Re: Some more setenv(3) update

[Nicolas Joly <njoly%pasteur.fr@localhost>]

On Tue, Oct 12, 2010 at 10:47:32PM +0100, Matthias Scheler wrote:
> On Tue, Oct 12, 2010 at 12:06:45PM +0200, Nicolas Joly wrote:
> > >        for (cc = name; *cc && *cc != '='; ++cc)        /* no `=' in name 
> > > */
> > >                continue;
> > > 
> > > before the initial test, and then check for *cc instead of scanning the
> > > name twice, once in strchr and a second time in strlen.
> 
> It will be cached the second time anyway.
> 
> > Right, but we then need to handle case for NULL name.
> 
> Which is exactly why your solution was better as it was simpler.
> 
> >> +  for (cc = name; cc && *cc && *cc != '='; ++cc)
> > +           continue;
> 
> It don't like this even a bit.

I'm not fond of it too; but wanted to give it a try.

> I would prefer this:
> 
>       if (name == NULL || value == NULL) {
>               errno = EINVAL;
>               return -1;
>       }
> 
>       size = strcspn(name, "=");
>       if (size == 0 || name[size] != '\0') {
>               errno = EINVAL;
>               return -1;
>       }
> 
> It only reads the string once and figures out the length and whether
> there is an equal sign in the name at the same time.

That's indeed much better, will use it. I didn't known about strcspn
function...

I plan to commit all of this in a few days unless someone objects. And
then have a look to putenv which require some care too.

Thanks.

-- 
Nicolas Joly

Biological Software and Databanks.
Institut Pasteur, Paris.