SAK-50696 - LTI Fix Non-Deterministic Unit Tests

[hermya]

Fixed flaky test cases in the module: ./lti/tsugi-util/

Test-cases details

Sr	Test-case	Point of failure	Fixed by
1	org.tsugi.lti13.GroupServiceTest# testConstructor	The bean of GroupService is converted to Json, and its string representation is asserted against expected value. JSON conversion is non-deterministic, and the order of attributes can be different on different jvms, resulting in different string representations	Converting the Json string to a Map and Asserting Map-equality, as it doesn't require matching key-ordering
2	org.tsugi.lti13.LTI13KeySetTest# testKeySets	In class org.tsugi.lti13.LTI13KeySetUtil at Line:60, rsaKey is being converted to a json-string, used further in a JsonObject. As rsaKey is a bean, its JSON conversion can result in non-deterministic string Same reason as 1	Converting the Json string to a Map and Asserting Map-equality, and a flexible contains check for asserting goodness of keySetJSON
3	org.tsugi.lti13.LTI13NimbusTest# testRSAFromString, testRSAPaulGray	Same reason as 2 for both test-cases	Adding a flexible contains check for asserting goodness of keySetJSON
4	org.tsugi.lti13.LTI13ObjectTes# testTwo, testThree	Same reason as 1 for both test-cases	Same fix as 1 for both test-cases

Fixed using NonDex

Suggested command to check all flaky tests :

mvn edu.illinois:nondex-maven-plugin:2.1.7:nondex

For the particular test class:

mvn -pl ./lti/tsugi-util edu.illinois:nondex-maven-plugin:2.1.7:nondex \
-DnondexMode=ONE \
-DnondexRuns=5  \
-Dtest=org.tsugi.lti13.GroupServiceTest#testConstructor

For more information: https://github.com/TestingResearchIllinois/NonDex

[csev]

@hermya Thanks for the patch - it looks good to me. We need to file JIRA in our issue tracker and have you fill out a Contribution Agreement. If your intent is to keep working on Sakai it would be worthwhile to go through all this effort - but if this is a one-time patch, I can take this fix over for you and do all the necessary stuff to get this into the code base. Sorry about the paperwork - open source projects need to do this.

[hermya]

Hello @csev,
I'll be happy to fill out the CLA and continue to contribute in future too! Can you please suggest how to do that? I tried logging in on Jira but it shows that I don't have access to create a ticket

[ottenhoff]

I believe there is a form linked from this page to create an individual CLA: https://form.jotform.com/60435795225156

[hermya]

Found it and submitted, thanks @ottenhoff!

[csev]

@ottenhoff - The technical bits look good to me. Once this is merged, I want to apply the approach to some new tests I just wrote this weekend :)

[hermya]

Hello All!
I just recently obtained the CLA, is there any other step I must perform?

[csev]

I think the one last step it so make a JIRA - @ottenhoff - WDYT?

[ottenhoff]

Agree and let the tests finish running!

[hermya]

Hello!
I've been trying to log in to sakai's jira using my email address that I used to create the Apereo license. Even after a couple days , I still can't get access to create a SAK-* ticket. I've mailed a person who responded regarding cla acceptance, but they haven't responded yet. Should I wait for longer (If I am supposed to create the ticket)? Please let me know if I am missing some steps

[csev]

I made a JIRA and edited the title and the CLA is applied for - so I am merging this. Thanks.

[ottenhoff]

Please try to create a new account at sakaiproject.atlassian.net, login via Google, or send your email address to ottenhoff at longsight dot com