[java] Fix issue #1736 and issue #2207

[YYoungC]

Describe the PR

Related issues

• Fixes [java] UseStringBufferForStringAppends: False positive if only one concatenation #1736
• Fixes [java] AvoidInstantiatingObjectsInLoops: False positive - should not flag objects when assigned to lists/arrays #2207

Ready?

• Complete build ./mvnw clean verify passes (checked automatically by travis)
• Added (in-code) documentation (if needed)

[pmd-test]

	1 Message
📖	This changeset introduces 85 new violations, 0 new errors and 0 new configuration errors, removes 241 violations, 2 errors and 2 configuration errors. Full report
✅	This changeset introduces 85 new violations, 0 new errors and 0 new configuration errors, removes 243 violations, 2 errors and 2 configuration errors. Full report
✅	This changeset introduces 0 new violations, 0 new errors and 0 new configuration errors, removes 819 violations, 2 errors and 2 configuration errors. Full report

Generated by 🚫 Danger

[adangel]

Hi @YYoungC , thanks for your pull request.
I'm afraid, I can't accept it as is.

Please have a look at my comments and create a unit test cases to proof, that we actually fixed the issues, we wanted to fix. Adding a unit test is easy:

• General documentation about rule testing: https://pmd.github.io/latest/pmd_userdocs_extending_testing.html
• A new test case needs to be added to either https://github.com/pmd/pmd/blob/master/pmd-java/src/test/resources/net/sourceforge/pmd/lang/java/rule/performance/xml/UseStringBufferForStringAppends.xml and https://github.com/pmd/pmd/blob/master/pmd-java/src/test/resources/net/sourceforge/pmd/lang/java/rule/performance/xml/AvoidInstantiatingObjectsInLoops.xml
• For the sample test code, look at the sample of [java] UseStringBufferForStringAppends: False positive if only one concatenation #1736 and [java] AvoidInstantiatingObjectsInLoops: False positive - should not flag objects when assigned to lists/arrays #2207

[adangel]

There are still many new violations, but these are ok - they have been actually false-negative, as this rule was not using rule chain previously but still short-circuited the visitor by returning data. Now all AllocationExpressions are visisted, even nested ones in a anonymous class allocation expression, when the anonymous class is not within a loop.

I'll merge that now.