api/routes.py

import json
import logging
from fastapi import FastAPI, HTTPException, Query
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
from typing import Optional, List
from core.database import (
    get_all_documents,
    get_document_by_id,
    search_documents,
    get_stats,
    init_db
)

logger = logging.getLogger(__name__)

# FastAPI app instance
api_app = FastAPI(
    title="Arabic Document Intelligence API",
    description="Query and retrieve processed Arabic documents",
    version="1.0.0"
)

api_app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_methods=["GET", "POST"],
    allow_headers=["*"]
)


class SearchRequest(BaseModel):
    query: str
    language: Optional[str] = None


@api_app.on_event("startup")
async def startup():
    init_db()
    logger.info("API started, DB initialized")


@api_app.get("/")
async def root():
    return {
        "service": "Arabic Document Intelligence API",
        "version": "1.0.0",
        "endpoints": ["/docs", "/stats", "/documents", "/documents/{id}", "/search"]
    }


@api_app.get("/stats")
async def stats():
    """Get database statistics."""
    data = get_stats()
    if not data:
        raise HTTPException(status_code=500, detail="Could not retrieve stats")
    return data


@api_app.get("/documents")
async def list_documents(
    limit: int = Query(default=20, ge=1, le=100),
    offset: int = Query(default=0, ge=0)
):
    """List all processed documents."""
    docs = get_all_documents(limit=limit, offset=offset)
    return {
        "total_returned": len(docs),
        "limit": limit,
        "offset": offset,
        "documents": docs
    }


@api_app.get("/documents/{document_id}")
async def get_document(document_id: str):
    """Get a specific document by ID."""
    doc = get_document_by_id(document_id)
    if not doc:
        raise HTTPException(status_code=404, detail=f"Document {document_id} not found")
    return doc


@api_app.post("/search")
async def search(request: SearchRequest):
    """Search documents by text content."""
    if not request.query or len(request.query) < 2:
        raise HTTPException(status_code=400, detail="Query must be at least 2 characters")

    results = search_documents(request.query, request.language)
    return {
        "query": request.query,
        "language_filter": request.language,
        "total_results": len(results),
        "results": results
    }


@api_app.get("/documents/{document_id}/json")
async def get_document_json(document_id: str):
    """Get the full structured JSON of a document."""
    doc = get_document_by_id(document_id)
    if not doc:
        raise HTTPException(status_code=404, detail=f"Document {document_id} not found")

    structured = doc.get("structured_data")
    if not structured:
        raise HTTPException(status_code=404, detail="Structured JSON not available")

    return structured