Skip to content

Merge release/6.4.0 into master #1621

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 33 commits into from
Nov 14, 2023
Merged

Merge release/6.4.0 into master #1621

merged 33 commits into from
Nov 14, 2023

Conversation

rjrudin
Copy link
Contributor

@rjrudin rjrudin commented Nov 14, 2023

Contains all changes since 6.3.0.

rjrudin and others added 30 commits September 11, 2023 08:59
This is failing on the assertion for all the docs being deleted in the `finally` block. Removing that for now to see if the test itself actually fails.
I was looking at this while analyzing support for 2-way SSL and I couldn't really understand what I wrote. So I refactored in the following way:

1. All construction for SSLContext/TrustManager is in buildSSLInputs.
2. There are 4 approaches, each clearly identified and implemented in its own method.
3. I moved the tests I use for verifying SSL support to a "test.ssl" package (no changes to the tests themselves).

This may actually fix a bug but I'm not certain. The bug would have been that "newCertificateAuthContext" was called before all the SSL-input logic had occurred, meaning that e.g. "default" as an sslProtocol value would not have impacted certificate authentication.
Refactored logic for determining SSL inputs
This catches the bug that I thought might have existed previous to the prior PR - it ensures that a user-defined TrustManager is used when an SSLContext is created based on a user-defined path for a PKCS12 file.
…-test

Added test for certificate auth with file and password
Removed everything that was clearly incorrect or out-of-date. Hoping to get Palamida to scan this correctly.
See the comment for why it's deprecated; it presumably should never have been added in the first place as the `directory` method that receives a boolean is sufficient.
#1603 Deprecating the `directory` method that receives a depth value
Wanted to make the distinction between "OneWay" and "TwoWay" SSL tests.
Renamed some things in SSL tests to improve clarity
This is all the code-generated stuff for DEVEXP-563. Will use a second PR for the manually-added stuff. 

Most of this is just copyright updates that were made by the code generator. 

Note that the code generator also updated PlanGeneratorTest. We have some manual mods in there so that some tests only run on MarkLogic 11, so I had to ensure those were retained.
Bumped the version to 6.4-SNAPSHOT as well so we can start using it in the Mule connector. 

The main value here is in TwoWaySSLTest, which can use these new properties instead of having to construct an SSLContext itself.
DEVEXP-563 Added geohashSubhashes to Optic plan builder
Was able to code-generate for the document format/root/permission queries, but DBQ-643 in the server prevents them from working. 

joinDocAndUri appears to be good-to-go.

I had to manually add the string constructors for cts.point and cts.polygon as the code generator does not yet generate those. Added comments to capture that fact. 

Finally, renamed the two XML TDEs that were using .tdex as an extension - .xml is preferred.
DEVEXP-563 Added support for other 11.1 Optic methods
DEVEXP-585 Can now configure 2-way SSL
Documenting that setMaxBatches works with an iterator, but not with a query.
Copyright updates, removed a dep.txt file that I didn't mean to add, and further cleanup of a test in QueryBatcherJobReportTest.
Small cleanup before 6.4.0 release
No idea yet why a couple tests in TwoWaySSLTest started failing - the error message is null instead of being an expected . And UpdateUseCasesTest runs fine locally but seems to be pulling in extra zipcodes in Jenkins.
Reverting the schemas from xml to tdex to see if that makes a difference in the Jenkins environment.
Reloading schemas to ensure that the previous name of schemas is not in the schemas database.
The test ran fine in a suite, but got an NPE when run by itself since it referenced `Common.client` without instantiating it.
Updated Jenkinsfile to use latest releases
Prepping for 6.4.0 release.
@rjrudin rjrudin merged commit 045acdc into master Nov 14, 2023
@rjrudin rjrudin deleted the release/6.4.0 branch November 14, 2023 18:04
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants