Fixes 9880

[BontaVlad]

Fixes #9880
I'm still learning nim, of course this PR needs adult supervision.

[krux02]

Can you add a test case please? Apart from that, I agree, the error should contain the bounds.

[narimiran]

I'm still learning nim, of course this PR needs adult supervision.

Your contributions (I'm including here your comment on #10198, which led to a pull-request and a fix) are very appreciated and welcome. Keep them coming as you learn Nim.

[narimiran]

From my understanding (and I might be wrong — please correct me), formatErrorIndexBound returns a valid index range (usually from 0 to len-1), which would mean that in several places where you pass someLength+1 you should pass someLength-1.

(Don't correct anything until somebody confirms this :))

[krux02]

@narimiran A test case that triggers these exceptions should clear that question up.

[BontaVlad]

I'm on board with having tests, I just need some directions on how do I trigger these exceptions and also how to cache the mentioned exceptions.

[krux02]

you should be able to raise the exception when you acces an arrar or a seq outside of its bounds. With try catch you can catch the exception and write an assert for the message.

[BontaVlad]

/cc @narimiran @krux02. I fixed the bounds, also I tried to add tests in the parts that I understood how to trigger the exceptions, for the others I tested with either echo commands or gdb, or common sense. If you know how to trigger the missing ones or you know how to play with vm.nim optcode better than nimscript feel free to ping me.

[timotheecour]

@BontaVlad failure on appveyor is unrelated to your PR, it's caused by #10359

[timotheecour]

friendly ping @BontaVlad

[BontaVlad]

What do I need to do on this PR to get it merged?

[Araq]

Don't use a generic compilerproc, follow what the compilerproc for the C backend does.

[BontaVlad]

Since #10610 was merged, can I close this PR?

[timotheecour]

@BontaVlad yes, I don't think there's any remaining items left out; thanks again!