modcomms/backend/app/main.py

import logging
from contextlib import asynccontextmanager

from fastapi import FastAPI, Depends
from fastapi.middleware.cors import CORSMiddleware

from app.config import settings
from app.dependencies.auth import get_current_user
from app.models.database import init_db, close_db
from app.api import router as api_router, kb_router, analysis_router
from app.services.gemini_service import GeminiService
from app.services.reference_docs import ReferenceDocsService
from app.services.analysis_service import AnalysisService
from app.services.knowledge_base_service import KnowledgeBaseService

# Configure logging
logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
    datefmt="%Y-%m-%d %H:%M:%S"
)
logger = logging.getLogger(__name__)


class HealthCheckFilter(logging.Filter):
    """Filter out health check endpoint logs from uvicorn access log."""
    def filter(self, record: logging.LogRecord) -> bool:
        if "GET /health" in record.getMessage():
            return False
        return True


logging.getLogger("uvicorn.access").addFilter(HealthCheckFilter())


@asynccontextmanager
async def lifespan(app: FastAPI):
    """Initialize services on startup and cleanup on shutdown."""
    # Validate settings
    settings.validate()

    # Initialize database
    print("Initializing database connection...")
    db_available = False
    try:
        await init_db()
        print("Database initialized successfully")
        db_available = True
    except Exception as e:
        logger.warning(f"Database initialization failed (may not be available): {e}")
        print("Warning: Database not available - running in stateless mode")

    # Initialize services
    print("Loading reference documents...")
    reference_docs = ReferenceDocsService(settings.REFERENCE_DOCS_PATH)

    if db_available:
        try:
            from app.models.database import async_session_factory
            async with async_session_factory() as session:
                print("Loading specs from database...")
                await reference_docs.load_specs_from_db(session)
        except Exception as e:
            logger.warning(f"Failed to load specs from DB (falling back to files): {e}")

    doc_summary = reference_docs.get_context_summary()
    print(f"  Brand documents: {len(doc_summary['brand_files'])} files ({doc_summary['brand_context_length']} chars)")
    print(f"  Channel documents: {len(doc_summary['channel_files'])} files ({doc_summary['channel_context_length']} chars)")

    print("Initializing Gemini service...")
    gemini_service = GeminiService(settings.GEMINI_API_KEY)

    print("Initializing analysis service...")
    analysis_service = AnalysisService(gemini_service, reference_docs)
    app.state.analysis_service = analysis_service

    # Initialize Knowledge Base service (requires LlamaParse API key)
    knowledge_base_service = None
    if settings.LLAMA_CLOUD_API_KEY:
        from app.services.llamaparse_service import LlamaParseService
        print("Initializing LlamaParse service...")
        llamaparse_service = LlamaParseService(settings.LLAMA_CLOUD_API_KEY, settings.LLAMA_CLOUD_BASE_URL)
        knowledge_base_service = KnowledgeBaseService(llamaparse_service, gemini_service, reference_docs)
        app.state.knowledge_base_service = knowledge_base_service
        print("Knowledge Base pipeline ready!")

        if db_available:
            import asyncio as _asyncio
            from app.models.database import async_session_factory as _session_factory
            from app.repositories.knowledge_base_repository import KnowledgeBaseRepository as _KBRepo
            try:
                async with _session_factory() as _session:
                    _repo = _KBRepo(_session)
                    _stale = await _repo.get_all_stale_active_jobs()
                    if _stale:
                        print(f"Resuming {len(_stale)} interrupted processing job(s)...")
                        for _job in _stale:
                            _kb = await _repo.get_knowledge_base(_job.knowledge_base_id)
                            if _kb:
                                _asyncio.create_task(
                                    knowledge_base_service.process_documents(
                                        kb_id=_job.knowledge_base_id,
                                        job_id=_job.id,
                                        agent_key=_kb.agent_key,
                                        user_id=_job.triggered_by_id,
                                        user_name=_job.triggered_by_name,
                                    )
                                )
                                print(f"  Resumed: {_kb.display_name} (job {_job.id}, was {_job.status})")
            except Exception as _e:
                logger.warning(f"Startup job recovery failed: {_e}")
    else:
        print("LLAMA_CLOUD_API_KEY not set - Knowledge Base processing pipeline disabled")

    print("Backend ready!")

    yield

    # Cleanup on shutdown
    print("Shutting down...")
    await close_db()


# Create FastAPI app
app = FastAPI(
    title="ModComms Proof Review API",
    description="AI-powered proof review backend for Barclays marketing materials",
    version="1.0.0",
    lifespan=lifespan,
)

# CORS middleware
app.add_middleware(
    CORSMiddleware,
    allow_origins=settings.CORS_ORIGINS.split(","),
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

# Include API routes
app.include_router(api_router, prefix="/api")
app.include_router(kb_router, prefix="/api")
app.include_router(analysis_router, prefix="/api")


@app.get("/health")
async def health_check():
    """Health check endpoint."""
    return {"status": "healthy", "service": "modcomms-backend"}


@app.get("/info")
async def info(user: dict = Depends(get_current_user)):
    """Get backend information. Requires authentication."""
    analysis_service = getattr(app.state, "analysis_service", None)
    if analysis_service:
        ref_docs = analysis_service.reference_docs
        doc_summary = ref_docs.get_context_summary()
        return {
            "status": "ready",
            "user": user.get("name", "Unknown"),
            "agents": ["Risk & Control Agent", "Brand Agent", "Channel Best Practices Agent", "Channel Tech Specs Agent"],
            "reference_docs": doc_summary,
        }
    return {"status": "initializing", "user": user.get("name", "Unknown")}