# arXiv Paper Curator - Environment Configuration (EXAMPLE)
# Copy this file to .env and adjust values as needed

# Application Settings
DEBUG=true
ENVIRONMENT=development

# PostgreSQL Database
POSTGRES_DATABASE_URL=postgresql+psycopg2://rag_user:rag_password@postgres:5432/rag_db

# External Services
OPENSEARCH_HOST=http://opensearch:9200
OPENSEARCH__HOST=http://opensearch:9200
OLLAMA_HOST=http://ollama:11434

# arXiv API Configuration  
ARXIV__MAX_RESULTS=15
ARXIV__BASE_URL=https://export.arxiv.org/api/query
ARXIV__PDF_CACHE_DIR=./data/arxiv_pdfs
ARXIV__RATE_LIMIT_DELAY=3.0
ARXIV__TIMEOUT_SECONDS=30
ARXIV__SEARCH_CATEGORY=cs.AI
ARXIV__DOWNLOAD_MAX_RETRIES=3
ARXIV__DOWNLOAD_RETRY_DELAY_BASE=5.0
ARXIV__MAX_CONCURRENT_DOWNLOADS=5
ARXIV__MAX_CONCURRENT_PARSING=1

# PDF Parser Configuration
PDF_PARSER__MAX_PAGES=30
PDF_PARSER__MAX_FILE_SIZE_MB=20
PDF_PARSER__DO_OCR=false
PDF_PARSER__DO_TABLE_STRUCTURE=true

# OpenSearch Configuration (Single hybrid index for all search types)
OPENSEARCH__INDEX_NAME=arxiv-papers
OPENSEARCH__CHUNK_INDEX_SUFFIX=chunks
OPENSEARCH__MAX_TEXT_SIZE=1000000

# Vector Search Settings
OPENSEARCH__VECTOR_DIMENSION=1024
OPENSEARCH__VECTOR_SPACE_TYPE=cosinesimil

# Hybrid Search Settings  
OPENSEARCH__RRF_PIPELINE_NAME=hybrid-rrf-pipeline
OPENSEARCH__HYBRID_SEARCH_SIZE_MULTIPLIER=2

# Text Chunking Configuration
CHUNKING__CHUNK_SIZE=600
CHUNKING__OVERLAP_SIZE=100
CHUNKING__MIN_CHUNK_SIZE=100
CHUNKING__SECTION_BASED=true

# Jina AI Embeddings (Required for hybrid search)
JINA_API_KEY=your_jina_api_key_here

# Ollama Configuration
OLLAMA_MODEL=llama3.2:1b
OLLAMA_TIMEOUT=300

# Langfuse v3 Tracing Configuration - Official SDK Standard (single underscore)
LANGFUSE_ENABLED=true
LANGFUSE_HOST=http://localhost:3001
LANGFUSE_PUBLIC_KEY=pk-lf-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
LANGFUSE_SECRET_KEY=sk-lf-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
LANGFUSE_FLUSH_AT=15
LANGFUSE_FLUSH_INTERVAL=1.0
LANGFUSE_DEBUG=true

# Langfuse Server Configuration (for docker-compose)
# WARNING: Change these values in production! These are development defaults.
LANGFUSE_NEXTAUTH_SECRET=changeme-v3-nextauth-secret-min-32-chars-recommended
LANGFUSE_SALT=changeme-v3-salt-min-32-chars-recommended-for-security
# REQUIRED: Generate a real key before starting: openssl rand -hex 32
LANGFUSE_ENCRYPTION_KEY=0000000000000000000000000000000000000000000000000000000000000000
LANGFUSE_REDIS_PASSWORD=langfuse_redis_password
LANGFUSE_MINIO_ACCESS_KEY=langfuse_minio
LANGFUSE_MINIO_SECRET_KEY=langfuse_minio_secret

# Redis Cache Configuration
REDIS__HOST=redis
REDIS__PORT=6379
# Leave empty since compose.yml Redis has no --requirepass configured
REDIS__PASSWORD=
REDIS__DB=0
REDIS__TTL_HOURS=6

# Telegram Bot Configuration (Week 7)
# Get your bot token from @BotFather on Telegram
TELEGRAM__ENABLED=true
TELEGRAM__BOT_TOKEN=your_telegram_bot_token_here

# Airflow Settings
AIRFLOW__CORE__EXECUTOR=LocalExecutor
AIRFLOW__CORE__LOAD_EXAMPLES=false
AIRFLOW__WEBSERVER__EXPOSE_CONFIG=true
AIRFLOW__HOME=/opt/airflow
AIRFLOW__DATABASE__SQL_ALCHEMY_CONN=postgresql+psycopg2://rag_user:rag_password@postgres:5432/rag_db