fix: resolve problem with the url sign up

app.py

@@ -12,8 +12,6 @@ from pathlib import Path
 import spaces
 from pydub.exceptions import CouldntEncodeError
 from transformers import pipeline
-import requests
-from bs4 import BeautifulSoup
 
 # Initialize text generation model
 model = pipeline('text-generation', model='EleutherAI/gpt-neo-125M')
@@ -23,6 +21,28 @@ OUTPUT_FOLDER = "separation_results/"
 INPUT_FOLDER = "input"
 download_path = ""
 
+class MyLogger:
+    def debug(self, msg):
+        # For compatibility with youtube-dl, both debug and info are passed into debug
+        if msg.startswith('[debug] '):
+            pass
+        else:
+            self.info(msg)
+
+    def info(self, msg):
+        pass
+
+    def warning(self, msg):
+        pass
+
+    def error(self, msg):
+        print(msg)
+
+def my_hook(d):
+    if d['status'] == 'finished':
+        print('Done downloading, now post-processing ...')
+
+
 def sanitize_filename(filename):
     """
     Remove special characters from filename to ensure it's valid across different file systems.
@@ -92,10 +112,27 @@ def standardize_title(input_title):
 
     return standardized_output.strip()
 
+def get_video_title(video_url):
+    ydl_opts = {
+        'logger': MyLogger(),
+        'progress_hooks': [my_hook],
+        'cookiefile': 'cookies.txt',
+        'quiet': True,  # Suppress output
+    }
+    
+    with yt_dlp.YoutubeDL(ydl_opts) as ydl:
+        # Extract video info using the provided URL
+        video_info = ydl.extract_info(video_url, download=False)
+        
+        # Get the video title
+        video_title = video_info['title']  # Get the video title
+        
+        return video_title
+
 def download_youtube_audio(youtube_url: str, output_dir: str = './download', delete_existing: bool = True, simulate: bool = False) -> str:
     """
     Downloads audio from a YouTube URL and saves it as an MP3 file with specified yt-dlp options.
-    
+
     Args:
         youtube_url (str): URL of the YouTube video.
         output_dir (str): Directory to save the downloaded audio file.
@@ -108,40 +145,40 @@ def download_youtube_audio(youtube_url: str, output_dir: str = './download', del
     if not os.path.exists(output_dir):
         os.makedirs(output_dir)
 
-    r = requests.get(youtube_url)
-    soup = BeautifulSoup(r.text)
+    title = get_video_title(youtube_url)
 
-    link = soup.find_all(name="title")[0]
-    title = str(link)
-    title = title.replace("<title>","")
-    title = title.replace("</title>","")
-        
     audio_file = os.path.join(output_dir, title)
 
     # Remove existing file if requested
     if delete_existing and os.path.exists(audio_file + '.mp3'):
         os.remove(audio_file + '.mp3')
-        
-    # Prepare yt-dlp command
-    command = [
-        'yt-dlp', '-x', youtube_url, '-f', 'bestaudio',
-        '-o', audio_file, '--audio-format', 'mp3', '--restrict-filenames',
-        '--extractor-retries', '10', '--force-overwrites', '--username', 'oauth2', '--password', '', '-v'
-    ]
-    
+
+    # Prepare yt-dlp options
+    ydl_opts = {
+        'logger': MyLogger(),
+        'progress_hooks': [my_hook],
+        'format': 'bestaudio',
+        'outtmpl': audio_file,
+        'postprocessors': [{
+            'key': 'FFmpegExtractAudio',
+            'preferredcodec': 'wav',
+        }],
+        'extractor_retries': 10,
+        'force_overwrites': True,
+        'cookiefile': 'cookies.txt', 
+        'verbose': True,
+    }
+
+
     if simulate:
-        command.append('-s')
+        ydl_opts['simulate'] = True
 
-    # Run the download command
-    process = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True)
+    # Download the audio using yt-dlp
+    with yt_dlp.YoutubeDL(ydl_opts) as ydl:
+        ydl.download([youtube_url])
+
+    return audio_file + '.wav'
 
-    for line in iter(process.stdout.readline, ''):
-        print(line, end='')  # Print download progress
-        
-    process.stdout.close()
-    process.wait()
-    
-    return audio_file + '.mp3'
 
 
 def handle_file_upload(file):
@@ -149,7 +186,7 @@ def handle_file_upload(file):
     Handle file upload, standardize the filename, change extension to .wav, and copy it to the input folder.
     
     Args:
-        file: Uploaded file object
+        file: Uploaded file object or file path string
     
     Returns:
         tuple: (input_path, formatted_title) or (None, error_message)
@@ -157,9 +194,15 @@ def handle_file_upload(file):
     if file is None:
         return None, "No file uploaded"
 
-    filename = os.path.basename(file.name)
-    
-    # Remove the original extension and standardize the filename
+    # Check if 'file' is an instance of a file object or a string
+    if isinstance(file, str):
+        filename = os.path.basename(file)  # If it's a string, use it directly
+        file_path = file  # The string itself is the file path
+    else:
+        filename = os.path.basename(file.name)  # If it's a file object
+        file_path = file.name
+
+
     formatted_title = standardize_title(os.path.splitext(filename)[0])  # Removing extension
     formatted_title = sanitize_filename(formatted_title.strip())
 
@@ -168,7 +211,7 @@ def handle_file_upload(file):
     os.makedirs(os.path.dirname(input_path), exist_ok=True)
 
     # Convert the input file to .wav if it's not already
-    audio = AudioSegment.from_file(file.name)
+    audio = AudioSegment.from_file(file_path)
     audio.export(input_path, format="wav")
 
     return input_path, formatted_title

cookies.txt

@@ -0,0 +1,24 @@
+# Netscape HTTP Cookie File
+# http://curl.haxx.se/rfc/cookie_spec.html
+# This is a generated file!  Do not edit.
+
+.youtube.com	TRUE	/	TRUE	0	YSC	7nJSJzFobwo
+.youtube.com	TRUE	/	TRUE	1761392077	__Secure-YEC	Cgtpckt1M25BSHZ1NCiF9s-3BjIiCgJGUhIcEhgSFhMLFBUWFwwYGRobHB0eHw4PIBAREiEgVA%3D%3D
+.youtube.com	TRUE	/	TRUE	1764879120	PREF	f4=4000000&f6=40000000&tz=Europe.Paris&f5=20000&f7=100
+.youtube.com	TRUE	/	FALSE	1764876802	HSID	AuFfOOSaK_BDmLYBM
+.youtube.com	TRUE	/	TRUE	1764876802	SSID	AJufaIAV7L5xI7IZG
+.youtube.com	TRUE	/	FALSE	1764876802	APISID	-K6RGlr9VwFUMBl9/AtYYpBGnmMzDFOfFb
+.youtube.com	TRUE	/	TRUE	1764876802	SAPISID	-eI2tTs085wmtRwX/AuP4Ex_mjgNsvPdhz
+.youtube.com	TRUE	/	TRUE	1764876802	__Secure-1PAPISID	-eI2tTs085wmtRwX/AuP4Ex_mjgNsvPdhz
+.youtube.com	TRUE	/	TRUE	1764876802	__Secure-3PAPISID	-eI2tTs085wmtRwX/AuP4Ex_mjgNsvPdhz
+.youtube.com	TRUE	/	TRUE	1764876802	LOGIN_INFO	AFmmF2swRgIhAMj8ifOeygBSr75_78gkYO4MBYuQGOatR6JgShhq7OqbAiEAmTb0D_nGF2knbj5BhNbrFAXldPZpofmO-C_233c-OEs:QUQ3MjNmeE9ES0pjT3R6TGtZaERHOGhQamZFX1JkZzB2cUE3VmVPV0kzTmk3QVZQNTE4T3ozMG5XZXpxMm1qd3RscS1zdk92b0t1LU96dnV3X3Y5MGN5dnBRV28weXcyV1hsRGEyX1A4RWkza1ZzVkhyc0h3VkxILVBhZjJDSnB6YWt2czNFcVFkUE1OQlRoY20zNnlmdE1LOHF2ckROSHd3
+.youtube.com	TRUE	/	FALSE	1764876802	SID	g.a000pghKW7HghO8DASpX11UdgeavxJzTkvkX0EEb_Tq3dG4M01z8F6EqmjMaxmCKBnVVeubZIgACgYKAYkSARESFQHGX2Mi89Z1jk1Q6ttIiCfGTAwemBoVAUF8yKqtk2by-PRBc4le2Q-T9djF0076
+.youtube.com	TRUE	/	TRUE	1764876802	__Secure-1PSID	g.a000pghKW7HghO8DASpX11UdgeavxJzTkvkX0EEb_Tq3dG4M01z8m-E1zyyDeV7mbmGrJECBLQACgYKAXUSARESFQHGX2MictGrh1xU6oa65-Th4ABj7BoVAUF8yKpJ6cGV1gdUs0WqSCOL2uYY0076
+.youtube.com	TRUE	/	TRUE	1764876802	__Secure-3PSID	g.a000pghKW7HghO8DASpX11UdgeavxJzTkvkX0EEb_Tq3dG4M01z8jr2VQE5yWKqnaouJW3XchQACgYKAYwSARESFQHGX2MiNzYWKdfei_hqLUOJWdWY1BoVAUF8yKr3WvYxsYfDW6msQqi4arWF0076
+.youtube.com	TRUE	/	TRUE	1745870521	VISITOR_PRIVACY_METADATA	CgJGUhIcEhgSFhMLFBUWFwwYGRobHB0eHw4PIBAREiEgNw%3D%3D
+.youtube.com	TRUE	/	TRUE	1746128009	NID	518=amrdfkufp86-FwA22dILcalS8OtzR48UdrK9cx5CqoZUwpYhd5G2lt8pY2qqQ4_6Ommx_y1gmjVKaOijgoStqmyhXD7rEbhwpUHtIndYZ713pnBaCaPDWmyNBYNeta6zOJtv6uLYCPzK_fA0UjpNGJHXqTErGZjRYIOtZN9c68FDxwWz3EBwyUn3Gsnkfwq7kHSgC3QNLGUSjY24ndwEnnaM8opinQ0d9pWcMbkZCrpZ2e9W4mnOdK_E29WqcKz7czlZ738WFwspYW-s1e62TvaUbsm3G9R1K7BG4Q
+.youtube.com	TRUE	/	TRUE	1761856347	__Secure-1PSIDTS	sidts-CjIBQT4rX_Sn5aNOtmawkPePf_mf7nCRzOSq5E1N9Eaf54yS2LqlA4UOrs21BRUtT9Dp5RAA
+.youtube.com	TRUE	/	TRUE	1761856347	__Secure-3PSIDTS	sidts-CjIBQT4rX_Sn5aNOtmawkPePf_mf7nCRzOSq5E1N9Eaf54yS2LqlA4UOrs21BRUtT9Dp5RAA
+.youtube.com	TRUE	/	FALSE	1761856599	SIDCC	AKEyXzVGGEoO2_ENgIKoZV9xMDuS02kb8VvffhPnD59YMriuUHwYUmkMtBxr5nR9euTg4mQEJNA
+.youtube.com	TRUE	/	TRUE	1761856599	__Secure-1PSIDCC	AKEyXzV8MMEEcTezw9GXyXnSlODcrGes3gNcmMZRTEUUQMRWQ1TRTCA2H6Fs7lfuBF6yELgQJz0
+.youtube.com	TRUE	/	TRUE	1761856599	__Secure-3PSIDCC	AKEyXzWkWpktP9VxK07V1AvAm6XODApfSbB1OT3WYrQ0nCvcghfdXzhIZyoHgYrvJswicG_3gg

requirements.txt

@@ -34,4 +34,5 @@ gradio
 transformers 
 datasets 
 accelerate
-spaces
+spaces
+bitsandbytes