Java: Add CompilationUnit.getATypeInScope()
#10498
Conversation
This predicate is mainly helpful for Javadoc queries and for queries which check whether the name of an element shadows another type.
Marcono1234
force-pushed
the
marcono1234/compilation-unit-simple-name-type
branch
from
ef7046a to
431aa2c
Compare
| or | ||
| exists(ImportStaticTypeMember importDecl | | ||
| importDecl.getCompilationUnit() = this and | ||
| result = importDecl.getATypeImport() | ||
| ) | ||
| or | ||
| exists(ImportType importDecl | | ||
| importDecl.getCompilationUnit() = this and | ||
| result = importDecl.getImportedType() | ||
| ) | ||
| or | ||
| exists(ImportStaticOnDemand importDecl | | ||
| importDecl.getCompilationUnit() = this and | ||
| result = importDecl.getATypeImport() | ||
| ) | ||
| or | ||
| exists(ImportOnDemandFromType importDecl | | ||
| importDecl.getCompilationUnit() = this and | ||
| result = importDecl.getAnImport() | ||
| ) | ||
| or | ||
| exists(ImportOnDemandFromPackage importDecl | | ||
| importDecl.getCompilationUnit() = this and | ||
| result = importDecl.getAnImport() | ||
| ) | ||
| or | ||
| // From same package or java.lang, see https://docs.oracle.com/javase/specs/jls/se17/html/jls-7.html |
There was a problem hiding this comment.
Add CompilationUnit.getAnImport() to make this much briefer?
There was a problem hiding this comment.
Do you mean as separate predicate: getAnImport(): Import?
There was a problem hiding this comment.
yes, so you don't need all the exists quantifiers but instead can do this.getAnImport().(ImportStaticOnDemand).getAnImport() for example
There was a problem hiding this comment.
I have simplified the code but decided against a separate predicate getAnImport() for know because I am not sure if it would be useful in any other situations.
I hope that is fine.
| * - The type is declared in the same package as this compilation unit | ||
| * - The type is declared in the package `java.lang` | ||
| */ | ||
| ClassOrInterface getATypeAvailableBySimpleName() { |
There was a problem hiding this comment.
Maybe AvailableBySimpleName -> InScope?
There was a problem hiding this comment.
In that case should the shadowing and simple name aspects be removed from the comments? Because a shadowed type would still be in scope (I think).
There was a problem hiding this comment.
I don't know off hand if the JLS formally defines "in scope"? In either case though explicitly mentioning this tricky corner seems useful, so I'd keep it.
| ClassOrInterface getATypeInScope() { | ||
| // See "Shadowing", https://docs.oracle.com/javase/specs/jls/se17/html/jls-6.html#jls-6.4.1 | ||
| // Currently shadowing is not considered | ||
| result.(TopLevelType).getCompilationUnit() = this |
There was a problem hiding this comment.
Redundant w.r.t. the same-package test
There was a problem hiding this comment.
You are right. Currently this structure matches (roughly?) the precedence order of shadowing, but I could drop this nonetheless if you want.
| | | ||
| exception.getAnAncestor().hasQualifiedName("java.lang", uncheckedException) | ||
| ) | ||
| predicate canThrow(Callable callable, Class exception) { |
There was a problem hiding this comment.
Same question re: throwing things by an interface type
There was a problem hiding this comment.
All exception types are classes though, so you cannot declare an interface type to be thrown, can you?
(neither in Kotlin, right?)
There was a problem hiding this comment.
Doh right, I had Exception in mind as an interface. Never mind.
Marcono1234
changed the title
CompilationUnit.getATypeAvailableBySimpleName()CompilationUnit.getATypeInScope()
Marcono1234
deleted the
marcono1234/compilation-unit-simple-name-type
branch
This predicate is mainly helpful for Javadoc queries and for queries which check whether the name of an element shadows another type.
Resolves #4119
It makes probably most sense to implement the logic requested there for
CompilationUnit, as done in this PR because there it is also possible to account for shadowing between different kinds of imports (though this is currently not implemented by this PR).This PR is a bit experimental and does not have a dedicated test; feedback is appreciated.