Improve remove_duplicate_tests.py

[llxia]

With #393, we have remove_duplicate_tests.py to remove duplicate tests. However, there are some enhancements that we can make:

• find a way to persevere comments (whether the license or actual comments within the root element). For example https://stackoverflow.com/questions/33573807/faithfully-preserve-comments-in-parsed-xml. The license can be different from repo to repo. Users should not need to provide --license-file. Also, actual comments in the playlist.xml should be preserved.
• We should compare the content instead of simply removing elements with the duplicated <testCaseName>. Only remove it if everything under <test> is the same. Otherwise, we should print a warning message (i.e., warning: found duplicated testCaseName xxx, but the content is not identical.)
• Update remove_duplicate_tests.py to take a git repo. remove_duplicate_tests.py should git clone the repo and check all subdirectories to find playlist.xml and remove duplicate tests if possible.
• create a GitHub workflow for this task. On a schedule (i.e., biweekly), run remove_duplicate_tests.py against all repos and create auto PR against specific repo if needed.
  https://github.com/adoptium/aqa-tests
  https://github.com/eclipse-openj9/openj9
  https://github.com/adoptium/temurin-build

[ArkaprabhaChakraborty]

Hi, can I get this assigned?

[ArkaprabhaChakraborty]

PS we might have to make several PRs to this completely issue ina good trackable way :)

[llxia]

@ArkaprabhaChakraborty yes, that is correct. This is a top-level issue to track the enhancements. Please divide the deliverables into PRs that are of a reasonable size and contain specific features. Thanks

[ArkaprabhaChakraborty]

Where can I add tests for the script? :) More of like unit tests you can say :)

[ArkaprabhaChakraborty]

Okay, I tried different ways to parse comments outside the "tree" but none worked with python's default XMLParser... StackOverflow discussion

A seemingly simple problem just got weirder :)