app/api/ingest.py

from fastapi import APIRouter, UploadFile, File, Header, HTTPException
from app.core.config import INTERNAL_API_KEY
from app.rag.chunker import chunk_text
from app.rag.store import save_vector_store
import io
import json

try:
    from pypdf import PdfReader
    from docx import Document
except ImportError:
    print("CRITICAL WARNING: Thiếu thư viện 'pypdf' hoặc 'python-docx'. Chức năng đọc file sẽ lỗi.")

router = APIRouter()

@router.post("/upload")
async def upload_document(
    file: UploadFile = File(...),
    x_api_key: str = Header(...)
):
    if x_api_key != INTERNAL_API_KEY:
        raise HTTPException(status_code=403, detail="Sai mã bảo mật (API Key).")

    filename = file.filename.lower()
    content_bytes = await file.read()
    text_content = ""

    try:
        if filename.endswith(".pdf"):
            reader = PdfReader(io.BytesIO(content_bytes))
            for page in reader.pages:
                extracted = page.extract_text()
                if extracted:
                    text_content += extracted + "\n"
        
        elif filename.endswith(".docx"):
            doc = Document(io.BytesIO(content_bytes))
            text_content = "\n".join([para.text for para in doc.paragraphs if para.text.strip()])
            
        elif filename.endswith(".json"):
            data = json.load(io.BytesIO(content_bytes))
            text_content = json.dumps(data, ensure_ascii=False, indent=2)
            
        elif filename.endswith((".txt", ".md")):
            text_content = content_bytes.decode("utf-8")
        
        else:
            return {"status": "error", "message": "Định dạng file không hỗ trợ. Chỉ nhận: PDF, DOCX, JSON, TXT."}

        # Validate nội dung
        if len(text_content.strip()) < 10:
            return {"status": "error", "message": "File không có nội dung văn bản đọc được."}

        # Pipeline xử lý
        chunks = chunk_text(text_content)
        save_vector_store(chunks)

        return {
            "status": "success",
            "filename": file.filename,
            "chunks_created": len(chunks),
            "message": "Đã học xong tài liệu mới!"
        }

    except Exception as e:
        print(f"Ingest Error: {e}")
        raise HTTPException(status_code=500, detail=f"Lỗi xử lý file: {str(e)}")