>> On 18 Nov 2006, at 18:26, Graham J Lee wrote: >> >>> >> I've got a few hours tomorrow, I'd like to have some go at >> implementing -stringForObjectValue: >> > > That took longer than I thought :-). Just to make it look like there's > still activity...I've done this now, I'm just going to do some tests and > compare with the OS X version so I should be able to provide a patch later > this week. >
OK, here's the patch to NSNumberFormatter, as well as a patch on the testsuite to verify that a variety of common use-cases work. I also fixed a problem with -attributedStringForObjectValue:withDefaultAttributes: where it would return attributedStringForZero for any number x, |x|<0.5. I didn't put too much effort into checking whether -stringForObjectValue: works with complex formatting strings, because as ratmice pointed out, on Cocoa the various -setFormat: methods actually sanitise the format string which -stringForObjectValue: actually sees. So that still needs to be done ;-) Cheers, Graham. P.S. I think it's possible to get away with using a lot fewer local variables than I needed :-(
nsnumberformatter-patch
Description: Binary data
nsnumberformatter-testsuite-patch
Description: Binary data
_______________________________________________ Bug-gnustep mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-gnustep
