Message Handling with gettext

'gettext()' is not a standard C language functionality but a widely-used message extracting mechanism. The basic idea is the same as the standard C language's message catalog but 'gettext()' uses the English message itself (or its transformed value) as the message id.

The following are the functions related to 'gettext()':

-------------------------------------------------------------------------
Function		    Description
-------------------------------------------------------------------------
gettext			    Gets a message
dgettext		    Gets a message
dcgettext		    Gets a message
ngettext		    Gets a message
dngettext		    Gets a message
dcngettext		    Gets a message
bindtextdomain		    Binds a message file with a domain
textdomain		    Selects a domain
bind_textdomain_codeset	    Sets a codeset of message
_			    Gets a message (macro)
N_			    Directs message extraction to a tool (macro)
gettext_noop		    Directs message extraction to a tool (macro)
-------------------------------------------------------------------------

The last three are macros. Their definitions are as follows:

#define _(String)   gettext(String)
#define N_(String)  gettext_noop(String)
#define gettext_noop(String) (String)

A text extracting tool searches for and extracts messages in these macro calls.

*** Gettext Steps ***

The basic concept is the same as the message catalog.
Please refer to message catalogs.
Below is an explanation of its differences from message catalogs.

Message File

Messages collected from programs can be described in .pot files.
This is a sample generated by World Wide Navi.

wwnavi_get.pot

# $NLS-WWNAVI 2008-05-24T17:48:21+0900
msgid ""
msgstr ""
"Project-Id-Version: PACKAGE VERSION\n"
"Report-Msgid-Bugs-To: \n"
"POT-Creation-Date: 2008-05-24T17:48:21+0900\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"Last-Translator: FULL NAME \n"
"Language-Team: LANGUAGE \n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=UTF-8\n" <-- *Encoding name
"Content-Transfer-Encoding: 8bit\n"

# Test.c
msgid "Failed to create fontset\n" <-- *Message ID
msgstr "Failed to create fontset\n" <-- *Message

msgid "Hello, how are you? This is sample for gettext. (Push .)"
msgstr "Hello, how are you? This is sample for gettext. (Push .)"

Compile Message Files.
- Compile .pot files into .mo files that can be loaded with the 'msgfmt' command.
```
msgfmt -o wwnavi_get.mo wwnavi_get.pot
```

Describe message loading.

Describe the following codes in the sources:

setlocale (LC_ALL, ""); <-- *Described in the main function. 
bindtextdomain("wwnavi_get", ".");<-- *Described in the main function.
textdomain("wwnavi_get");<-- *Described in the main function.
...
printf(_(N_("Failed to create fontset\n"))); <-- *Loading message.

Locations of the .mo files.
- The locations of the .mo files depend on the running environment like message catalogs.
  
  For example, run the program 'Test' according to the settings below.
```
export LANG=en_US.UTF-8
./Test
export LANG=ja_JP.UTF-8
./Test
```
  The locations of the .mo files will be the following:
```
./en/LC_MESSAGES/wwnavi_get.mo <-- *English message file
./ja/LC_MESSAGES/wwnavi_get.mo <-- *Japanase message file
```

This process can be checked with 'gettext' externalization of the sample 'linux_gettext' in World Wide Navi.

Go to Internationalization Programming Top