#312 DocBook Transclusion: allow 'prefix' in trans:idfixup

[Thomas Schraitle]

Problem

The "DocBook Transclusion" paper[1] defines some attributes (trans:idfixup, trans:suffix, and trans:linkscope) which are used for ID fixup in an (x)included XML document.

The trans:idfixup attribute can only have one of the values "none", "suffix", or "auto" (see B.5 "Manually assigned suffix"[2] for a more exhaustive example).

Using the keyword suffix in trans:idfixup, an transclusion processor places the value of trans:suffix after the resulting IDs. For example, lets assume the following xi:include element:

<xi:include href="x.xml" 
              trans:idfixup="suffix"
              trans:suffix="_bar_"/>

With the above trans attributes, a former ID foo is rewritten as foo_bar_. However, to get the opposite (_bar_foo) is not possible with the current specification.

Some writers maybe have a style guide which defines IDs in a certain way.

Solution

Allow an additional value of prefix for trans:idfixup with the following definition:

• value prefix
  The prefix property is set to the concatenation of the value specified in the trans:suffix attribute and the
  inherited suffix value.

Resources

[1] http://docbook.org/docs/transclusion/
[2] http://docbook.org/docs/transclusion/#ex.9
[3] https://lists.oasis-open.org/archives/docbook/201605/msg00007.html

[Jirka Kosek]

Hi,

during drafting stage there was originally prefix attribute instead of suffix. But discussion revealed that suffix is more friendly option as in many scenarios suffix will be autogenerated piece of uniquie garbage.

See: https://lists.oasis-open.org/archives/docbook/201411/msg00006.html
https://lists.oasis-open.org/archives/docbook-tc/201411/msg00003.html

IMHO, supporting both prefix and suffix will just put more burden on implementers, so I don't think it's good idea to introduce prefix.

Jirka