a compilation issue caused by the order of two 'requires' directives

[Alan Bateman]

On 13/01/2012 17:17, Dmitry Bessonov wrote:
> Hello,
>
> While playing with the recent JigSaw build
> I came across a strange compilation issue which exposes
> the inability to resolve module dependences in case
> when two 'requires' clauses are listed in particular order.
>
> The structure is the following - 3 modules are compiled
> and installed one after the other.
>
> module bar @ 1.0 {   }
>
> module foo @ 1.0 {
>    requires optional bar @>1.0;
> }
>
> module test @ 1.0 {
>     requires bar;
>     requires foo;
> }
>
> The compilation fails for the last module ('test').
> To compile module successfully it is just needed
> to interchange two 'requires' clauses in module 'test'.
>
> Best regards,
> Dmitry
This is indeed confusing and a reminder that the resolver will need good 
diagnostics so that javac and install/other time resolution can emit 
helpful messages.

I believe that when the 3 module-info files are compiled together that 
the resolver will be called with all 3 in the root query set and that it 
will fail as expected (because bar at 1.0 doesn't satisfy the needs of both 
foo and test).

When compiled individually then it does appear that there is an issue. I 
believe foo will compile because the configuration for foo will be 
computed  (won't fail because the bar@>1.0 is an optional dependency). 
However when it comes to test then it does appear that the order changes 
changes. Jon - is there any possibility that the order issue is due to 
the Catalog implementation in javac? If not then I guess this means 
going through the resolver trace to see what is going on.

-Alan