Re: [Pkg-isocodes-devel] iso-codes merging using trying to use ...

[Alastair McKinstry]

Hi Davi,

Thanks for your work, however I have some questions over the expected
usage of this information.

Different languages have different plural forms based on the number of
objects (and type), ie. you cannot safely use:

printf("%d %s", num, _("dinars");

English has simple rules with "singular == dinar, plural = dinars",
with other languages it can vary with context and number, eg.
the form in Irish varies with n < 7 and n > 7.
There is code to handle this; see the gettext info pages,
but plural forms are best avoided altogether and the interface
defined accordingly.

Regards
Alastair



Davi Leal wrote:
> The attached file keeps the modifications which the GNU Herds project has 
> realized to the SVN HEAD of the iso-codes packages:
>
>   Spanish
>   Italian
>   Portuguese
>   French
>
>   iso_639
>   iso_3166
>   iso_4217
>
> We want follow working to be in sync with the iso-codes package's PO files.  
> We have done a lot of changes at our files, however we propose several 
> changes to the isocodes PO files too, which I hope you take into 
> consideration:
>
>
>
> If the iso-codes package is going to be used by a lot of different 
> applications it should offer all the used word constructions.  For example, 
> at the ComboBoxes we group sets of entities to make it easier to the user to 
> find it.
>
>                 Dinars, Libia
>                 Dinars, Serbia
>                 Dinars, Sudán
>                 Dinars, Túnez
>                   ...
>                 Dollars, Bahamas
>                 Dollars, Barbados
>                 Dollars, Belize
>                   ...
>
> The ComboBoxes are sorted.
>
>
> When we expose it as a label, it is better to show it as:
>
>                "Libian Dinars"
>
>                   "Serbian Dinars"
>
>                      "Sudanese Dinars"
>
>
>
> The proposed structure for the POT iso_4217 files should have at least the 
> below elements:
>
>   #. currency_name for AED
>       msgid "UAE Dirham"
>       msgstr "Dirham of the Emiratos Arabes Unidos"
>
>       msgid "Dirham, United Arab Emirates"
>       msgstr "Dirham, Emiratos Arabes Unidos"
>
>   # plural
>       msgid "UAE Dirhams"
>       msgstr "Dirhams of the Emiratos Arabes Unidos"
>
>       msgid "Dirhams, United Arab Emirates"
>       msgstr "Dirhams, Emiratos Arabes Unidos"
>
>
>
> About the the iso_3166 files, we propose the below structure.  Note we have 
> already added several names to the attached PO iso_3166 files following the 
> below structure:
>
>   #. name for FJI
>       msgid "Fiji"
>       msgstr "Figi"
>
>       msgid "Fiji Islands"
>       msgstr "Isole Figi"
>
>       msgid "name -3-"
>
>       ...
>       
>       msgid "name -N-"
>
>   #. official_name for FJI
>       msgid "Republic of the Fiji Islands"
>       msgstr "Repubblica delle Isole Figi"
>
>
>
>
> P.S.: I have set uppercase the first character of the iso_639 languages. Maybe 
> I should not have done it. The GNU Herds project needs a French translator 
> maintainer.
>
> Best regards,
> Davi
>   
> ------------------------------------------------------------------------
>
> _______________________________________________
> Pkg-isocodes-devel mailing list
> address@hidden
> http://lists.alioth.debian.org/mailman/listinfo/pkg-isocodes-devel