poll.py

#!/usr/bin/env python3
"""Poll every Discord channel the bot can see and write a daily digest."""
from __future__ import annotations

import datetime as dt
import json
import os
import pathlib
import sys
import urllib.error
import urllib.parse
import urllib.request

API = "https://discord.com/api/v10"
TOKEN = os.environ["DISCORD_BOT_TOKEN"]
HEADERS = {
    "Authorization": f"Bot {TOKEN}",
    "User-Agent": "discord-niklas-digest (github actions, v1)",
}
LOOKBACK_HOURS = 24


def api_get(path: str):
    req = urllib.request.Request(API + path, headers=HEADERS)
    with urllib.request.urlopen(req, timeout=30) as r:
        return json.loads(r.read().decode("utf-8"))


def recent_messages(channel_id: str, since_iso: str):
    try:
        msgs = api_get(f"/channels/{channel_id}/messages?limit=100")
    except urllib.error.HTTPError as e:
        if e.code in (403, 404):
            return []
        raise
    return [m for m in msgs if m.get("timestamp", "") >= since_iso]


def main() -> int:
    now = dt.datetime.now(dt.timezone.utc)
    since = (now - dt.timedelta(hours=LOOKBACK_HOURS)).isoformat()
    today = now.date().isoformat()

    try:
        guilds = api_get("/users/@me/guilds")
    except urllib.error.HTTPError as e:
        body = e.read().decode("utf-8", errors="replace")
        print(f"ERROR fetching /users/@me/guilds: HTTP {e.code}\n{body}", file=sys.stderr)
        if e.code in (401, 403):
            print(
                "Bot token rejected or bot not in any server. Invite it:\n"
                "  https://discord.com/api/oauth2/authorize"
                "?client_id=1493705177050386562&scope=bot&permissions=66560",
                file=sys.stderr,
            )
        return 1

    lines = [f"# Discord digest — {today}", ""]
    total = 0

    if not guilds:
        lines.append("_Bot is not in any server yet. Invite it and re-run._")
    for g in guilds:
        lines.append(f"## {g['name']}")
        lines.append("")
        try:
            channels = api_get(f"/guilds/{g['id']}/channels")
        except urllib.error.HTTPError as e:
            lines.append(f"_(couldn't list channels: HTTP {e.code})_")
            lines.append("")
            continue
        for ch in channels:
            if ch.get("type") != 0:  # 0 = GUILD_TEXT
                continue
            msgs = recent_messages(ch["id"], since)
            if not msgs:
                continue
            lines.append(f"### #{ch['name']}")
            lines.append("")
            for m in reversed(msgs):  # oldest first
                author = m.get("author", {}).get("username", "?")
                content = (m.get("content") or "").strip()
                if not content:
                    continue
                ts = m.get("timestamp", "")[:16].replace("T", " ")
                lines.append(f"- **{author}** ({ts} UTC): {content}")
                total += 1
            lines.append("")

    if total == 0:
        lines.append("_No new messages in the last 24h._")
        lines.append("")

    body = "\n".join(lines)
    out = pathlib.Path("digests")
    out.mkdir(exist_ok=True)
    (out / "latest.md").write_text(body, encoding="utf-8")
    (out / f"{today}.md").write_text(body, encoding="utf-8")
    print(f"Wrote digests/latest.md and digests/{today}.md ({total} messages).")
    return 0


if __name__ == "__main__":
    sys.exit(main())