Update app.py
Browse files
app.py
CHANGED
|
@@ -46,134 +46,50 @@ if not OPENAI_API_KEY:
|
|
| 46 |
|
| 47 |
openai.api_key = OPENAI_API_KEY
|
| 48 |
|
| 49 |
-
|
| 50 |
-
try:
|
| 51 |
-
result = urlparse(url)
|
| 52 |
-
return all([result.scheme, result.netloc])
|
| 53 |
-
except ValueError:
|
| 54 |
-
logger.error(f"Invalid URL: {url}")
|
| 55 |
-
return False
|
| 56 |
-
|
| 57 |
-
def download_media(url):
|
| 58 |
-
logger.info(f"Attempting to download media from URL: {url}")
|
| 59 |
-
try:
|
| 60 |
-
if "youtube.com" in url or "youtu.be" in url:
|
| 61 |
-
yt = YouTube(url)
|
| 62 |
-
stream = yt.streams.filter(progressive=True, file_extension='mp4').first()
|
| 63 |
-
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as temp_file:
|
| 64 |
-
stream.download(output_path=os.path.dirname(temp_file.name), filename=temp_file.name)
|
| 65 |
-
logger.info(f"YouTube video downloaded: {temp_file.name}")
|
| 66 |
-
return temp_file.name
|
| 67 |
-
else:
|
| 68 |
-
response = requests.get(url)
|
| 69 |
-
content_type = response.headers.get('content-type', '')
|
| 70 |
-
if 'video' in content_type:
|
| 71 |
-
suffix = '.mp4'
|
| 72 |
-
elif 'audio' in content_type:
|
| 73 |
-
suffix = '.mp3'
|
| 74 |
-
else:
|
| 75 |
-
suffix = ''
|
| 76 |
-
with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as temp_file:
|
| 77 |
-
temp_file.write(response.content)
|
| 78 |
-
logger.info(f"Media downloaded: {temp_file.name}")
|
| 79 |
-
return temp_file.name
|
| 80 |
-
except Exception as e:
|
| 81 |
-
logger.error(f"Error downloading media: {str(e)}")
|
| 82 |
-
raise
|
| 83 |
-
|
| 84 |
-
def extract_audio(file_path):
|
| 85 |
-
logger.info(f"Extracting audio from video: {file_path}")
|
| 86 |
-
try:
|
| 87 |
-
if VideoFileClip is None:
|
| 88 |
-
raise ImportError("VideoFileClip is not available. Cannot extract audio.")
|
| 89 |
-
video = VideoFileClip(file_path)
|
| 90 |
-
audio = video.audio
|
| 91 |
-
audio_file = tempfile.NamedTemporaryFile(delete=False, suffix=".wav")
|
| 92 |
-
audio.write_audiofile(audio_file.name)
|
| 93 |
-
video.close()
|
| 94 |
-
audio.close()
|
| 95 |
-
logger.info(f"Audio extracted: {audio_file.name}")
|
| 96 |
-
return audio_file.name
|
| 97 |
-
except Exception as e:
|
| 98 |
-
logger.error(f"Error extracting audio: {str(e)}")
|
| 99 |
-
raise
|
| 100 |
-
|
| 101 |
-
def transcribe_audio(file_path):
|
| 102 |
-
logger.info(f"Transcribing audio: {file_path}")
|
| 103 |
-
try:
|
| 104 |
-
with open(file_path, "rb") as audio_file:
|
| 105 |
-
transcript = openai.Audio.transcribe("whisper-1", audio_file)
|
| 106 |
-
logger.info("Transcription completed successfully")
|
| 107 |
-
return transcript["text"]
|
| 108 |
-
except Exception as e:
|
| 109 |
-
logger.error(f"Error during transcription: {str(e)}")
|
| 110 |
-
raise
|
| 111 |
-
|
| 112 |
-
def process_media(contents, filename, url):
|
| 113 |
-
logger.info("Starting media processing")
|
| 114 |
-
try:
|
| 115 |
-
if contents:
|
| 116 |
-
content_type, content_string = contents.split(',')
|
| 117 |
-
decoded = base64.b64decode(content_string)
|
| 118 |
-
suffix = os.path.splitext(filename)[1]
|
| 119 |
-
with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as temp_file:
|
| 120 |
-
temp_file.write(decoded)
|
| 121 |
-
temp_file_path = temp_file.name
|
| 122 |
-
logger.info(f"File uploaded: {temp_file_path}")
|
| 123 |
-
elif url:
|
| 124 |
-
temp_file_path = download_media(url)
|
| 125 |
-
else:
|
| 126 |
-
logger.error("No input provided")
|
| 127 |
-
raise ValueError("No input provided")
|
| 128 |
-
|
| 129 |
-
if temp_file_path.lower().endswith(('.mp4', '.avi', '.mov', '.flv', '.wmv')):
|
| 130 |
-
logger.info("Video file detected, extracting audio")
|
| 131 |
-
audio_file_path = extract_audio(temp_file_path)
|
| 132 |
-
transcript = transcribe_audio(audio_file_path)
|
| 133 |
-
os.unlink(audio_file_path)
|
| 134 |
-
else:
|
| 135 |
-
logger.info("Audio file detected, transcribing directly")
|
| 136 |
-
transcript = transcribe_audio(temp_file_path)
|
| 137 |
-
|
| 138 |
-
os.unlink(temp_file_path)
|
| 139 |
-
return transcript
|
| 140 |
-
except Exception as e:
|
| 141 |
-
logger.error(f"Error in process_media: {str(e)}")
|
| 142 |
-
raise
|
| 143 |
|
| 144 |
app.layout = dbc.Container([
|
| 145 |
-
|
| 146 |
-
|
| 147 |
-
|
| 148 |
-
dcc.Upload(
|
| 149 |
-
id='upload-media',
|
| 150 |
-
children=html.Div([
|
| 151 |
-
'Drag and Drop or ',
|
| 152 |
-
html.A('Select Audio/Video File')
|
| 153 |
-
]),
|
| 154 |
-
style={
|
| 155 |
-
'width': '100%',
|
| 156 |
-
'height': '60px',
|
| 157 |
-
'lineHeight': '60px',
|
| 158 |
-
'borderWidth': '1px',
|
| 159 |
-
'borderStyle': 'dashed',
|
| 160 |
-
'borderRadius': '5px',
|
| 161 |
-
'textAlign': 'center',
|
| 162 |
-
'margin': '10px'
|
| 163 |
-
},
|
| 164 |
-
multiple=False
|
| 165 |
-
),
|
| 166 |
-
html.Div(id='file-info', className="mt-2"),
|
| 167 |
-
dbc.Input(id="media-url", type="text", placeholder="Enter audio/video URL or YouTube link", className="my-3"),
|
| 168 |
-
dbc.Button("Transcribe", id="transcribe-button", color="primary", className="w-100 mb-3"),
|
| 169 |
-
dbc.Spinner(html.Div(id="transcription-output", className="mt-3")),
|
| 170 |
-
html.Div(id="progress-indicator", className="text-center mt-3"),
|
| 171 |
-
dbc.Button("Download Transcript", id="download-button", color="secondary", className="w-100 mt-3", style={'display': 'none'}),
|
| 172 |
-
dcc.Download(id="download-transcript"),
|
| 173 |
-
dcc.Interval(id='progress-interval', interval=500, n_intervals=0, disabled=True)
|
| 174 |
])
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 175 |
])
|
| 176 |
-
])
|
| 177 |
|
| 178 |
@app.callback(
|
| 179 |
Output("file-info", "children"),
|
|
@@ -190,14 +106,16 @@ def update_file_info(filename, last_modified):
|
|
| 190 |
Output("download-button", "style"),
|
| 191 |
Output("progress-indicator", "children"),
|
| 192 |
Output("progress-interval", "disabled"),
|
|
|
|
| 193 |
Input("transcribe-button", "n_clicks"),
|
| 194 |
Input("progress-interval", "n_intervals"),
|
| 195 |
State("upload-media", "contents"),
|
| 196 |
State("upload-media", "filename"),
|
| 197 |
State("media-url", "value"),
|
|
|
|
| 198 |
prevent_initial_call=True
|
| 199 |
)
|
| 200 |
-
def update_transcription(n_clicks, n_intervals, contents, filename, url):
|
| 201 |
ctx = callback_context
|
| 202 |
if ctx.triggered_id == "transcribe-button":
|
| 203 |
if not contents and not url:
|
|
@@ -212,40 +130,43 @@ def update_transcription(n_clicks, n_intervals, contents, filename, url):
|
|
| 212 |
|
| 213 |
thread = threading.Thread(target=transcribe)
|
| 214 |
thread.start()
|
| 215 |
-
return html.Div("Processing..."), {'display': 'none'}, "", False
|
| 216 |
|
| 217 |
elif ctx.triggered_id == "progress-interval":
|
|
|
|
|
|
|
| 218 |
dots = "." * (n_intervals % 4)
|
| 219 |
-
return html.Div("Processing" + dots), {'display': 'none'}, "", False
|
| 220 |
|
| 221 |
thread = threading.current_thread()
|
| 222 |
if hasattr(thread, 'result'):
|
| 223 |
transcript = thread.result
|
| 224 |
if transcript and not transcript.startswith("An error occurred"):
|
| 225 |
logger.info("Transcription successful")
|
| 226 |
-
return
|
| 227 |
-
dbc.CardBody([
|
| 228 |
-
html.H5("Transcription Result"),
|
| 229 |
-
html.Pre(transcript, style={"white-space": "pre-wrap", "word-wrap": "break-word"})
|
| 230 |
-
])
|
| 231 |
-
]), {'display': 'block'}, "", True
|
| 232 |
else:
|
| 233 |
logger.error(f"Transcription failed: {transcript}")
|
| 234 |
-
return transcript, {'display': 'none'}, "", True
|
| 235 |
|
| 236 |
-
return dash.no_update, dash.no_update, dash.no_update, dash.no_update
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 237 |
|
| 238 |
@app.callback(
|
| 239 |
Output("download-transcript", "data"),
|
| 240 |
Input("download-button", "n_clicks"),
|
| 241 |
-
State("transcription-
|
| 242 |
prevent_initial_call=True
|
| 243 |
)
|
| 244 |
-
def download_transcript(n_clicks,
|
| 245 |
-
if not
|
| 246 |
raise PreventUpdate
|
| 247 |
-
|
| 248 |
-
transcript = transcription_output['props']['children'][0]['props']['children'][1]['props']['children']
|
| 249 |
return dict(content=transcript, filename="transcript.txt")
|
| 250 |
|
| 251 |
if __name__ == '__main__':
|
|
|
|
| 46 |
|
| 47 |
openai.api_key = OPENAI_API_KEY
|
| 48 |
|
| 49 |
+
# ... (keep all the helper functions as they are) ...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
|
| 51 |
app.layout = dbc.Container([
|
| 52 |
+
dbc.Row([
|
| 53 |
+
dbc.Col([
|
| 54 |
+
html.H1("Audio/Video Transcription App", className="text-center my-4"),
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
])
|
| 56 |
+
]),
|
| 57 |
+
dbc.Row([
|
| 58 |
+
dbc.Col([
|
| 59 |
+
dbc.Card([
|
| 60 |
+
dbc.CardBody([
|
| 61 |
+
dcc.Upload(
|
| 62 |
+
id='upload-media',
|
| 63 |
+
children=html.Div([
|
| 64 |
+
'Drag and Drop or ',
|
| 65 |
+
html.A('Select Audio/Video File')
|
| 66 |
+
]),
|
| 67 |
+
style={
|
| 68 |
+
'width': '100%',
|
| 69 |
+
'height': '60px',
|
| 70 |
+
'lineHeight': '60px',
|
| 71 |
+
'borderWidth': '1px',
|
| 72 |
+
'borderStyle': 'dashed',
|
| 73 |
+
'borderRadius': '5px',
|
| 74 |
+
'textAlign': 'center',
|
| 75 |
+
'margin': '10px'
|
| 76 |
+
},
|
| 77 |
+
multiple=False
|
| 78 |
+
),
|
| 79 |
+
html.Div(id='file-info', className="mt-2"),
|
| 80 |
+
dbc.Input(id="media-url", type="text", placeholder="Enter audio/video URL or YouTube link", className="my-3"),
|
| 81 |
+
dbc.Button("Transcribe", id="transcribe-button", color="primary", className="w-100 mb-3"),
|
| 82 |
+
dbc.Spinner(html.Div(id="transcription-output", className="mt-3")),
|
| 83 |
+
html.Div(id="progress-indicator", className="text-center mt-3"),
|
| 84 |
+
dbc.Button("Download Transcript", id="download-button", color="secondary", className="w-100 mt-3", style={'display': 'none'}),
|
| 85 |
+
dcc.Download(id="download-transcript"),
|
| 86 |
+
dcc.Store(id="transcription-store"),
|
| 87 |
+
dcc.Interval(id='progress-interval', interval=500, n_intervals=0, disabled=True)
|
| 88 |
+
])
|
| 89 |
+
])
|
| 90 |
+
], width=12)
|
| 91 |
])
|
| 92 |
+
], fluid=True)
|
| 93 |
|
| 94 |
@app.callback(
|
| 95 |
Output("file-info", "children"),
|
|
|
|
| 106 |
Output("download-button", "style"),
|
| 107 |
Output("progress-indicator", "children"),
|
| 108 |
Output("progress-interval", "disabled"),
|
| 109 |
+
Output("transcription-store", "data"),
|
| 110 |
Input("transcribe-button", "n_clicks"),
|
| 111 |
Input("progress-interval", "n_intervals"),
|
| 112 |
State("upload-media", "contents"),
|
| 113 |
State("upload-media", "filename"),
|
| 114 |
State("media-url", "value"),
|
| 115 |
+
State("transcription-store", "data"),
|
| 116 |
prevent_initial_call=True
|
| 117 |
)
|
| 118 |
+
def update_transcription(n_clicks, n_intervals, contents, filename, url, stored_transcript):
|
| 119 |
ctx = callback_context
|
| 120 |
if ctx.triggered_id == "transcribe-button":
|
| 121 |
if not contents and not url:
|
|
|
|
| 130 |
|
| 131 |
thread = threading.Thread(target=transcribe)
|
| 132 |
thread.start()
|
| 133 |
+
return html.Div("Processing..."), {'display': 'none'}, "", False, None
|
| 134 |
|
| 135 |
elif ctx.triggered_id == "progress-interval":
|
| 136 |
+
if stored_transcript:
|
| 137 |
+
return display_transcript(stored_transcript), {'display': 'block'}, "", True, stored_transcript
|
| 138 |
dots = "." * (n_intervals % 4)
|
| 139 |
+
return html.Div("Processing" + dots), {'display': 'none'}, "", False, None
|
| 140 |
|
| 141 |
thread = threading.current_thread()
|
| 142 |
if hasattr(thread, 'result'):
|
| 143 |
transcript = thread.result
|
| 144 |
if transcript and not transcript.startswith("An error occurred"):
|
| 145 |
logger.info("Transcription successful")
|
| 146 |
+
return display_transcript(transcript), {'display': 'block'}, "", True, transcript
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 147 |
else:
|
| 148 |
logger.error(f"Transcription failed: {transcript}")
|
| 149 |
+
return html.Div(transcript), {'display': 'none'}, "", True, None
|
| 150 |
|
| 151 |
+
return dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update
|
| 152 |
+
|
| 153 |
+
def display_transcript(transcript):
|
| 154 |
+
return dbc.Card([
|
| 155 |
+
dbc.CardBody([
|
| 156 |
+
html.H5("Transcription Result"),
|
| 157 |
+
html.Pre(transcript, style={"white-space": "pre-wrap", "word-wrap": "break-word"})
|
| 158 |
+
])
|
| 159 |
+
])
|
| 160 |
|
| 161 |
@app.callback(
|
| 162 |
Output("download-transcript", "data"),
|
| 163 |
Input("download-button", "n_clicks"),
|
| 164 |
+
State("transcription-store", "data"),
|
| 165 |
prevent_initial_call=True
|
| 166 |
)
|
| 167 |
+
def download_transcript(n_clicks, transcript):
|
| 168 |
+
if not transcript:
|
| 169 |
raise PreventUpdate
|
|
|
|
|
|
|
| 170 |
return dict(content=transcript, filename="transcript.txt")
|
| 171 |
|
| 172 |
if __name__ == '__main__':
|