Update app.py
Browse files
app.py
CHANGED
@@ -41,20 +41,28 @@ def download_youtube_audio(url):
|
|
41 |
}],
|
42 |
'outtmpl': output_path
|
43 |
}
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
|
|
|
|
|
|
|
|
54 |
|
55 |
@spaces.GPU(duration=60)
|
56 |
def transcribe_audio(file_path):
|
57 |
print(f"Starting transcription of file: {file_path}")
|
|
|
|
|
|
|
|
|
58 |
if file_path.endswith(('.mp4', '.avi', '.mov', '.flv')):
|
59 |
print("Video file detected. Extracting audio...")
|
60 |
try:
|
@@ -65,8 +73,8 @@ def transcribe_audio(file_path):
|
|
65 |
file_path = audio_path
|
66 |
except Exception as e:
|
67 |
print(f"Error extracting audio from video: {e}")
|
68 |
-
|
69 |
-
|
70 |
output_file = generate_unique_filename('.json')
|
71 |
command = [
|
72 |
"insanely-fast-whisper",
|
@@ -80,25 +88,33 @@ def transcribe_audio(file_path):
|
|
80 |
print(f"Executing command: {' '.join(command)}")
|
81 |
try:
|
82 |
result = subprocess.run(command, check=True, capture_output=True, text=True)
|
|
|
|
|
83 |
except subprocess.CalledProcessError as e:
|
84 |
print(f"Error running insanely-fast-whisper: {e}")
|
85 |
-
|
86 |
-
|
|
|
|
|
|
|
|
|
87 |
try:
|
88 |
with open(output_file, "r") as f:
|
89 |
transcription = json.load(f)
|
90 |
except json.JSONDecodeError as e:
|
91 |
print(f"Error decoding JSON: {e}")
|
92 |
-
|
93 |
-
|
|
|
|
|
94 |
if "text" in transcription:
|
95 |
result = transcription["text"]
|
96 |
else:
|
97 |
result = " ".join([chunk["text"] for chunk in transcription.get("chunks", [])])
|
98 |
-
|
99 |
cleanup_file(file_path)
|
100 |
cleanup_file(output_file)
|
101 |
-
|
102 |
return result
|
103 |
|
104 |
@spaces.GPU(duration=60)
|
@@ -121,16 +137,23 @@ def process_youtube(url):
|
|
121 |
return "Please enter a YouTube URL.", None
|
122 |
try:
|
123 |
audio_file = download_youtube_audio(url)
|
|
|
|
|
124 |
transcription = transcribe_audio(audio_file)
|
|
|
|
|
125 |
return transcription, None
|
126 |
except Exception as e:
|
127 |
return f"Processing error: {str(e)}", None
|
128 |
finally:
|
129 |
-
|
|
|
130 |
|
131 |
def process_uploaded_video(video_path):
|
132 |
try:
|
133 |
transcription = transcribe_audio(video_path)
|
|
|
|
|
134 |
return transcription, None
|
135 |
except Exception as e:
|
136 |
return f"Processing error: {str(e)}", None
|
|
|
41 |
}],
|
42 |
'outtmpl': output_path
|
43 |
}
|
44 |
+
try:
|
45 |
+
with yt_dlp.YoutubeDL(ydl_opts) as ydl:
|
46 |
+
ydl.download([url])
|
47 |
+
|
48 |
+
if os.path.exists(output_path):
|
49 |
+
print(f"Audio download completed. File saved at: {output_path}")
|
50 |
+
print(f"File size: {os.path.getsize(output_path)} bytes")
|
51 |
+
return output_path
|
52 |
+
else:
|
53 |
+
print(f"Error: File {output_path} not found after download.")
|
54 |
+
return None
|
55 |
+
except Exception as e:
|
56 |
+
print(f"Error downloading YouTube audio: {e}")
|
57 |
+
return None
|
58 |
|
59 |
@spaces.GPU(duration=60)
|
60 |
def transcribe_audio(file_path):
|
61 |
print(f"Starting transcription of file: {file_path}")
|
62 |
+
if not os.path.exists(file_path):
|
63 |
+
print(f"Error: File {file_path} does not exist.")
|
64 |
+
return None
|
65 |
+
|
66 |
if file_path.endswith(('.mp4', '.avi', '.mov', '.flv')):
|
67 |
print("Video file detected. Extracting audio...")
|
68 |
try:
|
|
|
73 |
file_path = audio_path
|
74 |
except Exception as e:
|
75 |
print(f"Error extracting audio from video: {e}")
|
76 |
+
return None
|
77 |
+
|
78 |
output_file = generate_unique_filename('.json')
|
79 |
command = [
|
80 |
"insanely-fast-whisper",
|
|
|
88 |
print(f"Executing command: {' '.join(command)}")
|
89 |
try:
|
90 |
result = subprocess.run(command, check=True, capture_output=True, text=True)
|
91 |
+
print(f"Standard output: {result.stdout}")
|
92 |
+
print(f"Error output: {result.stderr}")
|
93 |
except subprocess.CalledProcessError as e:
|
94 |
print(f"Error running insanely-fast-whisper: {e}")
|
95 |
+
print(f"Standard output: {e.stdout}")
|
96 |
+
print(f"Error output: {e.stderr}")
|
97 |
+
cleanup_file(file_path)
|
98 |
+
cleanup_file(output_file)
|
99 |
+
return None
|
100 |
+
|
101 |
try:
|
102 |
with open(output_file, "r") as f:
|
103 |
transcription = json.load(f)
|
104 |
except json.JSONDecodeError as e:
|
105 |
print(f"Error decoding JSON: {e}")
|
106 |
+
cleanup_file(file_path)
|
107 |
+
cleanup_file(output_file)
|
108 |
+
return None
|
109 |
+
|
110 |
if "text" in transcription:
|
111 |
result = transcription["text"]
|
112 |
else:
|
113 |
result = " ".join([chunk["text"] for chunk in transcription.get("chunks", [])])
|
114 |
+
|
115 |
cleanup_file(file_path)
|
116 |
cleanup_file(output_file)
|
117 |
+
|
118 |
return result
|
119 |
|
120 |
@spaces.GPU(duration=60)
|
|
|
137 |
return "Please enter a YouTube URL.", None
|
138 |
try:
|
139 |
audio_file = download_youtube_audio(url)
|
140 |
+
if audio_file is None:
|
141 |
+
return "Error downloading YouTube audio.", None
|
142 |
transcription = transcribe_audio(audio_file)
|
143 |
+
if transcription is None:
|
144 |
+
return "Error transcribing audio.", None
|
145 |
return transcription, None
|
146 |
except Exception as e:
|
147 |
return f"Processing error: {str(e)}", None
|
148 |
finally:
|
149 |
+
if audio_file:
|
150 |
+
cleanup_file(audio_file)
|
151 |
|
152 |
def process_uploaded_video(video_path):
|
153 |
try:
|
154 |
transcription = transcribe_audio(video_path)
|
155 |
+
if transcription is None:
|
156 |
+
return "Error transcribing video.", None
|
157 |
return transcription, None
|
158 |
except Exception as e:
|
159 |
return f"Processing error: {str(e)}", None
|