SummaryGenerates fragment nodes from all HTML headers which have an id attribute set
Short namefragments
Content typetext
Provided by bundlebuilt-in
API docWebgen::ContentProcessor::Fragments


This processor generates nested fragment nodes from all HTML headers h1 to h6 which have an id attribute set.

This is only done for the block named content or if no block is associated with the given render context! This is to ensure that fragment nodes are not created from multiple blocks of one file. So this content processor has no effect when used in the pipeline of a block that is not named ‘content’.


This processor is best only included in the processing pipeline of page files, not template files, because template files may have headers with ids for which no fragment nodes should be created (e.g. the website title or a website slogan).

The reason why only header tags with an id attribute are used is that only those can be referenced and linked to later.

Since only headers with an id attribute are used, you might want to use a markup processor like kramdown which automatically generates an id attribute for all headers. If you use another markup language or plain old HTML, you might need to set the id attributes by hand.

The generated fragment nodes can be used like any other node. So you can link to them and use them in a menu. Note that although the fragment nodes are nested according to the nesting level of their representative headers their ALCN and destination path are constructed as if they were child nodes of the file.


Assume that we have the following page file with the ALCN /example.en.html:

--- name:content pipeline:kramdown,fragments
# Heading 1

## Heading 11

## Heading 12

# Heading 2

## Heading 21

The “content” block gets processed by the content processor kramdown first which changes the content to:

<h1 id="heading-1">Heading 1</h1>

<h2 id="heading-11">Heading 11</h2>

<h2 id="heading-12">Heading 12</h2>

<h1 id="heading-2">Heading 2</h1>

<h2 id="heading-21">Heading 21</h2>

Then the fragments processor uses this result to generate the following fragment nodes (the nesting of the list items corresponds to the nesting of the nodes):

  • /example.en.html#heading-1
    • /example.en.html#heading-11
    • /example.en.html#heading-12
  • /example.en.html#heading-2
    • /example.en.html#heading-21

You can see that although the fragment nodes are nested their CNs are directly appended to the ALCN of the page file (which is the correct thing to do).