Correctly implement function equality testing within structures

[jwiegley]

This issue tracks the fact that the following tests in EvalTests.hs fail when MATCHING_TESTS=yes:

• case_function_equals1
• case_function_equals2

[matthewbauer]

Hmmm... I'm confused by the error:

    function equals2:
FAIL (0.02s)
      tests/TestCommon.hs:65:
      ([(x: x)] == [(x: x)])
      expected: "false\n"
       but got: "true\n"

But ([(x: x)] == [(x: x)]) should be true and the test case is constantEqualText "true" "([(x: x)] == [(y: y)])". Where does the false come from?

[ryantrinkle]

I think it comes from the the C++ implementation: ``` » nix-repl Welcome to Nix version 1.11.16. Type :? for help. nix-repl> ([(x: x)] == [(x: x)]) false ```

…

On Thu, Aug 9, 2018 at 3:39 PM, Matthew Bauer ***@***.***> wrote: Hmmm... I'm confused by the error: function equals2: FAIL (0.02s) tests/TestCommon.hs:65: ([(x: x)] == [(x: x)]) expected: "false\n" but got: "true\n" But ([(x: x)] == [(x: x)]) should be true and the test case is constantEqualText "true" "([(x: x)] == [(y: y)])". Where does the false come from? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#340 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABGlYD7s9-XSlYIsS0J8hzloBA30ROj_ks5uPI_dgaJpZM4VvaGW> .

[matthewbauer]

Very weird. Nix has very weird behaviors. I will fix it with the test mentioned in #323.

[matthewbauer]

Hnix isn't running Nix in those tests. I think I was just thinking about it backwards the left side is the actual, the right is the expected.