<headius> lopex: yeah that last commit to jcodings, I had < instead of <= for the final UTF-8 codepoint check
<headius> MRI had multiple tests dealing with that codepoint
<headius> specs did not appear to have any
<headius> lopex: have you looked at the case-folding improvements they made in MRI (2.3 or 2.4, I think the latter)
<headius> at the moment I'm not going to fix these issues but will tag as a group and file a bug
<lopex> yeah, pretty intensive
<headius> big changes?
<lopex> trying to recall
<lopex> headius: in the core (String#*case) it's mostly rb_str_casemap
<lopex> headius: which in turn uses enc->case_map
<lopex> so not very intensive for core
<lopex> but more for jcodings
<headius> I saw that getting used all over now, yeah
<lopex> headius: yeah, the biggest paint is the case_map field in enc
<lopex> since lots have specialised impl
<lopex> so pretty controllable
<lopex> and I dont see onigmo using it
<lopex> or they have amacro for enc->case_map
<lopex> dont see one though
<lopex> headius: yeah, for unicode joni should be ok
<lopex> headius: and for iso etc mri doesnt support much either
<headius> ok
<headius> we'll have to do that at some point
<lopex> headius: so only two things, case_map functions for encodings and a string.c
<lopex> quite noninvasie
<lopex> rb_str_casemap is quite small
<lopex> current_buffer = current_buffer->next
<lopex> doh, when I see those things..
<lopex> asdf
<lopex> 27 encoding impls for case_map
<lopex> other use either onigenc_ascii_only_case_map or onigenc_single_byte_ascii_only_case_map
<headius> hmmm
<headius> so a bit of porting there
https_GK1wmSU has joined #jruby
https_GK1wmSU has left #jruby [#jruby]
GitHub76 has joined #jruby
<GitHub76> jcodings/master 5046712 Charles Oliver Nutter: Return error code, not zero.
GitHub76 has left #jruby [#jruby]
<GitHub76> jcodings/master 1e1f98a Charles Oliver Nutter: MRI does not treat any single-byte encodings differently....
<GitHub76> [jcodings] headius pushed 2 new commits to master: https://git.io/v7BPA
<GitHub17> [jruby-openssl] kares reopened issue #134: X509ExtensionFactory generates incorrect extension for subjectAltName https://git.io/vHT6R
<GitHub188> [jruby] headius created test-integer-chr at 439f915 (+0 new commits): https://git.io/v7BXT
<GitHub12> [jruby] headius pushed 1 new commit to test-integer-chr: https://git.io/v7BXP
<GitHub12> jruby/test-integer-chr e343686 Charles Oliver Nutter: Various fixes for Integer#chr....
etehtsea has joined #jruby
GitHub33 has joined #jruby
GitHub33 has left #jruby [#jruby]
<GitHub33> jcodings/master 47d4524 Charles Oliver Nutter: [maven-release-plugin] prepare release jcodings-1.0.24
<GitHub33> [jcodings] headius pushed 1 new commit to master: https://git.io/v7BDc
GitHub37 has joined #jruby
GitHub37 has left #jruby [#jruby]
<GitHub37> [jcodings] headius tagged jcodings-1.0.24 at 08426e0: https://git.io/v7BDC
GitHub102 has joined #jruby
GitHub102 has left #jruby [#jruby]
<GitHub102> jcodings/master 509bce7 Charles Oliver Nutter: [maven-release-plugin] prepare for next development iteration
<GitHub102> [jcodings] headius pushed 1 new commit to master: https://git.io/v7BDW
<GitHub110> [jruby] headius pushed 1 new commit to ruby-2.4: https://git.io/v7BDu
<GitHub110> jruby/ruby-2.4 0f746f6 Charles Oliver Nutter: Squashed commit of the following:...
etehtsea has quit [Ping timeout: 260 seconds]
prasun has joined #jruby
olle has joined #jruby
rawra has joined #jruby
olle has quit [Quit: olle]
rawra has quit [Ping timeout: 260 seconds]
etehtsea has joined #jruby
shellac_ has joined #jruby
etehtsea has quit [Client Quit]
jensnockert has joined #jruby
jensnockert has quit [Client Quit]
rawra has joined #jruby
shellac_ has quit [Quit: Computer has gone to sleep.]
vtunka has joined #jruby
vtunka has quit [Quit: Leaving]
shellac has quit [Quit: Leaving]
shellac has joined #jruby
drbobbeaty has joined #jruby
vtunka has joined #jruby
drbobbeaty has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
olle has joined #jruby
drbobbeaty has joined #jruby
bbrowning_away is now known as bbrowning
olle has quit [Quit: olle]
olle has joined #jruby
rawra_ has joined #jruby
rawra has quit [Read error: Connection reset by peer]
prasun has quit [Ping timeout: 255 seconds]
vtunka has quit [Quit: Leaving]
vtunka has joined #jruby
olle_ has joined #jruby
olle has quit [Ping timeout: 268 seconds]
olle_ is now known as olle
vtunka has quit [Quit: Leaving]
rawra_ has quit [Remote host closed the connection]
rawra has joined #jruby
rawra_ has joined #jruby
rawra_ has quit [Remote host closed the connection]
rawra has quit [Ping timeout: 246 seconds]
vtunka has joined #jruby
rdubya has joined #jruby
vtunka has quit [Quit: Leaving]
rawra has joined #jruby
olle has quit [Quit: olle]
rawra has quit [Ping timeout: 260 seconds]
rawra has joined #jruby
prasun has joined #jruby
rawra has quit [Ping timeout: 240 seconds]
_whitelogger has joined #jruby
drbobbeaty has quit [Read error: Connection reset by peer]
drbobbeaty has joined #jruby
shellac_ has joined #jruby
shellac_ has quit [Quit: Computer has gone to sleep.]
rawra has joined #jruby
shellac_ has joined #jruby
rawra has quit [Ping timeout: 246 seconds]
subbu is now known as subbu|lunch
<GitHub7> [jruby] headius opened issue #4731: [2.4] New case-folding logic https://git.io/v70Mi
prasun has quit [Remote host closed the connection]
shellac_ has quit [Ping timeout: 246 seconds]
subbu|lunch is now known as subbu
<nirvdrum> enebo: I'm intent on making this thing not suck so much: https://github.com/jruby/jruby/blob/master/spec/ruby/core/encoding/converter/convpath_spec.rb#L6-L8
<enebo> nirvdrum: in performance or grossness?
<nirvdrum> Performance. It takes like 10s in jcodings and 0.25s in MRI.
<nirvdrum> Hmm... that may not be true. JRuby 9.1.12.0 is much faster.
<nirvdrum> Still slower than MRI, but only ~1s.
<enebo> how long does it take
<enebo> I see
<nirvdrum> We basically call into jcodings for this, so I wonder what we're doing wrong.
<enebo> could it be graal?
<enebo> something about how it is coded?
<nirvdrum> I'm doing all this in HotSpot.
<enebo> hmm
<headius> nirvdrum: perhaps it forces all the other encodings to load?
<enebo> headius: bingo
<nirvdrum> Wouldn't JRuby have to load them, too?
<enebo> nirvdrum: run it in a loop
<lopex> nirvdrum: do you have a profile ?
<nirvdrum> lopex: Nothing really sensible. The AST gets in the way.
<enebo> oh yeah I guess the 10s is a big deal :) unless you did something which somehow makes all encodings eager...even then 10s?
<nirvdrum> We eager loaded them at one point and I think that was 1 - 1.5s.
<nirvdrum> Well, we must be doing something dumb. I'll have to track it down.
rdubya has quit [Quit: Leaving.]
lanceball is now known as lance|afk
drbobbeaty has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
<lopex> headius, enebo: eventually it happened https://github.com/rep-movsd/see-phit
<enebo> lopex: it should be called CRB
<lopex> crb ?
bbrowning is now known as bbrowning_away
<enebo> lopex: like ERB
rawra has joined #jruby
rawra has quit [Ping timeout: 240 seconds]
https_GK1wmSU has joined #jruby
https_GK1wmSU has left #jruby [#jruby]
<GitHub9> [jruby] headius pushed 4 new commits to ruby-2.4: https://git.io/v7EuM
<GitHub9> jruby/ruby-2.4 aa17c14 Charles Oliver Nutter: Commit regenerated pom.xml.
<GitHub9> jruby/ruby-2.4 007ce3d Charles Oliver Nutter: Tag peculiar or MRI-specific failures.
<GitHub9> jruby/ruby-2.4 7734c91 Charles Oliver Nutter: Tag casecmp failures pending new case-folding logic (#4731)