|
By Akerbeltz - Tuesday, June 28, 2011
|
Hi guys,
Been a while and I just noticed to my shock my language had dropped below 50%. One thing I noticed is that there seem to be a lot more phrase generation strings such as {0} objects selected.
That works for English where you have one form of a noun for 1 and another form for anything more than one, including 0. But a lot of languages are more complex than that (apologies if I'm asking something that's already been sorted but I searched for plural rules and couldn't find any).
Compare the ruleset Mozilla (the folk who do Firefox) have: https://developer.mozilla.org/en/Localization_and_Plurals
Is there anything comparable on GenoPro (yet)? I just want to ask before I re-start translation.
Cheers
|
|
By Akerbeltz - Tuesday, July 12, 2011
|
Please don't all answer at the same time! 
|
|
By jcmorin - Tuesday, July 12, 2011
|
The translation use the same technology as the Phrase generator use by Ron and others to create nice report.
If you have special plural rules (0,1,2, 2+ elements). You can do it with conditional. Take your time and read this document: |
|
By Akerbeltz - Saturday, February 18, 2012
|
Ok, this is a bit over my head I'm afraid. Let's say we have this string:
The family tree has {0} individuals and is too big for the AutoArrange!
Gaelic requires different handling of nouns following this pattern:
1, 11 require form I (duine)
2, 12 require form II (dhuine)
3-10, 13-19 require form III (daoine)
0, 20+ require form IV (duine)
So how do I have to code that?
|
|
By GenoProSupport - Monday, February 20, 2012
|
|
Ideally, this has to be implemented inside GenoPro, unless you want to write your own function. Looking at https://developer.mozilla.org/en/Localization_and_Plurals, it would not be too hard to implement.
|
|
By Akerbeltz - Thursday, July 12, 2012
|
So how can I request that sort of thing is coded into GenoPro?
Addendum: Since response to this issue seems to be slow, would this workaround be ok? I found
[{?0=1} one more day][{!}{0} more days]
in the files. Which I re-modelled as
[{?0=1|11}{0} taigh][{?0=2|12}{0} thaigh][{?0=3|4|5|6|7|8|9|10|13|14|15|16|17|18|19|20}{0}taighean][{!}{0} taigh]
I'm not great shakes at code so feedback would be much appreciated.
|
|
By Akerbeltz - Wednesday, August 1, 2012
|
Ok, thanks jcmorin ... I'm a tiny little bit closer now 
I have entered this:
But in live, that gives me strange stuff if I select objects:
Since I can't add *all* number manually, I tried to do an "or else" function but that creates real havoc:
And I also does not like me using OR:
What am I doing wrong?
|
|
By genome - Tuesday, September 18, 2012
|
Having read your latest post, I had an attempt at this using the Phrase Editor. The Else {!} applies to the immediately preceding conditional, so theretically to get the result you want you need to nest the clauses and should be able to use
| Tha [{?0=1}{0} taigh][{!}[{?0=2}{0} thaigh][{!}[{?0=3}{0} taighean][{!}{0} taighaen]]] |
But whilst this appears to work in the GenoPro Phrase Editor, when actually applied in a localization then the phrase template is not converted correctly as you have already discovered.
This seems therefore to be a real bug in GenoPro and perhaps is more easily fixed than the 'limitation' your refer too.
Perhaps Dan or JC could comment on this. |
|
By Akerbeltz - Friday, September 28, 2012
|
Just found another related bug.I was working on this:
The registration key '{0}' is not valid[ because it had expired [{?1=1}yesterday][{!}{1} days ago]].
and since Gaelic does not allow "days ago" until you get past 2, I tried
[{?1=1}an-dè][{?1=2}a-bhòin-dè][{!}{1} làithean air ais]
I get the right forms (ignoring the bad plurals) but for yesterday I now get this weird one:
an-dè1 làithean air ais
|
|
By genome - Saturday, September 29, 2012
|
As before you need to nest the 'else' phrases| [{?1=1}an-dè][{!}[{?1=2}a-bhòin-dè][{!}{1} làithean air ais]] |
|
|
By Akerbeltz - Saturday, September 29, 2012
|
|
Thanks. It seems to produce the right outcome in the editor but I'm not sure if it produces the right outcome in the UI, given the above problem.
|
|
By Akerbeltz - Sunday, February 3, 2013
|
|
There still doesn't seem to be a solution that actually works in the live system. Given it affects a large number of languages, shouldn't this become an in-program feature in such a way that it can handle various plurals? A lot of programs have them, Mozilla does, Wikis do, LibreOffice and OpenOffice also handle it - it's quite common and not that hard I'm given to understand (for a programmer that is).
|
|
By Akerbeltz - Wednesday, September 17, 2014
|
|
Any news? I take it there are still developers around?
|