Skip to content

Configuration

Configuration is done through a config file in TOML format. The complete config options can be found in the reference.

Use the command line option -c or --config to specify the location of the config file.

pinka run -c /etc/pinka/config.toml

Activity Pub

Pinka needs to know the base URL in order to construct the full actor URL or object URL. It also needs to know the webfinger host so federated server can use webfinger to discover local actors, for example @[email protected] on Mastodon.

[activity_pub]
base_url = "https://social.example.org"
webfinger_at_host = "social.example.org"

Feed

Pinka can periodically parse RSS/Atom/JSON feeds and publish new feed items as activity. Multiple feeds can be configured and attribute to different users. The feed content can be transformed or decorated with a minijinja template.

[feeds.social-example-org]
uid = "username"
feed_url = "https://social.example.org/feed.xml"
base_url = "https://social.example.org"
template = '''
<p>{{ title.content|to_text }}</p>

<p>{{ content.body|excerpt(100) }}</p>

{% if id is startingwith "http" %}
<p><a href="{{ id }}">{{ id }}</a></p>
{% elif links %}
<p><a href="{{ links[0].href }}">{{ links[0].href }}</a></p>
{% endif %}
'''

In addition to the builtin minijinja filters, two custom filters are available:

  • to_text(string) strips all HTML tags and convert the content to plain text.
  • excerpt(string, n) Truncate a string to a certain number of words. Each CJK character is counted as one word. Last word is not truncated.