Calling a method that returns 'itself' on a multi-value should not lose multi-value structure?

[radeusgd]

I think that indeed it is undesirable to lose the multi value structure in methods such as:

type My_Type
    method self = if self.some_property then self else 0

This is demonstrated in ffe1172.

FYI:

• Let EnsoMultiValue.to_text delegate to first type to_text #11827
• multi_value.to doesn't work on second & further elements of intersection type #11935

[radeusgd]

The relevant tests are

group_builder.specify "calling a method on one of the types should not lose the intersection type" <|
    ab = make_a_and_b

    ab.a_id_unchecked . is_a A . should_be_true
    ab.a_id_unchecked . is_a B . should_be_true
    ab.a_id_unchecked.a_method . should_equal "A method"
    ab.a_id_unchecked.b_method . should_equal "B method"

    # Checked variant can hide the B part
    ab.a_id . is_a A . should_be_true
    ab.a_id . is_a B . should_be_false
    # But it can be uncovered via explicit cast
    b = (ab.a_id):B
    b.is_a A . should_be_false
    b.is_a B . should_be_true

    new_ab = (ab.a_id):A&B
    new_ab.is_a A . should_be_true
    new_ab.is_a B . should_be_true
    new_ab.a_method . should_equal "A method"
    new_ab.b_method . should_equal "B method"

    # The same should apply to the B part
    ab.b_id_unchecked . is_a A . should_be_true
    ab.b_id_unchecked . is_a B . should_be_true
    ab.b_id_unchecked.a_method . should_equal "A method"
    ab.b_id_unchecked.b_method . should_equal "B method"

    ab.b_id . is_a A . should_be_false
    ab.b_id . is_a B . should_be_true
    new_ab_2 = (ab.b_id):A&B
    new_ab_2.is_a A . should_be_true
    new_ab_2.is_a B . should_be_true

group_builder.specify "calling `.catch` on an intersection type should not lose the refinements" <|
    ab = make_a_and_b
    r = ab.catch Any _->"catched"
    r.is_a A . should_be_true
    r.is_a B . should_be_true
    r.a_method . should_equal "A method"
    r.b_method . should_equal "B method"

group_builder.specify "calling `.throw_on_warning` on an intersection type should not lose the refinements" <|
    ab = make_a_and_b
    r = ab.throw_on_warning
    r.is_a A . should_be_true
    r.is_a B . should_be_true
    r.a_method . should_equal "A method"
    r.b_method . should_equal "B method"

group_builder.specify "attaching warnings to an intersection type should not lose the refinements" <|
    ab = make_a_and_b
    ab_with_warning = Warning.attach (Illegal_State.Error "my warning") ab
    ab_with_warning.is_a A . should_be_true
    ab_with_warning.is_a B . should_be_true
    ab_with_warning.a_method . should_equal "A method"
    ab_with_warning.b_method . should_equal "B method"
    Problems.expect_only_warning Illegal_State ab_with_warning

group_builder.specify "removing warnings from an intersection type should not lose the refinements" <|
    ab = make_a_and_b
    ab_with_warning = Warning.attach (Illegal_State.Error "my warning") ab

    ab_without_warning = ab_with_warning.remove_warnings
    Problems.assume_no_problems ab_without_warning
    ab_without_warning.is_a A . should_be_true
    ab_without_warning.is_a B . should_be_true
    ab_without_warning.a_method . should_equal "A method"
    ab_without_warning.b_method . should_equal "B method"

    ab2 = ab.remove_warnings
    ab2.is_a A . should_be_true
    ab2.is_a B . should_be_true
    ab2.a_method . should_equal "A method"
    ab2.b_method . should_equal "B method"

To be honest, based on @JaroslavTulach's workaround 9bb73ae led me to believe that all of these tests will fail. Somehow, in my tests I see:

    - [FAILED] calling a method on one of the types should not lose the intersection type [46ms]
        Reason: Expected False to be True (at Base_Tests\src\Semantic\Multi_Value_As_Type_Refinement_Spec.enso:189:13-55).

    - calling `.catch` on an intersection type should not lose the refinements [15ms]
    - calling `.throw_on_warning` on an intersection type should not lose the refinements [37ms]
    - attaching warnings to an intersection type should not lose the refinements [75ms]
    - removing warnings from an intersection type should not lose the refinements [52ms]

So only the first one fails but others pass already. That suggests to me that these tests are inexhaustive, because the workaround that @JaroslavTulach added was for throw_on_warning. For now I was not yet able to find a good repro.

[radeusgd]

Eureka!

In b8407da I was able to add tests that are failing. When on a walk, I've realized what was the difference between my test cases and the -> Column example. The difference was that the 'refinement' part was hidden.

Apparently, the hidden part of the type gets cut off at dispatch of e.g. remove_warnings or catch.

I think it should not disappear as that will be very confusing to our users. catch or remove_warnings are relatively common user-facing operations.

[radeusgd]

Additionally, the tests from b8407da uncover a related assertion failure:

Execution finished with an error: java.lang.AssertionError: Dispatch (MultiType{types=[A, A]} and extra MultiType{types=[]} should be disjoin!
        at <java> org.enso.runtime/org.enso.interpreter.runtime.data.EnsoMultiValue$NewNode.newValue(EnsoMultiValue.java:98)
        at <java> org.enso.runtime/org.enso.interpreter.node.typecheck.AllOfTypesCheckNode.executeCheckOrConversion(AllOfTypesCheckNode.java:84)
        at <java> org.enso.runtime/org.enso.interpreter.node.typecheck.TypeCheckValueNode.handleCheckOrConversion(TypeCheckValueNode.java:56)
        at <java> org.enso.runtime/org.enso.interpreter.node.typecheck.TypeCheckExpressionNode.executeGeneric(TypeCheckExpressionNode.java:23)
        at <java> org.enso.runtime/org.enso.interpreter.node.callable.thunk.ForceNodeGen.executeGeneric(ForceNodeGen.java:46)
        at <java> org.enso.runtime/org.enso.interpreter.node.scope.AssignmentNodeGen.executeGeneric(AssignmentNodeGen.java:51)
        at <java> org.enso.runtime/org.enso.interpreter.node.callable.function.BlockNode.executeGeneric(BlockNode.java:70)
        at <java> org.enso.runtime/org.enso.interpreter.node.ClosureRootNode.execute(ClosureRootNode.java:85)
        at <enso> Multi_Value_As_Type_Refinement_Spec.add_specs.Multi_Value_As_Type_Refinement_Spec.add_specs<arg-1>(Internal)
        at <enso> case_branch.case <internal-216>(Group.enso:30:61-64)
        at <enso> Helpers.execute_spec_code<arg-1>(Helpers.enso:57:18-34)
        at <enso> Panic.catch(Internal)
        at <enso> Panic.type.recover(Panic.enso:189-194)
        at <enso> Helpers.execute_spec_code(Helpers.enso:56-60)
        at <enso> case_branch<arg-3>(Helpers.enso:48:37-63)
        at <enso> case_branch<arg-1>(Helpers.enso:48:13-64)
        at <enso> Runtime.no_inline(Internal)
        at <enso> Duration.type.time_execution(Duration.enso:131:18-43)
        at <enso> case_branch(Helpers.enso:47-48)
        at <enso> Helpers.run_spec(Helpers.enso:46-49)
        at <enso> case_branch.test_results(Helpers.enso:27:24-36)
        at <enso> Array_Like_Helpers.map.Array_Like_Helpers.map(Array_Like_Helpers.enso:232:56-77)
        at <enso> Array_Like_Helpers.vector_from_function(Internal)
        at <enso> case_branch(Helpers.enso:24-30)
        at <enso> Helpers.run_specs_from_group(Helpers.enso:21-36)
        at <enso> case_branch(Suite.enso:97:35-92)
        at <enso> Suite.run_with_filter.junit_sb_builder(Suite.enso:95-101)
        at <enso> Array_Like_Helpers.each.Array_Like_Helpers.each(Array_Like_Helpers.enso:272:9-24)
        at <enso> Range.each.go<arg-1>(Range.enso:263:36-51)
        at <enso> Range.each.go<arg-2>(Range.enso:263:21-52)
        at <enso> Range.each.go(Range.enso:262-264)
        at <enso> Range.each<arg-2>(Range.enso:265:13-25)
        at <enso> Range.each(Range.enso:259-265)
        at <enso> Array_Like_Helpers.each(Array_Like_Helpers.enso:271-272)
        at <enso> Vector.each(Vector.enso:817:9-38)
        at <enso> Suite.run_with_filter<arg-1>(Suite.enso:92-101)
        at <enso> Test_Reporter.wrap_junit_testsuites(Test_Reporter.enso:29:14-19)
        at <enso> Suite.run_with_filter(Suite.enso:91-101)
        at <enso> Multi_Value_As_Type_Refinement_Spec.main(Multi_Value_As_Type_Refinement_Spec.enso:336:5-32)