Re: Signal names

[Mark . Burgess]

That's not a bad idea. Actually the main problem is to find these
definitions. They are not always in signal.h. Often they are
in a sub file like sys/signal.h, but as you would expect this is
completely irrational.

M

On 23 Apr, Michael Beattie wrote:
> On Mon, Apr 22, 2002 at 05:13:07PM -0700, Tim Auckland wrote:
>> signal.h does not give you the strings for the signal names, just the
>> #defines for the numbers.  You have to put in a bit more work to get the
>> string-to-number mapping, but it's still do-able.
>> 
>> Something like:
> ...
>> although there's probably a neater way.
>> 
>> Cfengine also uses an enum somewhere in the signal name mapping code,
>> which would have to be dealt with similarly.
> 
> 
> How about just using the #define's to give the right signal, when the
> name is used?
> 
> signal = hup
> 
> wherever it parses that, use SIGHUP. Then you can map symbolic names
> cfengine style, to SIGxxx #define values.
> 
> I would have thought that was the sensible method.
> 
> If I sound incoherent, thats because I usually am.
> 
> Mike.



~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Work: +47 22453272            Email:  address@hidden
Fax : +47 22453205            WWW  :  http://www.iu.hio.no/~mark
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~