SOLR-18234 | Move packagemanager under CLI and stop test-framework from depending on CLI

[jaykay12]

Description

Context: #4663 (comment)

This is a pre-requisite change needed before we attempt: CLI module movement outside core.

Solution

Used Cursor help in formulating the approach, code is reviewed & vetted by me.

Tests

Please describe the tests you've developed or run to confirm this patch implements the feature or solves the problem.

Checklist

Please review the following and check all that apply:

• I have reviewed the guidelines for How to Contribute and my code conforms to the standards described there to the best of my ability.
• I have created a Jira issue and added the issue ID to my pull request title.
• I have given Solr maintainers access to contribute to my PR branch. (optional but recommended, not available for branches on forks living under an organisation)
• I have developed this patch against the main branch.
• I have run ./gradlew check.
• I have added tests for my changes.
• I have added documentation for the Reference Guide
• I have added a changelog entry for my change

[dsmiley]

Thanks for remembering to carry this forward!

[dsmiley]

nice change

[dsmiley]

ToolRuntime (in CLI) implements this, i see. What implements this within solr core?

[jaykay12]

Disclaimer - Reply via Cursor

Packagemanager uses UserIO directly; it doesn’t implement console output.

PackageManager / RepositoryManager take a UserIO and call print / printSuccess / printError. PackageUtils uses the color constants on the interface. That is how we dropped the CLI dependency.

Implementations:

CLI: ToolRuntime (passed in from PackageTool)
tests: UserIO.NONE
There is no production core path that constructs PackageManager outside the CLI, so core doesn’t need its own console impl.

[dsmiley]

That's weird! If UserIO is only implemented in the CLI package yet it's defined outside of that package... it's weird; no? Why are those parts of PackageManager that use UserIO located in PackageManager when they could be in some CLI side place?

[jaykay12]

That’s a fair point - it does look inverted.

PackageManager still owns the operation flow today, so the print calls sit next to that logic.
UserIO is only a sink so those calls don’t import org.apache.solr.cli: CLI passes ToolRuntime, and tests use NONE.

I’d be happy to move the user-facing output into PackageTool and leave PackageManager as throw/return. That’s a larger split than this PR was aiming for; I was treating this change as just cutting the CLI dependency.

Can I take that up in a follow-up PR, if that's okay to you.

[dsmiley]

I think you could just move the packagemanager package to org.apache.solr.cli.packagemanager and then you wouldn't need the UserIO idea. That's a simpler change.

[jaykay12]

Thanks for the suggestion - this is much cleaner.

I’ve moved packagemanager to org.apache.solr.cli.packagemanager and dropped UserIO, so it can use ToolRuntime again. That should also make the later CLI module extraction simpler.

Appreciate you taking the time to steer this. ✨

[jaykay12]

self-review ✅

@dsmiley can you please retrigger the workflows, i will keep an eye on the crave tests.