app.py

import gradio as gr
import pdfplumber
import os
from langchain.schema import Document
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.embeddings import OpenAIEmbeddings
from langchain.vectorstores import Pinecone
import pinecone
import pandas as pd
import time
from pinecone.grpc import PineconeGRPC as Pinecone
from pinecone import ServerlessSpec

# OpenAI API key
openai_api_key = os.getenv("OPENAI_API_KEY")

# Initialize Pinecone with PineconeGRPC
pinecone_api_key = os.getenv("PINECONE_API_KEY")
pc = Pinecone(api_key=pinecone_api_key)

# Define index name and parameters
index_name = "italy-kg"

# Create index if it doesn't exist
if index_name not in pinecone.list_indexes():
    pc.create_index(
        name=index_name,
        dimension=1536,
        metric="cosine",
        spec=ServerlessSpec(
            cloud="aws",
            region="us-east-1"
        ),
        deletion_protection="disabled"
    )

# Embedding using OpenAI
embeddings = OpenAIEmbeddings(api_key=openai_api_key)

# Gradio Blocks app with PDF uploader and table for logs
def process_pdf(file):
    # Extract text from PDF using pdfplumber
    with pdfplumber.open(file.name) as pdf:
        text = ""
        for page in pdf.pages:
            text += page.extract_text()
    
    # Split text using RecursiveCharacterTextSplitter
    documents = [Document(page_content=text)]
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=300, chunk_overlap=50)
    docs = text_splitter.split_documents(documents)
    
    # Add documents to Pinecone Vector Store
    vectorstore = Pinecone.from_existing_index(index_name, embeddings)
    vectorstore.add_documents(docs)
    
    # Prepare log data
    log_data = {
        "File Name": [file.name],
        "File Size (KB)": [os.path.getsize(file.name) / 1024],
        "Number of Chunks": [len(docs)],
        "Timestamp": [time.strftime("%Y-%m-%d %H:%M:%S")]
    }
    
    # Create a DataFrame for logs
    df_logs = pd.DataFrame(log_data)
    
    return "PDF processed successfully!", df_logs

# Gradio Interface
with gr.Blocks() as demo:
    gr.Markdown("# PDF Uploader to Pinecone with Logs")
    
    with gr.Row():
        with gr.Column():
            pdf_input = gr.File(label="Upload PDF", type="file")
            process_button = gr.Button("Process PDF")
        
        with gr.Column():
            output_text = gr.Textbox(label="Status", interactive=False)
            log_table = gr.DataFrame(label="Logs", interactive=False)

    # Define action on button click
    process_button.click(process_pdf, inputs=pdf_input, outputs=[output_text, log_table])

# Launch the Gradio app
demo.launch()