[
https://issues.apache.org/jira/browse/GROOVY-12561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18135267#comment-18135267
]
ASF GitHub Bot commented on GROOVY-12561:
-----------------------------------------
codeconsole commented on PR #3090:
URL: https://github.com/apache/groovy/pull/3090#issuecomment-6526951328
@daniellansun
1. Either order. The only textual conflict with #3089 was that both PRs
appended a test at the end of `ClosuresStaticCompileTest`; this PR now adds its
test earlier in the file (e608a44e04), and `git merge-tree` merges the two
branches cleanly both ways, to the same tree. Neither depends on the other's
tests: on a merge of both, the `groovy.transform.stc` and
`org.codehaus.groovy.classgen.asm.sc` packages pass. The STC metadata does not
overlap: this PR records PARAMETER_TYPE on a named constructor argument's key,
#3089 on call arguments. A closure assigned to a property becomes a cast here
before the setter call is written, so #3089's writer sees an argument of the
interface type and does not convert it again.
2. Intentional. A Closure already is a `Runnable`, a `Callable` and a
`GroovyCallable`, and `DefaultTypeTransformation.castToType` returns an object
unchanged when it is an instance of the target type, so a cast would leave the
Closure anyway; the check only skips a cast that does nothing. That is also
what an assignment to a `Callable` variable and dynamic Groovy do.
`ClosuresSTCTest.testClosureAssignedToACallableOrRunnableSubscriptOrPropertyStaysAClosure`
assigns closures to a `Callable` subscript, property and named constructor
argument and a `Runnable` property, and checks each stays a Closure and works,
under `@TypeChecked` and `@CompileStatic` (8b73244656). Added a sentence to the
description.
3. Added
`ClosuresSTCTest.testValueAssignedToAFunctionalSubscriptOrPropertyIsStoredAsItIsIfNotAClosure`:
a `Supplier` variable, a method returning a `Supplier`, and an Elvis and a
conditional whose branches are such values, assigned to a subscript and a
property of type `Supplier`. Each stored value is the same object as the
right-hand side, under `@TypeChecked` and `@CompileStatic` alike (8b73244656).
Also merged master. The test changes are the only changes in this round; the
Checkstyle and CodeNarc gates pass.
> SC: a closure assigned to a subscript or property of a functional interface
> type is stored unconverted
> ------------------------------------------------------------------------------------------------------
>
> Key: GROOVY-12561
> URL: https://issues.apache.org/jira/browse/GROOVY-12561
> Project: Groovy
> Issue Type: Bug
> Components: Static compilation
> Affects Versions: 4.0.33, 5.0.8, 6.0.0
> Reporter: Scott Murphy Heiberg
> Priority: Major
>
> Under {{@CompileStatic}}, a closure assigned to a subscript or a property
> whose type is a functional interface is stored as the closure itself when the
> method that stores it declares its parameter with a type parameter: {{putAt}}
> of a {{Map<String, Function<String, String>>}} or {{List<Supplier<String>>}},
> or the setter or field of a {{Holder<T>}} property for a
> {{Holder<Supplier<String>>}}. A later call of the interface method fails:
> {code:groovy}
> import java.util.function.Function
> @groovy.transform.CompileStatic
> Map<String, Function<String, String>> make() {
> Map<String, Function<String, String>> map = [:]
> map['shout'] = { String s -> s + '!' }
> map
> }
> make().shout.apply('hi') // MissingMethodException: No signature of method:
> apply ... (a Closure was stored)
> {code}
> The same holds for {{map?['k'] = ...}}, {{map.k = ...}}, {{holder.value =
> ...}}, {{box.@field = ...}} and the named constructor argument {{new
> Holder<Supplier<String>>(value: { ... })}}. Assigning to a variable of the
> interface type converts the closure.
> Expected: the closure is converted to the type of the subscript or property,
> as for a variable, and the value of the assignment is the converted closure.
> PR: https://github.com/apache/groovy/pull/3090
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
