app.py

from flask import Flask, request, jsonify
import chromadb

# Initialize Flask app
app = Flask(__name__)

# Initialize Chroma DB client
client = chromadb.Client()

# Function to get collection by name or create if it doesn't exist
def get_or_create_collection(collection_name: str):
    if collection_name in client.list_collections():
        return client.get_collection(collection_name)
    else:
        return client.create_collection(collection_name)

@app.route("/create_collection/", methods=["POST"])
def create_collection():
    data = request.json
    collection_name = data.get("collection_name")
    
    if not collection_name:
        return jsonify({"error": "Collection name is required"}), 400

    collection = get_or_create_collection(collection_name)
    return jsonify({"status": f"Collection '{collection_name}' is ready."}), 200

@app.route("/add_embedding/", methods=["POST"])
def add_embedding():
    data = request.json
    collection_name = data.get("collection_name")
    document = data.get("document")
    embedding = data.get("embedding")
    
    if not collection_name or not document or not embedding:
        return jsonify({"error": "Collection name, document, and embedding are required"}), 400

    # Get or create the collection
    collection = get_or_create_collection(collection_name)
    
    # Add the document and its embedding to Chroma DB
    collection.add([document], [{"source": document}], [embedding])
    
    return jsonify({"status": "success"}), 200

@app.route("/query_embedding/", methods=["POST"])
def query_embedding():
    data = request.json
    collection_name = data.get("collection_name")
    query_embedding = data.get("embedding")
    n_results = data.get("n_results", 5)
    
    if not collection_name or not query_embedding:
        return jsonify({"error": "Collection name and embedding are required"}), 400

    # Get or create the collection
    collection = get_or_create_collection(collection_name)

    # Query the collection for nearest neighbors
    results = collection.query(query_embeddings=[query_embedding], n_results=n_results)
    
    return jsonify({"results": results["documents"]}), 200


if __name__ == "__main__":
    app.run(host="0.0.0.0", port=7860