Case suggested by exhaustivity checker is wrong #7504
Comments
|
Comment author: @Octachron Good catch, this case is indeed mishandled; a fix is incoming in #1106. |
|
Comment author: @gasche I have memories of working on a very similar issue in the recent months -- maybe of having already changed the formulation in 4.05+dev. (Is this just jetlag?) |
|
Comment author: @Octachron @gasche, maybe some remininescences of #501 |
|
Comment author: @gasche Yes, exactly, thanks. We should have noticed the current issue if we had been more careful thinking about the corner cases of the last patch. (I think one could argue that "{ _ }" should be accepted by the language when no field is matched upon, and that using it there could make the code more regular, but your patch is just fine.) |
|
Comment author: @Octachron Would it not be unpratical to type |
|
Comment author: @gasche Solved by merging octachron's proposed fix. Thanks for the patch, and thanks for the report. |
Original bug ID: 7504
Reporter: @johnwhitington
Assigned to: @Octachron
Status: resolved (set by @gasche on 2017-03-14T23:16:34Z)
Resolution: fixed
Priority: normal
Severity: tweak
Version: 4.04.0
Target version: 4.05.0 +dev/beta1/beta2/beta3/rc1
Fixed in version: 4.05.0 +dev/beta1/beta2/beta3/rc1
Category: language features
Related to: #7169
Bug description
Consider:
type t =
{a : int;
b : string}
let f x y =
match x with {a = c}, 0 -> 0
Which gives:
Warning 8: this pattern-matching is not exhaustive.
Here is an example of a case that is not matched:
({}, 1)
{} isn't legal pattern match syntax. Shouldn't it read "(_, 1)"?
The text was updated successfully, but these errors were encountered: