corpus_collection_engine/config.py

"""
Configuration settings for the Corpus Collection Engine
"""

import os
from pathlib import Path
from typing import List, Dict

# Project paths
PROJECT_ROOT = Path(__file__).parent.parent
DATA_DIR = PROJECT_ROOT / "data"
MODELS_DIR = PROJECT_ROOT / "models"
CACHE_DIR = PROJECT_ROOT / ".cache"

# Supported Indic languages
SUPPORTED_LANGUAGES: Dict[str, str] = {
    'hi': 'Hindi',
    'bn': 'Bengali', 
    'ta': 'Tamil',
    'te': 'Telugu',
    'ml': 'Malayalam',
    'kn': 'Kannada',
    'gu': 'Gujarati',
    'mr': 'Marathi',
    'pa': 'Punjabi',
    'or': 'Odia',
    'en': 'English'
}

# Activity types
ACTIVITY_TYPES: List[str] = [
    'meme',
    'recipe', 
    'folklore',
    'landmark'
]

# AI model configurations
AI_CONFIG = {
    'text_model': 'sarvamai/sarvam-1',
    'vision_model': 'microsoft/DiT-base',
    'max_tokens': 512,
    'temperature': 0.7
}

# Database configuration
DATABASE_CONFIG = {
    'local_db': 'sqlite:///corpus_collection.db',
    'remote_db': os.getenv('DATABASE_URL', ''),
    'batch_size': 100
}

# PWA and offline configuration
PWA_CONFIG = {
    'cache_version': 'v1.0.0',
    'offline_timeout': 5000,  # milliseconds
    'sync_interval': 300000,  # 5 minutes in milliseconds
    'max_offline_storage': 50 * 1024 * 1024  # 50MB
}

# Content validation settings
VALIDATION_CONFIG = {
    'min_text_length': 10,
    'max_text_length': 5000,
    'max_image_size': 10 * 1024 * 1024,  # 10MB
    'allowed_image_types': ['jpg', 'jpeg', 'png', 'webp']
}

# Create necessary directories
for directory in [DATA_DIR, MODELS_DIR, CACHE_DIR]:
    directory.mkdir(exist_ok=True)