compiler error parsing in kdevelop

Hi,

it seems I found the problem, the regexp for getting the
filename from the 
compiler error message is the following:

ErrorFormat( "([^: t]+):([0-9]+):([^0-9]+)", 1,
2, 3 ),

where the filename is "([^: t]+)"
Now if the filename has spaces (which may very well be) this
doesn't work 
correctly. I could simply remove the " " from the
regexp. This will fix it 
for me. Are there error messages where this might break the
parsing ?

Alex

_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel

On Sunday 20 May 2007 21:36, Alexander Neundorf wrote:
> Hi,
>
> it seems I found the problem, the regexp for getting
the filename from the
> compiler error message is the following:
>
> ErrorFormat( "([^: t]+):([0-9]+):([^0-9]+)",
1, 2, 3 ),
>
> where the filename is "([^: t]+)"
> Now if the filename has spaces (which may very well be)
this doesn't work
> correctly. I could simply remove the " " from
the regexp. This will fix it
> for me. Are there error messages where this might break
the parsing ?
>
> Alex
>
> 

We won't be able to find out until you test it, since it
sounds like you'll be 
the one writing the patch. However, I don't think there is
any error message 
where your fix would cause another bug.
-- 
Matt

_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel

On Monday 21 May 2007 8:06:46 am Alexander Neundorf wrote:
> it seems I found the problem, the regexp for getting
the filename from the
> compiler error message is the following:
>
> ErrorFormat( "([^: t]+):([0-9]+):([^0-9]+)",
1, 2, 3 ),
>
> where the filename is "([^: t]+)"
> Now if the filename has spaces (which may very well be)
this doesn't work
> correctly. I could simply remove the " " from
the regexp. This will fix it
> for me. Are there error messages where this might break
the parsing ?

Sort of OT but im curious... How good is the support for
error parsing? Can a 
user easily extend it?

I am looking forward to using kdevelop with compilers other
than gcc but i am 
not sure how well the error parser in the output view will
accept it?
-- 
Cheers!
kitts

_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel

On 21.05.07 21:01:48, kitts wrote:
> On Monday 21 May 2007 8:06:46 am Alexander Neundorf
wrote:
> > it seems I found the problem, the regexp for
getting the filename from the
> > compiler error message is the following:
> >
> > ErrorFormat( "([^:
t]+):([0-9]+):([^0-9]+)", 1, 2, 3 ),
> >
> > where the filename is "([^: t]+)"
> > Now if the filename has spaces (which may very
well be) this doesn't work
> > correctly. I could simply remove the " "
from the regexp. This will fix it
> > for me. Are there error messages where this might
break the parsing ?
> 
> Sort of OT but im curious... How good is the support
for error parsing? Can a 
> user easily extend it?

Not without changing the source and have a goood idea about
regexps.

> I am looking forward to using kdevelop with compilers
other than gcc but i am 
> not sure how well the error parser in the output view
will accept it?

We already try to support compilers besides gcc, see for
example recent
fixes to support Sun's compiler.

Andreas

-- 
Your lucky number is 3552664958674928.  Watch for it
everywhere.

_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel

On Monday 21 May 2007 9:43:34 pm Andreas Pakulat wrote:
> > Sort of OT but im curious... How good is the
support for error parsing?
> > Can a user easily extend it?
>
> Not without changing the source and have a goood idea
about regexps.

Hmmm...

> > I am looking forward to using kdevelop with
compilers other than gcc but
> > i am not sure how well the error parser in the
output view will accept
> > it?
>
> We already try to support compilers besides gcc, see
for example recent
> fixes to support Sun's compiler.

OK. I'll take a look. My interest in this case is for
embedded systems. Most 
commercial compilers that i look forward to come with their
own IDE's but i 
would be glad if we would not have to support many IDE's.
But now that i 
think about it... debugging would anyway need those
IDE's...
-- 
Cheers!
kitts

_______________________________________________
KDevelop-devel mailing list
KDevelop-develkdevelop.org
https://barney.cs.uni-potsdam.de/mailman/listinf
o/kdevelop-devel