[Digikam-devel] xml2pot / docmessages / docbook bug

[Burkhard Lück]

Hi,

thanks to Chusslove's commit #987834 for extract-doc-assoc-ui-cats.py errors 
are printed properly to stderr now, so I noticed some problems creating these 
docmessages templates:

extragear-utils/krusader/java.docbook
extragear-utils/krusader/installation.docbook              

playground-office/kmymoney/details-widgets.docbook

extragear-graphics/digikam/credits-annex.docbook
extragear-graphics/digikam/sidebar.docbook

The xml parser called in extract-doc-assoc-ui-cats.py throughs this error 
message:
"parse error: junk after document element: line xxx, column yy"

Reason is the wrong splitting of the documentaion with a structure like e.g in 
digikam/sidebar.docbook:
<sect2>
content1 ...
</sect2>
<sect2>
content2 ...
</sect2>
<sect2>
content3 ...
</sect2>

xml2pot extracts only the messages from the first <sect2> ... </sect2> and 
ignores the messages from the other <sect2> (content2 / content3) silently.
That means, a few hundred messages are not extracted and remain untranslated 
in english in the language docbooks.

There is no problem with checkXML and meinproc4, the html documentation is 
properly generated building the english documentation or a language 
documentation. All language documentations only have some untranslatable 
parts, which are displayed as english text.

(/me is surprised, that no one noticed this so far, sidebar.docbook is more 
than two years old ...)

There are two ways to solve this problem:

1. change xml2pot (I'm not able to do that)

2. change the relevant english docbooks (in consultation with the devel teams 
of course), but that means all translated messages from these docbooks will 
get lost.

I'd prefer solution no 2, this is the proper way.

Any objections?

-- 
Burkhard Lück