| | 1 | \section{Internationalization}\label{internationalization} |
| | 2 | |
| | 3 | Although internationalization of an application (i18n for short) involves far |
| | 4 | more than just translating its text messages to another message -- date, time and |
| | 5 | currency formats need changing too, some languages are written left to right |
| | 6 | and others right to left, character encoding may differ and many other things |
| | 7 | may need changing too -- it is a necessary first step. wxWidgets provides |
| | 8 | facilities for message translation with its |
| | 9 | \helpref{wxLocale}{wxlocale} class and is itself fully translated into several |
| | 10 | languages. Please consult wxWidgets home page for the most up-to-date |
| | 11 | translations - and if you translate it into one of the languages not done |
| | 12 | yet, your translations would be gratefully accepted for inclusion into the |
| | 13 | future versions of the library! |
| | 14 | |
| | 15 | The wxWidgets approach to i18n closely follows GNU gettext package. wxWidgets uses the |
| | 16 | message catalogs which are binary compatible with gettext catalogs and this |
| | 17 | allows to use all of the programs in this package to work with them. But note |
| | 18 | that no additional libraries are needed during the run-time, however, so you |
| | 19 | have only the message catalogs to distribute and nothing else. |
| | 20 | |
| | 21 | During program development you will need the gettext package for |
| | 22 | working with message catalogs. {\bf Warning:} gettext versions < 0.10 are known |
| | 23 | to be buggy, so you should find a later version of it! |
| | 24 | |
| | 25 | There are two kinds of message catalogs: source catalogs which are text files |
| | 26 | with extension .po and binary catalogs which are created from the source ones |
| | 27 | with {\it msgfmt} program (part of gettext package) and have the extension .mo. |
| | 28 | Only the binary files are needed during program execution. |
| | 29 | |
| | 30 | The program i18n involves several steps: |
| | 31 | |
| | 32 | \begin{enumerate}\itemsep=0pt |
| | 33 | \item Translating the strings in the program text using |
| | 34 | \helpref{wxGetTranslation}{wxgettranslation} or equivalently the |
| | 35 | \helpref{\_()}{underscore} and \helpref{ngettext}{ngettext} macros. |
| | 36 | \item Extracting the strings to be translated from the program: this uses the |
| | 37 | work done in the previous step because {\tt xgettext} program used for string |
| | 38 | extraction recognises the standard \_() as well as (using its {\tt -k} option) |
| | 39 | our wxGetTranslation and extracts all strings inside the calls to these |
| | 40 | functions. Alternatively, you may use {\tt -a} option to extract all the |
| | 41 | strings, but it will usually result in many strings being found which don't |
| | 42 | have to be translated at all. This will create a text message catalog -- a .po |
| | 43 | file. |
| | 44 | \item Translating the strings extracted in the previous step to other |
| | 45 | language(s). It involves editing the .po file. |
| | 46 | \item Compiling the .po file into .mo file to be used by the program. |
| | 47 | \item Setting the appropriate locale in your program to use the strings for the |
| | 48 | given language: see \helpref{wxLocale}{wxlocale}. |
| | 49 | \end{enumerate} |
| | 50 | |
| | 51 | See also the GNU gettext documentation linked from {\tt docs/html/index.htm} in |
| | 52 | your wxWidgets distribution. |
| | 53 | |
| | 54 | See also \helpref{Writing non-English applications}{nonenglishoverview}. |
| | 55 | It focuses on handling charsets related problems. |
| | 56 | |
| | 57 | Finally, take a look at the \helpref{i18n sample}{sampleinternat} which shows |
| | 58 | to you how all this looks in practice. |
| | 59 | |
projects / wxWidgets.git / blame_incremental