When registering Ident use, consider the original span for synthetics

[natsukagami]

Fix #19970

• Consider the original span for synthetics for registering Ident use
• Clarify that tryFindSynthetic is for Apply | TypeApply tree
• Update tests

[rochala]

I looked at this, and I think it didn't solve the issue I've had, but the reason is completely on me. I made a mistake when I reported it. Likewise, I have updated the comment on the issue with the real reproduction. The fix should be similar, though.

[rochala]

I think that the solution has to be a bit different.

Right now we're logging the occurence of local1, but we should instead go to the RHS of the ValDef from lifting. Then we should go over all calls and register their usage (they have correct spans).

Example:

object T:
  def enrichedCompilerCompletions() =
    val compilerCompletions: Seq[String] = ???

    compilerCompletions
      .toList
      .flatMap(toCompletionValues)
      .filterInteresting()


  def toCompletionValues(x: String): List[String] = ???

  extension (l: List[String])
    def filterInteresting(qualType: String = ""): List[String] = ???

results in:

Occurrences:
[3:7..3:8) <= _empty_/T.
[4:6..4:33) <= _empty_/T.enrichedCompilerCompletions().
[5:8..5:27) <= local0
[5:29..5:32) => scala/package.Seq#
[5:33..5:39) => scala/Predef.String#
[5:43..5:46) => scala/Predef.`???`().
[7:4..9:34) => local1
[10:7..10:24) => _empty_/T.filterInteresting().
[13:6..13:24) <= _empty_/T.toCompletionValues().
[13:25..13:26) <= _empty_/T.toCompletionValues().(x)
[13:28..13:34) => scala/Predef.String#
[13:37..13:41) => scala/package.List#
[13:42..13:48) => scala/Predef.String#
[13:52..13:55) => scala/Predef.`???`().
[15:13..15:14) <= _empty_/T.filterInteresting$default$2().(l)
[15:13..15:14) <= _empty_/T.filterInteresting().(l)
[15:16..15:20) => scala/package.List#
[15:21..15:27) => scala/Predef.String#
[16:8..16:25) <= _empty_/T.filterInteresting().
[16:26..16:34) <= _empty_/T.filterInteresting().(qualType)
[16:36..16:42) => scala/Predef.String#
[16:50..16:54) => scala/package.List#
[16:55..16:61) => scala/Predef.String#
[16:65..16:68) => scala/Predef.`???`().

But it should rather be:

Occurrences:
[3:7..3:8) <= _empty_/T.
[4:6..4:33) <= _empty_/T.enrichedCompilerCompletions().
[5:8..5:27) <= local0
[5:29..5:32) => scala/package.Seq#
[5:33..5:39) => scala/Predef.String#
[5:43..5:46) => scala/Predef.`???`().
[7:4..7:22) => local0
[8:6..8:12] => .toList
[9:6..9:xx] => .flatMap
[9:xx..9:yy] => toCompletionValue
[10:7..10:24) => _empty_/T.filterInteresting().
[13:6..13:24) <= _empty_/T.toCompletionValues().
[13:25..13:26) <= _empty_/T.toCompletionValues().(x)
[13:28..13:34) => scala/Predef.String#
[13:37..13:41) => scala/package.List#
[13:42..13:48) => scala/Predef.String#
[13:52..13:55) => scala/Predef.`???`().
[15:13..15:14) <= _empty_/T.filterInteresting$default$2().(l)
[15:13..15:14) <= _empty_/T.filterInteresting().(l)
[15:16..15:20) => scala/package.List#
[15:21..15:27) => scala/Predef.String#
[16:8..16:25) <= _empty_/T.filterInteresting().
[16:26..16:34) <= _empty_/T.filterInteresting().(qualType)
[16:36..16:42) => scala/Predef.String#
[16:50..16:54) => scala/package.List#
[16:55..16:61) => scala/Predef.String#
[16:65..16:68) => scala/Predef.`???`().

[tgodzik]

Suggested change

	def recursive/*<-example::NamedApplyBlockMethods.recursive().*/ = foo/*->example::NamedApplyBlockMethods.foo().*//*->local1*/(local/*->example::NamedApplyBlockMethods.local.*/, c/*->example::NamedApplyBlockMethods.foo().(c)*/ = foo/*->example::NamedApplyBlockMethods.foo().*/(local/*->example::NamedApplyBlockMethods.local.*/, c/*->example::NamedApplyBlockMethods.foo().(c)*/ = 3))
	def recursive/*<-example::NamedApplyBlockMethods.recursive().*/ = foo/*->example::NamedApplyBlockMethods.foo().*/(local/*->example::NamedApplyBlockMethods.local.*/, c/*->example::NamedApplyBlockMethods.foo().(c)*/ = foo/*->example::NamedApplyBlockMethods.foo().*/(local/*->example::NamedApplyBlockMethods.local.*/, c/*->example::NamedApplyBlockMethods.foo().(c)*/ = 3))

That local doesn't exist in the source code so we should not add it to occurrences

[natsukagami]

Superceded by #21856