Validation of number of cols for tables
Problem description
Created a table including 3 columns. Set tgroup attribute cols="2". No problem for webhelp output. But pdf build isn't possible. The reason is unknown.
Expected behavior
It would be great to get an informational output from daps. Otherwise the search for the may take a while.
Steps to reproduce the behavior
Insert following table in doc:
<table>
<title>Beschreibung der Übersichtspanels</title>
<tgroup cols="2" align="left" colsep='1' rowsep='1'>
<thead>
<row>
<entry align="center">Panel</entry>
<entry align="center">Beschreibung</entry>
<entry align="left">...</entry>
</row>
</thead>
<tbody>
<row>
<entry align="left">Suche</entry>
<entry align="left">...</entry>
<entry align="left">...</entry>
</row>
<row>
<entry align="left">Warenkorb</entry>
<entry align="left">... </entry>
<entry align="left">...</entry>
</row>
<row>
<entry align="left">Artikelkategorien</entry>
<entry align="left">... </entry>
<entry align="left">...</entry>
</row>
</tbody>
</tgroup>
</table>
When processing the pdf build you will get something like that:
user@slgh99-dev01:/srv/www/htdocs/docs05.sd.de> daps -m "xml/SD-docs05.xml" --styleroot "/usr/share/xml/docbook/stylesheet/synergy2015-ns" -v3 pdf
[...]
Making portrait pages on A4 paper (210mmx297mm)
Height: 88.454
(cd /srv/www/htdocs/docs05.sd.de/build/.tmp; ln -sf /usr/share/xml/docbook/stylesheet/synergy2015-ns/images)
Successfully remade target file '/srv/www/htdocs/docs05.sd.de/build/.tmp/SD-docs05-fop_color_de.fo'.
Must remake target '/srv/www/htdocs/docs05.sd.de/build/SD-docs05/SD-docs05_color_de.pdf'.
(cd /srv/www/htdocs/docs05.sd.de/build/.tmp/ && /usr/share/daps/libexec/daps-fop /srv/www/htdocs/docs05.sd.de/build/.tmp/SD-docs05-fop_color_de.fo /srv/www/htdocs/docs05.sd.de/build/SD-docs05/SD-docs05_color_de.pdf )
/usr/share/daps/make/pdf.mk:174: recipe for target '/srv/www/htdocs/docs05.sd.de/build/SD-docs05/SD-docs05_color_de.pdf' failed
make: *** [/srv/www/htdocs/docs05.sd.de/build/SD-docs05/SD-docs05_color_de.pdf] Error 1
Right, we have that issue ourselves every couple of weeks. Unfortunately, it is not possible with pure-RNG to check for this issue. You can use Schematron to validate this. The DocBook 5-RNG (I think) actually even includes the relevant Schematron rules for tables. And jing (which is what DAPS uses) can validate with Schematron. Unfortunately, jing cannot validate Schematron that is embedded in RNG, it seems. And this is all we have deducted on the topic for the moment. :(
Table validation was introduced in df01d4a34