news

news/main.py

@@ -0,0 +1,116 @@
+from __future__ import annotations
+
+import asyncio
+import logging
+import os
+from typing import Final, Optional, List
+
+import discord
+from dotenv import load_dotenv
+import re
+from pool import PlaywrightPool
+import trafilatura
+import io
+
+from playwright.async_api import async_playwright, Browser, BrowserContext, Page
+
+load_dotenv()
+
+DISCORD_TOKEN: Final[str] = os.getenv("DISCORD_TOKEN")
+
+ROLE_NAME = "Newsulizer"
+
+intents = discord.Intents.default()
+intents.message_content = True
+
+bot = discord.Client(intents=intents)
+
+LOGGER = logging.getLogger("Newsulizer")
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(name)s: %(message)s",
+)
+
+async def send_text_file(channel: discord.abc.Messageable, content: str, filename: str = "article.md") -> None:
+    fp = io.BytesIO(content.encode("utf-8"))
+    file = discord.File(fp, filename=filename)
+    await channel.send("📄 Full article attached:", file=file)
+
+
+async def handle_article_url(message: discord.Message, url: str) -> None:
+    """
+    Placeholder: download + analyse the article here.
+
+    Currently just acknowledges receipt so you can verify the event flow.
+    """
+    LOGGER.info("Received URL from %s: %s", message.author, url)
+
+    try:
+        html = await PlaywrightPool.fetch_html(url)
+        # TODO: parse `html`, summarise, etc.
+        await message.channel.send(f"✅ Article downloaded – {len(html):,} bytes.")
+        await send_text_file(message.channel, trafilatura.extract(html, output_format='markdown', include_images=True, include_formatting=True, include_comments=False, favor_recall=True))
+    except:
+        LOGGER.exception("Playwright failed")
+        await message.channel.send("❌ Sorry, I couldn't fetch that page.")
+
+
+def extract_first_url(text: str) -> Optional[str]:
+    """Return the first http(s)://… substring found in *text*, or None."""
+
+    match = re.search(r"https?://\S+", text)
+    return match.group(0) if match else None
+
+@bot.event
+async def on_ready() -> None:
+    LOGGER.info("Logged in as %s (id=%s)", bot.user, bot.user.id)
+    await PlaywrightPool.start()
+    LOGGER.info("Playwright pool ready")
+    LOGGER.info("------")
+
+@bot.event
+async def on_message(message: discord.Message) -> None:
+    # Ignore our own messages
+    if message.author == bot.user:
+        return
+
+    is_dm = message.guild is None
+
+    overwrite = False
+    if not is_dm:
+        role = discord.utils.get(message.guild.roles, name=ROLE_NAME)
+        if role is None:
+            # The role doesn't even exist in this server
+            await message.channel.send(f"Warning! Role **{ROLE_NAME}** not found in this server.")
+            return
+
+        overwrite = role in message.channel.overwrites
+
+    # Either a DM or a channel message that mentions the bot
+
+    is_mention = (bot.user in message.mentions if message.guild else False) or overwrite
+
+    if not (is_dm or is_mention):
+        return
+
+    url = extract_first_url(message.content)
+    if not url:
+        await message.channel.send("Please send me a link to a news article.")
+        return
+
+    await message.channel.send(f"🔗 Thanks, <@{message.author.id}>! I’ve queued that article for analysis.")
+
+    # Launch the processing task without blocking Discord’s event loop
+    asyncio.create_task(handle_article_url(message, url))
+
+def main() -> None:
+    if DISCORD_TOKEN is None:
+        raise RuntimeError("Set the DISCORD_TOKEN environment variable or add it to a .env file.")
+
+    try:
+        bot.run(DISCORD_TOKEN)
+    finally:
+        asyncio.run(PlaywrightPool.stop())
+
+if __name__ == "__main__":
+    main()

news/pool.py

@@ -0,0 +1,57 @@
+from playwright.async_api import async_playwright, Browser, BrowserContext, Page
+from typing import Final, Optional
+import asyncio
+
+class PlaywrightPool:
+    _pw = None           # playwright instance
+    _browser: Optional[Browser] = None
+    _ctx: Optional[BrowserContext] = None
+    _sema: asyncio.Semaphore  # limit parallel pages
+
+    @classmethod
+    async def start(cls, max_concurrency: int = 4) -> None:
+        if cls._pw is not None:
+            return
+
+        cls._pw = await async_playwright().start()
+        cls._browser = await cls._pw.chromium.launch(
+            headless=True,
+            args=["--disable-blink-features=AutomationControlled"],
+        )
+        cls._ctx = await cls._browser.new_context(
+            user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0 Safari/537.36",
+            viewport={"width": 1280, "height": 800},
+        )
+        cls._sema = asyncio.Semaphore(max_concurrency)
+
+    @classmethod
+    async def new_context(cls) -> None:
+        if cls._ctx:
+            await cls._ctx.close()
+        cls._ctx = await cls._browser.new_context(
+            user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0 Safari/537.36",
+            viewport={"width": 1280, "height": 800},
+        )
+
+    @classmethod
+    async def stop(cls) -> None:
+        if cls._ctx:
+            await cls._ctx.close()
+        if cls._browser:
+            await cls._browser.close()
+        if cls._pw:
+            await cls._pw.stop()
+
+    @classmethod
+    async def fetch_html(cls, url: str) -> str:
+        if cls._browser is None:
+            await cls.start()
+
+        async with cls._sema:                       # throttle concurrency
+            page: Page = await cls._ctx.new_page()
+            try:
+                await page.goto(url, wait_until="load", timeout=60_000)
+                html = await page.content()
+                return html
+            finally:
+                await page.close()

news/test.py

@@ -0,0 +1,16 @@
+from pool import PlaywrightPool
+import asyncio
+import trafilatura
+
+article = "https://financialpost.com/news/economy/ontario-and-quebec-economies-to-be-hardest-hit-this-year-by-trade-war-deloitte"
+
+async def main() -> None:
+    await PlaywrightPool.start()
+
+    html = await PlaywrightPool.fetch_html(article)
+
+    print(trafilatura.extract(html))
+
+    await PlaywrightPool.stop()
+
+asyncio.run(main())