Re: [Discuss-gnustep] Localization

[Pascal J. Bourguignon]

> From: Lele Gaifax <lele@seldati.it>
> Date: Fri, 22 Sep 2000 10:08:45 +0200 (CEST)

> >>>>> On Fri, 22 Sep 2000 07:05:07 +0200 (CEST), "Pascal J. Bourguignon" 
> >>>>> <pjb@imaginet.fr> said:
> 
>     PJB> The process is as follow:
> 
>     PJB> 1- You put these lines in a source file:
> 
>     PJB> <snip>
> 
>     PJB> /* From a.c */ 
>     PJB> "cheese" = "queso"; /* The generic name for
>     PJB> the fermented milk stuff */ 
>     PJB> "cheese" = "patata"; /* What you
>     PJB> say to smile in a picture */
> 
>     PJB>    Which file you put in your Spain.lproj subdirectory.
> 
> 
>     PJB> 4- You compile your source file. The macros expand to :
> 
>     PJB>    NSString* cheese1= [[NSBundle mainBundle]
>     PJB> localizedStringForKey:( @"cheese" ) value:@"" table: (id)0 ]
>     PJB> ; NSString* cheese2= [[NSBundle mainBundle]
>     PJB> localizedStringForKey:( @"cheese" ) value:@"" table: (id)0 ]
> 
> I still miss the point: in which way lSFK does differentiate between
> the two "cheese"? The only way I see is when in some way it uses also
> the given comment to select the right translation. Shouldn't
> NSLocalizedString() pass also the comment to
> localizedStringForKey:value:table:?

I'll shamefully  acknowledge that  my examples and  understanding come
from  NeXTSTEP,  and  you're  making  me  realize  that  the  NeXTSTEP
localization macros (NXLocalizedString*)  have one more parameter than
their OPENSTEP conterparts.

Anyway, you're right: If the key is the same (and the table and bundle
too), then only  one entry will be selected. I've  not thought of that
consciously,  but  perhaps it's  for  this  reason  I didn't  use  the
"original language word or sentense" as key, but the name of the macro
I was defining to hold the localized string :

#ifdef NeXTSTEP_3_3
#define sE_PoliceNonDisponible NXLocalizedStringFromTable(@"Mention",\
        @"sE_PoliceNonDisponible",\
        @"La police Postscript '%s' n'est pas disponible. L'affichage "\
        @"du texte de la mention se fera dans la police par défaut.",\
        @"Message indiquant l'absence d'une police Postscript impliquant "\
        @"l'affichage du texte de la mention dans la police par défaut. "\
        @"Le champ %s sera remplacé par le nom de la police Postcript "\
        @"manquante.") 

In our cheese example, that would be:

#define s_cheese_as_food  NXLocalizedString(@"s_cheese_as_food",\
                           @"cheese",\
                           @"The generic name for the fermented milk stuff")
#define s_cheese_as_smile NXLocalizedString(@"s_cheese_as_smile",\
                           @"cheese",\
                           @"The generic name for the fermented milk stuff")

Of course,  there was  then no problem,  because the strings  file was
generated as:

/* From a.c */ 
"s_cheese_as_food" = "cheese"; /* The generic name for the fermented milk stuff 
*/ 
"s_cheese_as_smile" = "cheese"; /* What you say to smile in a picture */

Which would be translated to:

/* From a.c */ 
"s_cheese_as_food" = "queso"; /* The generic name for the fermented milk stuff 
*/ 
"s_cheese_as_smile" = "patato"; /* What you say to smile in a picture */

#endif // NeXTSTEP_3_3


Now, with the OPENSTEP macros  for localization, since the value is no
more taken  as parameter,  we have a  problem. I don't  understand why
they removed it, because it prevent the programmer to insert a default
language in  the source file,  and as you  pointed, the key  cannot be
used because of the homonyms.


Perhaps they  did that to make one  more reason to keep  and write the
comment  everytime, because  now,  the programmer  should  put in  the
comment  the intended default  language sentense,  in addition  to the
context description:

#ifdef OPENSTEP
#define s_cheese_as_food  NSLocalizedString(@"s_cheese_as_food",\
                @"'cheese' The generic name for the fermented milk stuff")
#define s_cheese_as_smile NSLocalizedString(@"s_cheese_as_smile",\
                @"'cheese' The generic name for the fermented milk stuff")

generated as:

/* From a.c */ 
"s_cheese_as_food" = "s_cheese_as_food"; /* 'cheese' The generic name for the 
fermented milk stuff */ 
"s_cheese_as_smile" = "s_cheese_as_smile"; /* 'cheese' What you say to smile in 
a picture */

Which would be translated to the "default" language as:

/* From a.c */ 
"s_cheese_as_food" = "cheese"; /* 'cheese' The generic name for the fermented 
milk stuff */ 
"s_cheese_as_smile" = "cheese"; /* 'cheese' What you say to smile in a picture 
*/

Which would be translated to other language as:

/* From a.c */ 
"s_cheese_as_food" = "queso"; /* 'cheese' The generic name for the fermented 
milk stuff */ 
"s_cheese_as_smile" = "patato"; /* 'cheese' What you say to smile in a picture 
*/

#endif // OPENSTEP


> I think event gettext is weak on this: its tools even compact the
> "strings" equivalent by eliminating duplicated entries, since AFAICT
> it cannot make any difference between two messages like above...
> 
> thanx, and bye,
> lele.


-- 
__Pascal Bourguignon__    PGP Key ID:      0xEF5E9966
mailto:pjb@imaginet.fr    PGP fingerprint: 00 F5 7B DB CA 51 8A AD 04 5B 
http://www.imaginet.fr/~pjb/               6C DE 32 60 16 8E EF 5E 99 66

() Join the ASCII ribbon campaign against html email and Microsoft attachments.
/\ Software patents are endangering the computer industry all around the world.
   Join the LPF:     http://lpf.ai.mit.edu/      http://petition.eurolinux.org/