app.py

from config.settings import OPENAI_API_KEY
from flask import Flask, jsonify, request
from flask_cors import CORS
import logging
from services import DocumentService, CodebaseIndexService, ChatService
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.vectorstores import Chroma
import os
from langchain.document_loaders import TextLoader
from langchain.text_splitter import CharacterTextSplitter
from database import DatabaseHandler

app = Flask(__name__)
# app.config["CORS_HEADERS"] = "Content-Type"

# Allow CORS for your frontend origin
CORS(app, origins="*")

# Initialize the OpenAIEmbeddings and DeepLake database instances

embeddings2 = OpenAIEmbeddings()
# Embed and store the texts
# Supplying a persist_directory will store the embeddings on disk

seed_data = './memory'  # todo update this path


vectordb_instance = DatabaseHandler()

# Initialize the DocumentService and CodebaseIndexService
document_service = DocumentService(vectordb_instance)

# codebase_index_service = CodebaseIndexService().index_codebase()


retriever = vectordb_instance.vectordb.as_retriever()
retriever.search_kwargs['distance_metric'] = 'cos'
retriever.search_kwargs['fetch_k'] = 100
retriever.search_kwargs['maximal_marginal_relevance'] = True
retriever.search_kwargs['k'] = 14

# Initialize ChatService
chat_service = ChatService(retriever)

logging.getLogger('flask_cors').level = logging.DEBUG


@app.route('/')
def hello():
    return 'Hello, World!'


@app.route('/api/index_codebase', methods=['POST'])
def index_codebase():
    data = request.json
    repo_url = data['git_url']
    use_existing_index = data.get('use_existing_index', False)

    texts = CodebaseIndexService().index_codebase(
        repo_url, use_existing_index)
    # print(texts)
    document_strings = [str(doc) for doc in texts]

    vectordb_instance.vectordb.add_texts(document_strings)
    print(1)
    return jsonify({"success": True, "message": "Codebase indexed successfully."})


@app.route('/api/chat', methods=['POST'])
def chat():

    # After initializing vectordb, add the following line:
    data = request.json
    question = data['question']
    chat_history = data.get('chat_history', [])

    answer, chat_history = chat_service.ask(question, chat_history)

    response = {
        'answer': answer,
        'chat_history': chat_history,
    }

    return jsonify(response)


if __name__ == '__main__':
    app.run(debug=True, port=8000)