Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Refine type for opaque (remain user_type) #276

Merged
merged 1 commit into from
Feb 10, 2020
Merged

Refine type for opaque (remain user_type) #276

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions src/typechecker.erl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3268,6 +3268,10 @@ refine_ty({Tag1, _, M}, {Tag2, _, N}, _TEnv)
refine_ty(Ty1, Ty2, _TEnv) when ?is_int_type(Ty1),
?is_int_type(Ty2) ->
gradualizer_int:int_type_diff(Ty1, Ty2);
refine_ty({user_type, Anno, Name, Args}, {user_type, Anno, Name, Args}, _TEnv) ->
% After being normalized, it's because it's defined as opaque.
% If it has the same annotation, name and args, it's the same.
type(none);
Comment on lines +3271 to +3274
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

There is a clause like this above:

refine_ty(?type(T, A), ?type(T, A), _) ->
    type(none);

Perhaps both these could be replaced by simply one clause like this:

refine_ty(Same, Same, _TEnv) ->
    %% If it's the same, it's the same
    type(none);

What do you think? Future improvement?

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The only thing is this would also catch a few other cases I haven't put thoughts into. From looking at gradualizer_type.erl, The other stuff that would pass:

  • ann_type: Are these supposed to be removed at that point? I think I saw a PR go by mentioning this but haven't looked into it enough.
  • var: Honestly have no clue, I have barely touched type variables at this point.
  • remote_type: Pretty sure normalize catches this and either expand it in a type, or a user_type if it's opaque.
  • user_type: D'uh. After normalization, the only way it's still a user_type is an opaque type from another module. Seems fine and would remove the case I wrote.

I created my other PR #277 but not sure I wanna include them. I merged this because I'm using our existant codebase to test my record refinement branch and I needed this change 😄

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

OK!

Anything minus itself would be none() I guess, even a type variable. (Then, perhaps we also may want to add other cases such as SomeType \ TypeVar and vice versa and add constraints.)

refine_ty(Ty1, Ty2, TEnv) ->
case glb(Ty1, Ty2, TEnv) of
{?type(none), _} -> throw(disjoint); %% disjoint
Expand Down
13 changes: 13 additions & 0 deletions test/should_pass/opaque.erl
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
-module(opaque).

-export([external/2, internal/2]).

-spec external(user_types:my_opaque(), integer() | undefined) -> integer().
external(_, undefined) -> 0;
external(_, I) -> I.

-opaque my_opaque() :: integer().

-spec internal(my_opaque(), integer() | undefined) -> integer().
internal(_, undefined) -> 0;
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'd expect a test that the type is still visible from inside the module, i.e. that the first parameter is used in this function.

internal(_, I) -> I.