`UNREACHABLE_CODE` is weird.
If we look into its
positioning strategy,
we'll see that this diagnostic
attempts to split the source
of its element into multiple ranges.
This is bad, because such smaller
ranges can easily overlap with other
diagnostics in such a way that
they are no longer hierarchical.
See the
`compiler/tests-spec/testData/diagnostics/notLinked/dfa/neg/3.kt` test
for an example of this behavior
(allow reporting `UNREACHABLE_CODE`
first). The problem will appear at
the condition of `case_7`.
New tests `testHmppLibAndConsumer` and `testMultiModulesHmpp` were added
The first is a test with MPP library and its consumers, it checks import
and also highlighting before and after the library was published
The second is a project which contains three multiplatform modules and
two platform modules and checks import and highlighting.
Original commit: https://jetbrains.team/p/ij/repositories/intellij/revision/15582b03530eba4fa0e3d0b512bb0a4c1fa3aa4c
Previously CodeMetaInfoParser saved MatchResults for each meta info in
text (and for each meta info it was different string with text because
of extracting parsed part from original text), so it led to O(n^2)
memory complexity (where n is a size of original text).
Such complexity caused OOME in case of big testdata (~1k meta infos
for text of 4k lines)
Nikolay Lunyak