Re: A common base exception object for Perl - RFC

[Nick Ing-Simmons <nik@tiuk.ti.com>]

Pete Jordan <pjordan1@email.mot.com> writes:
>Nick Ing-Simmons wrote:
> 
>> If C code just does what it does now all the way down to the
>> hunk(s) that do(es) the longjmp() to actually "thow" the exception
>> then we can just convert the string at that point into an object.
>> Sounds very do-able
>
>Yeah, that should be straightforward once I've sussed out how to create
>an object in C. 

SV *MakeObject(SV *raw, char *class)
{
 SV *sv = newRV_noinc(raw);  /* typically you don't want to inc refcount */
 HV *stash = gv_stashpv(class, TRUE); /* Find stash for class name */
 return sv_bless(sv, stash);  /* bless raw into class via new RV */
}


>What I was thinking about was providing a call to allow
>XS programmers to raise a non-default exception. Syntax really, given
>C's limitations.

Yes passing desired class down through the croak/die/... chain 
would be a pain. And given the fact it is going to die the 
way I usually do it (re-bless on return) is a non-starter.

So we would need a new API which had extra "desired class" args.



>
>Pete
-- 
Nick Ing-Simmons <nik@tiuk.ti.com>
Via, but not speaking for: Texas Instruments Ltd.

---

Follow-Ups from:

Pete Jordan <pjordan1@email.mot.com>

References to:

"Redford, John" <John.Redford@fmr.com>
Pete Jordan <pjordan1@email.mot.com>
Stephen Zander <gibreel@pobox.com>
Pete Jordan <pjordan1@email.mot.com>
Nick Ing-Simmons <nick@ing-simmons.net>
Pete Jordan <pjordan1@email.mot.com>

---

[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index][Thread Index][Top&Search][Original]

---