Mapping voices onto faces

[Pierre Lorenzon]

From: Milan Zamazal <address@hidden>
To: address@hidden
Subject: Re: Mapping voices onto faces
Date: Mon, 17 Mar 2008 21:35:27 +0100

> >>>>> "PL" == Pierre Lorenzon <devel at pollock-nageoire.net> writes:
> 
>     PL> When testing (line 963 in speechd.el (cvs version)), match is
>     PL> done against result of (get-text-property point 'face) assuming
>     PL> that it returns a face ! But it seems that in certain cases a
>     PL> list of faces might be returned e.g. in a LaTeX file I have :
>     PL> (get-text-property (point) 'face) => (font-lock-keyword-face
>     PL> font-latex-sedate-face).
> 
> [...]
> 
>     PL> So what should we do ? Improve slightly the code in speechd.el
>     PL> so that it can handle list of faces ? Or are these list of faces
>     PL> something pathological that should not append ?
> 
> According to Elisp manual the value of the `face' text property may
> indeed be a list which can contain (among other objects) face names as
> symbols or strings.  So this is a bug in speechd-el.  It should test for
> both a simple value and a list and additionally it should recognize not
> only symbols as face names but also strings.
> 
> Well, providing a patch would probably be the fastest way to get it
> fixed. ;-)

  OK I'll have a look at this point if I have enough time among
  all the things I have to do ! If I write a patch I'll
  surely communicate it !

  Regards

  Pierre