drbobbeaty has quit [Ping timeout: 246 seconds]
ur5us_ has joined #jruby
Specialist has quit [Ping timeout: 264 seconds]
mistergibson has joined #jruby
ur5us has joined #jruby
ur5us_ has quit [Ping timeout: 264 seconds]
ur5us_ has joined #jruby
ur5us has quit [Ping timeout: 272 seconds]
<byteit101[m]> hmm, why is PositionAware of a method the line before it's defined? Is it 0-based?
<headius[m]> Yeah it is
ur5us_ has quit [Ping timeout: 260 seconds]
<byteit101[m]> awesome, I'll just add one then
<byteit101[m]> at org.jruby.java.proxies.ConcreteJavaProxy.splitInitialized(ConcreteJavaProxy.java:351)
<byteit101[m]> at rubyobj.Umm.Paintbrush.Balloons.<init>(test.rb:130)
<byteit101[m]> fancy stack traces!
<byteit101[m]> Ok, I think the last major component is for AST/IR fixups
<headius[m]> Oooh nice!
<byteit101[m]> I have AST most of the way, but I need time with enebo to figure it out. Any idea what time he'll be around?
ur5us has joined #jruby
<headius[m]> I believe he is working this week but I'm not sure in what capacity
<headius[m]> I will mention that you are looking for him
ur5us has quit [Ping timeout: 260 seconds]
ur5us has joined #jruby
ur5us has quit [Ping timeout: 265 seconds]
Specialist has joined #jruby
mistergibson has quit [Read error: Connection reset by peer]
mistergibson has joined #jruby
mistergibson has quit [Quit: Leaving]
<enebo[m]> byteit101: howdy...I"M BACK
enebo[m] has joined #jruby
<headius[m]> Woohoo
subbu is now known as subbu|lunch
ur5us has joined #jruby
ur5us has quit [Ping timeout: 264 seconds]
subbu|lunch is now known as subbu
<byteit101[m]> enebo: yay! give a look to the colorized block of code around https://github.com/jruby/jruby/pull/6422#issuecomment-716337104 That is the AST transform I'm trying to do, except I think I should be doing it in IR instead of how I'm doing it now (https://github.com/jruby/jruby/pull/6422/files#diff-ee5b3e6e296d6bf1315743db55b99867762ade5bdfedec34c19bef47ffc8ae98R357-R418 and
<byteit101[m]> *around the super-splitter headline
subbu is now known as subbu|away
ur5us has joined #jruby
<enebo[m]> I will read up a bit
<enebo[m]> byteit101: so what is the behavior if someone calls super/zsuper explicitly? Error if it cannot match a real constructor?
<enebo[m]> I guess I believe this is any super which exists will appropriately just call the specified constructor on java side assuming it can bind it properly
<enebo[m]> In looking at this current WIP I am guessing this does not work yet? That desugar() method in IRMethod will be null by the time you try and JIT
subbu|away is now known as subbu
<enebo[m]> I am done today but I will see anything written and I will be on in the morning. AST is not where I would like to see this as we would have to retain the AST but a pass for IR might mean thinking quite a bit about accounting for values to super and which instr is for super (since we actually will sometimes just replace super with the method call name when we know it).
subbu is now known as subbu|away
Specialist has quit [Ping timeout: 260 seconds]
<byteit101[m]> enebo: thanks!, Yes, see https://github.com/jruby/jruby/pull/6422#issuecomment-720055420 for the decompiled bytecode. This does mostly work right now, but yes, that's one reason why I know I need some help
<byteit101[m]> enebo: been testing with https://matrix.org/_matrix/media/r0/download/matrix.org/tWWwyaCbYJSysxWscQhgiqwh/message.txt and it works as you'd expect today
<byteit101[m]> > I guess I believe this is any super which exists will appropriately just call the specified constructor on java side assuming it can bind it properly
<byteit101[m]> With this work there is nothing to bind to, as we have to pass the args up to the java ctor (the caller of initialize) before resuming the method
<byteit101[m]> see the uninitializedThis discussion on https://logs.jruby.org/jruby/2020-11-02 for the why
ur5us has quit [Ping timeout: 260 seconds]