Kill survivng mutation in DefaultComesLastCheck.java

[ThatSneakyCoder]

Kill mutation:

checkstyle/config/pitest-suppressions/pitest-coding-1-suppressions.xml

Lines 30 to 37 in 54df97c

	<mutation unstable="false">
	<sourceFile>DefaultComesLastCheck.java</sourceFile>
	<mutatedClass>com.puppycrawl.tools.checkstyle.checks.coding.DefaultComesLastCheck</mutatedClass>
	<mutatedMethod>findNextSibling</mutatedMethod>
	<mutator>org.pitest.mutationtest.engine.gregor.mutators.NonVoidMethodCallMutator</mutator>
	<description>removed call to com/puppycrawl/tools/checkstyle/api/DetailAST::getNextSibling</description>
	<lineContent>node = node.getNextSibling();</lineContent>
	</mutation>

[ThatSneakyCoder]

@romani
please approve

[ThatSneakyCoder]

I'm on it.

This issue is open for other interested candidates to work on as I am busy with other important PRs.

[jxr98]

I will work on this issue.

[jxr98]

checkstyle/src/main/java/com/puppycrawl/tools/checkstyle/checks/coding/DefaultComesLastCheck.java

Lines 215 to 228 in 2e99aca

	private static DetailAST findNextSibling(DetailAST ast, int tokenType) {
	DetailAST token = null;
	DetailAST node = ast.getNextSibling();
	while (node != null) {
	if (node.getType() == tokenType) {
	token = node;
	break;
	}
	node = node.getNextSibling();
	}
	return token;
	}
	}

The pitest will fail after timeout or out of memory if the code((Line 223) is changed to node = node;. I think this suppressed mutation cannot be removed because it is kind of the halting problem to add code to check if the loop is infinite or not. Additionally, can't kill this mutation without redundant code. I don't think it is worth to add one more condition to while only for one mutation test.

I guess this issue can be closed.

[romani]

Pitest catching infinite loops and out of memory, this tool is well aware of damages it might do.
We need to kill mutation.

[nrmancuso]

the code((Line 223) is changed to node = node;.

@jxr98 please explain how you concluded that this is what the code is mutated to. According to the documentation for this mutator, it should be mutated to node = null; however this is an unstable mutator so we could be getting different results. If the code is mutated to node = null, then it looks like all the tests we have only cause us to loop one time in the code snippet you shared. Try to find a case where we must loop more than once.

[jxr98]

I thought this mutator just removed getNextSibling. Thanks for pointing this out.

[rnveach]

Fix was merged