Skip to content

Posts rss feeds or top reddit posts to telegram channels

License

Notifications You must be signed in to change notification settings

zenwarr/telegram-feed-bot

Repository files navigation

This bot posts rss feeds to telegram channels. It can also collect top posts from a subreddit and post them to a telegram channel hourly.

It can save original post formatting (to the extent supported by Telegram).

License

Feed configuration

Configuration is stored in data/config.yaml file. Example configuration:

feeds:
- url: https://xkcd.com/rss.xml # RSS feed url

  # Handle of a telegram channel to post to.
  # You can reuse same channel for multiple feeds.
  channel: "@xkcd"

  # Optional, `generic` by default
  # Filter to convert a feed entry to a telegram message
  filter: xkcd

  # Optional, none by default
  # Regexp to not post messages not matching it.
  # Matching is done against an already generated telegram message, not source html provided in a feed entry.
  should_match: "^xkcd"

  # Optional, none by default
  # Regexp to not post messages matching it. It can be an array of patterns.
  # Matching is done against an already generated telegram message, not source html provided in a feed entry.
  should_not_match: "^xkcd"
  
  # Optional, none by default
  # Prefix is added to the message title and can be useful if you post multiple feeds to the same telegram channel
  # Title with a prefix looks like this: `[XKCD] Original entry title`
  title_prefix: "XKCD"
  
  # Optional, true by default
  # If set to false, disables link previews for telegram messages 
  link_preview: true
  
  # Optional, true by default
  # If set to true, text in the title is going to be a clickable link to an original post.
  # Setting this to true automatically disables footer.
  title_link: true
  
  # Optional, false by default
  # If true, telegram messages will have footer with clickable link to an original post
  footer: true
  
  # Optional, none by default
  # If set, links to original posts for this feed are going to be replaced with links to
  # t.me/iv?url=original_link&rhash=instant_view_rhash, so you can use your own instant view templates
  instant_view_rhash: "some_rhash"

reddit:
- name: funny # Name of a subreddit
  channel: "@funny" # Handle of a telegram channel to post to. You can reuse same channel for multiple subreddits.

Filters are functions that convert incoming rss feed entry to telegram message. generic filter is used by default. It tries its best to extract content from a feed entry and convert it to simple markdown Telegram understands.

There are two types of filters: built-in and custom. Built-in filters are located in src/builtin_filters directory — each file here exports a function named content_filter and the name of the file is a name of the filter. If a filter with given name is not found in src/builtin_filter, application tries to find it in src/custom_filters directory. You can use it to create your custom filters (take a look at a filter in src/builtin_filter for example).

If you are using default docker-compose.yml config for deploy, you do not need to modify code in the repository. filters directory is mounted into src/custom_filters, so you can just place your custom filters there.

Application does not need to be restarted to pick up a new configuration. Updated config is going to be used on next update.

Reusing config

You can reuse configuration for multiple feeds with YAML links:

configs:
  feed: &feed
    channel: "-129837198239123"
    link_preview: no

feeds:
- url: https://example.com/rss.xml
  title_prefix: Example
  << : *feed

Bot configuration

You have to set bot token in data/.env file. For example:

TELEGRAM_BOT_TOKEN=my_telegram_bot_token

Deployment

You can use an example provided in prod.docker-compose.yml file for deployment. Create docker-compose.yml file and paste contents from prod.docker-compose.yml file into it. You should also create data (for keeping data/.env and data/config.yaml) and filters (for keeping custom filters) directories.

After setting up feeds in config and configuring bot token, you can start the bot with docker-compose.

docker-compose up --build