Gradle

Clone Tools
  • last updated a few seconds ago
Constraints
Constraints: committers
 
Constraints: files
Constraints: dates
REVIEW-2482: Improved DSL docs for IvyPublication - Moved javadoc comment 'setArtifacts' into getter so that they are rendered correctly in DSL doc - Removed separate 'setArtifacts' method from DSL doc

Further spec updates for multiple publications from a single project

Update publication-model.md

    • -80
    • +89
    /design-docs/publication-model.md
Started to tidy up the way that ExternalResourceResolver uses the cache-manager - Moved parsing of module descriptor files out of RepositoryArtifactCache and into ExternalResourceResolver - Tweaked the implementation so it does just what is required - Native dependency resolvers continue to use the CacheManager as-is

Declare the distribution zips as inputs to the integ test task of the distributions project.

REVIEW-2440: Updated test for now auto-imported packages

REVIEW-2441: Check existence of lib.exe for VisualCpp toolchain availability

Changed DefaultIvyDependencyPublisher.publish() so that it accepts a ModuleVersionPublishMetaData instead of a ModuleDescriptor, so that it no longer needs to use the magic extra attributes. This in turn means we no longer need to set the extra attribute when resolving and can get rid of the duplicate PublishModuleDescriptorConverter.

REVIEW-2302 Started using strictlyEquals in one of the unit tests.

REVIEW-2297 Simplified the unit test

Improved algorithm for resolving a dynamic version when a repository contains multiple ivyPatterns - VersionList now contains the ResourcePattern where a version was loaded from, as well as the version itself. - ExternalResourceRepository lists all available versions before sorting them and assessing which is the 'latest'

HTTP-aware version test for resolving 'latest.milestone' against an ivy repository with multiple ivyPatterns

Recent change that avoids eager resolution of testCompile and testRuntime configurations also fixed a minor problem with '--no-rebuild' switch. Updated the integ test to reflect that.

REVIEW-1702 Added more integ test coverage for parallel execution.

ExternalResourceResolver no longer implements ivy DependencyResolver - Introduced separate methods to RepositoryArtifactCache to replace the RepositoryCacheManager methods for ExternalResourceResolver - The existing implementations will still be used for native ivy resolvers - Wrap ExternalResourceResolver in a DependencyResolver for use in parsing (until we switch parsing to Gradle APIs) - Removed a lot of cruft from ExternalResourceResolver that is no longer required

REVIEW-2455 Removed unnecessary method from MultiProcessSafePersistentIndexedCache.

1. I needed to relax the validation of getLock() method. Currently, it is possible that the close operation is invoked on a cache that does not have any owner thread (see the unit test). I'm not sure if this approach is good - please review.

2. Reworked the DefaultCacheAccess cache a little bit to make above change possible (cannot clear the lock owner thread before closing the lock).

Revert "REVIEW-2455 Removed unnecessary method from MultiProcessSafePersistentIndexedCache."

The change was not quite ready and made some tests upset.

This reverts commit bc87de9d3a5ec8409abc2102a0450c0cc75e0dae.

Integ test for resolving Ivy dynamic versions with multiple ivyPatterns specified

REVIEW-2455 Attempted to unify the implementation of FileLockContentionHandlers.

1. The attempt has failed. Down the road we want to have a single FileLockManager per process. Getting to that point is not trivial however. The default file lock manager opens a datagram port before creating a lock and requires the clients of the locks to provide reasonable callbacks when lock contention happens. At the moment, we use the default file lock manager in few places where the lock contention handling is absent and adding it is tricky (I'm open to suggestions). Hence perhaps we allow the no-op implementation of lock contention handler in those places for the time being until we have a better reason to push harder for a single file lock manager per process.

2. Rename job and refactoring. Added dedicated method for enabling file lock contention handling on the file lock manager.

  1. … 8 more files in changeset.
cleanup some archive related tests

- REVIEW-2446/REVIEW-2451: Add more assertions to the duplicateStrategy integration tests

- REVIEW-2446: fix assertZipContainsOnly and move to ZipTestFixture

- REVIEW-2446: create explicit TarTestFixture

- REVIEW-2446: create explicit JarTestFixture and close jarFile properly

REVIEW-2455 Updated javadoc.

REVIEW-2455 Removed unnecessary method from MultiProcessSafePersistentIndexedCache.

Reworked the DefaultCacheAccess cache a little bit to make above change possible (cannot clear the lock owner thread before closing the lock).

Changed ProjectDependencyResolver to use the artifact info provided by the publish meta data, rather than looking for a magic extended attribute.

REVIEW-2455 Injection of the executor factory. Technically, it's not injected ATM, as there doesn't seem to be a good reason.

REVIEW-2455 Tweaked the javadoc to reflect the recent changes.

Fixed the problem with with testRuntime and testCompile configurations resolved eagerly, even when no tests were present.

1. Added @SkipWhenEmpty annotation where needed. Removed @Input annotation because it didn't add any value. Fixed the integration tests.

2. This change may improve performance and heap consumption of certain builds (large multi project builds with little tests).

GRADLE-2666: Don't process Gradle plugins and extensions when calculating pom model

Spec update.

    • -0
    • +4
    /design-docs/tooling-api-improvements.md
Introduced BuildableModuleVersionPublishMetaData and changed ArtifactsToModuleDescriptorConverter to attach artifacts to the publish meta data rather than the module descriptor.

Fix bug in HTML dependency report generation - insight dependencies were matched with a "contains" spec instead of an "equals" spec