#!/usr/bin/env python3
"""
SneakerPicks — Deeplink/Affiliate Link Checker

Checks if affiliate links still work by sending HTTP HEAD requests.

Usage:
    python3 scripts/deeplink-audit.py                      # check 100 random offers
    python3 scripts/deeplink-audit.py --sample 500         # check 500
    python3 scripts/deeplink-audit.py --feed "JD Sports"   # check specific feed
    python3 scripts/deeplink-audit.py --fix                # mark broken as out-of-stock
"""

import argparse
import os
import sys
import time
from concurrent.futures import ThreadPoolExecutor, as_completed
from threading import Semaphore
from urllib.parse import urlparse

import psycopg2
import requests

TIMEOUT = 10
MAX_WORKERS = 5
USER_AGENT = "SneakerPicks Link Checker/1.0"
# Rate limit: max 5 requests/sec via semaphore + sleep
rate_semaphore = Semaphore(MAX_WORKERS)


def get_connection():
    url = os.environ.get("DATABASE_URL")
    if not url:
        print("ERROR: DATABASE_URL environment variable not set")
        sys.exit(1)
    return psycopg2.connect(url)


def check_link(offer_id: int, url: str) -> dict:
    """Check a single link. Returns result dict."""
    rate_semaphore.acquire()
    try:
        try:
            resp = requests.head(
                url,
                timeout=TIMEOUT,
                allow_redirects=True,
                headers={"User-Agent": USER_AGENT},
            )
            status = resp.status_code
            final_url = resp.url

            if status in (404, 410):
                result = "broken"
            elif status >= 500:
                result = "error"
            elif status in (301, 302, 303, 307, 308) or final_url != url:
                result = "redirected"
            elif 200 <= status < 400:
                result = "ok"
            else:
                result = "error"

            return {
                "offer_id": offer_id,
                "url": url,
                "status": status,
                "final_url": final_url,
                "result": result,
            }
        except requests.RequestException as e:
            return {
                "offer_id": offer_id,
                "url": url,
                "status": None,
                "final_url": None,
                "result": "error",
                "error": str(e),
            }
    finally:
        time.sleep(0.2)  # ~5 req/s across 5 workers
        rate_semaphore.release()


def run(sample: int, feed: str | None, fix: bool):
    conn = get_connection()
    cur = conn.cursor()

    # Build query
    query = "SELECT id, product_url FROM product_offers WHERE in_stock = true"
    params: list = []
    if feed:
        query += " AND feed_id = (SELECT id FROM feeds WHERE name = %s LIMIT 1)"
        params.append(feed)
    query += " ORDER BY RANDOM() LIMIT %s"
    params.append(sample)

    cur.execute(query, params)
    offers = cur.fetchall()
    print(f"Checking {len(offers)} offers" + (f" from feed '{feed}'" if feed else "") + "...")

    results = {"ok": 0, "redirected": 0, "broken": 0, "error": 0}
    broken_ids: list[int] = []

    with ThreadPoolExecutor(max_workers=MAX_WORKERS) as executor:
        futures = {executor.submit(check_link, oid, url): oid for oid, url in offers}
        for i, future in enumerate(as_completed(futures), 1):
            r = future.result()
            results[r["result"]] += 1
            if r["result"] == "broken":
                broken_ids.append(r["offer_id"])
                print(f"  BROKEN [{r['status']}] {r['url']}")
            elif r["result"] == "error":
                err = r.get("error", f"HTTP {r['status']}")
                print(f"  ERROR  {r['url'][:80]} — {err}")
            if i % 50 == 0:
                print(f"  ... {i}/{len(offers)} checked")

    print(f"\nResults: ✓ ok={results['ok']}  ↪ redirected={results['redirected']}  "
          f"✗ broken={results['broken']}  ⚠ error={results['error']}")

    if broken_ids and fix:
        cur.execute("""
            UPDATE product_offers SET in_stock = false
            WHERE id = ANY(%s)
        """, (broken_ids,))
        conn.commit()
        print(f"  ✓ Marked {cur.rowcount} broken offers as out-of-stock")
    elif broken_ids and not fix:
        print(f"  {len(broken_ids)} broken offers found. Use --fix to mark as out-of-stock.")

    cur.close()
    conn.close()


if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="SneakerPicks deeplink audit")
    parser.add_argument("--sample", type=int, default=100, help="Number of offers to check (default: 100)")
    parser.add_argument("--feed", type=str, default=None, help="Filter by feed name")
    parser.add_argument("--fix", action="store_true", help="Mark broken links as out-of-stock")
    args = parser.parse_args()
    run(sample=args.sample, feed=args.feed, fix=args.fix)