[publican-list] [Bug 616142] New: RFC: building only part of a book by altering xrefs

[Douglas Silas]

Hi,

On 08/02/2010 02:03 AM, Jeffrey Fearn wrote:
> Darrin Mison wrote:
>> Actually I would like this feature for a different usecase:
>> debugging.
>>
>> It's useful to be able to comment out includes from the root document
>> xml to quickly narrow down the source of the problem, but with
>> projects
>> that have lots of xrefs it's a nightmare.
>
> Any reason you can't use conditions as per my original response?

I believe Jeff's suggestion would be implemented like this:

<para
   condition="RPM">For more information on checking package 
signatures, refer to <xref
     linkend="s1-check-rpm-sig"/>.</para>
<para
   condition="NoRPM">For more information on checking package 
signatures, refer to the forthcoming &NEXTVER; edition of the 
&MAJOROS; <citetitle
     condition="NoRPM">&BOOKTITLE;</citetitle>.</para>

Obviously, calling the conditions "ch1", "ch2" and so on would be a 
bad idea; instead, they should be meaningfully echo the names of 
chapters or sections.

I had overlooked that multiple conditions can be set simultaneously 
(there's only a brief mention of that in one example in the Publican 
Users Guide).

Jeff's solution works, and conforms with the DocBook DTD. However:

* it can't be done ad-hoc (it would probably be overkill just for 
debugging)
* it busies up the XML markup
* in the case of large books, would add 40-50+ conditions to the 
build (two for each chapter, such as "RPM" and "NoRPM" above)
* renaming a chapter (which happens occasionally) would suggest 
renaming the associated conditions everywhere they appear in the 
chapters
* would force you to change the conditions set in publican.cfg quite 
often
* would probably make maintaining and debugging the xrefs a nightmare

Question: is there an upper limit on the number of conditions that 
can be set? This would be good to know & document.

On the whole, using conditions for this purpose is probably not 
worth the effort and trouble. That is why I was looking for a more 
Publican-centric solution.

Thanks,

Silas
>
> Modifying the DTD to be incompatible with upstream is a huge step in
> the wrong direction IMHO. Doubly so since using conditions is a
> pretty simple work around, again IMHO.
>
> Cheers, Jeff.
>