Allow to define parent first artifacts in fast jar mode

[essobedo]

fixes #17391

Motivation

In case a dependency of the application contains classes that are part of the JDK, we can end up with LinkageError. This is typically what happens when we have a dependency that contains the class javax.xml.namespace.QName, because this class is still part of the JDK.

The official workaround is to define the artifact with quarkus.class-loading.parent-first-artifacts but it is ignored in fast jar mode.

Modifications

• Relies on ClassLoadingConfig to extract the parent first artifacts that are defined by the enduser to build a collection of artifact keys to define as parent first artifacts in the model.
• Adds an integration test to ensure that it works as expected in build mode whatever the target packing type
• Adds an integration test to ensure that it works as expected in dev mode

[essobedo]

@stuartwdouglas it is the PR for the problem that I described here. WDYT of it?

[quarkus-bot]

This workflow status is outdated as a new workflow run has been triggered.

Failing Jobs - Building 32d5711

Status	Name	Step	Test failures	Logs	Raw logs
✖	Initial JDK 11 Build	Build	⚠️ Check →	Logs	Raw logs

[famod]

@essobedo CI says there is a formatting issue. Please check/fix.

[essobedo]

@essobedo CI says there is a formatting issue. Please check/fix.

@famod Fixed, thx

[quarkus-bot]

This workflow status is outdated as a new workflow run has been triggered.

🚫 This workflow run has been cancelled.

Failing Jobs - Building 0d900a7

⚠️ Artifacts of the workflow run were not available thus the report misses some details.

Status	Name	Step	Test failures	Logs	Raw logs
⌛	Initial JDK 11 Build	Build	⚠️ Check →	Logs	Raw logs

[quarkus-bot]

This workflow status is outdated as a new workflow run has been triggered.

Failing Jobs - Building 8aeea2a

Status	Name	Step	Test failures	Logs	Raw logs
✔️	JVM Tests - JDK 11
✖	JVM Tests - JDK 11 Windows	Build	Test failures	Logs	Raw logs
✔️	JVM Tests - JDK 16
✖	MicroProfile TCKs Tests	Verify	Test failures	Logs	Raw logs

Full information is available in the Build summary check run.

Test Failures

⚙️ JVM Tests - JDK 11 Windows #

📦 extensions/grpc/deployment

✖ io.quarkus.grpc.server.scaling.SingleGrpcVerticleTest. - More details - Source on GitHub

📦 extensions/resteasy-reactive/jaxrs-client-reactive/deployment

✖ io.quarkus.jaxrs.client.reactive.deployment.test.UniInvokerTestCase.testInjection line 48 - More details - Source on GitHub

---

⚙️ MicroProfile TCKs Tests #

📦 tcks/microprofile-config

✖ org.eclipse.microprofile.config.tck.ConfigPropertiesTest.testConfigPropertiesDefaultOnBean line 172 - More details - Source on GitHub

✖ org.eclipse.microprofile.config.tck.ConfigPropertiesTest.testConfigPropertiesNoPrefixOnBean line 149 - More details - Source on GitHub

✖ org.eclipse.microprofile.config.tck.ConfigPropertiesTest.testConfigPropertiesNoPrefixOnBeanThenSupplyPrefix line 156 - More details - Source on GitHub

✖ org.eclipse.microprofile.config.tck.ConfigPropertiesTest.testConfigPropertiesPlainInjection line 106 - More details - Source on GitHub

✖ org.eclipse.microprofile.config.tck.ConfigPropertiesTest.testConfigPropertiesWithPrefix line 115 - More details - Source on GitHub

✖ org.eclipse.microprofile.config.tck.ConfigPropertiesTest.testConfigPropertiesWithoutPrefix line 132 - More details - Source on GitHub

✖ org.eclipse.microprofile.config.tck.ConfigPropertiesTest.testNoConfigPropertiesAnnotationInjection line 165 - More details - Source on GitHub

[stuartwdouglas]

LGTM. With our internal parent first stuff we differentiate between dev/test and prod, but I don't think it really matters here because you can control it via profiles.

[quarkus-bot]

Failing Jobs - Building 8aeea2a

Status	Name	Step	Test failures	Logs	Raw logs
✖	Maven Tests - JDK 11	Build	Test failures	Logs	Raw logs

Full information is available in the Build summary check run.

Test Failures

⚙️ Maven Tests - JDK 11 #

📦 integration-tests/maven

✖ io.quarkus.maven.it.DevMojoIT.testThatNewResourcesAreServed line 849 - More details - Source on GitHub

[stuartwdouglas]

Thanks for the patch

[essobedo]

@stuartwdouglas Thx for reviewing and merging it, do you want a backport for the branch 1.13?