Skip to content

Configuration

To produce a valid RSS feed, the plugin uses:

MkDocs configuration#

Setting Expected level Corresponding RSS element
site_description required description
site_name required title and also as source URL label for each feed item
site_url required link
---- ---- ----
repo_url recommended docs
site_author recommended managingEditor
---- ---- ----
copyright optional copyright
theme/locale (or theme/language for Material theme) optional language

Automatic elements#

Variable / value Corresponding RSS element
MkDocs build timestamp optional channel elements lastBuildDate and pubDate
MkDocs RSS plugin - v1.0.0 optional channel elements generator

Page attributes#

Basically, each page is an item element in the RSS feed.

Attribute Expected level Corresponding RSS element
page.canonical_url required and optional mandatory item element link and also used as guid
page.meta.title required title
page.meta.description if present, else extract headlines from the content. See below the item_description_length option. required description
creation or last update datetime according git log. Can be overridden by dates in page.meta. If not, then it uses MkDocs build date recommended pubDate
page.meta.image optional item element enclosure. Some HTTP requests can be performed to retrieve remote images length.
page.meta.authors or page.meta.author. Accepted value types: str list, tuple.
To comply with the standard, the page writer is responsible to fill this field following this syntax: john@doe.com (John Doe) (read this SO).		 optional author
page.meta.tags or any tags value (for example page.meta.categories...). Accepted value types: str list. optional category

Item image (enclosure)#

To add an image to a feed item as enclosure, the page writer is responsible to fill the page.meta.image. The plugin tries to retrieve the image length and mime-type to complete the enclosure tag.

Accepted keys:

  • image: preferred
  • illustration: alternative to make it easier to comply with some themes

Accepted values:

  • remote image URL: must be reachable from the build environment.
  • relative path to the image: the plugin adds the site_url to the path to ensure that image will be reachable by external feed readers once the site is published.

Examples#

Local image#

mkdocs.yml :

site_url: https://blog.mydomain.com

Page:

---
[...]
image: "featured_images.png"
---

# Page h1 title

Some page text.

Output:

<item>
  [...]
  <title>Page h1 title</title>
  <enclosure url="https://blog.mydomain.com/featured_images.png" type="image/png" length="219753"/>
  [...]
</item>
Remote image#
---
[...]
image: "http://example.com/image.jpg"
---

# Page h1 title

Some page text.

Output:

<item>
  [...]
  <title>Page h1 title</title>
  <enclosure url="http://example.com/image.jpg" type="image/jpg" length="19753"/>
  [...]
</item>

Plugin options#

For a sample see homepage.

enabled: enabling/disabling the plugin#

You can use the enabled option to optionally disable this plugin. A possible use case is local development where you might want faster build times. It's recommended to use this option with an environment variable together with a default fallback (introduced in mkdocs v1.2.1, see docs). Example:

plugins:
  - rss:
      enabled: !ENV [MKDOCS_ENABLE_RSS_PLUGIN, True]

Which allows you to disable the plugin locally using:

export MKDOCS_ENABLE_RSS_PLUGIN=false
mkdocs serve

json_feed_enabled: enabling/disabling export to JSON Feed#

Set it to false if you want to only export to RSS.

Default: true.

rss_feed_enabled: enabling/disabling export to RSS#

Set it to false if you want to only export to JSON feed.

Default: true.

abstract_chars_count: item description length#

Used, in combination with abstract_delimiter, to determine each item description element:

  • If this value is set to -1, then the articles' full HTML content will be filled into the description element.
  • If you want to customize the description per each Markdown page, refer to the example below.
  • Otherwise, the plugin first tries to retrieve the value of the keyword description from the page metadata.
  • If that fails and abstract_delimiter is found in the page, the article content up to (but not including) the delimiter is used.
  • If the above has failed, then the plugin retrieves the first number of characters of the page content defined by this setting. Retrieved content is the raw markdown converted roughly into HTML.

Be careful: if set to 0 and there is no description, the feed's compliance is broken (an item must have a description).

Override feed description per page#

To customize the value of the RSS description per each page and override the value of site_description and plugins.rss.feed_description, you can modify the value per each page as you see in the example below:

---
date: 2024-06-24
description: >-
  This is the SEO description.
rss:
  feed_description: >-
    And I want to have customized RSS description.
---

abstract_chars_count: number of characters to use as item description.

Default: 150

abstract_delimiter: abstract delimiter#

Please see abstract_chars_count for how this setting is used. A value of "" (the empty string) disables this step.

abstract_delimiter: string to mark where the description ends.

Default: <!-- more -->

cache_dir: folder where to store plugin's cached files#

The plugin implements a caching mechanism, ensuring that a remote media is only get once during its life-cycle on remote HTTP server (using Cache Control under the hood). It is normally not necessary to specify this setting, except for when you want to change the path within your root directory where HTTP body and metadata files are cached.

If you want to change it, use:

plugins:
  - rss:
      cache_dir: my/custom/dir

It's strongly recommended to add the path to your .gitignore file in the root of your project:

.gitignore
.cache

Default: .cache/plugins/rss.

categories: item categories#

categories: list of page metadata values to use as RSS item categories.

Default: None.

Example#

In configuration:

- rss:
    categories:
      - tags        # will look into page.meta.tags
      - categories  # will also look into page.meta.categories

In page 1:

---
title: "Lorem Ipsum 1"
tags:
  - tag x
  - tag Y
---

[...]

In page 2

---
title: "Page 2"
categories: ["Release notes", "test"]
---

[...]

Output:

  [...]
  <item>
    <title>Lorem Ipsum 1</title>
    <category>tag x</category>
    <category>tag Y</category>
    [...]
  </item>
  <item>
    <title>Page 2</title>
    <category>Release notes</category>
    <category>test</category>
    [...]
  </item>
  [...]

comments_path: item comments path#

comments_path: path to add to each item URL pointing.

Default: None.

For example, if you're using Material for Mkdocs with comment integration (Disqus or Isso), the comment block is under the div id __comments, so you can set: comments_path: "#__comments" and the output will be:

<item>
  <title>This page title is a perfect clickbait!</title>
  <link>https://website.com/articles/best_article/</link>
  <comments>https://website.com/articles/best_article/#__comments</comments>
  [...]

</item>

date_from_meta: override dates from git log with page.meta#

Basically, the plugin aims to retrieve creation and update dates from git log. But sometimes, it does not match the content workflow: markdown generated from sources, .

So, it's possible to use the dates manually specified into the page metadata through the YAML frontmatter.

  • as_creation: meta tag name (or a dot-separated tag name) to use as creation date. Default to False.
  • as_update: meta tag name (or a dot-separated tag name) to use as update date. Default to False.
  • datetime_format: datetime format. Default to "%Y-%m-%d %H:%M".
  • default_timezone: timezone to use by default to make aware datetimes. Default to UTC. Introduced in version 1.3.0 with PR 142.
  • default_time: time to use if page contains only a date. Useful to avoid the 'midnight syndrome' or have to specify hour in every single page. Default to None. 24h-clock format is expected: %H:%M. Example: "14:20". Introduced in version 1.4.0 with PR 145.

Example#

For example, in your best_article.md created in 2019, you can write the front-matter like this:

---
title: "This page title is a perfect clickbait!"
authors:
  - "Julien M."
date: "2020-10-22 17:18"
---

# This plugin will change your MkDocs life

Lorem ipsum [...]

So in your mkdocs.yml you will have:

plugins:
  - rss:
      date_from_meta:
        as_creation: "date"
        as_update: "git"
        datetime_format: "%Y-%m-%d %H:%M"
        default_timezone: Europe/Paris

---
title: "This page title is a perfect clickbait!"
authors:
  - "Julien M."
date:
  created: "2020-10-22 17:18"
---

# This plugin will change your MkDocs life

Lorem ipsum [...]

So in your mkdocs.yml you will have:

plugins:
  - rss:
      date_from_meta:
        as_creation: "date.created"
        as_update: "git"
        datetime_format: "%Y-%m-%d %H:%M"
        default_timezone: Europe/Paris

At the end, into the RSS you will get:

<item>
  <title>This page title is a perfect clickbait!</title>
  <link>https://website.com/articles/best_article/</link>
  <pubDate>Thu, 22 Oct 2020 17:18:00 +0200</pubDate>
  [...]

</item>

Timezone dependencies

The timezones data depends on the Python version used to build: - for Python >= 3.9, it uses the standard library and ships tzdata only on Windows which do not provide such data - for Python < 3.9, pytz is shipped.

feed_description: override site description#

This option allows you to override the default MkDocs site description for the description tag in this feed. This is useful if you have multiple instances of this plugin for multiple feeds. (For example, one feed for the blog, and a second for documentation updates.)

This setting is optional. If you do not include it, the default site description will be used.

plugins:
  - rss:
      feed_description: The best blog from the best site

Default: Use the default MkDocs site_description:.

feed_title: override site title#

This option allows you to override the default MkDocs site name for the title tag in this feed. This is useful if you have multiple instances of this plugin for multiple feeds. (For example, one feed for the blog, and a second for documentation updates.)

This setting is optional. If you do not include it, the default site name will be used.

plugins:
  - rss:
      feed_title: My awesome blog feed

Default: Use the default MkDocs site_name:.

feed_ttl: feed's cache time#

feed_ttl: number of minutes to be cached. Inserted as channel ttl element. See: W3C RSS 2.0 documentation.

Default: 1440 (= 1 day)

Output:

<ttl>1440</ttl>

feeds_filenames: customize the output feed URL#

Since version 1.13.0.

Customize every feed filenames generated by the plugin:

mkdocs.yml with custom RSS and JSON feeds names.
plugins:
  - rss:
      feeds_filenames:
        json_created: feed.json
        json_updated: feed-updated.json
        rss_created: rss.xml
        rss_updated: rss-updated.xml

Default:

  • JSON feed for created items: feed_json_created.json
  • JSON feed for updated items: feed_json_updated.json
  • RSS feed for created items: feed_rss_created.json
  • RSS feed for updated items: feed_rss_updated.json

image: set the channel image#

image: URL to image to use as feed illustration.

Default: None.

Example#

plugins:
  - rss:
      image: https://upload.wikimedia.org/wikipedia/commons/thumb/4/43/Feed-icon.svg/128px-Feed-icon.svg.png

Output:

<image>
  <url>
    https://upload.wikimedia.org/wikipedia/commons/thumb/4/43/Feed-icon.svg/128px-Feed-icon.svg.png
  </url>
  <title>MkDocs RSS Plugin</title>
  <link>https://guts.github.io/mkdocs-rss-plugin/</link>
</image>

length: number of items to include in feed#

length: number of pages to include as feed items (entries).

Default: 20

match_path: filter pages to include in feed#

This adds a match_path option which should be a regex pattern matching the path to your files within the docs_dir. For example if you had a blog under docs/blog where docs_dir is docs you might use:

plugins:
  - rss:
      match_path: "blog/.*"

Since match_path gives you all the power of regular expressions you can have more complex patterns to include multiple directories. For example, to include all pages under both release-notes and articles:

plugins:
  - rss:
      match_path: "(release-notes|articles)/.*"

Default: .*.

pretty_print: prettified XML#

By default, the output file is minified, using Jinja2 strip options and manual work. It's possible to disable it and prettify the output using pretty_print: true.

plugins:
  - rss:
      pretty_print: true

Default: False.

url_parameters: additional URL parameters#

This option allows you to add parameters to the URLs of the RSS feed items. It works as a dictionary of keys/values that is passed to Python urllib.parse.urlencode. One possible use case is the addition of Urchin Tracking Module (UTM) parameters:

plugins:
  - rss:
      url_parameters:
        utm_source: "documentation"
        utm_medium: "RSS"
        utm_campaign: "feed-syndication"

Will result in:

<?xml version="1.0" encoding="UTF-8" ?>
<rss>
    [...]
    <item>
      [...]
      <link>https://guts.github.io/mkdocs-rss-plugin/?utm_source=documentation&amp;utm_medium=RSS&amp;utm_campaign=feed-syndication</link>
      [...]
    </item>
    [...]
  </channel>
</rss>

Default: None.

use_git: enable/disable git log#

If false, the plugin does not try use the git log nor does not check if this is a valid git repository and use informations exclusively from page.meta (YAML frontmatter).

Useful if you build your documentation in an environment where you can't easily install git.

Default: true.

use_material_social_cards: enable/disable integration with Material Social Cards plugin#

If false, the integration with Social Cards is disabled.

Default: true.

See the related section in integrations page.