On Mon, Oct 27, 2008 at 10:03 AM, Lyman wrote:

Hi,

I noticed some translation flaw of pacman and wish to correct it. I followed the instruction as this link

http://www.archlinux.org/pacman/translation-help.html

and get the updated po file (to be exact, zh_CN.po).

Here's the problem when I open this po file: I don't know how to identify parameter with sequence. For example, when I need to translate a sentence

msgid ":: Replace %s with %s/%s?"

into the form of

msgstr ":: Use %s/%s to replace %s?"

How could tell which %s is which. This does happens when translating English into Chinese and there are such bugs in zh_CN.po right now.

I've searched for a while and found that different projects might use different policies.

In KDE people uses

msgid "%1 articles match rule %2" msgstr "匹配规则 %2 的文章有 %1 个"

and in Gnome people uses

msgid "%d articles match rule %d" msgstr "匹配规则 %2$d 的文章有 %1$d 个"

What about pacman? Thanks.

If these two projects also use gettext, then both ways should be possible. However, in the first case, you have to modify the original message (msgid), and you as a translator have no control over it. You can only modify the translated message (msgstr) which only leaves you with the second option. I found only one example of its usage, in lib/libalpm/po/hu.po but well it should be a good sign that it works as intended : #, c-format msgid "cannot resolve \"%s\", a dependency of \"%s\"\n" msgstr "nem sikerült a(z) \"%2$s\" csomag \"%1$s\" függőségét feloldani\n"