app.py

from flask import Flask, send_from_directory, request, jsonify
import requests
import os
import json
import base64
from datetime import datetime, timezone

app = Flask(__name__, static_folder='.')

TAVILY_KEY = os.environ.get("TAVILY_KEY", "")
HF_TOKEN   = os.environ.get("HF_TOKEN", "")
HF_DATASET = os.environ.get("HF_DATASET", "")  # e.g. "AnesNT/surfgo-cache"

# ─── HuggingFace Dataset helpers ────────────────────────────────────────────

def hf_headers():
    return {"Authorization": f"Bearer {HF_TOKEN}", "Content-Type": "application/json"}

def hf_get_cache():
    """Download cache.json from HF dataset. Returns dict."""
    url = f"https://huggingface.co/datasets/{HF_DATASET}/resolve/main/cache.json"
    r = requests.get(url, headers={"Authorization": f"Bearer {HF_TOKEN}"}, timeout=10)
    if r.status_code == 200:
        return r.json()
    return {}

def hf_put_cache(cache: dict):
    """Upload updated cache.json to HF dataset."""
    content  = json.dumps(cache, ensure_ascii=False, indent=2)
    encoded  = base64.b64encode(content.encode()).decode()

    # Get current file SHA for update
    sha = None
    r = requests.get(
        f"https://huggingface.co/api/datasets/{HF_DATASET}/tree/main",
        headers=hf_headers(), timeout=10
    )
    if r.status_code == 200:
        for f in r.json():
            if f.get("path") == "cache.json":
                sha = f.get("oid")
                break

    payload = {
        "message": f"surfgo cache — {datetime.now(timezone.utc).strftime('%Y-%m-%d %H:%M UTC')}",
        "content": encoded,
    }
    if sha:
        payload["sha"] = sha

    requests.put(
        f"https://huggingface.co/api/datasets/{HF_DATASET}/upload/main/cache.json",
        headers=hf_headers(), json=payload, timeout=15
    )

def normalize(q: str) -> str:
    return q.strip().lower()

# ─── Routes ─────────────────────────────────────────────────────────────────

@app.route('/')
def index():
    return send_from_directory('.', 'index.html')

@app.route('/search', methods=['POST'])
def search():
    if not TAVILY_KEY:
        return jsonify({"error": "TAVILY_KEY not set"}), 500

    body  = request.get_json()
    query = body.get("query", "").strip()
    key   = normalize(query)
    use_hf = bool(HF_TOKEN and HF_DATASET)

    # ── 1. Check cache first ─────────────────────────────────────────────────
    if use_hf:
        try:
            cache = hf_get_cache()
            if key in cache:
                entry = cache[key]
                # Bump hit counter
                cache[key]["hits"] = entry.get("hits", 0) + 1
                try: hf_put_cache(cache)
                except: pass
                result = entry["result"]
                result["_cached"]    = True
                result["_hits"]      = cache[key]["hits"]
                result["_stored_at"] = entry.get("stored_at", "")
                return jsonify(result)
        except Exception as e:
            print(f"[cache read error] {e}")

    # ── 2. Cache miss → call Tavily ──────────────────────────────────────────
    payload = {**body, "api_key": TAVILY_KEY}
    resp = requests.post("https://api.tavily.com/search", json=payload, timeout=20)
    data = resp.json()

    # ── 3. Store new result in HF ────────────────────────────────────────────
    if use_hf and resp.status_code == 200:
        try:
            cache = hf_get_cache()
            cache[key] = {
                "query":     query,
                "result":    data,
                "hits":      1,
                "stored_at": datetime.now(timezone.utc).isoformat(),
            }
            hf_put_cache(cache)
        except Exception as e:
            print(f"[cache write error] {e}")

    return jsonify(data), resp.status_code

@app.route('/debug')
def debug():
    info = {
        "TAVILY_KEY_set": bool(TAVILY_KEY),
        "HF_TOKEN_set":   bool(HF_TOKEN),
        "HF_DATASET":     HF_DATASET or "NOT SET",
    }
    if HF_TOKEN and HF_DATASET:
        try:
            cache = hf_get_cache()
            info["hf_read"]          = "OK"
            info["cache_keys_count"] = len(cache)
            info["cache_keys"]       = list(cache.keys())[:20]
        except Exception as e:
            info["hf_read"] = f"ERROR: {e}"
        try:
            r = requests.get(
                f"https://huggingface.co/api/datasets/{HF_DATASET}/tree/main",
                headers=hf_headers(), timeout=10
            )
            info["hf_tree_status"] = r.status_code
            info["hf_files"] = [f.get("path") for f in r.json()] if r.ok else r.text[:200]
        except Exception as e:
            info["hf_tree"] = f"ERROR: {e}"
    else:
        info["hf_status"] = "HF_TOKEN or HF_DATASET not set — cache disabled"
    return jsonify(info)

@app.route('/<path:path>')
def static_files(path):
    return send_from_directory('.', path)

if __name__ == '__main__':
    port = int(os.environ.get('PORT', 7860))
    app.run(host='0.0.0.0', port=port, debug=False)