From 3cdeca91ed57820cba3abecb306e3f620cfc5a02 Mon Sep 17 00:00:00 2001
From: "Karina J. Kwiatek" <kkwiatek@studentrobotics.org>
Date: Tue, 6 Aug 2024 01:14:38 +0200
Subject: [PATCH] Send blog posts from RSS feed into text channel

---
 requirements.in  |  2 ++
 requirements.txt |  8 +++++++
 src/bot.py       |  2 +-
 src/constants.py |  4 ++++
 src/main.py      | 13 ++++++++++-
 src/rss.py       | 58 ++++++++++++++++++++++++++++++++++++++++++++++++
 6 files changed, 85 insertions(+), 2 deletions(-)
 create mode 100644 src/rss.py

diff --git a/requirements.in b/requirements.in
index 733ecc5..c1aa6c0 100644
--- a/requirements.in
+++ b/requirements.in
@@ -1,2 +1,4 @@
 python-dotenv
 discord
+feedparser
+beautifulsoup4
diff --git a/requirements.txt b/requirements.txt
index 2e61730..fe925c5 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -14,10 +14,14 @@ async-timeout==4.0.3
     # via aiohttp
 attrs==24.1.0
     # via aiohttp
+beautifulsoup4==4.12.3
+    # via -r requirements.in
 discord==2.3.2
     # via -r requirements.in
 discord-py==2.4.0
     # via discord
+feedparser==6.0.11
+    # via -r requirements.in
 frozenlist==1.4.1
     # via
     #   aiohttp
@@ -30,5 +34,9 @@ multidict==6.0.5
     #   yarl
 python-dotenv==1.0.1
     # via -r requirements.in
+sgmllib3k==1.0.0
+    # via feedparser
+soupsieve==2.5
+    # via beautifulsoup4
 yarl==1.9.4
     # via aiohttp
diff --git a/src/bot.py b/src/bot.py
index 01ad9db..9b747cc 100644
--- a/src/bot.py
+++ b/src/bot.py
@@ -4,7 +4,7 @@
 import logging
 
 from constants import *
-from typing import Tuple, AsyncGenerator
+from typing import Tuple, AsyncGenerator, Optional
 
 
 class BotClient(discord.Client):
diff --git a/src/constants.py b/src/constants.py
index de56ee5..0ea1d5a 100644
--- a/src/constants.py
+++ b/src/constants.py
@@ -17,3 +17,7 @@
 SPECIAL_ROLE = "unverified-volunteer"
 
 PASSWORDS_CHANNEL_NAME = "role-passwords"
+
+FEED_URL = "https://studentrobotics.org/feed.xml"
+FEED_CHANNEL_NAME = "blog"
+FEED_CHECK_INTERVAL = 10  # seconds
diff --git a/src/main.py b/src/main.py
index 9c159bd..e38e489 100644
--- a/src/main.py
+++ b/src/main.py
@@ -1,9 +1,11 @@
+import asyncio
 import os
 import sys
 import logging
 from dotenv import load_dotenv
 
 from bot import BotClient
+from rss import post_check_timer
 
 logger = logging.getLogger('srbot')
 logger.setLevel(logging.INFO)
@@ -13,4 +15,13 @@
 
 load_dotenv()
 bot = BotClient(logger=logger)
-bot.run(os.getenv('DISCORD_TOKEN'))
+loop = asyncio.get_event_loop()
+
+try:
+    loop.create_task(post_check_timer(bot))
+    loop.run_until_complete(bot.start(os.getenv('DISCORD_TOKEN')))
+except KeyboardInterrupt:
+    loop.run_until_complete(bot.close())
+    # cancel all tasks lingering
+finally:
+    loop.close()
diff --git a/src/rss.py b/src/rss.py
new file mode 100644
index 0000000..710bb8d
--- /dev/null
+++ b/src/rss.py
@@ -0,0 +1,58 @@
+import asyncio
+from typing import Optional
+
+import discord
+import feedparser
+from bs4 import BeautifulSoup
+from feedparser import FeedParserDict
+
+from src.bot import BotClient
+from src.constants import FEED_URL, FEED_CHECK_INTERVAL, FEED_CHANNEL_NAME
+
+
+def get_feed_channel(bot: BotClient) -> discord.TextChannel:
+    for channel in bot.get_all_channels():
+        if channel.name == FEED_CHANNEL_NAME:
+            return channel
+
+
+async def get_last_blog_post(channel: discord.TextChannel) -> str | None:
+    # TODO: This doesn't work when the bot is restarted, store the URL instead
+    last_message: Optional[discord.Message] = channel.last_message
+    if last_message is not None and len(last_message.embeds) > 0:
+        return last_message.embeds[0].url
+
+    return None
+
+
+async def check_posts(bot: BotClient):
+    feed = feedparser.parse(FEED_URL)
+    channel = get_feed_channel(bot)
+    post = feed.entries[0]
+    newest_post_url = post.link
+    last_message_url = await get_last_blog_post(channel)
+    if newest_post_url != last_message_url:
+        await channel.send(embed=create_embed(post))
+
+
+def create_embed(post: FeedParserDict) -> discord.Embed:
+    soup = BeautifulSoup(post.content[0].value, 'html.parser')
+
+    embed = discord.Embed(
+        title=post.title,
+        type="article",
+        url=post.link,
+        description=soup.p.text,
+    )
+
+    if len(post.media_thumbnail) > 0:
+        embed.set_image(url=post.media_thumbnail[0]['url'])
+
+    return embed
+
+
+async def post_check_timer(bot: BotClient):
+    await bot.wait_until_ready()
+    while True:
+        await check_posts(bot)
+        await asyncio.sleep(FEED_CHECK_INTERVAL)