import gradio as gr
from langchain.vectorstores import FAISS
from langchain.embeddings import HuggingFaceEmbeddings
import zipfile
import os

# Percorsi per il primo file ZIP
zip_path_m = "faiss_manual_index.zip"  # File ZIP per l'indice manuale
extraction_dir_m = "./extracted_models/manual_index"  # Sottocartella per estrazione manuale
testm_dir = "./extracted_models/manual_index/faiss_manual_index"  # Cartella finale

# Percorsi per il secondo file ZIP
zip_path_p = "faiss_problems_index.zip"  # File ZIP per l'indice problemi
extraction_dir_p = "./extracted_models/problems_index"  # Sottocartella per estrazione problemi
testp_dir = "./extracted_models/problems_index/faiss_problems_index"  # Cartella finale

# Estrai il primo file ZIP se non esiste già
if not os.path.exists(testm_dir):
    with zipfile.ZipFile(zip_path_m, 'r') as zip_ref:
        zip_ref.extractall(extraction_dir_m)
    print(f"Indice Manuale estratto correttamente nella cartella {extraction_dir_m}")
else:
    print(f"Indice Manuale già presente in {testm_dir}")

# Estrai il secondo file ZIP se non esiste già
if not os.path.exists(testp_dir):
    with zipfile.ZipFile(zip_path_p, 'r') as zip_ref:
        zip_ref.extractall(extraction_dir_p)
    print(f"Indice Problemi estratto correttamente nella cartella {extraction_dir_p}")
else:
    print(f"Indice Problemi già presente in {testp_dir}")


    
# Carica il modello di embedding
embedding_model = HuggingFaceEmbeddings(model_name="sentence-transformers/LaBSE")

# Carica i vectorstore FAISS salvati
vectorstore = FAISS.load_local("faiss_index", embedding_model, allow_dangerous_deserialization=True)
manual_vectorstore = FAISS.load_local("faiss_manual_index", embedding_model, allow_dangerous_deserialization=True)
problems_vectorstore = FAISS.load_local("faiss_problems_index", embedding_model, allow_dangerous_deserialization=True)

def search_query(query):
    # Cerca nei manuali
    manual_results = manual_vectorstore.similarity_search(query, k=2)
    manual_output = "\n\n".join([doc.page_content for doc in manual_results])

    # Cerca nei problemi
    problems_results = problems_vectorstore.similarity_search(query, k=2)
    problems_output = "\n\n".join([doc.page_content for doc in problems_results])

    # Restituisce i risultati come output diviso
    return manual_output, problems_output
    
examples = [
    ["How to change the knife?"],
    ["What are the safety precautions for using the machine?"],
    ["How can I get help with the machine?"]
]

# Interfaccia Gradio
iface = gr.Interface(
    fn=search_query,
   inputs=gr.Textbox(lines=2, placeholder="Enter your question here..."),
    outputs=[
    gr.Textbox(label="Manual Results"),
    gr.Textbox(label="Issues Results")
],
    examples=examples, 
     title="Manual Querying System",
    description="Enter a question to get relevant information extracted from the manual and the most common related issues."
)

# Avvia l'app
iface.launch()