"error-type" element needed in grammar

[Matthieu Caneill]

I use the Coccinelle [1] static analyzer (which I've instrumented to
support Firehose output), and it can process different kind of
(possibly user-defined) tests. It could be useful to add a variable in
the issue element to categorize more precisely the type of error; and
any other tool which supports different kind of tests would benefit
from such an extension. Such an element could be called "error-type"
or "check-name". What do you think about this? I can submit a pull
request if you agree with this.


--
Matthieu

[David Malcolm]

On Fri, 2013-07-05 at 17:56 +0200, Matthieu Caneill wrote:
>  I use the Coccinelle [1] static analyzer (which I've instrumented
> to
>  support Firehose output), and it can process different kind of
>  (possibly user-defined) tests. It could be useful to add a variable in
>  the issue element to categorize more precisely the type of error; and
>  any other tool which supports different kind of tests would benefit
>  from such an extension. Such an element could be called "error-type"
>  or "check-name". What do you think about this? I can submit a pull
>  request if you agree with this. 
Isn't this just the test-id attribute of <issue> ?

[Matthieu Caneill]

On Fri, Jul 5, 2013 at 7:18 PM, David Malcolm <dmalcolm(a)redhat.com&gt; wrote:
>  On Fri, 2013-07-05 at 17:56 +0200, Matthieu Caneill wrote:
> > I use the Coccinelle [1] static analyzer (which I've instrumented to
> > support Firehose output), and it can process different kind of
> > (possibly user-defined) tests. It could be useful to add a variable in
> > the issue element to categorize more precisely the type of error; and
> > any other tool which supports different kind of tests would benefit
> > from such an extension. Such an element could be called "error-type"
> > or "check-name". What do you think about this? I can submit a pull
> > request if you agree with this.
>
>  Isn't this just the test-id attribute of <issue> ? 
I didn't see this like an error type, but after re-reading its
description it seems that it's the corresponding item. Thanks!

--
Matthieu

[David Malcolm]

On Fri, 2013-07-05 at 19:27 +0200, Matthieu Caneill wrote:
>  On Fri, Jul 5, 2013 at 7:18 PM, David Malcolm
> <dmalcolm(a)redhat.com&gt; wrote:
>  > On Fri, 2013-07-05 at 17:56 +0200, Matthieu Caneill wrote:
>  >> I use the Coccinelle [1] static analyzer (which I've instrumented to
>  >> support Firehose output), and it can process different kind of
>  >> (possibly user-defined) tests. It could be useful to add a variable in
>  >> the issue element to categorize more precisely the type of error; and
>  >> any other tool which supports different kind of tests would benefit
>  >> from such an extension. Such an element could be called "error-type"
>  >> or "check-name". What do you think about this? I can submit a pull
>  >> request if you agree with this.
>  >
>  > Isn't this just the test-id attribute of <issue> ?
>  
>  I didn't see this like an error type, but after re-reading its
>  description it seems that it's the corresponding item. Thanks! 
Hmmmm... perhaps the attribute needs renaming, or a better description?

[Matthieu Caneill]

On Fri, Jul 5, 2013 at 7:30 PM, David Malcolm <dmalcolm(a)redhat.com&gt; wrote:
>  On Fri, 2013-07-05 at 19:27 +0200, Matthieu Caneill wrote:
> > On Fri, Jul 5, 2013 at 7:18 PM, David Malcolm <dmalcolm(a)redhat.com&gt; wrote:
> > > Isn't this just the test-id attribute of <issue> ?
> >
> > I didn't see this like an error type, but after re-reading its
> > description it seems that it's the corresponding item. Thanks!
>
>  Hmmmm... perhaps the attribute needs renaming, or a better description? 
Maybe "error-id"? It's the "test" that confused me at first.

--
Matthieu

[Stefano Zacchiroli]

On Fri, Jul 05, 2013 at 07:33:32PM +0200, Matthieu Caneill wrote:
>  > Hmmmm... perhaps the attribute needs renaming, or a better
> description?
>  Maybe "error-id"? It's the "test" that confused me at first.
I'd say that both "error" and "test" are prone to be fairly
ambiguous in
this content. To my reading, "check" would be clearer (but I'm not a
native speaker either, so take it with a pinch of salt). Also, "id" make
it sound like something numerical, or anyway something defined elsewhere
and only referenced here. Maybe "check-name" or "check-type" would be
a
little bit more self-describing?

Cheers.
-- 
Stefano Zacchiroli  . . . . . . .  zack(a)upsilon.cc . . . . o . . . o . o
Maître de conférences . . . . . http://upsilon.cc/zack . . . o . . . o o
Former Debian Project Leader  . . @zack on identi.ca . . o o o . . . o .
« the first rule of tautology club is the first rule of tautology club »