E0005; should all constructors be listed? #56379

Activity

1. Centril

   added

   A-diagnosticsArea: Messages for errors, warnings, and lintsArea: Messages for errors, warnings, and lints

   T-compilerRelevant to the compiler team, which will review and decide on the PR/issue.Relevant to the compiler team, which will review and decide on the PR/issue.

   on Nov 30, 2018
2. Centril

   assigned

   estebank

   on Nov 30, 2018
3. estebank commented on Nov 30, 2018

   

   estebank

   on Nov 30, 2018

   Contributor

   More actions

   Maybe we could do that with an upper bound, as having more than ~3 listed ranges will get really noisy. Imagine if you had a check for every other integer in u16, you'd have 32K ranges! I'd be fine with it as long as this is handled gracefully.

   It'll also be nice to mention when a range between two values involves std::{integer}::MAX or MIN.

   👍React with 👍2Reacted by varkor and Mazdak Farrokhzad
4. Centril commented on Nov 30, 2018

   

   Centril

   on Nov 30, 2018

   ContributorAuthor

   More actions

   @estebank Yeah an upper bound makes sense.
5. varkor commented on Nov 30, 2018

   

   varkor

   on Nov 30, 2018

   Last edited by varkor

   Contributor

   More actions

   It'll also be nice to mention when a range between two values involves std::{integer}::MAX or MIN.

   Yeah, this was a case I was thinking about too — it'll probably be fairly common case, so it'd be good to handle sensibly.

   Edit: I've filed #56393 for this specific suggestion.
6. estebank commented on Oct 10, 2019

   

   estebank

   on Oct 10, 2019

   Contributor

   More actions

   @Centril is there still work left to do on this front?
7. varkor commented on Oct 10, 2019

   

   varkor

   on Oct 10, 2019

   Contributor

   More actions

   The match algorithm still short-circuits and reports only the first unencountered constructor in some cases. You can see this in some of the test cases for #65160, for instance.

   👍React with 👍1Reacted by Esteban Kuber
8. crlf0710

   added

   C-enhancementCategory: An issue proposing an enhancement or a PR with one.Category: An issue proposing an enhancement or a PR with one.

   on Jun 11, 2020
9. Nadrieril commented on Nov 22, 2020

   

   Nadrieril

   on Nov 22, 2020

   Last edited by Nadrieril

   Member

   More actions

   Couldn't find an example with missing integer ranges like the one mentioned in the OP; I suspect we've fixed those. Here's a different one (play):

   enum Foo {
       A(bool),
       B(bool),
       C(bool),
   }
   fn main() {
       match Foo::A(true) {
           Foo::A(true) => {},
           Foo::B(true) => {},
           Foo::C(true) => {},
       }
   }

   This only errors about A(false) not being covered.

   I'm adding my favorite label so I can find this later.
   @rustbot modify labels: +A-exhaustiveness-checking

   👍React with 👍1Reacted by Esteban Kuber
10. rustbot

    added

    A-exhaustiveness-checkingRelating to exhaustiveness / usefulness checking of patternsRelating to exhaustiveness / usefulness checking of patterns

    on Nov 22, 2020
11. bors

    closed this as completedin 1f5bc17on Dec 19, 2020