from datetime import datetime as dt, timezone as tz
from flask import current_app
from langchain_mistralai import MistralAIEmbeddings
from langchain_openai import OpenAIEmbeddings
from langchain_community.document_loaders.pdf import PyPDFLoader
from langchain_community.vectorstores.chroma import Chroma
from langchain_text_splitters import CharacterTextSplitter
import os

from eveai_app import celery
from ..utils.database import Database
from ..models.document import DocumentVersion, EmbeddingMistral, EmbeddingSmallOpenAI
from .. import db


@celery.task(name='create_embeddings', queue='embeddings')
def create_embeddings(tenant_id, document_version_id, embedding_model_def):
    current_app.logger.info(f'Creating embeddings for tenant {tenant_id} on document version {document_version_id} '
                            f'with model {embedding_model_def}')
    Database(tenant_id).switch_schema()
    document_version = DocumentVersion.query.get(document_version_id)
    if document_version is None:
        current_app.logger.error(f'Cannot create embeddings for tenant {tenant_id}. '
                                 f'Document version {document_version_id} not found')
        return
    db.session.add(document_version)

    # start processing
    document_version.processing = True
    document_version.processing_started_at = dt.now(tz.utc)
    db.session.commit()

    embedding_provider = embedding_model_def.rsplit('.', 1)[0]
    embedding_model = embedding_model_def.rsplit('.', 1)[1]
    # define embedding variables
    match (embedding_provider, embedding_model):
        case ('openai', 'text-embedding-3-small'):
            embedding_model = EmbeddingSmallOpenAI()
        case ('mistral', 'text-embedding-3-small'):
            embedding_model = EmbeddingMistral()

    match document_version.file_type:
        case 'pdf':
            pdf_file = os.path.join(current_app.config['UPLOAD_FOLDER'],
                                    document_version.file_location,
                                    document_version.file_path)
            loader = PyPDFLoader(pdf_file)

    # We
    text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
    documents = text_splitter.split_documents(loader.load())

    pass


@celery.task(name='ask_eveAI', queue='llm_interactions')
def ask_eve_ai(query):
    # Interaction logic with LLMs like GPT (Langchain API calls, etc.)
    pass