feat: API Sync by GitHub Action (2023-08-15)

[github-actions]

This API Sync PR was automated through GitHub Actions workflow_displatch
on 2023-08-15.

• latest Swagger is fetched
• patches have been applied
• generated client has been updated

[github-actions]

Failed to generate code from latest patched spec. Someone with write access must fix this PR manually and then convert it from Draft status to Ready for Review.

[displague]

for diff in patches/spec.fetched.json/05-remove-because-java-generator-lack-of-support-for-primitive-anyof.patch patches/spec.fetched.json/08-make-vrf-required-param-in-vrf-metal-gateway.patch patches/spec.fetched.json/09-set-additional-prop-to-false-in-href.patch patches/spec.fetched.json/20230407-Device-schema-HwReservation-or-href.patch; do \
	patch --no-backup-if-mismatch -N -t -p1 -i $diff; \
done
patching file spec/oas3.patched/openapi/public/components/schemas/FacilityInput.yaml
Hunk #1 succeeded at 11 (offset 1 line).
patching file spec/oas3.patched/openapi/public/components/schemas/IPReservationOrHref.yaml
patching file spec/oas3.patched/openapi/public/components/schemas/MetalGateway.yaml
patching file spec/oas3.patched/openapi/public/components/schemas/MetalGatewayList.yaml
Hunk #1 succeeded at 2 with fuzz 2.
patching file spec/oas3.patched/openapi/public/components/schemas/VrfIpReservationOrHref.yaml
patching file spec/oas3.patched/openapi/public/components/schemas/VrfMetalGateway.yaml
patching file spec/oas3.patched/openapi/public/components/schemas/Href.yaml
patching file spec/oas3.patched/openapi/public/components/schemas/Device.yaml
[main a163723] sync: apply patches to 2023-08-15 spec 18 files changed, 1[9](https://github.com/equinix-labs/metal-java/actions/runs/5871156270/job/15919829071#step:6:10)3 insertions(+), 39 deletions(-) create mode [10](https://github.com/equinix-labs/metal-java/actions/runs/5871156270/job/15919829071#step:6:11)0644 spec/oas3.patched/openapi/public/components/schemas/DedicatedPortCreateInput.yaml create mode 100644 spec/oas3.patched/openapi/public/components/schemas/VlanFabricVcCreateInput.yaml rename spec/oas3.patched/openapi/public/components/schemas/{InterconnectionCreateInput.yaml => VrfFabricVcCreateInput.yaml} (54%)

Error:  COMPILATION ERROR : 
[INFO] -------------------------------------------------------------
Error:  /home/runner/work/metal-java/metal-java/equinix-openapi-metal/src/test/java/com/equinix/openapi/metal/v1/api/InterconnectionsApiTest.java:[74,89] error: incompatible types: InterconnectionCreateInput cannot be converted to CreateOrganizationInterconnectionRequest
Error:  /home/runner/work/metal-java/metal-java/equinix-openapi-metal/src/test/java/com/equinix/openapi/metal/v1/api/InterconnectionsApiTest.java:[91,79] error: incompatible types: InterconnectionCreateInput cannot be converted to CreateOrganizationInterconnectionRequest

[displague]

@cprivitere and I walked through this and made this change to the InterconnectionApiTest with the intent to address the failing test message:

incompatible types: InterconnectionCreateInput cannot be converted to CreateOrganizationInterconnectionRequest

It seems that the openapitools-generator may have a blindspot to requestBodies with oneOf schema bodies. Java is neither of our strong suits, so we're not sure. The tests pass with this change, but the change to use CreateOrganizationInterconnectionRequest may be incorrect. We're not sure how to trigger the TypeAdapterFactory for the abstract (generic?) type InterconnectionCreateInput in the correct way. Using CreateOrganizationInterconnectionRequest may be the wrong type, but since the value used for testing is null, maybe it doesn't matter?

[displague]

We noticed that one of the OAS3 changes introduced createOrganizationInterconnection_request which has an unusual naming pattern.

[displague]

Also, we didn't see any issues pop-out based on our searches

[cprivitere]

My local environment was JDK 20 and if I ran make docker_run twice, it'd work. Our CI is JDK 11, however, so I'm working on setting up a JDK 11 local environment to get a better test.

It would be good to know if folks would like to move to the newer LTS of 17 or not, as I'm not sure how we arrived at using 11.