[systemd-devel] [PATCH 1/3] Add helper for fnmatch over strv

[Zbigniew Jędrzejewski-Szmek]

On Mon, Feb 16, 2015 at 02:12:38PM +0100, Lennart Poettering wrote:
> On Sat, 14.02.15 00:53, Zbigniew Jędrzejewski-Szmek (zbyszek at in.waw.pl) wrote:
> 
> > No functional change intended.
> 
> I like this simplification!
> 
> >  
> >          if (match_host && !condition_test(match_host))
> >                  return false;
> > @@ -117,49 +112,17 @@ bool net_match_config(const struct ether_addr *match_mac,
> >          if (match_mac && (!dev_mac || memcmp(match_mac, dev_mac, ETH_ALEN)))
> >                  return false;
> >  
> > -        if (!strv_isempty(match_paths)) {
> > -                if (!dev_path)
> > -                        return false;
> > +        if (!strv_isempty(match_paths))
> > +                return strv_fnmatch(dev_path, match_paths, 0);
> 
> Can't this be shortened further by combining the stv_isempty() with
> the strv_fnmatch? 
This code is changed in 2/3. I believe it is broken in the original
version (and after the change above, which does not change functionality).


> > +bool strv_fnmatch(const char *s, char* const* patterns, int flags);
> > +
> > +static inline bool strv_fnmatch_or_empty(const char *s, char* const* patterns, int flags) {
> > +        assert(s);
> > +        return strv_isempty(patterns) ||
> > +               strv_fnmatch(s, patterns, flags);
> > +}
> 
> Wouldn't the order of arguments be more natural if we specified the
> strv ("haystack") first, and the string ("needle") second? After all,
> it's kinda an OO interface, where the first object should come first?
Yeah, like strv_find and friends. I'll do that.

> Anyway, this all looks like a great simplification. If this doesn't
> change behaviour I love the idea, please apply!
I'll wait for some feedback on 2/3 from Tom.

Zbyszek