Re: [Bison-Announce] Bison 3.5.90 released [beta]

Hi Frank,

> Le 19 avr. 2020 à 17:25, Frank Heckenbach <[email protected]> a écrit :
> 
>>> it seems I have to define both
>>> YYENABLE_NLS and ENABLE_NLS. The former is mentioned in the manual
>>> (but only in conjunction with autoblah), the latter not at all.
>>> 
>>> So if I got it correctly now, this seems to be the minimum required
>>> for i18n, apart from the bindtextdomain call. Is that correct?
>>> 
>>> %define parse.error detailed
>>> %{
>>> #define YYENABLE_NLS 1
>>> #define ENABLE_NLS 1
>>> #define N_
>>> static auto _ (const char *s) { return dgettext ("mydomain", s); }
>>> %}
>> 
>> I don't think this is the right approach.  I think you are trying
>> to exploit the fact that the skeleton is actually already having
>> support to translate its own messages ("syntax error, expected...").
>> 
>> Rather I had something like the attachement in mind.  I am very
>> afraid about really installing this in the source tree, because I
>> expect tons of portability issues here: the examples are meant to
>> be self-contained, independent of bison's build, so I cannot use
>> the same tricks to be portable.
>> 
>> But I agree it would be better to actually demonstrate the whole
>> process in bistromathic.  I'll see what I can do.
> 
> I'm not sure I follow you here. I think what I do is rather close to
> what you suggest here. Maybe there was some confusion because I
> hadn't mentioned the steps that were unproblematic (or well-known)
> to me already. So for completeness, I had this already (with my own
> domain, of course):
> 
>>  # include <libintl.h>
>>  setlocale (LC_ALL, "");
>>  bindtextdomain ("bistromathic", LOCALEDIR);
> 
> But don't you also need:
> 
>  bindtextdomain ("bison-runtime", BISON_LOCALEDIR);
> 
> Or is it there already? I can't seem to find it.

You are right, I missed the (obvious) fact that anyway you *should*
be using this.

> Anyway, I had put
> that in my code, and FWIW I didn't need autoblah and m4 to get
> BISON_LOCALEDIR, I just use the following my Makefile:
> 
>  -DBISON_LOCALEDIR='"$(shell $(BISON) $(BISON_FLAGS) --print-localedir)"'

Yes, that's exactly the right way to do it!  Thanks for reminding
me this...  I'll have to emphasize this in the I18n section in the
doc.

> So the only thing missing before it worked for me was:
> 
>  #define YYENABLE_NLS 1
>  #define ENABLE_NLS 1

Yes, you are right, you need both.

> I suppose they're normally set by autoblah, so I need to do this
> manually. But maybe I should rather set them on the command-line
> than in the source!?

I would probably keep them close to BISON_LOCALEDIR, so out
of the file, but...

> Anyway, not a big deal,

yes, both are ok.

> though I think it should at least be briefly
> mentioned, unless you argue using gettext without autoblah is a
> non-supported mode of operation (which I've been doing in other
> projects for a long time and it seems to be working well, and
> doesn't require defining ENABLE_NLS).

Well, you may well use another system that gettext for i18n, it is
not mandatory.  The point of this stack of macros is to provide
magic _if_ the user wants to use what's already inside the generated
code.  But you may well define YY_ yourself, and not use all the
ENABLE_NLS stuff (provided you take care of all the #includes
yourself).  That's why what you need to do looks so weird.

So you are in a way "in between": not using the Autotools magic,
but willing to use the machinery inside.

I'm saying it's wrong.  It's only that I had not thought about it.
So, yes, what you do is right.  Not only is it right, it's also
what I should be doing in bistromathic.  I'll have to complete
the doc.

Thanks!


> BTW:
> 
>>  textdomain ("bistromathic");
>>  # define _(Msgid)  gettext (Msgid)
> 
> This works for an example (or a "main" program), but I think the
> more general approach (as required e.g. in a library containing a
> parser, where the main program might use its own domain) would be
> no "textdomain" call and rather:
> 
>  #define _(Msgid) dgettext ("bistromathic", Msgid)

Yes, that is correct.  But I don't want this example to be a showcase
for gettext, only for Bison :)  Most users need to see the simple
case (i18n of programs), for more complex scenarios, they have to
read gettext's manual.

> PS:
> 
>>    examples: bistromathic: demontrate internationalization
> 
> "demonstrate" ;)

Yeah, I already caught that one too :)  Thanks!