)
else:
song = load_track(
chapter_path,
f"{img_pth}.png",
info["artists"],
info["album_name"],
chapters[i][2],
)
os.remove(chapter_path)
else:
print(f"[processing] loading {title}")
# load album image info = get_track_info(title)
r = requests.get(info["album_image"]) r = requests.get(info["album_image"])
img_pth = str( img_pth = str(
settings.MEDIA_ROOT settings.MEDIA_ROOT
+ f"/{info['album_image'].split('/')[-1]}_{str(randint(100, 999))}.png" + f"/{info['album_image'].split('/')[-1]}_{str(randint(100, 999))}"
) )
with open(img_pth, "wb") as f: with open(img_pth, "wb") as f:
f.write(r.content) f.write(r.content)
@ -54,13 +148,26 @@ def download_from_youtube_link(link: str) -> Song:
im.save(str(f"{img_pth}.png")) im.save(str(f"{img_pth}.png"))
os.remove(img_pth) os.remove(img_pth)
if "genre" in info:
song = load_track(
path,
f"{img_pth}.png",
info["artists"],
info["album_name"],
title,
genre=info["genre"],
)
else:
song = load_track(
path,
f"{img_pth}.png",
info["artists"],
info["album_name"],
title,
)
os.remove(path)
load_track(path, img_pth, info["artists"], info["album_name"]) return song
except Exception as e:
print(e)
que.name = e
que.error = True
que.save()
def search_channel(name): def search_channel(name):

View File

@ -66,3 +66,8 @@ class Meta:
def __str__(self): def __str__(self):
return self return self
class UserNotification:
# TODO: add notification system
...

View File

@ -407,9 +407,9 @@
# https://docs.celeryq.dev/en/stable/userguide/configuration.html#std:setting-result_serializer # https://docs.celeryq.dev/en/stable/userguide/configuration.html#std:setting-result_serializer
CELERY_RESULT_SERIALIZER = "json" CELERY_RESULT_SERIALIZER = "json"
# https://docs.celeryq.dev/en/stable/userguide/configuration.html#task-time-limit # https://docs.celeryq.dev/en/stable/userguide/configuration.html#task-time-limit
CELERY_TASK_TIME_LIMIT = 5 * 60 CELERY_TASK_TIME_LIMIT = 20 * 60
# https://docs.celeryq.dev/en/stable/userguide/configuration.html#task-soft-time-limit # https://docs.celeryq.dev/en/stable/userguide/configuration.html#task-soft-time-limit
CELERY_TASK_SOFT_TIME_LIMIT = 60 CELERY_TASK_SOFT_TIME_LIMIT = 10 * 60
# https://docs.celeryq.dev/en/stable/userguide/configuration.html#beat-scheduler # https://docs.celeryq.dev/en/stable/userguide/configuration.html#beat-scheduler
CELERY_BEAT_SCHEDULER = "django_celery_beat.schedulers:DatabaseScheduler" CELERY_BEAT_SCHEDULER = "django_celery_beat.schedulers:DatabaseScheduler"

511
poetry.lock generated

File diff suppressed because it is too large Load Diff

View File

@ -73,7 +73,6 @@ pydotplus = "^2.0.2"
preview-generator = "^0.29" preview-generator = "^0.29"
uuid = "^1.30" uuid = "^1.30"
mutagen = "^1.46.0" mutagen = "^1.46.0"
pytube = "^12.1.3"
ytmusicapi = "^1.0.2" ytmusicapi = "^1.0.2"
pydub = "^0.25.1" pydub = "^0.25.1"
python-mpd2 = "^3.0.5" python-mpd2 = "^3.0.5"
@ -97,6 +96,10 @@ django-ipware = "^5.0.0"
fastapi = {extras = ["all"], version = "^0.101.0"} fastapi = {extras = ["all"], version = "^0.101.0"}
sqlalchemy = "^2.0.19" sqlalchemy = "^2.0.19"
pydantic-settings = "^2.0.2" pydantic-settings = "^2.0.2"
yt-dlp = "^2023.7.6"
pytube = "^15.0.0"
urllib3 = ">=1.26"
requests = ">=2.25"
[build-system] [build-system]