app.py

import os
import tempfile
import shutil
import chromadb
from langchain_community.document_loaders import PyPDFLoader
from langchain_community.vectorstores import Chroma
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain.chains import RetrievalQA
from langchain_community.llms import HuggingFaceHub
import gradio as gr

DB_DIR = "chroma_db"
os.makedirs(DB_DIR, exist_ok=True)

def load_and_index_pdf(pdf_file):
    with tempfile.TemporaryDirectory() as tmpdir:
        pdf_path = os.path.join(tmpdir, pdf_file.name)
        shutil.copy(pdf_file.name, pdf_path)
        loader = PyPDFLoader(pdf_path)
        documents = loader.load_and_split()
        embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
        vectordb = Chroma.from_documents(documents, embedding=embeddings, persist_directory=DB_DIR)
        vectordb.persist()
    return "✅ PDF تمت معالجته بنجاح! يمكنك الآن طرح الأسئلة."

def answer_question(question):
    if not os.path.exists(DB_DIR) or not os.listdir(DB_DIR):
        return "❌ الرجاء رفع ملف PDF أولًا."
    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
    vectordb = Chroma(persist_directory=DB_DIR, embedding_function=embeddings)
    retriever = vectordb.as_retriever()
    llm = HuggingFaceHub(repo_id="mistralai/Mistral-7B-Instruct-v0.2", model_kwargs={"temperature": 0.5, "max_new_tokens": 512})
    qa = RetrievalQA.from_chain_type(llm=llm, chain_type="stuff", retriever=retriever)
    return qa.run(question)

with gr.Blocks() as demo:
    gr.Markdown("## 🤖 Smart PDF Assistant - مساعدك الذكي في قراءة وفهم ملفات PDF")
    
    with gr.Tab("📁 تحميل PDF"):
        pdf_input = gr.File(label="ارفع ملف PDF", file_types=[".pdf"])
        upload_output = gr.Textbox(label="حالة المعالجة")
        upload_btn = gr.Button("📄 معالجة الملف")
        upload_btn.click(fn=load_and_index_pdf, inputs=pdf_input, outputs=upload_output)

    with gr.Tab("❓ اسأل سؤالك"):
        question = gr.Textbox(label="اكتب سؤالك هنا")
        answer = gr.Textbox(label="الإجابة", lines=5)
        ask_btn = gr.Button("🔍 إرسال السؤال")
        ask_btn.click(fn=answer_question, inputs=question, outputs=answer)

demo.launch()