Re: [Chicken-hackers] [PATCH] Fix line numbers in error messages inside modules

[Felix]

>> If the module form is the result of a macro-expansion, then (at least)
>> the export list will be invalid, as these need to be pure, unaliased
>> names.
> 
> I was wondering and unsure about that.  How should this work?
> For example, if I put a module definition in a LET statement, it doesn't
> work at all, neither inside the LET nor outside (with or without my patch).
> I'm not sure that's even supposed to work.  The s48-modules tests don't
> work in compiled mode with the old or new versions either.  I think we
> could have a look at that after 4.8.0 is released, since these are
> rather unusual use-cases.

"module" is not supposed to work anywhere besides in a toplevel context
(this is documented, AFAIK). The s48-modules tests appear to work for me.
Or do you mean whjen put in a "let" context?

> 
> I guess the module name itself should be stripped as well as the exported
> symbols, since modules are "global".  I guess that's also the reason
> strip-syntax was being used.

That's correct. I will add this now.

> 
>> I have applied your patch and modified the processing of ##core#module
>> to strip the export list.
> 
> Thank you!

You're very welcome.


cheers,
felix