Propagating errors in C code to Perl

missingthepoint has asked for the wisdom of the Perl Monks concerning the following question:

Greetings, monks.

I've been tinkering with Inline::C recently, and found myself printing error messages from C directly. I consider this bad juju for all but trivial uses of C-in-Perl... So what is the best way to propagate errors at the C level to Perl? I think some of Perl's builtins (the ones which do syscalls?) set $! on error. Can I set this variable from C? Is there a better way to report errors that occur in C code?

Awaiting your wisdom,

mtp

email: perl -e 'print reverse map { chr( ord($_)-1 ) } split //, "\x0bufo/hojsfufqAofc";'

Comment on Propagating errors in C code to Perl Download Code

Replies are listed 'Best First'.
Re: Propagating errors in C code to Perl by betterworld (Curate) on Sep 05, 2008 at 13:42 UTC
Afaik you can die from C code using the "croak" function. Here's an example how to do it and set `$@`. Probably setting `$!` works ~~similar~~ (Update:) by setting errno (see Joost's post). If you throw exceptions from C code, you just have to make sure that all your dynamically allocated memory is freed, which can be a problem in callbacks from external libraries.	[reply] [d/l] [select]
Re: Propagating errors in C code to Perl by Joost (Canon) on Sep 05, 2008 at 13:50 UTC
As betterworld said: you can just croak() from C and it'll throw an exception in the calling perl code. This, or returning undef/false, is almost always what you want to do. You can set $! from C, but you should probably only do so under very limited circumstances. $! is tied directly to the C macro errno which is almost exclusively used for IO errors. Look it up. Note that errno is a number. "What should it profit a man, if he should win a flame war, yet lose his cool?"	[reply]
Re^2: Propagating errors in C code to Perl by missingthepoint (Friar) on Sep 06, 2008 at 07:08 UTC
Thanks, betterworld and Joost. Right now to return undef I'm saying: `return (SV *) &PL_sv_undef;` and to return false I'm saying: `return newSViv(0);` are these correct? Again, is there a better way to do that? email: perl -e 'print reverse map { chr( ord($_)-1 ) } split //, "\x0bufo/hojsfufqAofc";'	[reply] [d/l] [select]
Re^3: Propagating errors in C code to Perl by Joost (Canon) on Sep 07, 2008 at 12:39 UTC
That looks correct, though you can probably just `return 0` for false if your function has an `int function_name(...)` prototype. "What should it profit a man, if he should win a flame war, yet lose his cool?"	[reply] [d/l] [select]
Re^4: Propagating errors in C code to Perl by missingthepoint (Friar) on Sep 08, 2008 at 00:26 UTC
Re: Propagating errors in C code to Perl by sasdrtx (Friar) on Sep 05, 2008 at 22:35 UTC
Well, I thought this post was going to be about propagating coding errors while translating a program from C to Perl. I have experience doing that. :-) missingthepoint as the author makes for nice irony. sas	[reply]
Re^2: Propagating errors in C code to Perl by missingthepoint (Friar) on Sep 06, 2008 at 07:01 UTC
:) I originally chose the nick because I say that phrase all the time... But quickly realised that I'm leaving myself wide open to jokes if I post something silly. Ah, well. :o) email: perl -e 'print reverse map { chr( ord($_)-1 ) } split //, "\x0bufo/hojsfufqAofc";'	[reply]