Re: string not clear for a translator

[Lukas-Fabian Moser]

Hi Federico,

Am 07.11.22 um 23:51 schrieb Federico Bruni:
> I'm updating the italian language and I can't understand what this 
> dash is about here:
>
> $ git grep -n -A4 -B4 "Re-defining dash" lily/
> lily/parser.yy-3316-// 
> -----------------------------------------------------------------
> lily/parser.yy-3317-// obsoletion handling, may be removed at some 
> point (e.g. for 2.26)
> lily/parser.yy-3318- } else if (scm_is_string (s)) {
> lily/parser.yy-3319- string s_string = ly_scm2string (s);
> lily/parser.yy:3320: @$.warning (_f ("Re-defining dash%s using a 
> string is deprecated. \
> lily/parser.yy-3321-Please try replacing \"%s\" by \\%s or run 
> convert-ly.",
> lily/parser.yy-3322- ly_scm2string ($1), s_string, s_string));
> lily/parser.yy-3323- Music *a = MY_MAKE_MUSIC ("ArticulationEvent", @$);
> lily/parser.yy-3324- set_property (a, "articulation-type", 
> scm_string_to_symbol (s));
>
> BTW, no space between dash and %s?

That was me :-).

> commit 61cd3bc1f3254b430bf04acd587c4082253602d4
> Author: Lukas-Fabian Moser <lfm@gmx.de>
> Date:   Mon Dec 27 01:25:43 2021 +0100
>
>     Make articulation-type a symbol? instead of a string?
>
>     This was an ancient TODO (see note from 2003 in 
> scm/define-music-properties.scm).
>     It does not seem to make much sense to rely on string comparison for
>     distinguishing a well-defined set of articulation types. Also, the
>     articulation-type is used as an alist index, for which symbols are 
> much more
>     natural than strings.
>
>     This also makes it easier to (for instance) let users enter lists 
> of articulation
>     types to use in custom code, since nowadays the parser turns
>
>         types = tenuto, fermata, espressivo
>
>     into a symbol-list?.
>
>     This commit also starts an obsoletion cycle for the outdated 
> articulation
>     definition style: As a legacy feature, it used to be possible to do
>
>         dashDash = "tenuto"
>
>     instead of the modern variants
>
>         dashDash = \tenuto
>         dashDash = #(make-articulation 'tenuto)
>
>     which are more in line with current LilyPond syntax principles.
>
>     The legacy input still works but emits an obsoletion warning that 
> may be
>     removed later (e.g. for 2.26).
>
>     Lastly, the code handling scripts is made more robust: Choosing 
> non-existing
>     scripts now tends to emit warnings/programming errors instead of 
> terminating.
In plain English: If you want to re-define the meaning of commands like 
"--" (tenuto), "-." (staccato) etc., you formerly could set them to a 
string, which is not at all in line with current LilyPond syntax 
principles. This is now officially deprecated (but still supported).

So, dash%s is displayed as "dashHat", "dashDot", "dashDash" etc.

Lukas