Intlize is intended for use in conjunction with xgettext and msgmerge.
The internationalization process takes four steps.
- Extract the translatable strings from the sources with xgettext. Please refer to the xgettext manual for details. The file produced by xgettext should be named
NAME is a unique name for your application.
- Merge the translated strings from
NAME.LANG and the extracted strings from
NAME.pot with msgmerge. Or, if
NAME.LANG does not exist, copy
LANG is the language code, e. g. it for italian or de_AT for german in austria.
NAME.LANG must be in the same directory.
NAME.LANG. The intlize runtime framework evaluates the header fields PACKAGE, VERSION and CHARSET. Of course, you want to fill in the missing msgstrs and revise the fuzzy ones.
- Run intlize to produce the catalog. Intlize writes to
LANG/NAME.msg. The subdirectory
LANG is created if necessary.
Repeat steps 2 ... 4 for each language. If you are using intlize native format you will need a default translation. To create this default translation, run intlize once with argument -C on one of the
Command line options
Intlize is invoked with:
intlize [OPTIONS] POFILE
POFILE is the edited
NAME.LANG file mentioned in the previous chapter.
OPTIONS is one or more of the following:
KEYWORD: Search for keyword
KEYWORD instead of
ENCODING: Store encoding
ENCODING instead of the
POFILE's CHARSET header field.
Possible values for
- iso-8859-1 ... iso-8859-16
- koi8-r / koi8-u / koi8-t
Other values are not yet supported but can be registered on request.
VERSIONID: Store the ID
VERSIONID instead of the
POFILE's PACKAGE and VERSION fields.
-C : (intlize native only) Create a default catalog from the POFILE's msgids in addition to the catalog created from the msgstrs. The default message catalog has the language code C.
-fc : Produce output in catgets format.
-fi : Produce output in intlize native format.
intlize -v prints its version and exits.