#!/usr/bin/env python3
"""TaskToAgent - RSS to Discord. Python 3.9+, standard library only.
Watches RSS/Atom feeds and posts new items (optionally only those matching keywords) to a Discord webhook.
First run only records what's already there (no flood); later runs post what's new."""
import os, sys, json, time, html, re, urllib.request, urllib.error, xml.etree.ElementTree as ET

FEEDS = [f.strip() for f in os.getenv("FEEDS", "https://hnrss.org/frontpage,https://www.rockpapershotgun.com/feed").split(",") if f.strip()]
KEYWORDS = [k.strip().lower() for k in os.getenv("KEYWORDS", "").split(",") if k.strip()]  # empty = post everything
WEBHOOK = os.getenv("DISCORD_WEBHOOK_URL", "")
STATE = os.getenv("STATE_FILE", os.path.join(os.path.dirname(os.path.abspath(__file__)), "seen_rss.json"))
MAX_PER_RUN = int(os.getenv("MAX_PER_RUN", "10"))
UA = {"User-Agent": "TaskToAgent-RSS-to-Discord/1.0 (+https://tasktoagent.com/agents/)"}
A = "{http://www.w3.org/2005/Atom}"

def clean(s): return re.sub(r"\s+", " ", html.unescape(re.sub(r"<[^>]+>", " ", s or ""))).strip()

def items(url):
    with urllib.request.urlopen(urllib.request.Request(url, headers=UA), timeout=20) as r: root = ET.fromstring(r.read())
    for it in root.iter("item"):
        yield {"id": (it.findtext("guid") or it.findtext("link") or "").strip(), "title": clean(it.findtext("title")),
               "link": (it.findtext("link") or "").strip(), "text": clean(it.findtext("description"))}
    for e in root.iter(A + "entry"):
        l = e.find(A + "link[@rel='alternate']")
        if l is None: l = e.find(A + "link")
        href = l.get("href", "") if l is not None else ""
        yield {"id": (e.findtext(A + "id") or href).strip(), "title": clean(e.findtext(A + "title")), "link": href,
               "text": clean(e.findtext(A + "summary") or e.findtext(A + "content"))}

def matches(i): return not KEYWORDS or any(k in (i["title"] + " " + i["text"]).lower() for k in KEYWORDS)

def post(i):
    msg = {"username": "RSS Bot", "content": f"**{i['title'][:200]}**\n{i['link']}"}
    if not WEBHOOK or "--dry-run" in sys.argv: print(msg["content"]); return
    req = urllib.request.Request(WEBHOOK, json.dumps(msg).encode(), {**UA, "Content-Type": "application/json"})
    for attempt in range(3):
        try: urllib.request.urlopen(req, timeout=20).read(); return
        except urllib.error.HTTPError as e:
            if e.code == 429: time.sleep(float(json.loads(e.read() or b"{}").get("retry_after", 2)) + 0.5); continue
            raise

def main():
    first = not os.path.exists(STATE); seen = set() if first else set(json.load(open(STATE)))
    new = []
    for f in FEEDS:
        try: new += [i for i in items(f) if i["id"] and i["id"] not in seen]
        except Exception as e: print(f"warn: {f}: {e}", file=sys.stderr)
    if first and "--post-existing" not in sys.argv and "--dry-run" not in sys.argv:
        print(f"first run: remembered {len(new)} existing item(s), nothing posted")
    else:
        hits = [i for i in new if matches(i)][:MAX_PER_RUN]
        for i in hits: post(i); time.sleep(0.6)
        print(f"{len(hits)} new matching item(s) posted")
    if "--dry-run" not in sys.argv: json.dump(sorted(seen | {i["id"] for i in new})[-5000:], open(STATE, "w"))

if __name__ == "__main__": main()