Pluralizing Translations in i18n | Hacker News Reader

Pluralizing Translations in i18n

2n.pl

19 pointspan_sarin8 comments

For those confused (like I was at first); judging by the post, this is I18n pluralization in Ruby on Rails.

Pluralization is really old, gnu gettext solve this long time ago (1995).

http://www.gnu.org/software/gettext/manual/gettext.html

But you are right, many projects not support this, many don't know it or they are not interessted in other languages.

Yep, gettext has been working well for decades (and ported to many languages), though the syntax is a little clunky I find.

Polish has 3 plural forms: 1: When the number is 1 2: When the number ends in 2,3,4, but not 12,13,14 3: For all other numbers

It beats me how translators would use "few" for plural form 2 (which includes 22, 33, 44). "Zero" and "other" are actually the same form. A good translation tool would spare the effort of entering the same information twice.

How would this work for Chinese, where all numbers are treated the same?

I would never present these keywords to translators, but instead show the full explanation like in the plural forms above.

Translators should know both the source and target language well enough to know the rules for the different plural forms. Using a name here is probably just to have a key for them (and frankly, better than gettext's approach). The correct name apparently would be singular, plural, genitive plural, if I'm reading Wikipedia right, but that's a bit cumbersome to type every time.

Interesting to see the pluralization rules for so many languages expressed so succinctly.

Apparently in Polish 11 gets the "many" form (:other in this Ruby library), but 12 to 14 get the "few" form, like 2 to 4. But in Russian and a few other Slavic languages, 11 gets the "one" form.

You probably mean useful. And known.

+1: OP probably should have checked the English grammar and spelling when writing about... grammar and spelling

Reply on news.ycombinator.com