Spaces:
Runtime error
Runtime error
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,153 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
from transformers import pipeline
|
| 3 |
+
import torch
|
| 4 |
+
import moviepy.editor as mp
|
| 5 |
+
import os
|
| 6 |
+
from googleapiclient.discovery import build
|
| 7 |
+
from google.oauth2 import credentials
|
| 8 |
+
from google_auth_oauthlib.flow import InstalledAppFlow
|
| 9 |
+
from google.auth.transport.requests import Request
|
| 10 |
+
from googleapiclient.http import MediaIoBaseDownload
|
| 11 |
+
from googleapiclient.errors import HttpError
|
| 12 |
+
import uuid
|
| 13 |
+
|
| 14 |
+
# Device setup
|
| 15 |
+
device = "cpu" # Using CPU to avoid CUDA memory issues
|
| 16 |
+
print(f"Using device: {device}")
|
| 17 |
+
|
| 18 |
+
# Preload models
|
| 19 |
+
try:
|
| 20 |
+
transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-large-v2", device=device)
|
| 21 |
+
summarizer = pipeline("summarization", model="facebook/bart-large-cnn", device=device)
|
| 22 |
+
except Exception as e:
|
| 23 |
+
print(f"Error loading models: {e}. Make sure you have internet connectivity and sufficient resources.")
|
| 24 |
+
transcriber = None
|
| 25 |
+
summarizer = None
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
# Google Drive API setup
|
| 30 |
+
SCOPES = ['https://www.googleapis.com/auth/drive.readonly']
|
| 31 |
+
CLIENT_SECRET_FILE = 'client_secrets.json' # Relative path
|
| 32 |
+
APPLICATION_NAME = 'MyMeetNotes'
|
| 33 |
+
|
| 34 |
+
def get_credentials():
|
| 35 |
+
if not os.path.exists(CLIENT_SECRET_FILE):
|
| 36 |
+
raise FileNotFoundError(
|
| 37 |
+
"The 'client_secrets.json' file is missing. Please ensure it’s in the same directory as the script. See: "
|
| 38 |
+
"https://developers.google.com/drive/api/v3/quickstart/python"
|
| 39 |
+
)
|
| 40 |
+
creds = None
|
| 41 |
+
if os.path.exists('token.json'):
|
| 42 |
+
creds = credentials.Credentials.from_authorized_user_file('token.json', SCOPES)
|
| 43 |
+
if not creds or not creds.valid:
|
| 44 |
+
if creds and creds.expired and creds.refresh_token:
|
| 45 |
+
creds.refresh(Request())
|
| 46 |
+
else:
|
| 47 |
+
flow = InstalledAppFlow.from_client_secrets_file(CLIENT_SECRET_FILE, SCOPES)
|
| 48 |
+
creds = flow.run_local_server(port=0)
|
| 49 |
+
with open('token.json', 'w') as token:
|
| 50 |
+
token.write(creds.to_json())
|
| 51 |
+
return creds
|
| 52 |
+
|
| 53 |
+
def list_google_meet_recordings():
|
| 54 |
+
try:
|
| 55 |
+
creds = get_credentials()
|
| 56 |
+
service = build('drive', 'v3', credentials=creds)
|
| 57 |
+
results = service.files().list(
|
| 58 |
+
pageSize=10, fields="nextPageToken, files(id, name, mimeType)",
|
| 59 |
+
q="name contains 'Meeting recording' and mimeType contains 'video/'").execute()
|
| 60 |
+
items = results.get('files', [])
|
| 61 |
+
if not items:
|
| 62 |
+
return ["No Google Meet recordings found."]
|
| 63 |
+
return [f"{item['name']} (ID: {item['id']})" for item in items]
|
| 64 |
+
except HttpError as error:
|
| 65 |
+
return [f"An error occurred: {error}"]
|
| 66 |
+
except Exception as e:
|
| 67 |
+
return [f"An unexpected error occurred: {e}"]
|
| 68 |
+
|
| 69 |
+
def download_google_meet_recording(file_id):
|
| 70 |
+
try:
|
| 71 |
+
creds = get_credentials()
|
| 72 |
+
service = build('drive', 'v3', credentials=creds)
|
| 73 |
+
file_name = f"temp_recording_{file_id}_{uuid.uuid4()}.mp4" # removed os.path.join(os.getcwd(), ...) for simplicity in local usage
|
| 74 |
+
print(f"Downloading file ID: {file_id} to {file_name}")
|
| 75 |
+
|
| 76 |
+
request = service.files().get_media(fileId=file_id)
|
| 77 |
+
with open(file_name, 'wb') as fh:
|
| 78 |
+
downloader = MediaIoBaseDownload(fh, request)
|
| 79 |
+
done = False
|
| 80 |
+
while not done:
|
| 81 |
+
status, done = downloader.next_chunk()
|
| 82 |
+
print(f"Download progress: {int(status.progress() * 100)}%")
|
| 83 |
+
|
| 84 |
+
if os.path.exists(file_name):
|
| 85 |
+
print(f"Download complete. File saved as: {file_name}")
|
| 86 |
+
return file_name
|
| 87 |
+
else:
|
| 88 |
+
return f"Error: File '{file_name}' not found after download."
|
| 89 |
+
except HttpError as error:
|
| 90 |
+
return f"HTTP error during download: {error}"
|
| 91 |
+
except Exception as e:
|
| 92 |
+
return f"Unexpected error during download: {e}"
|
| 93 |
+
|
| 94 |
+
def generate_meeting_notes(file_id):
|
| 95 |
+
if transcriber is None or summarizer is None:
|
| 96 |
+
return "Error: Models failed to load. Check your internet connection and resources."
|
| 97 |
+
try:
|
| 98 |
+
download_result = download_google_meet_recording(file_id)
|
| 99 |
+
if isinstance(download_result, str) and "Error" in download_result:
|
| 100 |
+
return download_result
|
| 101 |
+
file_name = download_result
|
| 102 |
+
|
| 103 |
+
if not os.path.exists(file_name):
|
| 104 |
+
return f"Error: Downloaded file '{file_name}' not found."
|
| 105 |
+
|
| 106 |
+
print(f"Extracting audio from {file_name}")
|
| 107 |
+
video = mp.VideoFileClip(file_name)
|
| 108 |
+
audio_file = f"temp_audio_{uuid.uuid4()}.wav"
|
| 109 |
+
video.audio.write_audiofile(audio_file)
|
| 110 |
+
print("Audio extraction complete.")
|
| 111 |
+
|
| 112 |
+
print("Starting transcription...")
|
| 113 |
+
transcription = transcriber(audio_file)["text"]
|
| 114 |
+
print("Transcription complete.")
|
| 115 |
+
|
| 116 |
+
print("Starting summarization...")
|
| 117 |
+
summary = summarizer(transcription, max_length=1024, min_length=100, do_sample=False)
|
| 118 |
+
meeting_notes = summary[0]["summary_text"]
|
| 119 |
+
print("Summarization complete.")
|
| 120 |
+
|
| 121 |
+
return meeting_notes
|
| 122 |
+
|
| 123 |
+
except Exception as e:
|
| 124 |
+
return f"Error in processing: {str(e)}"
|
| 125 |
+
finally:
|
| 126 |
+
try:
|
| 127 |
+
if 'video' in locals() and video: # Check if video exists
|
| 128 |
+
video.close()
|
| 129 |
+
if 'audio_file' in locals() and os.path.exists(audio_file):
|
| 130 |
+
os.remove(audio_file)
|
| 131 |
+
if 'file_name' in locals() and os.path.exists(file_name): # Check if file_name exists
|
| 132 |
+
os.remove(file_name)
|
| 133 |
+
except Exception as cleanup_error:
|
| 134 |
+
print(f"Cleanup error: {cleanup_error}")
|
| 135 |
+
|
| 136 |
+
def launch_gradio():
|
| 137 |
+
with gr.Blocks() as iface:
|
| 138 |
+
gr.Markdown("# Automatic Meeting Notes Generator from Google Meet")
|
| 139 |
+
with gr.Row():
|
| 140 |
+
with gr.Column():
|
| 141 |
+
file_id_input = gr.Textbox(label="Google Meet Recording File ID", placeholder="Enter the Google Drive File ID")
|
| 142 |
+
generate_button = gr.Button("Generate Notes")
|
| 143 |
+
recording_list_button = gr.Button("List Recordings")
|
| 144 |
+
with gr.Column():
|
| 145 |
+
output_text = gr.Textbox(label="Meeting Notes")
|
| 146 |
+
recording_list = gr.Textbox(label="Available Recordings")
|
| 147 |
+
|
| 148 |
+
generate_button.click(fn=generate_meeting_notes, inputs=file_id_input, outputs=output_text)
|
| 149 |
+
recording_list_button.click(fn=list_google_meet_recordings, inputs=None, outputs=recording_list)
|
| 150 |
+
iface.launch(debug=True)
|
| 151 |
+
|
| 152 |
+
if __name__ == "__main__":
|
| 153 |
+
launch_gradio()
|