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

Preserve empty variant constructor in data-deps. #7303

Merged
merged 3 commits into from
Sep 2, 2020
Merged

Preserve empty variant constructor in data-deps. #7303

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
31a694c
Preserve empty variant constructor in data-deps.
sofiafaro-da Sep 2, 2020
1cca2ae
Add special case for single constructor variants
sofiafaro-da Sep 2, 2020
59926ba
Add test for special case
sofiafaro-da Sep 2, 2020
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
3 changes: 3 additions & 0 deletions compiler/damlc/daml-compiler/src/DA/Daml/Compiler/DataDependencies.hs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -479,6 +479,9 @@ generateSrcFromLf env = noLoc mod

convConDetails :: LF.Type -> Gen (HsConDeclDetails GhcPs)
convConDetails = \case
-- empty variant constructor (see issue #7207)
Copy link
Contributor

Choose a reason for hiding this comment

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

"Empty" is such an overloaded word:

Suggested change
-- empty variant constructor (see issue #7207)
-- variant constructor without argument (see issue #7207)

Or "payload" instead of "argument".

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Right, but "variant constructor without argument" is such a clumsy phrase, and what else could "empty variant constructor" mean?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

I guess the right answer here is "nullary variant constructor".

Copy link
Contributor

Choose a reason for hiding this comment

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

I like "nullary variant constructor". 👍

LF.TUnit ->
Copy link
Contributor

Choose a reason for hiding this comment

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

This means that a variant where I explicilty specify a Unit argument will also be decompiled to a Prefix con right? I think that’s perfectly reasonable but probably worth documenting in a comment and perhaps warning on in the preprocessor.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Yes, that's what we decided on in the issue and I'm preparing the warning in a separate PR.

pure $ PrefixCon []

-- variant record constructor
LF.TConApp LF.Qualified{..} _
Expand Down
43 changes: 43 additions & 0 deletions compiler/damlc/tests/src/DA/Test/Packaging.hs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1667,6 +1667,49 @@ dataDependencyTests Tools{damlc,repl,validate,davlDar,oldProjDar} = testGroup "D
withCurrentDirectory (tmpDir </> "proj") $
callProcessSilent damlc ["build"]

, testCaseSteps "Empty variant constructors" $ \step -> withTempDir $ \tmpDir -> do
-- This test checks that variant constructors without argument
-- are preserved. This is a regression test for issue #7207.
step "building project with type definition"
createDirectoryIfMissing True (tmpDir </> "type")
writeFileUTF8 (tmpDir </> "type" </> "daml.yaml") $ unlines
[ "sdk-version: " <> sdkVersion
, "name: type"
, "source: ."
, "version: 0.1.0"
, "dependencies: [daml-prim, daml-stdlib]"
]
writeFileUTF8 (tmpDir </> "type" </> "Foo.daml") $ unlines
[ "module Foo where"
, "data A = B | C Int"
Copy link
Contributor

Choose a reason for hiding this comment

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

Should we add a third case D () and check that it behaves like D when imported as data-dependency?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

I don't think so. D () is something we're deprecating, and the fact that it's treated the same as D in the current compiler is mostly a mistake.

Copy link
Contributor

Choose a reason for hiding this comment

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

I generally agree but would feel more comfortable to have a test as long as users can still write it, even though they get warned about it.

Copy link
Contributor Author

@sofiafaro-da sofiafaro-da Sep 2, 2020 •
edited
Loading

Choose a reason for hiding this comment

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

I added a test for D () in the single constructor case, which is something we had talked about adding as a special case (since we can't distinguish single-constructor variants from enums any other way, and there's no ambiguity in the single constructor case). (This actually also came up in one of the other tests, but better to add one here as well.)

]
withCurrentDirectory (tmpDir </> "type") $
callProcessSilent damlc ["build", "-o", "type.dar"]

step "building project that uses it via data-dependencies"
createDirectoryIfMissing True (tmpDir </> "proj")
writeFileUTF8 (tmpDir </> "proj" </> "daml.yaml") $ unlines
[ "sdk-version: " <> sdkVersion
, "name: proj"
, "source: ."
, "version: 0.1.0"
, "dependencies: [daml-prim, daml-stdlib]"
, "data-dependencies: "
, " - " <> (tmpDir </> "type" </> "type.dar")
]
writeFileUTF8 (tmpDir </> "proj" </> "Main.daml") $ unlines
[ "module Main where"
, "import Foo"
, "mkA : A"
, "mkA = B"
, "matchA : A -> Int"
, "matchA a ="
, " case a of"
, " B -> 0"
, " C n -> n"
]
withCurrentDirectory (tmpDir </> "proj") $
callProcessSilent damlc ["build"]
]

-- | Check that the given file exists in the dar in the given directory.
Expand Down