This is going to be the smallest PR to ever be posed on this Repo, it changes a whopping one character 😄
Nevertheless, there was an actual bug in def staticInitializer: Option[Method], the piece of code that retrieves the static initializer for a ClassFile. Due to this bug, the code would only return the correct static initializer method if it was the first method in the Sequence of methods - this is not always the case, especially for Scala-Compiled JVM code, where many method names begin with "$", which is lexicographically smaller than <clinit>.
As a result, the analysis that normally marks static initializers as reachable during call graph construction missed some initializers - i confirmed that this issue is now fixed when applying this PR.
This is going to be the smallest PR to ever be posed on this Repo, it changes a whopping one character 😄
Nevertheless, there was an actual bug in
def staticInitializer: Option[Method], the piece of code that retrieves the static initializer for aClassFile. Due to this bug, the code would only return the correct static initializer method if it was the first method in the Sequence of methods - this is not always the case, especially for Scala-Compiled JVM code, where many method names begin with "$", which is lexicographically smaller than<clinit>.As a result, the analysis that normally marks static initializers as reachable during call graph construction missed some initializers - i confirmed that this issue is now fixed when applying this PR.