"""Fetch AI/quant research RSS: ArXiv cs.AI + quant blogs."""
import json, os, feedparser
from datetime import datetime, timezone, timedelta

OUTPUT = "/var/www/nav/data/research.json"
tz = timezone(timedelta(hours=8))

FEEDS = [
    # ArXiv AI (via arxiv-rss.appspot.com or direct)
    {"name": "ArXiv cs.AI", "url": "https://rss.arxiv.org/rss/cs.AI", "lang": "en"},
    {"name": "ArXiv q-fin", "url": "https://rss.arxiv.org/rss/q-fin", "lang": "en"},
    # Quant blogs
    {"name": "Quantitative Finance Stack Exchange", "url": "https://quant.stackexchange.com/feeds", "lang": "en"},
    {"name": "Alpha Vantage Blog", "url": "https://blog.alphavantage.co/rss.xml", "lang": "en"},
    # 机器之心（中文 AI 资讯）
    {"name": "机器之心", "url": "https://www.jiqizhixin.com/rss", "lang": "zh"},
    # Towards Data Science
    {"name": "Towards Data Science", "url": "https://towardsdatascience.com/feed", "lang": "en"},
    # AI Alignment Forum
    {"name": "AI Alignment Forum", "url": "https://www.alignmentforum.org/feed.xml", "lang": "en"},
]

MAX_PER_FEED = 20
MAX_TOTAL = 150


def fetch_feed(feed_info):
    try:
        feed = feedparser.parse(feed_info["url"])
        items = []
        for entry in feed.entries[:MAX_PER_FEED]:
            items.append({
                "title": entry.get("title", ""),
                "link": entry.get("link", ""),
                "summary": entry.get("summary", "")[:400],
                "published": entry.get("published", ""),
                "source": feed_info["name"],
                "lang": feed_info["lang"],
            })
        print(f"  {feed_info['name']}: {len(items)} items")
        return items
    except Exception as e:
        print(f"  {feed_info['name']}: ERROR - {e}")
        return []


def fetch_research():
    items = []
    for feed_info in FEEDS:
        items.extend(fetch_feed(feed_info))

    seen = set()
    unique = []
    for item in items:
        key = item["title"][:30]
        if key not in seen:
            seen.add(key)
            unique.append(item)

    result = {
        "updated_at": datetime.now(tz).strftime("%Y-%m-%d %H:%M"),
        "total": len(unique),
        "items": unique[:MAX_TOTAL],
        "sources": [f["name"] for f in FEEDS],
    }

    os.makedirs(os.path.dirname(OUTPUT), exist_ok=True)
    with open(OUTPUT, "w", encoding="utf-8") as f:
        json.dump(result, f, ensure_ascii=False, indent=2)

    print(f"\nTotal: {len(unique)} unique research items -> {OUTPUT}")
    return result


if __name__ == "__main__":
    print(f"Fetching research RSS at {datetime.now(tz).strftime('%Y-%m-%d %H:%M')}")
    fetch_research()