Translating with .po files - Support scripts
Brought to you by:
canajun2eh,
yalnifj
Menu
▾
▴
#560 Translating with .po files - Support scripts
I manage the Spanish translation using -po files, one per array (pgv_lang, factarray, countries, etc.) These are the scripts I use:
pgvxgettext traverses the source code and extracts translation keys generating what amounts to a skeleton po template file containing source cross references that are useful during translation. This process is fragile and will both fail to detect some references and match at spurious places. This problem cannot be solved without chamges to the PGV source code.
pgv2po takes a set of translation files and creates a .po file that can be msgmerge'd with the template file to add source cross references.
po2pgv takes a .po file, a source language file and produces a target language file as a copy of the source language file with all strings replaced by those in the .po file
lanza.pgv2po and lanza.po2pgv are sample shell scripts that use the above building blocks.
Is this script currently available?
Last edit: xiaofo 2018-09-25
No, this method was never accepted into PhpGedView.
It's best to use the Translator Tools that are part of PhpGedView. To use the Language File Edit tool (the most useful of the bunch), you need to set the permissions of the /languages directory, and all its contents, to 777.
The Language File Edit tool matches the English version of a file with whatever language you're trying to translate. You work on each English text entry and the corresponding Chinese (or whatever) text individually.
You can choose to work on only the texts that don't have a translation, or you can choose to work on even texts that are already translated.
The Language File Edit tool currently has a link to Google Translate for each entry. Don't use that -- it will be removed in the next SVN update and will not be part of version 4.3.1 when that's released (whenever that will be!)
I personally edit the language files off-line, using a suitable text editor. My method is cumbersome, but it produces reliable results where the resulting language file matches the English version exactly, even reproducing the order of the English entries in the edited language file.