[PATCH] XML validity of kcfg files

[Christian Mueller]

Am Sonntag, 31. Oktober 2004 16:40 schrieb Zack Rusin:
> I don't see the point... At all... Do you use pclint to validate all the 
> C++ files that are in CVS? Unless you start doing that, then making 
> other source files reside in CVS on special rules is silly. kcfg files 
> are just source files. I really don't get the whole discussion. I mean 
> checking the validitiy of the kconfig xt files is really very simply:
> - compile the app 
>    -> it compiles - the file format is correct as far as we're 
> concerned,
>    -> it doesn't compile - the file format is not correct as far as 
> we're concerned.
> And there it is validated.


Michael Pyne raised a valid point, IMHO: 

http://lists.kde.org/?l=kde-core-devel&m=109921486830566&w=2

But sure, it's a tradeoff between catching hidden errors 
and effort. 


 
> > Sure, kcfgcreator should be fixed before the old results are.
> >
> > But people *are* going to hand-edit those files.
> 
> How many people besides Ian and me have you ever seen editting Qt ui 
> files by hand? And really non-compilation of an app is a rather good 
> indication of the ui file non-conforming to the syntax.

I was talking about kcfg files.  And I'm sure people are hand-editing 
them because I've seen one in kdebase that was not even well-formed
(this has been fixed by my commit earlier today).  I'm surprised 
kconfig_compiler works on that. 


 
> Let me put it this way: I don't care. I'm not going to spent even a 
> second on this. If you feel that validating XML is worthy of your time, 
> knock yourself out. But if you're validating I'd suggest using pclint 
> over all files in the CVS, this way you could actually fix a bug.


That's a clear answer. Thanks.  Given my limited C++/Qt/KDE devel experience 
(and also time, I can only do so much, that's why I picked those small 
things to do) I surely won't touch either kcfgcreator or kconfig_compiler 
without the support of the author or with him thinking the whole effort 
is not even worthwhile. 

The case is closed as far as I am concerned. 
Benjamin, would you please remove this check from your test suite? 
Or at least add a note that explains the situation. 


Christian. 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://mail.kde.org/pipermail/kde-core-devel/attachments/20041031/f3cff5be/attachment.sig>