nightly macro gettext-macros

A few proc-macros to help internationalizing Rust applications

5 releases (breaking)

✓ Uses Rust 2018 edition

0.5.0 May 16, 2019
0.4.0 Mar 5, 2019
0.3.0 Jan 28, 2019
0.2.0 Jan 20, 2019
0.1.0 Jan 20, 2019

#48 in Internationalization (i18n)

Download history 176/week @ 2019-01-31 66/week @ 2019-02-07 97/week @ 2019-02-14 67/week @ 2019-02-21 64/week @ 2019-02-28 53/week @ 2019-03-07 176/week @ 2019-03-14 362/week @ 2019-03-21 155/week @ 2019-03-28 132/week @ 2019-04-04 72/week @ 2019-04-11 69/week @ 2019-04-18 2/week @ 2019-04-25 31/week @ 2019-05-02 40/week @ 2019-05-09

432 downloads per month

GPL-3.0 license

587 lines

Gettext macros

A few proc-macros to help you internationalize your Rust apps.

It uses gettext under the hood. The idea is that just by wrapping strings in macros, they are added to a template for translation files (.pot file). Then, the actual translation files (.po) are generated for each language, and you can upload them to Weblate/Crowdin/POedit/whatever to have them translated. And finally, they get transformed into binary translation files (.mo) that you can embed in your app.

How does it works?

There are five main macros:

  • init_i18n, that should be called first. It tells the domain to use for the current crate, and the supported locales.
  • compile_i18n, that should be called at the end of your It updates translation files and compile them.
  • include_i18n, that will embed translations in your binary, making it easier to distribute. It should be called after compile_i18n to work correctly.
  • i18n, that translates a given message.
  • t, that works like i18n, but doesn't actually translate the message, just adds it to the list of strings to translate.

The advantage of these macros is that they allow you to work with multiple translation domains (for instance, one for each of your workspace's crate), and that they automatically generate a .pot file for these domains.


use gettext_macros::*;

// The translations for this crate are stored in the "my_app" domain.
// Translations for all the listed langages will be available.
init_i18n!("my_app", ar, de, en, fr, it, ja, ru);

// Generate or update .po from .pot, and compile them to .mo

fn main() {
	let catalog = cat();

    println!("{}", i18n!(catalog, "Hello, world!"));
    let name = "Jane";
    println!("{}", i18n!(catalog, "Hello, {}!"; name));
    let message_count = 42;
    println!("{}", i18n!(catalog, "You have one new message", "You have {0} new messages"; message_count));

fn cat() -> gettext::Catalog {
	// include_i18n! embeds translations in your binary.
    // It gives a Vec<(&'static str, Catalog)> (list of catalogs with their associated language).
    let catalog = include_i18n!()[0]

Order of the macros

The macros should be called in a certain order to work properly. This order doesn't depend on the program flow, but of the Rust parser flow. Rust will execute macros in the same order they are written in your code. For instance:


Or, for projects with modules:

// In


mod a;

// In


So, for the macros provided by this crate, the order to follow is:

  1. init_i18n!
  2. i18n! and t!, as many times as you want
  3. compile_i18n!
  4. include_i18n!

Because some of these macros depends on files written by the previous ones to work properly.


~34K SLoC