samphippen has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
mister_solo has quit [Ping timeout: 244 seconds]
subbu|away has quit [Ping timeout: 252 seconds]
<rtyler>
nirvdrum: the thinking I had being, if these docs are in jruby/jruby (for example), or any other source repo with a travis build, it'll be easy to make the travis build process validate all the source code
blaines has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
blaines has joined #jruby
blaines has quit [Read error: Connection reset by peer]
mdedetrich has joined #jruby
Hobogrammer has joined #jruby
bf4 has joined #jruby
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] eam opened pull request #3093: isOpen()/close() is racy. Simply close and ignore EBADF. (jruby-1_7...jruby-1_7) http://git.io/vti6O
<JRubyGithub>
[jruby] DougEverly opened issue #3094: NoMethodError being reported as "UndefinedValue.java:52:in `undefinedOperation': java.lang.RuntimeException: IR compiler/interpreter bug: org.jruby.ir.operands” in jruby-9.0.0.0.pre2 http://git.io/vtMyE
JRubyGithub has left #jruby [#jruby]
skade has quit [Quit: Computer has gone to sleep.]
kfpratt has quit [Ping timeout: 248 seconds]
kfpratt has joined #jruby
subbu has joined #jruby
kfpratt has quit [Remote host closed the connection]
<dfr|work>
enebo, actually you might be right that it's a bad regexp.
<dfr|work>
for a longer one MRI seems to hang as well. It's about 5 times as quick for a smaller set, but that just means constant is smaller
<dfr|work>
so not gonna file bug after all :)
<enebo>
dfr|work: seemingly the right regexp would not do a single char capture and then + the capture
<enebo>
it should be /\w+(\w)\z/
<enebo>
At least I am assuming the cost somehow involves the capture + the ‘+’
<dfr|work>
enebo, the code is not even using the capture group, so parens are useless (original regexp in the wild was more like (\w-_) but may as well use (?: instead to avoid capturing overhead
<dfr|work>
so yea, I guess the regexp is bad :)
<enebo>
dfr|work: does using ?: fix perf?
<enebo>
dfr|work: just curious
<dfr|work>
enebo, yes
<dfr|work>
enebo, it's still not instant, but much faster (like maybe a second)
<dfr|work>
interesting, MRI was giving me warning: -e:1: warning: nested repeat operator '?' and '+' was replaced with '*' in regular expression: /(?:\w?)+$/
<enebo>
dfr|work: no doubt some harmless cases can be translated into some simpler bytecode in oniguruma but it is a language and one where idiomatic usage is important so most engines probably don’t work around poor regexps
<lopex>
dfr|work, enebo this might be checked against 1.9.3 or any pre onigmo mri
<lopex>
dfr|work, enebo
<enebo>
dfr|work: I know lopex said there were a lot of outstanding fixes made to oni which is not in joni
<lopex>
there's been some potimizations in onigmo past two years
<dfr|work>
my MRI is 2.2.1p85 so I'm assuming it's pretty late
<dfr|work>
anyhow, I probably agree that the problem is with regexp rather than joni :)
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] mkristian pushed 1 new commit to jruby-1_7: http://git.io/vtyuD
<JRubyGithub>
jruby/jruby-1_7 71f95f7 Christian Meier: use jossl-0.9.8.dev to run test before releasing gem
JRubyGithub has left #jruby [#jruby]
<enebo>
mri22 -e '"aaa" =~ /(a)+/; p $~'
<enebo>
#<MatchData "aaa" 1:"a">
<lopex>
dfr|work: joni is a bit behind and we cannot rule out a bug atm
<enebo>
funny
<lopex>
so both
<enebo>
lopex: so this regexp could be optimized by converting regexp to /a*(a)/
<enebo>
lopex: I was thinking it was keeping all of the captures or something
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius pushed 2 new commits to master: http://git.io/vtyV3
<JRubyGithub>
jruby/master 9093145 Charles Oliver Nutter: Fix hanging in modified FIFO read/write spec, tagged in 04db9d8d....
<JRubyGithub>
jruby/master 363175b Charles Oliver Nutter: Revert "This fails on MacOS (at least) - seems to work on ci/linux"...
JRubyGithub has left #jruby [#jruby]
mcclurmc has joined #jruby
bf4 has quit [Ping timeout: 255 seconds]
<lopex>
enebo: dark matter!
kfpratt has quit [Read error: Connection reset by peer]
<enebo>
lopex: definitely seems like repeated captures is one thing which could get optimized
<enebo>
lopex: not saying I want to figure it out or anything :)
<enebo>
lopex: they are usually in nested captures too
<dfr|work>
enebo, lopex for what it's worth, I'm fixing my code with /^[a-zA-Z]([\w\-\_]?)+\z/ -> /^[a-zA-Z][\w\-\_]*\z/
<dfr|work>
although I'm a bit confused about difference of \z and $, but that's just because I'm a regexp noob.
<enebo>
dfr|work: yeah I prefer to use \z but I think one will match on \0 or \n and the other will go to the end of string
kfpratt has joined #jruby
<enebo>
I think \z is considered more deadly but I hate a random ‘$’ in my code
mcclurmc has quit [Remote host closed the connection]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] enebo pushed 1 new commit to master: http://git.io/vtSWQ
<JRubyGithub>
jruby/master 9d2d570 Thomas E. Enebo: Remove some dead getIRScope usages. Use IRScope from IC rather than ask staticScope for it (closer proximity to ic than ss from interpreterengine's perspective)
JRubyGithub has left #jruby [#jruby]
mcclurmc has joined #jruby
havenn is now known as havenwood
mcclurmc_ has joined #jruby
mcclurmc has quit [Read error: Connection reset by peer]
mcclurmc_ has quit [Remote host closed the connection]
mcclurmc has joined #jruby
phlebas_ has quit [Ping timeout: 256 seconds]
mje113 has quit [Quit: Connection closed for inactivity]
subbu|lunch is now known as subbu
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius pushed 1 new commit to master: http://git.io/vtSVK
<JRubyGithub>
jruby/master 3f44a10 Charles Oliver Nutter: Kill unused method invoker logic and related classes....
<xardion>
For a more complete picture, I'm running a rack app using puma on jruby, and I'm getting a "not opened for writing" IOException trying to write to the stream.
<xardion>
The IO object I'm getting the stream from is the one that gets passed to the rack.hijack lambda
<xardion>
I don't have this problem when using the IO object itself, just the output stream
lanceball has joined #jruby
yfeldblum has joined #jruby
cristianrasch has quit [Quit: Leaving]
lanceball has quit [Changing host]
lanceball has joined #jruby
lanceball has quit [Client Quit]
lanceball has joined #jruby
lanceball has quit [Changing host]
lanceball has joined #jruby
samphippen has joined #jruby
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] enebo closed issue #3055: Cannot delete file after requiring it on Windows in 9.0.0.0.rc1 http://git.io/vLquP
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] enebo pushed 1 new commit to master: http://git.io/vtSb4
<JRubyGithub>
jruby/master 3e75be4 Thomas E. Enebo: Better explain why irScope is a field in StaticScope
<rtyler>
the fact that I wasn't able to replicate this with with 1.7.21-SNAPSHOT on my machine has me very puzzled
<blaines>
1.7.20 => 25 gems installed
<blaines>
both tests above were made with jdk1.7.0_51.jdk
<rtyler>
so have you isolated this to the --local option on `gem install`?
<blaines>
no
<blaines>
I've isolated it to the list of gems to install
<blaines>
jdk1.7.0_51.jdk jruby-complete-1.7.21-SNAPSHOT.jar => ERROR: While executing gem ... (NameError) missing class name (`org.jruby.util.Classpath')
mcclurmc has joined #jruby
skade has quit [Quit: Computer has gone to sleep.]
<xardion>
Well, I switched to using getChannel() (to_channel), and wrapping that with an output stream using Channels.newOutputStream() and explicitly casting it as a WritableByteChannel. Appears to have worked.
mcclurmc has quit [Remote host closed the connection]
blaines has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
mister_solo has quit [Ping timeout: 246 seconds]
camlow32_ has joined #jruby
camlow325 has quit [Read error: Connection reset by peer]