Re: g_str_hash

[Tim Janik <timj gtk org>]

On Fri, 11 Feb 2000, Karl Nelson wrote:

> >the only think left to note is that you can spare the first
> >hash step, since 0 needs no propagation. so we get:
> >
> >guint
> >g_str_hash (gconstpointer key)
> >{
> >  const gchar *p = key;
> >  guint h = *p;
> >
> >  for (p = key + 1; *p != '\0'; p += 1)
> >    h = (h << 5) - h + *p;
> >
> >  return h;
> >}
> 
> One objection, it segfaults on 0 :-)

that's a feature, not a bug.
we'll simply document that case!
ok i'm dump, from hereon i rpomise to continue posting untested
code in the future ;)

> 
> --Karl
> 

---
ciaoTJ