Re: generation of gnu/java/locale/*.uni

[Mark Wielaard]

Hi,

On Mon, 2002-02-18 at 21:05, Eric Blake wrote:
> 
> I just committed an update based on Artur's code.  And since the
> database is no longer a binary file, you should have no problems using
> it straight out of CVS for another run through Mauve.

Running it through mauve (jikes 1.15 + orp 1.0.9) still gives lots of
failures for the Character tests: 168221 of 3603824 tests failed.
Here are a couple of failures that are repeated often:

FAIL: gnu.testlet.java.lang.Character.unicode: Character 0:UNDEFINED has
wrong numeric value of -1 instead of 35 (number 1) (24209 times)

FAIL: gnu.testlet.java.lang.Character.unicode: Character 0:UNDEFINED
incorectly reported as javaidentifierpart (number 1) (14007 times)

FAIL: gnu.testlet.java.lang.Character.unicode: Character 0:UNDEFINED
incorectly reported as unicodeidentifierpart (number 1) (13976 times)

FAIL: gnu.testlet.java.lang.Character.unicode: Character 24:UNDEFINED
incorectly reported as javaindetifierstart (number 1) (13164 times)

FAIL: gnu.testlet.java.lang.Character.unicode: Character 41:UNDEFINED
incorectly reported as unicodeidentifierstart (number 1) (13122 times)

FAIL: gnu.testlet.java.lang.Character.unicode: Character 1bb:UNDEFINED
is reported to be type Lo instead of Cn (number 1) (11506 times)

These could easily be bugs in Mauve though. Mauve uses its own
UnicodeData.txt file.

Cheers,

Mark