Erlang gettext
t__ is an implementation of the Gettext system in Erlang.
gettext is an internationalization and localization system commonly used for writing multilingual programs on Unix-like computer operating systems.
t__ is also available on HEX
- ?T__("I have a joke about Erlang, but it requires a prologue.")
- Fast.
- Simple. Reading this file should be enough for understanding most use case scenarios.
- Documented source code with comments and specs.
- Highly configurable with multiple PO sources and languages per Erlang application or process.
- Supports contexts.
- Supports interpolation using familiar Erlang format control sequences (from io:format).
- Supports translating singular term with or without interpolation and with or without context.
- Supports translating plural term with or without interpolation and with or without context.
- Supports all plural terms formulas defined by UNICODE CLDR.
- Supports ETS tables based caching.
- Erlang Gettext tools for multi-lingual capabilities: https://github.com/etnt/gettext
| # | Feature | t__ | gettext |
|---|---|---|---|
| 1 | Single macro that handles everything | YES: T__ | NO: TXT & TXT2 |
| 2 | gettext contexts | YES | NO |
| 3 | gettext plural terms | YES | NO |
| 4 | Separate configurations for each application started at the same node | YES | NO |
| 5 | Different language sources | YES: using application 'repositories' | YES: using different languages servers |
| 6 | Developer mode/monitoring PO file changes | YES: automatically detecting PO changes | NO: requires manual reloading of the PO files |
| 7 | Cache | YES: ETS reads/writes on the calling process (faster) | YES: ETS reads/writes inside the translation gen_server |
- Gettext plural-forms formulas are hardcoded using a database. Work is underway to bypass this limitation and provide a full interpreter of any arbitrary gettext C formula.
- Github pages: https://ergenius.github.io/t__/
- '/doc' subdirectory (generated by ex_doc)
You can find a full demo application on Github t__ demoapp. The application demonstrate most functionalities, including using multiple repositories and t__ being able to monitor repositories PO changes and reload them in real time. You can test this feature by modifying any PO files used by the demoapp while you run the application in console mode. A gen_server constinously translating strings on a timer is provided, so you can easily check the updates.
It is highly recommended to set the calling process language in order to avoid specifying the language for each ?T__ macro call. The specified language will be used by all ?T__ macros and functions for the current process if no explicit language is specified as a macro or function parameter.
?T__LANGUAGE("en").
?T__LANGUAGE("en_GB").
?T__LANGUAGE("ro").Get the calling process language.
When you properly setup the language for the process, the expected time complexity for the current implementation of this macro is O(1) and the worst case time complexity is O(N), where N is the number of items in the process dictionary.
The function will perform the following steps in order to determine the default language:
- call erlang:get(t__language)
- call application:get_env(t__language)
- call application:get_env(t__, t__language)
- if no t__language environment key was set for both the current application or the t__ application we return t__ default language defined in t__.hrl file (which is "en").
Language = ?T__LANGUAGE().?T__("I have a joke about Erlang, but it requires a prologue.").You can use binaries and atoms, however it is not recommended.
?T__(<<"Erlang is user-friendly, it’s just picky about its friends!">>).
?T__('Why can\'t you trust atoms? Because they make up everything!').Context is useful for the translators to distinguish in between identical strings.
?T__({"menu", "Save"}).
?T__({"menu", "Quit"}).
?T__({"button", "Save"}).
?T__({"button", "Cancel"}).Context can also be used to create proper translations based on grammatical gender.
?T__({"female", "The cat belong to him/her"}).
?T__({"male", "The cat belong to him/her"}).Repositories are used to have different translations sources directories for the same application. For example let's assume your application has many HTML templates, each with his own translation directory.
?T__({"template1", {"Simple term from repository template1"}}).
?T__({"template2", {"Simple term from repository template2"}}).Repository can be combined with context also.
?T__({"template1", {"menu", "Save"}}).?T__("~4.2f", [3.56]). Context can also be used to create the proper translation based on grammatical gender combined with interpolation:
?T__({"female", "Her/his name is ~s"}, ["Marry"]).
?T__({"male", "Her/his name is ~s"}, ["John"]).?T__(["~B user", "~B users"], [3]).?T__({"female", ["~B file belongs to her/him", "~B files belong to her/him"]}, [3]).
?T__({"male", ["~B file belongs to her/him", "~B files belong to her/him"]}, [3]).?T__({"template1", {"female", ["~B file belongs to her/him", "~B files belong to her/him"]}}, [3]).
?T__({"template1", {"male", ["~B file belongs to her/him", "~B files belong to her/him"]}}, [3]).You can also specify everything using the #t__p{} record as a single parameter to the T__ macro. This is actually the performance wise way of doing it. You will save some extra functions calls necessary to understand your tuples. All #t__p{} fields except msg are optional.
?T__(#t__p{msg = "Hello world"}).
?T__(#t__p{msg = "Her name is ~s", data = ["Marry"]}).
?T__(#t__p{language = "ro", context = "female", msg = "Her/his name is ~s", data = ["Marry"]}).
?T__(#t__p{repository = "module1", language = "ro", context = "male", msg = "Her/his name is ~s", data = ["John"]}).
?T__(#t__p{application=myapp, repository = "module1", language = "ro", context = "female", msg = "Her/his name is ~s", data = ["Marry"]}).For your convenience a macro also exists with all possible parameters:
?T__(Application, Repository, Language, Context, Msg, Data).Short answer: Yes, if you trust CLDR Project.
For maintaining plural formulas database we started another project here: gettext-po-samples
We use the following sources and tools for compiling the database:
Formulas are automated generated by PHP gettext language list from CLDR. Each new release is manually checked and compared with the sources mentioned above. Data is compiled into plural-forms.eterm file. The same plural-forms.eterm file is used to generate src/t__plural.erl file.
- Continuously fixing bugs and tuning performance.
- Writing more testing units.
- Add more features.
t__ officially supports OTP release 21 and later.
Development takes place using OTP 27 release and tests are done on:
- 27.0.1
- 26.2.5
- 25.3.2
- 24.3.4
- 23.3.4
- 22.3.4
- 21.3.8
Unofficially, you may be able to use t__ with older Erlang versions. No guarantee included.
None.
- Madalin Grigore-Enescu (ergenius) - Github ergenius.com
t__ is available under the MIT license (see LICENSE).