Joey
Joey
Thanks for the encouragement @elliott5 , yes indeed it is quite the esoteric problem, with this out of the way I'm now working on priority number 1 of getting std...
@elliott5 Well done, now how do we move forward on this, should you or I try and look into the internals and fix it, or submit an issue to go/tools...
@elliott5 ah sorry should've been more clear, I think it's a bug that recv is null not TypeParams/results/args etc. If I add an argument into the function like so: ```go...
@elliott5 Updated to the latest version. At the bottom of the ``go.mod`` This line: ``` replace golang.org/x/tools => ./tools ``` Is used to replace tools with the fork of tools...
@elliott5 It's using the package inside go2hx's directory named "tools" https://github.com/go2hx/tools/blob/master/go/packages/packages.go#L252
This is a stellar and thorough job well done. I'm fine with leaving the panic() in case it can be used to quickly catch a case that slips through the...
@elliott5 Of course. Unfortunately if I remove line 148 to 175, the unit tests fail on https://github.com/go2hx/go2hx/blob/master/tests/unit/complex1.go I'll look into it more tomorrow. At quick glance (full speculation mode) I...
@elliott5 Ah no worries, Do you by chance have a code example for this? > When generating code for a call, if the signature type has a nil recv value,...
Also I believe you've tested this so no speculation required! > "full speculation mode": this will always be the method's named type rather than a pointer to it. If recv...
I have removed the type patcher. You could submit a bug report with that test example. > What I wonder is if it would be helpful to also capture the...