62 lines
2.5 KiB
Markdown
62 lines
2.5 KiB
Markdown
|
|
# Intllib developer documentation
|
|
|
|
## Enabling internationalization
|
|
|
|
In order to enable internationalization for your mod, you will need to copy the
|
|
file `lib/intllib.lua` into the root directory of your mod, then include this
|
|
boilerplate code in files needing localization:
|
|
|
|
-- Load support for intllib.
|
|
local MP = minetest.get_modpath(minetest.get_current_modname())
|
|
local S, NS = dofile(MP.."/intllib.lua")
|
|
|
|
You will also need to optionally depend on intllib, to do so add `intllib?`
|
|
to an empty line in your `depends.txt`. Also note that if intllib is not
|
|
installed, the getter functions are defined so they return the string
|
|
unchanged. This is done so you don't have to sprinkle tons of `if`s (or
|
|
similar constructs) to check if the lib is actually installed.
|
|
|
|
Once you have the code in place, you need to mark strings that need
|
|
translation. For each translatable string in your sources, use the `S`
|
|
function (see above) to return the translated string. For example:
|
|
|
|
minetest.register_node("mymod:mynode", {
|
|
-- Simple string:
|
|
description = S("My Fabulous Node"),
|
|
-- String with insertions:
|
|
description = S("@1 Car", "Blue"),
|
|
-- ...
|
|
})
|
|
|
|
The `NS` function is the equivalent of `ngettext`. It should be used when the
|
|
string to be translated has singular and plural forms. For example:
|
|
|
|
-- The first `count` is for `ngettext` to determine which form to use.
|
|
-- The second `count` is the actual replacement.
|
|
print(NS("You have one item.", "You have @1 items.", count, count))
|
|
|
|
## Generating and updating catalogs
|
|
|
|
This is the basic workflow for working with [gettext][gettext]
|
|
|
|
Each time you have new strings to be translated, you should do the following:
|
|
|
|
cd /path/to/mod
|
|
/path/to/intllib/tools/xgettext.sh file1.lua file2.lua ...
|
|
|
|
The script will create a directory named `locale` if it doesn't exist yet,
|
|
and will generate the file `template.pot` (a template with all the translatable
|
|
strings). If you already have translations, the script will proceed to update
|
|
all of them with the new strings.
|
|
|
|
The script passes some options to the real `xgettext` that should be enough
|
|
for most cases. You may specify other options if desired:
|
|
|
|
xgettext.sh -o file.pot --keyword=blargh:4,5 a.lua b.lua ...
|
|
|
|
NOTE: There's also a Windows batch file `xgettext.bat` for Windows users,
|
|
but you will need to install the gettext command line tools separately. See
|
|
the top of the file for configuration.
|
|
|
|
[gettext]: https://www.gnu.org/software/gettext/
|