from dotenv import load_dotenv

# langchain libraries
from langchain.document_loaders import DirectoryLoader, TextLoader
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.text_splitter import CharacterTextSplitter, RecursiveCharacterTextSplitter
from langchain.vectorstores import FAISS, Pinecone

import pinecone
import openai
import os

load_dotenv()
openai.api_key = os.getenv("OPENAI_API_KEY")
PINECONE_API_KEY = os.getenv("PINECONE_API_KEY")


def generate_pincone_vector_store(index_name='btc-chat-bot'):
    pinecone.init()
    pinecone.create_index("test-index", dimension=1536, metric='cosine')
    pinecone.list_indexes()
    result = Pinecone.from_documents(documents, embeddings, index_name)
    return result


def load_local_vector_store(index_name='hr_faiss_index'):
    embeddings = OpenAIEmbeddings()
    try:
        vector_store = FAISS.load_local(index_name, embeddings)
        print("Local VectorDB Found.")
        return vector_store
    except Exception as e:
        print(e)
        return None


def load_local_documents():
    doc_dir = os.path.join(os.getcwd() + '/docs', 'processed')
    loader = DirectoryLoader(doc_dir)
    documents = loader.load()
    assert len(documents) > 0
    return documents


def generate_new_vector_store(index_name='hr_faiss_index'):
    print("No Local VectorDB Found. Generating new one...")
    documents = load_local_documents()
    text_splitter = RecursiveCharacterTextSplitter(
        chunk_size=1000, chunk_overlap=0, separators=["\n", "\r\n", "\r", " "])
    documents = text_splitter.split_documents(documents)

    embeddings = OpenAIEmbeddings()
    vector_store = FAISS.from_documents(documents, embeddings)
    vector_store.save_local(index_name)
    return vector_store


def get_or_create_vector_store(index_name='hr_faiss_index'):
    vector_store = load_local_vector_store(index_name)
    if vector_store is None:
        vector_store = generate_new_vector_store(index_name)
    return vector_store


if __name__ == "__main__":
    vector = get_or_create_vector_store()
    print(vector)