Quote:

Please, provide the original .cd (catalog descriptor or something like that) instead of an already translated .cd

The file you're providing isn't complete enough for producing a catalog and is in fact a .ct, not a .catalog which is a binary file.

The .cd file contains important information for building the catalog. Each token line (MSG_FOOBAR) ends with '(//)' hint.

This '(//)' hint helps with producing consistent catalogs. This hint format is composed of 3 numbers (1/2/3), each of them instructing the characteristics of the translated string.

The most important one is the last number (//X) which limits the length of the string. This one is mostly used when the application is making use of a fixed length array for storing the string (like for example, app description which is limited to 40 bytes).

The first one, is used for assigning a starting id to the string. It can be used to insert 'free addressing space' in the catalog.

The middle one has never been used to my knowledge and is used for the defining a minimum length...

Also, for UTF8 encoded catalogs, you need to put some '## pragma'

Some of the (should be) supported pragma:

## version $VER: foo.catalog 1.1 (dd.mm.yyyy)
## language <lang>
## codeset 1
## chunk AUTH <name>

That's for feeding the .catalog with some usefull information.

The codeset pragma is important as it defines that the .ct is UTF8 encoded (used by locale.library to know about the file encoding format, else it will load the catalog as a ISO8859-X encoded resulting in garbled accented chars for example)

You might use flexcat from Aminet for building the .catalog.

To build an empty .ct file:

Code:

Edit foo.ct in your UTF8 enabled editor (use FlowStudio from SDK for ex.)

To build a .catalog file that will be used by the application:

Code:

There is some options to flexcat, see its documentation.

The root file for any translation is the '.cd' file.




[ Edited by Tcheko 02.02.2024 - 08:57 ]