Hi
Since I'm still seeing issues when working with multiple numeral forms in
translations, I'd like to re-iterate why this important and how such a string is
correctly translated.
Most languages have at least two numeral forms - one singular and a plural one.
Which numbers are regarded as singular and which ones are regarded as plural
differs from language to language. A lot seem to include zero in the plural form
(like English, German or Finnish), others, like French, treat both zero and one
as a singular and everything else as a plural.
Russian for instance has four forms: a "one"-form (for numbers like 1, 21, 31,
... , 1001, ... but not 11!), a "few"-form (for numbers like 2-4, 22-24, 32-34,
102, 1002, ...), a "many"-form (for numbers like 0, 5-19(!), 100, 1000, 10000,
...) and a last "other"-form (for numbers like 0.0-1.5, 10.0, ...)
If we want to display a string with an embedded arbitrary number, then of course
this most be correctly modeled according to the target language's grammar rules.
This covers the "why" part.
I've attached two screenshots of Qt Linguist with a numerus-form translation
string selected.
In the first screenshot (ending in -1), the application was maximized and you
can easily see that there are two fields for the German translation - one for
the singular form and a second one for the plural form. In this screenshot, the
source string was incorrectly translated with a plural form in the singular
field and the plural field left empty.
Compare this to the second screenshot (ending in -2): it's the same window, but
not maximized. This time, due to bad luck, only the first translation form is
displayed. If the window is layout like this, it's easy to miss the second
plural case, which can only be reached via scrolling (note the scroll bar).
If you see a source string containing "%n", please be extra-careful and remember
to scroll down when translating such entries, since you typically need to
provide more than one translation string.
Mihai