Follow-up changes for fixing RepositoryResolver install lists for tolerated dependencies #19963
Conversation
Correct the starting scenario by listing all the features which are meant to be installed.
The sets featuresMissing and requirementsFoundForOtherProducts are disjoint. If we're checking that featuresMissing contains a feature name, we don't need to also check that requirementsFoundForOtherProducts doesn't contain it. Previously we just had a check to ensure it wasn't in requirementsFoundForOtherProducts. This wasn't correct either as it's possible for the feature to be in neither set if it conflicted with another feature.
The copyright headers should have been updated in OpenLiberty#19871
Code analysis and actionsDO NOT DELETE THIS COMMENT.
|
| /** | ||
| * Map from symbolic name to feature for all features returned as resolved by the kernel resolver | ||
| * <p> | ||
| * Keyset is mutually exclusive with {@link #featuresMissing} and {@link #requirementsFoundForOtherProducts} | ||
| */ | ||
| Map<String, ProvisioningFeatureDefinition> resolvedFeatures; | ||
|
|
||
| /** | ||
| * List of requested features that were reported missing by the kernel resolver and weren't found in the repository applicable to another product. | ||
| * <p> | ||
| * May include sample names if they were missing | ||
| * <p> | ||
| * Mutually exclusive with {@link #resolvedFeatures} and {@link #requirementsFoundForOtherProducts} | ||
| */ | ||
| List<String> featuresMissing; |
There was a problem hiding this comment.
It's fairly straightforward to validate that these sets (and requirementsFoundForOtherProducts) are mutually exclusive as they only have items added to them within the resolveFeatures() method.
The kernel resolver returns a list of resolved features and a list of missing features, and then the missing features are split between the featuresMissing and requirementsFoundForOtherProducts sets.
|
#build |
|
Your personal build request is at https://wasrtc.hursley.ibm.com:9443/jazz/resource/itemOid/com.ibm.team.build.BuildResult/_2PjnsIAsEeykJKbDLxYjRw Target locations of links might be accessible only to IBM employees. |
|
The build Azquelt-19963-20220128-0434 For help analyzing your personal build, go to https://cognitive.hursley.ibm.com/buildAnalysis.html?uuid=_2PjnsIAsEeykJKbDLxYjRw |
|
The one failure in the build is a known problem. |
A few follow up fixes:
Follow up for #19860