On Tue, 23 Aug 2016 12:15:17 +0200
Ricardo Wurmus <rekado@elephly.net> wrote:

> Eric Bavier <ericbavier@openmailbox.org> writes:
> 
> > On Mon, 22 Aug 2016 22:03:58 +0200
> > Ricardo Wurmus <rekado@elephly.net> wrote:
> >  
> >> John Darrington <john@darrington.wattle.id.au> writes:
> >>   
> >> > These somehow crept in, but are an explicit violation of GNU policy and
> >> > coding standards.    
> >> 
> >> Good catch!  Thanks for fixing this.
> >> 
> >> ~~ Ricardo  
> >
> > How about the attached patch to catch such things earlier?  
> 
> I like it.  Thanks, Eric!
> 
> Not sure if we should use “for-each” here (going through the string
> once for each character) or if we could just go through the string once,
> checking for any character match.
> 
> Also, should we replace “sign” with “character”?

In this updated patch I addressed both of these concerns.  It's also
about 14% faster (on a single benchmark :).

`~Eric