Split the JUnit 5 `OutputCapture` class into separate `OutputExtension`
and `CapturedOutput` classes. The JUnit 5 callback methods are now
contained only in the `OutputExtension` class so no longer pollute the
public API that users will interact with.
The `CapturedOutput` class has also been updated to capture System.err
and System.out separately to allow distinct assertions if required.
Closes gh-17029
This commit our Neo4j OGM dependency with the Spring Data Neo4j
snapshots that are currently included in snapshots of Spring Data Kay.
It switches to using Neo4j's Bolt driver by default, aligning it with
the default of the latest Spring Data Neo4j 5 snapshots.
It also contains a workaround for a Neo4j OGM issue [1] and a change
to Neo4jDataAutoConfigurationTests that prevents the entire classpath
from being scanned.
See gh-8687
[1] https://github.com/neo4j/neo4j-ogm/issues/340
This commit polihes the original Neo4j contribution in several areas.
Rather than providing the packages to scan, this commit rearranges the
`EntityScan` and `EntityScanRegistrar` so that the logic can be shared
for other components. If no package is provided, scanning now defaults to
the "auto-configured" package(s) and a `@NodeEntityScan` annotation
allows to override that.
The configuration has also been updated to detect the driver based on the
`uri` property. If the embedded driver is available we use that by
default. If it is not available, we're trying to connect to a Neo4j
server running on localhost. It is possible to disable the embedded mode
or set the `uri` parameter explicitly to deviate from these defaults.
The sample no longer relies on the embedded driver for licensing reason:
rather it expects an instance running on localhost (like other
data-related samples) and gracefully ignore any connection error. A
README has been added in the sample to further explain the available
options;
Closes gh-5458
Andy Wilkinson
Phillip Webb
Madhura Bhave
Stephane Nicoll
Michael Hunger