RSS Aggregator - For RSS or HTML feeds

Simple aggregator for RSS & HTML feeds for my own pleasure
You can chose to use RSS aggregator to a single RSS feed by using the rss_aggregator.py. If you however are working with HTML blog posts (such as Microsoft blogs) you can instead chose to use html_aggregator.py. Note that you have the options to chose between aggregation or appending to a aggregated_feed.xml.
hitem

Howto

1. Create a new GitHub repository and upload files:
   Create a new public GitHub repository (e.g., "rss-aggregator"). You'll store the Python script, aggregated RSS feed and the workflow file in this repository.

2. Set up GitHub Pages:
   Go to the repository settings, scroll down to the GitHub Pages section, and choose the "main" branch as the source. Save the changes, and you'll get a URL for your GitHub Pages site (e.g., https://<username>.github.io/rss-aggregator/).

3. Update the 'link' field in the script: Replace the 'link' field in the rss_aggregator.py or html_aggregator.py script with your GitHub Pages URL:

   etree.SubElement(channel, "link").text = "https://<username>.github.io/<repo name>/aggregated_feed.xml"

4. Chose and change to RSS or HTML feeds accordingly in rss_aggregator.yml

   For HTML:

   run: |
       python html_aggregator.py

   For RSS:

   run: |
      python rss_aggregator.py

5. Change the github workflow timer accordingly in rss_aggregator.yml
   The Cron job is the main one (how often it runs here on github actions). But one more such setting is that links are only stored for 365 days under name: Update processed links file in the yml to prevent processed_links.txt to grow to big.

6. Chose if you want to aggregate or append aggregated_feed.xml by setting append_mode values to true or false in RSS or HTML *.py script.
   Aggregated (default)
   This is used to ingest the latest news where the ingestion is triggered elsewhere, such as teams or slack).

   append_mode = False

   Persistent/Appending
   This is to persist processed links in aggregated_feed.xml (Used for feeds such as feedly to see all entries.
   To avoid aggregated_feed.xml to grow to big, default time to save the links is 365 days, you can adjust according to your needs.

   append_mode = True
   max_age_days = 365

7. Then you take the link generated earlier to aggregated_feed.xml and paste it in to your RSS hook or powerautomate flow (and ingest frequenzy to match your cron configuration in rss_aggregator.yml)
   Example from teams:
   
   Example from powerautomate flows:
   

Note: You may also need to set up github access token for the repo in question. Else the github action workflow will not be allowed to checkout and make pullrequests (and merge). By default it uses GITHUB_TOKEN that can be configured on your repository project: https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/enabling-features-for-your-repository/managing-github-actions-settings-for-a-repository#setting-the-permissions-of-the-github_token-for-your-repository
I set these permissions in the workflow yml file:

permissions:
 contents: write
 pages: write
 id-token: write

Include as few permissions as possible needed for your project.

Current Behavior

Based on default mode (append_mode = False)

• Script Execution: The cron job triggers the script every hour. The script fetches and processes RSS or HTML feed entries from the last 2 hours.
• Link Processing: The script writes new links to processed_links.txt and skips links that are already present, it also creates aggregated_feed.xml
• Ingestion: Depending on your settings, the ingestions timer then comes around (in teams or where you have set it up) and post the links present in aggregated_feed.xml.

Customizing Timing and Frequency

This script is set up to collect links from Microsoft blog pages based on specific timing configurations. Here is how to customize frequenzy and how to avoid duplicates to ensure complete coverage.

Timing Configuration

• Collection Window (time_threshold): Collects entries from the past 2 hours, its set up in the main *.py script.

  time_threshold = datetime.datetime.utcnow() - datetime.timedelta(hours=2)

• Cron Job Interval: Runs every hour ('0 */1 * * *'), ensuring regular updates, set up in the github workflow *.yml file.
• Ingestion Frequency: Runs every hour, aligning with the cron job to process collected entries. (This is set by you, in the ingestion part, such as teams hooks or powerautomate flow).

Recommended Settings

• Standard (Hourly Updates):

time_threshold: 2 hours 
Cron Job Interval: 1 hour 
Ingestion Frequency: 1 hour

• Extended (Monthly Updates):

time_threshold: 60 days 
Cron Job Interval: 30 days
Ingestion Frequency: 30 days

Note: First run will actually gather 60 days worth of news (or what you set time_treshold to), but every subsquent run there is filters for links that are not already present. Time_treshold need to overlap the cron job and ingestion so you dont miss anything.

if pub_date >= time_threshold and link not in processed_links:

Other

For your own sanity, if you follow this repo or deploy your own, make sure to:

If you run every hour it will be very chatty :)