import 'dotenv/config';

// Setup LLM
import {Document, Settings, storageContextFromDefaults, VectorStoreIndex} from "llamaindex";
import {OpenAI, OpenAIEmbedding} from "@llamaindex/openai";
import path from "path";
import fs from "fs";
import {QdrantVectorStore} from "@llamaindex/qdrant";

// Initialize Qdrant vector store for plants

Settings.llm = new OpenAI({
    model: process.env.OPENAI_MODEL,
    apiKey: process.env.OPENAI_API_KEY,
});

Settings.embedModel = new OpenAIEmbedding({
    apiKey: process.env.OPENAI_API_KEY,
});

const vectorStore = new QdrantVectorStore({
    url: 'http://localhost:6333',
    collectionName: 'plants-rag',
});

const storageContext = await storageContextFromDefaults({
    vectorStore
});

try {
    const storageDir = path.join(process.cwd(), 'storage', 'plants');

    // Read all markdown files from the storage directory
    const files = fs.readdirSync(storageDir).filter(file => file.endsWith('.md'));

    const documents = [];
    for (const [index, file] of files.entries()) {
        console.log(`Processing file ${index + 1} of ${files.length}: ${file}`);
        const filePath = path.join(storageDir, file);
        const content = fs.readFileSync(filePath, 'utf-8');

        documents.push(new Document({
            text: content,
            metadata: {
                source: file
            }
        }));
    }

    // Create vector store index
    await VectorStoreIndex.fromDocuments(documents, {
        storageContext: storageContext
    });


} catch (error) {
    console.error('Error initializing plants RAG:', error);

}