<
headius[m]>
Array is 100% on 2.6 branch
<
lopex>
that BitSet in #difference just for memory locality ?
<
headius[m]>
actually I didn't give it much thought... it's probably a wash
<
headius[m]>
I have never actually looked into BitSet perf
<
headius[m]>
lopex: look into BitSet perf :-)
<
headius[m]>
better optimization would be to split out the 1-argument path so it's not creating these extra arrays in that case
<
headius[m]>
I just didn't want to get into a perf quagmire when I'm working on compat
<
lopex>
but it's just an arg lookup
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
<
headius[m]>
actually it's probably better as boolean[]
<
headius[m]>
BitSet uses a long[]
<
lopex>
but it only pays off if you do a big number of args to it
<
headius[m]>
Dir is done
<
headius[m]>
Enumerator and Enumerable are done
<
headius[m]>
and Enumerator::Chain and Lazy
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
ur5us has joined #jruby
<
lopex>
headius on a rampage
travis-ci has left #jruby [#jruby]
travis-ci has joined #jruby
<
headius[m]>
woohoo, 70 fails
<
headius[m]>
and 51 of those are nunmeric conversion methods taking exception: false
<
headius[m]>
the end is in sight
ur5us has quit [Ping timeout: 256 seconds]
<
headius[m]>
ok, that's it for me tonight
<
headius[m]>
if we I can knock out those numeric exception specs tomorrow I might just tag off the rest and call it good
<
headius[m]>
we're close
havenwood has joined #jruby
havenwood has quit [Changing host]
havenwood has joined #jruby
havenwood has joined #jruby
havenwood has quit [Changing host]
havenwood has joined #jruby
ur5us has joined #jruby
ur5us has quit [Quit: Leaving]
ur5us has joined #jruby
ur5us has quit [Ping timeout: 240 seconds]
sagax has quit [Read error: Connection reset by peer]
sagax has joined #jruby
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
xardion has quit [Remote host closed the connection]
xardion has joined #jruby
KeyJoo has joined #jruby
<
headius[m]>
huh why do we have a BitSet class in joni?
<
headius[m]>
hmm, maybe the version isn't even in here
<
lopex>
headius[m]: oh, yeah
<
lopex>
there's emoji version too afaik
<
lopex>
headius[m]: bitset is for ascii char classes
<
headius[m]>
yes but why not use jdk BitSet
<
lopex>
faster copying for compilation
<
headius[m]>
fair enough
<
headius[m]>
golly when did we last update joni unicode tabes
<
headius[m]>
tables
<
lopex>
it's 12.1.0
<
headius[m]>
I can hardcode a property in JRuby for now but I'm looking at joni source so I can add that somewhere
<
lopex>
it should be in jcodings though
<
headius[m]>
you mean it's already there or it should be added? I could not find it there
<
lopex>
hmm I think it shuld be parser by generator from name2ctype.h
<
lopex>
ONIG_UNICODE_VERSION_STRING and ONIG_UNICODE_EMOJI_VERSION_STRING
<
headius[m]>
yeah perfect
<
headius[m]>
where's the generator code now
<
lopex>
havent tested it on osx though
<
headius[m]>
oh weird, idea doesn't show that dir
<
headius[m]>
er wait
<
headius[m]>
I don't have that dir either
<
lopex>
it's in jcodings
<
headius[m]>
oh lord, coffee hasn't kicked in
<
headius[m]>
I'm looking at joni why
<
headius[m]>
ah yes
<
headius[m]>
hardcoded in generator right now
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
lucasb has joined #jruby
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
<
headius[m]>
scooty scoot
<
headius[m]>
Kernel is done
<
headius[m]>
should be below 20 fails now
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
travis-ci has joined #jruby
travis-ci has left #jruby [#jruby]
<
headius[m]>
woah, another new syntax
<
headius[m]>
t[42, &blk] ||= 42
<
headius[m]>
that's bizarre
<
headius[m]>
hah and there's a test for that syntax error output with ^ pointing out the bad token
ur5us has joined #jruby
<
headius[m]>
hmm some of these unicode constant tests use characters that JDK doesn't think are caps
<
headius[m]>
the combined Dz character does not pass isUpperCase in Java
<
headius[m]>
actually this one
<
headius[m]>
yeah probably not going to be able to do anything about these
<
headius[m]>
U+01F1 does report as caps but 1F2 does not
<
headius[m]>
1F2 is in the tests
<
headius[m]>
gods we emit a lot of bytecode for defined?
ur5us has quit [Ping timeout: 240 seconds]