Ability to skip elements in tuple type annotation #2671

Activity

1. Mohamed Hegazy (mhegazy)

   added

   SuggestionAn idea for TypeScriptAn idea for TypeScript

   on Apr 8, 2015
2. DanielRosenwasser commented on Apr 8, 2015

   

   Daniel Rosenwasser (DanielRosenwasser)

   on Apr 8, 2015

   Member

   More actions

   If I took your example and later added x as in the following:

   var [x, a, b]: [, string, number] = [0, "a", 5];

   What is the desired behavior?
3. zhuravlikjb commented on Apr 8, 2015

   

   Anton Lobov (zhuravlikjb)

   on Apr 8, 2015

   Author

   More actions

   I suppose that x here should have type of right-hand-side[0], or 'any', if the right-hand-side is absent.

   Exactly the same way as for ordinary type annotations, when they are omitted:
   var x = 0 => x: number
   var x => x: any

   The same in the case of non-destructuring declarations:

   var x: [, string, number] = [0, "a", 5];     // x is [number, string, number]
   var y: [, string, number];  // y is [any, string, number]
4. Mohamed Hegazy (mhegazy)

   added

   In DiscussionNot yet reached consensusNot yet reached consensus

   on Dec 10, 2015
5. RyanCavanaugh commented on Apr 20, 2016

   

   Ryan Cavanaugh (RyanCavanaugh)

   on Apr 20, 2016

   Member

   More actions

   We had a question when discussing this issue -- what are the kinds of cases where the inferred type is "not correct" ? In other words, what's the value of writing these annotations vs letting them be inferred?
6. RyanCavanaugh commented on Jun 13, 2016

   

   Ryan Cavanaugh (RyanCavanaugh)

   on Jun 13, 2016

   Member

   More actions

   Anton Lobov (@zhuravlikjb) any thoughts on the above question?
7. zhuravlikjb commented on Jul 4, 2016

   

   Anton Lobov (zhuravlikjb)

   on Jul 4, 2016

   Author

   More actions

   Sorry for the delay with answer. This question is just about consistency: JavaScript allows empty items and elision in many syntax elements. If it's allowed for array literals, why not to allow it for type annotations? Nothing critical here if the existing behaviour will remain. Just a "why not" thing. :)
8. Ryan Cavanaugh (RyanCavanaugh)

   added

   DeclinedThe issue was declined as something which matches the TypeScript visionThe issue was declined as something which matches the TypeScript vision

   and removed

   In DiscussionNot yet reached consensusNot yet reached consensus

   on Sep 28, 2016
9. RyanCavanaugh commented on Sep 28, 2016

   

   Ryan Cavanaugh (RyanCavanaugh)

   on Sep 28, 2016

   Member

   More actions

   Doesn't seem to be a lot of desire for this in practice, plus it introduces new weird cases as Daniel pointed out.
10. Ryan Cavanaugh (RyanCavanaugh)

    closed this as completedon Sep 28, 2016
11. microsoft

    locked and limited conversation to collaborators on Jun 18, 2018