Re: howto exclude a subtree in pdf generation?

[Ross Gardler]

EMMEL Thomas wrote:
Just one additional note:
I tracked that down to a lot of
<a name="123456"></a> lines in my output of doxygen.
This will end in cocoon-errors like
internal-destination or external-destination must be specified in 
basic-link

from fo:

<fo:basic-link external-destination="" text-decoration="underline" 
color="#0F3660"/>

Although I cannot prevent as far as I see to let doxygen create this
it is still an option for me to exculde the whole tree from wholesite.pdf.

There is no *really* easy way to exclude a single section from 
wholesite. That functionality is to provide, well, the whole site.

There are a few approaches you can take:

- Use a tab to define the docs you want in the aggregate PDF [1]

- add your own custom aggregate matcher to your project xmap based on 
the current wholesite.xml matcher in sitemap.xmap (which delegates to 
aggregate.xmap). What you need to do is add a new transformation step 
after the aggregation. This new transformation needs to remove the bit 
you don't want. I'd put it between the following two transformations:

<map:transform src="{lm:transform-aggregate.book.cinclude}">
  <map:parameter name="ignore" value="{1}"/>
  <map:parameter name="title" value="{conf:project-name}"/>
</map:transform>
<map:transform type="cinclude"/>

- Create a master document that uses XInclude each of the docs you want 
in the aggregate (would need to manually maintain the inclusion list)

With respect to your problem with empty <a> elements, this could be 
solved in main/webapp/skins/common/xslt/fo/document-to-fo.xsl - either 
strip the empty anchors or add a space when it is empty.

Ross


[1] http://forrest.apache.org/docs_0_90/howto/howto-pdf-tab.html


Kind regards

Thomas

EMMEL Thomas schrieb:
Hi,

I wonder whether there is an easy way to exclude a part of the 
documentation that is still generated for html-output for the 
wholesite.pdf?

In my particular case I put a doxygen-generated html-tree to forrest
which itself generates a lot of output and it is much easier to take the
latex-output from doxygen to generate a pdf than using the wholesite.pdf
in this case.
And in this case as I generate a lot of dependency graph forrest will 
just fail to create the wholesite.pdf file since it encouters the bug 
in fo(?) that prevents to add pictures to tall to pdf.
(No flow in page-sequence)

Kind regards

Thomas