[freeciv-i18n] Re: Grammatical cases in translations.

[Reinier Post <rp@xxxxxxxxxx>]

On Tue, Sep 23, 2003 at 11:37:02PM +0100, Gregory Berkolaiko wrote:
> On Tue, 23 Sep 2003, Christian Knoke wrote:
> 
> > On Tue, Sep 23, 2003 at 11:39:53PM +0200, Reinier Post wrote:
> > > On Tue, Sep 23, 2003 at 10:06:41PM +0100, Duarte Loreto wrote
[on the problem of noun inflection]
> > > Why not may English another target language, and mark up strings for case
> > > in the source language (which would be "almost, but not quite English")?
> > 
> > Because the case of a noun differs between languages?

The source language would have the cases. You'd have "Romans1" and
"Romans4" as translatble strings, or something like that.
Freeciv already has the Q_ macro, I thought it was designed for
cases like this.

> This must be the case.  In Indo-European languages cases might be 
> compatible, but I suspect when you step out, it all breaks.  And you don't 
> have to step very far, Finnish and Hungarian are examples.
> 
> The only way I could invent was to mark cases in already translated 
> strings, like in
> 
> "%s1 give %s2 the map of the world"
> 
> and then have something like a table for each name, like
> 
> 1, Romans
> 2, Romans
> 
> 1, We
> 2, Us  /* This being the only example of cases in English I know of */

I think with the present syntax it can be

  "%s give %s the map of the world",
    Q_("?case1:" . player1->name),
    Q_("?case4:" . player2->name)

provided you can design a universal case system that would cover enough
practical cases in enough languages, and except that, of course, we are
in C so there is no string concatenation operator and all workarounds
for that make me want to scream.

> But all this is just curiosity on my part.  I believe that localisation is 
> evil anyway.  Let's corrupt English instead!
> 
> G.

-- 
Reinier