wiki/tasks/snowbike-rag/server.py

#!/usr/bin/env python3
"""
Snowbike RAG - Flask API
Порт: 5557
GET /search?q={query}&topics={topic_ids}&limit={limit}
"""

import os
import sys
import json
from pathlib import Path

# Добавляем папку scripts в путь
sys.path.insert(0, str(Path(__file__).parent / "scripts"))

from flask import Flask, request, jsonify, Response, render_template

app = Flask(__name__)


def load_env():
    """Загружаем переменные окружения из ~/.openclaw/.env"""
    env_file = Path.home() / ".openclaw" / ".env"
    if env_file.exists():
        with open(env_file) as f:
            for line in f:
                line = line.strip()
                if line and not line.startswith("#") and "=" in line:
                    key, _, val = line.partition("=")
                    os.environ.setdefault(key.strip(), val.strip())


load_env()


@app.route("/", methods=["GET"])
def index():
    """Главная страница — веб-интерфейс поиска."""
    return render_template("index.html")


@app.route("/api/search", methods=["GET"])
def api_search_endpoint():
    """Алиас /search → /api/search (новый роут)."""
    return search_endpoint()


@app.route("/api/stats", methods=["GET"])
def api_stats_endpoint():
    """Алиас /stats → /api/stats (новый роут)."""
    return stats()


@app.route("/health", methods=["GET"])
def health():
    return jsonify({"status": "ok", "service": "snowbike-rag"})


@app.route("/search", methods=["GET"])
def search_endpoint():
    """
    Гибридный поиск по базе сообщений Telegram-группы.
    
    Параметры:
        q (str): поисковый запрос (обязательный)
        topics (str): список topic_id через запятую (опционально)
        limit (int): максимальное количество источников (по умолчанию 15)
    
    Ответ:
        {
          "query": "...",
          "answer": "...",
          "sources": [...],
          "count": N,
          "time_ms": N
        }
    """
    query = request.args.get("q", "").strip()
    if not query:
        return jsonify({"error": "Параметр 'q' обязателен"}), 400
    
    # Парсим topic_ids
    topics_param = request.args.get("topics", "")
    topic_ids = None
    if topics_param:
        try:
            topic_ids = [int(t.strip()) for t in topics_param.split(",") if t.strip()]
        except ValueError:
            return jsonify({"error": "Параметр 'topics' должен содержать числа через запятую"}), 400
    
    # Лимит
    try:
        limit = int(request.args.get("limit", 15))
        limit = max(1, min(limit, 30))  # от 1 до 30
    except ValueError:
        limit = 15
    
    try:
        from search import search
        result = search(query, topic_ids=topic_ids, limit=limit)
        return Response(
            json.dumps(result, ensure_ascii=False, indent=2),
            mimetype="application/json; charset=utf-8"
        )
    except Exception as e:
        import traceback
        return jsonify({
            "error": str(e),
            "traceback": traceback.format_exc()
        }), 500


@app.route("/topics", methods=["GET"])
def list_topics():
    """Список доступных топиков."""
    try:
        meta_file = "/home/node/.openclaw/workspace/data/telegram-collector/raw/1242788123/meta.json"
        with open(meta_file) as f:
            meta = json.load(f)
        
        topics = [
            {"id": int(k), "title": v}
            for k, v in meta.get("topics", {}).items()
        ]
        topics.sort(key=lambda x: x["id"])
        
        return jsonify({"topics": topics})
    except Exception as e:
        return jsonify({"error": str(e)}), 500


@app.route("/stats", methods=["GET"])
def stats():
    """Статистика индексов."""
    result = {}
    
    # Meilisearch — используем прямой HTTP запрос без прокси
    try:
        import requests as req
        s = req.Session()
        s.trust_env = False
        r = s.get("http://127.0.0.1:7700/indexes/snowbike_messages/stats", timeout=5)
        r.raise_for_status()
        data = r.json()
        result["meilisearch"] = {
            "status": "ok",
            "documents": data.get("numberOfDocuments", 0),
        }
    except Exception as e:
        result["meilisearch"] = {"status": "error", "error": str(e)}
    
    # ChromaDB
    try:
        import chromadb
        chroma_path = str(Path(__file__).parent / "data" / "chromadb")
        client = chromadb.PersistentClient(path=chroma_path)
        collection = client.get_collection("snowbike_embeddings")
        result["chromadb"] = {
            "status": "ok",
            "documents": collection.count(),
        }
    except Exception as e:
        result["chromadb"] = {"status": "error", "error": str(e)}
    
    return jsonify(result)


if __name__ == "__main__":
    print("=" * 50)
    print("Snowbike RAG API запускается...")
    print("URL: http://localhost:5557")
    print("Эндпоинты:")
    print("  GET /health")
    print("  GET /search?q=<запрос>&topics=<id1,id2>&limit=<N>")
    print("  GET /topics")
    print("  GET /stats")
    print("=" * 50)
    
    app.run(host="0.0.0.0", port=5557, debug=False)