varun324242's picture
Upload folder using huggingface_hub
fe2a0f2 verified
import pandas as pd
import time
from groq import Groq
import os
import csv
from tqdm import tqdm
import logging
from datetime import datetime
import json
import sys
import requests
import aiohttp
import asyncio
import google.auth
from google.oauth2.credentials import Credentials
from google_auth_oauthlib.flow import InstalledAppFlow
from google.auth.transport.requests import Request
from google.oauth2 import service_account
from googleapiclient.discovery import build
from googleapiclient.http import MediaFileUpload, MediaIoBaseDownload
import io
# OAuth 2.0 credentials
CLIENT_ID = "483287191355-udtleajik8ko1o2n03fqmimuu47n3hba.apps.googleusercontent.com"
CLIENT_SECRET = "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"
SCOPES = ['https://www.googleapis.com/auth/drive.file']
def authenticate_google():
"""Authenticate with Google Drive using OAuth 2.0"""
creds = None
# Load credentials from client_secret.json if exists
if os.path.exists('client_secret.json'):
flow = InstalledAppFlow.from_client_secrets_file(
'client_secret.json', SCOPES)
creds = flow.run_local_server(port=0)
else:
# Create credentials manually if client_secret.json not found
flow = InstalledAppFlow.from_client_config(
{
"installed": {
"client_id": CLIENT_ID,
"client_secret": CLIENT_SECRET,
"redirect_uris": ["urn:ietf:wg:oauth:2.0:oob"],
"auth_uri": "https://accounts.google.com/o/oauth2/auth",
"token_uri": "https://oauth2.googleapis.com/token"
}
},
SCOPES
)
creds = flow.run_local_server(port=0)
# Save credentials
with open('token.json', 'w') as token:
token.write(creds.to_json())
return creds
def mount_drive():
"""Mount Google Drive with authentication"""
try:
# Authenticate
creds = authenticate_google()
# Build drive service
service = build('drive', 'v3', credentials=creds)
logging.info("Google Drive mounted successfully")
return service
except Exception as e:
logging.error(f"Error mounting drive: {str(e)}")
raise
def setup_logging():
"""Setup enhanced logging configuration"""
timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')
log_dir = 'logs'
# Create logs directory structure
os.makedirs(f"{log_dir}/api", exist_ok=True)
os.makedirs(f"{log_dir}/process", exist_ok=True)
os.makedirs(f"{log_dir}/error", exist_ok=True)
# Configure logging with multiple handlers
logging.basicConfig(
level=logging.INFO,
format='%(asctime)s | %(levelname)s | %(message)s',
handlers=[
# Console handler
logging.StreamHandler(sys.stdout),
# Main process log
logging.FileHandler(f'{log_dir}/process/process_{timestamp}.log'),
# API interactions log
logging.FileHandler(f'{log_dir}/api/api_{timestamp}.log'),
# Error log
logging.FileHandler(f'{log_dir}/error/error_{timestamp}.log')
]
)
logging.info("""
=================================================================
Starting Message Processing System
=================================================================
Time: {timestamp}
Log Directory: {log_dir}
=================================================================
""")
return timestamp
def initialize_groq():
"""Initialize Groq API client"""
try:
groq_client = Groq(api_key="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")
logging.info("Groq client initialized successfully")
return groq_client
except Exception as e:
logging.error(f"Failed to initialize Groq client: {str(e)}")
raise
def log_api_details(message_id, original_message, converted_message, processing_time, status):
"""Log detailed API interaction information"""
api_log = {
'message_id': message_id,
'timestamp': datetime.now().strftime('%Y-%m-%d %H:%M:%S.%f'),
'original_message': original_message,
'converted_message': converted_message,
'processing_time_seconds': processing_time,
'status': status
}
logging.debug(f"API Details: {json.dumps(api_log, indent=2)}")
def convert_to_ham_message(groq_client, scam_info, message_id):
"""Convert scam message to legitimate message using Groq with detailed logging"""
start_time = time.time()
try:
logging.info(f"[Message {message_id}] Starting processing")
logging.debug(f"[Message {message_id}] Original message: {scam_info}")
# Special handling for specific message IDs
if message_id in [1589, 1597]:
# Skip API call and return original message for these IDs
processing_time = time.time() - start_time
logging.info(f"[Message {message_id}] Using original message")
log_api_details(
message_id=message_id,
original_message=scam_info,
converted_message=scam_info,
processing_time=processing_time,
status='success'
)
return scam_info, processing_time
prompt = f"""
Convert the following potential scam message into a legitimate, non-fraudulent message
while maintaining similar context but removing any fraudulent elements:
{scam_info}
Generate only the converted message without any additional remarks or characters.
"""
logging.info(f"[Message {message_id}] Sending request to Groq API")
logging.debug(f"[Message {message_id}] Prompt: {prompt}")
# List of models to try in order of preference
models = [
"llama-3.2-90b-text-preview",
"llama-3.2-11b-vision-preview",
"llama-3.2-3b-preview",
"llama-3.2-1b-preview",
"llama3-8b-8192",
"llama-3.1-70b-versatile",
"llama-3.1-8b-instant",
"llama3-70b-8192",
"llama-3.2-90b-text-preview", # Default model
"llama-3.2-90b-vision-preview",
"llama3-groq-70b-8192-tool-use-preview",
"llama3-groq-8b-8192-tool-use-preview",
"llama-guard-3-8b",
"gemma-7b-it",
"gemma2-9b-it"
]
for model in models:
try:
completion = groq_client.chat.completions.create(
messages=[
{
"role": "user",
"content": prompt
}
],
model=model,
temperature=0.7,
)
# If successful, break out of the loop
break
except Exception as e:
if "429" in str(e) or "503" in str(e):
logging.warning(f"API error with model {model}, trying next model...")
continue
else:
raise e
else:
# If we've exhausted all models
raise Exception("All models failed with rate limit or service errors")
processing_time = time.time() - start_time
converted_message = completion.choices[0].message.content.strip()
logging.info(f"[Message {message_id}] Conversion successful using model {model}")
logging.debug(f"""
[Message {message_id}] Conversion details:
- Processing time: {processing_time:.2f} seconds
- Original length: {len(scam_info)}
- Converted length: {len(converted_message)}
- Original message: {scam_info}
- Converted message: {converted_message}
- Model used: {model}
""")
log_api_details(
message_id=message_id,
original_message=scam_info,
converted_message=converted_message,
processing_time=processing_time,
status='success'
)
return converted_message, processing_time
except Exception as e:
error_msg = f"[Message {message_id}] Error in API call: {str(e)}"
logging.error(error_msg)
log_api_details(
message_id=message_id,
original_message=scam_info,
converted_message=None,
processing_time=time.time() - start_time,
status=f'error: {str(e)}'
)
return None, time.time() - start_time
def process_csv(input_file, output_file):
"""Process CSV file with simplified logic to handle only the first 5 messages"""
try:
# Mount Google Drive using OAuth
drive_service = mount_drive()
logging.info("Google Drive mounted successfully")
logging.info(f"Reading input CSV file: {input_file}")
df = pd.read_csv(input_file, encoding='latin-1')
# Only keep the first 5 rows (index 0 to 4)
df = df.iloc[4565:5803]
total_messages = len(df)
logging.info(f"Processing first 5 rows ({total_messages} rows)")
# Check and get the correct column name
if ',crimeaditionalinfo' in df.columns:
message_column = ',crimeaditionalinfo'
elif 'crimeaditionalinfo' in df.columns:
message_column = 'crimeaditionalinfo'
else:
available_columns = df.columns.tolist()
logging.error(f"Required column not found. Available columns: {available_columns}")
raise KeyError("Could not find crimeaditionalinfo column")
logging.info(f"Using column: {message_column}")
# Process only the first 5 messages
for i in range(total_messages):
message_id = i + 1
scam_info = df.iloc[i][message_column]
try:
# Process the message
ham_message, proc_time = convert_to_ham_message(
groq_client, scam_info, message_id
)
# Log success
if ham_message:
logging.info(f"Success for message ID {message_id}")
except Exception as e:
logging.error(f"Error processing message {message_id}: {str(e)}")
logging.info("Processed 5 messages successfully.")
except Exception as e:
logging.error(f"Critical error in process_csv: {str(e)}", exc_info=True)
raise
def generate_summary(category, stats, timestamp, summary_file):
"""Generate and append summary for each category"""
summary = f"""
=================================================================
Category: {category}
Processed at: {timestamp}
=================================================================
Processing Statistics:
- Total Messages: {stats['total']}
- Successfully Processed: {stats['processed']}
- Errors: {stats['errors']}
- Success Rate: {stats['success_rate']:.2f}%
Processing Time:
- Total Runtime: {stats['runtime']:.2f} seconds
- Average Time per Message: {stats['avg_time']:.2f} seconds
Files Generated:
- Output File: {stats['output_file']}
=================================================================
"""
# Append to summary file
with open(summary_file, 'a', encoding='utf-8') as f:
f.write(summary)
logging.info(f"Summary updated for category: {category}")
def setup_drive():
"""Setup Google Drive using OAuth2"""
# Update scopes to match
SCOPES = [
'https://www.googleapis.com/auth/drive.readonly',
'https://www.googleapis.com/auth/drive.file',
'https://www.googleapis.com/auth/drive.install',
'https://www.googleapis.com/auth/userinfo.email',
'https://www.googleapis.com/auth/userinfo.profile',
'https://www.googleapis.com/auth/gmail.readonly',
'openid'
]
try:
# Delete existing token
if os.path.exists('token.json'):
os.remove('token.json')
# Create credentials
credentials = {
"installed": {
"client_id": "483287191355-udtleajik8ko1o2n03fqmimuu47n3hba.apps.googleusercontent.com",
"client_secret": "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",
"auth_uri": "https://accounts.google.com/o/oauth2/auth",
"token_uri": "https://oauth2.googleapis.com/token",
"auth_provider_x509_cert_url": "https://www.googleapis.com/oauth2/v1/certs",
"redirect_uris": ["http://localhost:8080/"]
}
}
# Create flow
flow = InstalledAppFlow.from_client_config(
credentials,
SCOPES
)
# Get credentials
creds = flow.run_local_server(port=8080)
# Save credentials
with open('token.json', 'w') as token:
token.write(creds.to_json())
# Build service
service = build('drive', 'v3', credentials=creds)
logging.info("Google Drive service initialized successfully")
return service
except Exception as e:
logging.error(f"Error setting up Drive service: {str(e)}")
raise
def get_files_in_drive(service):
"""List all files in Google Drive"""
try:
results = service.files().list(
pageSize=1000,
fields="nextPageToken, files(id, name, mimeType)",
q="mimeType='application/vnd.google-apps.spreadsheet' or mimeType='text/csv'"
).execute()
files = results.get('files', [])
logging.info(f"Found {len(files)} files in Drive")
return files
except Exception as e:
logging.error(f"Error listing files: {str(e)}")
raise
def download_file(service, file_id, file_name):
"""Download a file from Google Drive"""
try:
request = service.files().get_media(fileId=file_id)
file = io.BytesIO()
downloader = MediaIoBaseDownload(file, request)
done = False
while done is False:
status, done = downloader.next_chunk()
logging.info(f"Download {int(status.progress() * 100)}%")
file.seek(0)
with open(file_name, 'wb') as f:
f.write(file.read())
logging.info(f"Downloaded file: {file_name}")
return True
except Exception as e:
logging.error(f"Error downloading file {file_name}: {str(e)}")
return False
def upload_file(service, file_path, file_name=None, parent_id=None):
"""Upload a file to Google Drive"""
try:
file_metadata = {
'name': file_name or os.path.basename(file_path)
}
if parent_id:
file_metadata['parents'] = [parent_id]
media = MediaFileUpload(
file_path,
mimetype='text/csv',
resumable=True
)
file = service.files().create(
body=file_metadata,
media_body=media,
fields='id'
).execute()
logging.info(f"Uploaded file: {file_name or os.path.basename(file_path)}")
return file.get('id')
except Exception as e:
logging.error(f"Error uploading file {file_path}: {str(e)}")
return None
def get_category_paths():
"""Get files from Google Drive"""
try:
# Setup drive service
service = setup_drive()
logging.info("Drive service setup complete")
# Define file mappings with correct paths from subcategory_messages
file_mappings = {
# Cyber Attack Dependent Crimes
# Cryptocurrency Crime
"Fraud_CallVishing_messages.csv": "subcategory_messages/Online_Financial_Fraud",
}
# Get all files from Drive
all_files = get_files_in_drive(service)
# Map files to their IDs
files = {}
for file_name, folder_path in file_mappings.items():
query = f"name='{file_name}'"
if folder_path:
# Get folder ID first
folder_results = service.files().list(
q=f"name='{folder_path}' and mimeType='application/vnd.google-apps.folder'",
spaces='drive',
fields='files(id)'
).execute()
folder_id = folder_results.get('files', [])[0]['id'] if folder_results.get('files') else None
if folder_id:
query += f" and '{folder_id}' in parents"
# Search for file
results = service.files().list(
q=query,
spaces='drive',
fields='files(id, name)'
).execute()
files_found = results.get('files', [])
if files_found:
file_id = files_found[0]['id']
files[file_name] = {
'id': file_id,
'name': file_name,
'path': folder_path
}
logging.info(f"Found file: {file_name} in {folder_path} (ID: {file_id})")
else:
logging.warning(f"File not found: {file_name} in {folder_path}")
files[file_name] = None
return files, service
except Exception as e:
logging.error(f"Error accessing Drive files: {str(e)}")
raise
def process_file(service, groq_client, file_info):
"""Process a single file with detailed logging"""
# Ensure file_info has the correct structure
if not all(key in file_info for key in ['name', 'id', 'path']):
logging.error(f"File info is missing required keys: {file_info}")
return
file_name = file_info['name']
file_id = file_info['id']
folder_path = file_info['path']
logging.info(f"""
=================================================================
Starting File Processing
=================================================================
File: {file_name}
Location: {folder_path}
Time: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
=================================================================
""")
try:
# Download file
temp_input = f"temp_input_{file_name}"
if not download_file(service, file_id, temp_input):
raise Exception("Failed to download file")
# Read CSV and take only first 5 rows
df = pd.read_csv(temp_input, encoding='latin-1')
df = df.iloc[4565:5803]
# Check for column name
if 'crimeaditionalinfo' not in df.columns:
logging.error(f"Column 'crimeaditionalinfo' not found. Available columns: {df.columns.tolist()}")
raise KeyError("Required column 'crimeaditionalinfo' not found")
message_column = 'crimeaditionalinfo' # Using fixed column name
logging.info(f"Using column: {message_column}")
total_messages = len(df)
logging.info(f"""
File Statistics:
- Processing first 5 messages
- Columns: {', '.join(df.columns)}
""")
# Setup output
output_name = f"converted4_{file_name}"
processed_count = 0
error_count = 0
# Create output CSV
fieldnames = ['message_id', 'original_message', 'converted_message',
'processing_time', 'model_used', 'timestamp']
with open(output_name, 'w', newline='', encoding='utf-8') as f:
writer = csv.DictWriter(f, fieldnames=fieldnames)
writer.writeheader()
# Process messages
for idx, row in df.iterrows():
try:
message_id = idx + 1
original_message = row[message_column]
# Convert message using Groq
converted_message, proc_time = convert_to_ham_message(
groq_client,
original_message,
message_id
)
if converted_message:
writer.writerow({
'message_id': message_id,
'original_message': original_message,
'converted_message': converted_message,
'processing_time': proc_time,
'model_used': 'groq-mixtral',
'timestamp': datetime.now().strftime('%Y-%m-%d %H:%M:%S')
})
processed_count += 1
logging.info(f"Processed message {message_id} successfully")
else:
error_count += 1
logging.error(f"Failed to process message {message_id}")
except Exception as e:
logging.error(f"Error processing message {message_id}: {str(e)}")
error_count += 1
continue
# Upload processed file back to Drive
try:
# Ensure the parent_id is a folder ID
folder_id = get_folder_id(service, "subcategory_messages")
file_id = upload_file(service, output_name, output_name, folder_id)
logging.info(f"Uploaded converted file to Drive with ID: {file_id}")
except Exception as e:
logging.error(f"Error uploading converted file: {str(e)}")
logging.info(f"""
=================================================================
Completed Processing: {file_name}
Messages Processed Successfully: {processed_count}
Errors: {error_count}
Output File: {output_name}
=================================================================
""")
except Exception as e:
logging.error(f"Error processing file {file_name}: {str(e)}")
raise
finally:
# Cleanup
if os.path.exists(temp_input):
os.remove(temp_input)
if os.path.exists(output_name):
os.remove(output_name)
def get_folder_id(service, folder_name):
"""Retrieve the folder ID for a given folder name"""
folder_query = f"name='{folder_name}' and mimeType='application/vnd.google-apps.folder'"
folder_results = service.files().list(
q=folder_query,
spaces='drive',
fields='files(id, name)'
).execute()
folders = folder_results.get('files', [])
if not folders:
raise FileNotFoundError(f"Folder not found: {folder_name}")
return folders[0]['id']
def get_and_process_files():
"""Get and process files from Google Drive one at a time"""
try:
# Setup drive service and Groq client
service = setup_drive()
groq_client = initialize_groq()
logging.info("Drive service and Groq API initialized successfully")
# Get files and process them
files, _ = get_category_paths()
for file_name, file_info in files.items():
if file_info is None:
logging.warning(f"Skipping {file_name} - Not found")
continue
try:
process_file(service, groq_client, file_info)
except Exception as e:
logging.error(f"Error processing file {file_name}: {str(e)}")
continue
except Exception as e:
logging.error(f"Error in get_and_process_files: {str(e)}")
raise
def main():
# Setup logging
timestamp = setup_logging()
try:
# Process files
get_and_process_files()
except Exception as e:
logging.error(f"Error in main: {str(e)}")
raise
if __name__ == "__main__":
main()