Spaces:
Paused
Paused
taslim19
commited on
Commit
·
eef327e
1
Parent(s):
07027a4
feat: Add downloader utility for advanced downloading
Browse files- DragMusic/utils/downloader.py +201 -0
DragMusic/utils/downloader.py
ADDED
@@ -0,0 +1,201 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import asyncio
|
2 |
+
import aiohttp
|
3 |
+
import aiofiles
|
4 |
+
import os
|
5 |
+
import re
|
6 |
+
from typing import Optional, Union, Dict
|
7 |
+
from yt_dlp import YoutubeDL
|
8 |
+
from config import API_URL, API_KEY
|
9 |
+
|
10 |
+
USE_API = bool(API_URL and API_KEY)
|
11 |
+
_logged_api_skip = False
|
12 |
+
CHUNK_SIZE = 8192
|
13 |
+
RETRY_DELAY = 2
|
14 |
+
cookies_file = "cookies.txt"
|
15 |
+
download_folder = "downloads"
|
16 |
+
os.makedirs(download_folder, exist_ok=True)
|
17 |
+
|
18 |
+
|
19 |
+
def extract_video_id(link: str) -> str:
|
20 |
+
if "v=" in link:
|
21 |
+
return link.split("v=")[-1].split("&")[0]
|
22 |
+
return link.split("/")[-1].split("?")[0]
|
23 |
+
|
24 |
+
|
25 |
+
def safe_filename(name: str) -> str:
|
26 |
+
return re.sub(r"[\\/*?\"<>|]", "_", name).strip()[:100]
|
27 |
+
|
28 |
+
|
29 |
+
def file_exists(video_id: str) -> Optional[str]:
|
30 |
+
for ext in ["mp3", "m4a", "webm"]:
|
31 |
+
path = f"{download_folder}/{video_id}.{ext}"
|
32 |
+
if os.path.exists(path):
|
33 |
+
print(f"[CACHED] Using existing file: {path}")
|
34 |
+
return path
|
35 |
+
return None
|
36 |
+
|
37 |
+
|
38 |
+
async def api_download_song(link: str) -> Optional[str]:
|
39 |
+
global _logged_api_skip
|
40 |
+
|
41 |
+
if not USE_API:
|
42 |
+
if not _logged_api_skip:
|
43 |
+
print("[SKIPPED] API config missing — using yt-dlp only.")
|
44 |
+
_logged_api_skip = True
|
45 |
+
return None
|
46 |
+
|
47 |
+
video_id = extract_video_id(link)
|
48 |
+
song_url = f"{API_URL}/song/{video_id}?api={API_KEY}"
|
49 |
+
|
50 |
+
try:
|
51 |
+
async with aiohttp.ClientSession() as session:
|
52 |
+
while True:
|
53 |
+
async with session.get(song_url) as response:
|
54 |
+
if response.status != 200:
|
55 |
+
print(f"[API ERROR] Status {response.status}")
|
56 |
+
return None
|
57 |
+
|
58 |
+
data = await response.json()
|
59 |
+
status = data.get("status", "").lower()
|
60 |
+
|
61 |
+
if status == "downloading":
|
62 |
+
await asyncio.sleep(RETRY_DELAY)
|
63 |
+
continue
|
64 |
+
elif status == "error":
|
65 |
+
print(f"[API ERROR] Status=error for {video_id}")
|
66 |
+
return None
|
67 |
+
elif status == "done":
|
68 |
+
download_url = data.get("link")
|
69 |
+
break
|
70 |
+
else:
|
71 |
+
print(f"[API ERROR] Unknown status: {status}")
|
72 |
+
return None
|
73 |
+
|
74 |
+
fmt = data.get("format", "mp3").lower()
|
75 |
+
path = f"{download_folder}/{video_id}.{fmt}"
|
76 |
+
|
77 |
+
async with session.get(download_url) as file_response:
|
78 |
+
async with aiofiles.open(path, "wb") as f:
|
79 |
+
while True:
|
80 |
+
chunk = await file_response.content.read(CHUNK_SIZE)
|
81 |
+
if not chunk:
|
82 |
+
break
|
83 |
+
await f.write(chunk)
|
84 |
+
|
85 |
+
return path
|
86 |
+
except Exception as e:
|
87 |
+
print(f"[API Download Error] {e}")
|
88 |
+
return None
|
89 |
+
|
90 |
+
|
91 |
+
def _download_ytdlp(link: str, opts: Dict) -> Optional[str]:
|
92 |
+
try:
|
93 |
+
with YoutubeDL(opts) as ydl:
|
94 |
+
info = ydl.extract_info(link, download=False)
|
95 |
+
ext = info.get("ext", "webm")
|
96 |
+
vid = info.get("id")
|
97 |
+
path = f"{download_folder}/{vid}.{ext}"
|
98 |
+
if os.path.exists(path):
|
99 |
+
return path
|
100 |
+
ydl.download([link])
|
101 |
+
return path
|
102 |
+
except Exception as e:
|
103 |
+
print(f"[yt-dlp Error] {e}")
|
104 |
+
return None
|
105 |
+
|
106 |
+
|
107 |
+
async def yt_dlp_download(link: str, type: str, format_id: str = None, title: str = None) -> Optional[str]:
|
108 |
+
loop = asyncio.get_running_loop()
|
109 |
+
|
110 |
+
if type == "audio":
|
111 |
+
opts = {
|
112 |
+
"format": "bestaudio/best",
|
113 |
+
"outtmpl": f"{download_folder}/%(id)s.%(ext)s",
|
114 |
+
"quiet": True,
|
115 |
+
"no_warnings": True,
|
116 |
+
"cookiefile": cookies_file,
|
117 |
+
"noplaylist": True,
|
118 |
+
"concurrent_fragment_downloads": 5,
|
119 |
+
}
|
120 |
+
return await loop.run_in_executor(None, _download_ytdlp, link, opts)
|
121 |
+
|
122 |
+
elif type == "video":
|
123 |
+
opts = {
|
124 |
+
"format": "best[height<=?720][width<=?1280]",
|
125 |
+
"outtmpl": f"{download_folder}/%(id)s.%(ext)s",
|
126 |
+
"quiet": True,
|
127 |
+
"no_warnings": True,
|
128 |
+
"cookiefile": cookies_file,
|
129 |
+
"noplaylist": True,
|
130 |
+
"concurrent_fragment_downloads": 5,
|
131 |
+
}
|
132 |
+
return await loop.run_in_executor(None, _download_ytdlp, link, opts)
|
133 |
+
|
134 |
+
elif type == "song_video" and format_id and title:
|
135 |
+
safe_title = safe_filename(title)
|
136 |
+
opts = {
|
137 |
+
"format": f"{format_id}+140",
|
138 |
+
"outtmpl": f"{download_folder}/{safe_title}.mp4",
|
139 |
+
"quiet": True,
|
140 |
+
"no_warnings": True,
|
141 |
+
"prefer_ffmpeg": True,
|
142 |
+
"merge_output_format": "mp4",
|
143 |
+
"cookiefile": cookies_file,
|
144 |
+
}
|
145 |
+
await loop.run_in_executor(None, lambda: YoutubeDL(opts).download([link]))
|
146 |
+
return f"{download_folder}/{safe_title}.mp4"
|
147 |
+
|
148 |
+
elif type == "song_audio" and format_id and title:
|
149 |
+
safe_title = safe_filename(title)
|
150 |
+
opts = {
|
151 |
+
"format": format_id,
|
152 |
+
"outtmpl": f"{download_folder}/{safe_title}.%(ext)s",
|
153 |
+
"quiet": True,
|
154 |
+
"no_warnings": True,
|
155 |
+
"prefer_ffmpeg": True,
|
156 |
+
"cookiefile": cookies_file,
|
157 |
+
"postprocessors": [{
|
158 |
+
"key": "FFmpegExtractAudio",
|
159 |
+
"preferredcodec": "mp3",
|
160 |
+
"preferredquality": "192",
|
161 |
+
}],
|
162 |
+
}
|
163 |
+
await loop.run_in_executor(None, lambda: YoutubeDL(opts).download([link]))
|
164 |
+
return f"{download_folder}/{safe_title}.mp3"
|
165 |
+
|
166 |
+
return None
|
167 |
+
|
168 |
+
|
169 |
+
async def download_audio_concurrent(link: str) -> Optional[str]:
|
170 |
+
video_id = extract_video_id(link)
|
171 |
+
|
172 |
+
existing = file_exists(video_id)
|
173 |
+
if existing:
|
174 |
+
return existing
|
175 |
+
|
176 |
+
if not USE_API:
|
177 |
+
return await yt_dlp_download(link, type="audio")
|
178 |
+
|
179 |
+
yt_task = asyncio.create_task(yt_dlp_download(link, type="audio"))
|
180 |
+
api_task = asyncio.create_task(api_download_song(link))
|
181 |
+
|
182 |
+
done, _ = await asyncio.wait([yt_task, api_task], return_when=asyncio.FIRST_COMPLETED)
|
183 |
+
|
184 |
+
for task in done:
|
185 |
+
try:
|
186 |
+
result = task.result()
|
187 |
+
if result:
|
188 |
+
return result
|
189 |
+
except Exception as e:
|
190 |
+
print(f"[Download Task Error] {e}")
|
191 |
+
|
192 |
+
for task in [yt_task, api_task]:
|
193 |
+
if not task.done():
|
194 |
+
try:
|
195 |
+
result = await task
|
196 |
+
if result:
|
197 |
+
return result
|
198 |
+
except Exception as e:
|
199 |
+
print(f"[Fallback Task Error] {e}")
|
200 |
+
|
201 |
+
return None
|