Translation issues in KUnitConversion::Currency

[Petri Damstén]

On Thursday 03 December 2009 00:23:39 John Layt wrote:
> On Tuesday 01 December 2009 08:16:06 Petri Damstén wrote:
> > Yes, and I think that SI units like i18nc("mass unit symbol", "kg")
> > should not be translated either?
> 
> Not sure about that, I'll check with the translators, but I think they do
>  get translated.  The ISO Currency Codes are designed as a data interchange
>  standard and need to stay constant, hence why they are not normally
>  translated.  I'd guess the SI unit abbreviations being in wider user are
>  likely to have local versions, e.g. I don't know if Arabic speakers would
>  still use kg?  I'll check.

ok

> > I think they should still be case sensitive. Mg and mg are different
> > units. But it's true that there's no need for multiple versions. Lower
> > casing can be handled in runner. Runner now first tries to find perfect
> > match and then startsWith match that can result multiple hits. If
> > startsWith is CaseInsensitive it should work.
> 
> Cool.  So keep only those with the correct case, remove any duplicates with
> 'incorrect' case, and let runner do the work? 

Yes.

> I'm no expert on measurements, but I think I have found some other
>  problems.
> 
> timeunit.cpp:
> 
> julian year => Julian year
> 
> 
> temperature.cpp
> 
> celsiuses and fahrenheits are definitely not real words and should be
>  removed, and I don't think  delisles, reaumurs and romers are either. 
>  newtons I don't think is right in this context.

I think "%1 degree Newton" is good here too.
Newton temperature scale:
http://en.wikipedia.org/wiki/Newton_scale
 
> celsius => Celsius
> fahrenheit => Fahrenheit
> rankine => Rankine
> delisle => Delisle
> newton => Newton
> reaumur => Reamur
> romer => Romer
> 
> I think it should be:
> "%1 Rankine" => "%1 degree Rankine"
> "%1 Rankines" => "%1 degrees Rankine"
> And an extra synonym "Ra"?

Yes, °Ra and Ra. Although ° is incorrect I think it should match it (probably 
in Kelvin too).

> 
> velocity.cpp
> 
> beaufort => Beaufort
> 
> "%1 beauforts" => "Force %1 on the Beaufort scale" or just "Force %1"
> (I've never heard it used as a unit, it's a scale like Celsius)

If "1km/h > b" results "Force 0.479727" it might be confusing so "Force %1 on 
the Beaufort scale" is better.

 
> "%1 speed of light" doesn't quite seem right to me. How is this used in
> practice, given nothing is faster than the speed of light so the value will
>  be between 0 and 1?  For example I think I've only seen 0.99c written as
>  "0.99 of the speed of light" or "0.99 the speed of light" or "99% of the
>  speed of light"?  Not sure.

I think, although impossible (at least with the current knowledge :-), 
"10000000000 km/h > c" should give some answer (result = 9.2c)  

> I can ask for exemptions for these if you agree.

Yes
 
> > So this could be:
> >     i18nc("unit synonyms for matching user input",
> >           "rand;rands;%1;%2;%3",
> >           "ZAR",
> >           KGlobal::locale()->countryCodeToName("za"),
> >           KCurrencyCode::currencyCodeToName("ZAR"))
> 
> Even better :-)
> 
> > This leaves "South Africa Rands" unmatched though.
> 
> Well, the current match string is "rand;rands;ZAR;zar;south africa", so I'm
> not sure how "South Africa Rands" would currently match to that but not the
> new "rand;rands;ZAR;South Africa;South African Rand"?

It's not currently matched, so I think it's fine without the plural.

Petri
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kde-core-devel/attachments/20091203/e4be9db3/attachment.sig>