Using package google.cloud.bigquery.v2 for resource group BigQuery
As part of moving resources from Terraform controllers to direct controllers (Epic #5954), we need to create the Go types for BigQueryDatasetAccess.

Currently, BigQueryDatasetAccess is managed by the Terraform controller (marked with tf2crd=true). The goal is to create the Go types in apis/bigquery/v1beta1/ so that we can eventually migrate the controller implementation to the "direct" approach.

Instructions

1. Create a generate.sh:
   Create apis/bigquery/v1beta1/generate.sh which includes BigQueryDatasetAccess.
   It likely maps to something like google.cloud.bigquery.v2.
   Example:

   go run . generate-types \
     --service google.cloud.bigquery.v2 \
     --api-version bigquery.cnrm.cloud.google.com/v1beta1 \
     --resource BigQueryDatasetAccess:DatasetAccess \
     --include-skipped-output
   
   go run . generate-mapper \
     --service google.cloud.bigquery.v2 \
     --api-version bigquery.cnrm.cloud.google.com/v1beta1 \
     --include-skipped-output

2. Set the write permission on the new apis/bigquery/v1beta1/generate.sh file. You should do this by running both chmod +x apis/bigquery/v1beta1/generate.sh and git add --chmod=+x apis/bigquery/v1beta1/generate.sh.

3. Generate Scaffolding:
   Run apis/bigquery/v1beta1/generate.sh. This should create apis/bigquery/v1beta1/datasetaccess_types.go.

4. Iterate on Types:
   Compare the generated CRD with the existing one using dev/tasks/diff-crds.
   Modify apis/bigquery/v1beta1/datasetaccess_types.go until the CRD matches the existing one at config/crds/resources/apiextensions.k8s.io_v1_customresourcedefinition_bigquerydatasetaccesss.bigquery.cnrm.cloud.google.com.yaml.

   Acceptance Criteria:

   • Running dev/tasks/diff-crds should not show differences (or minimal acceptable ones like descriptions).
   • Ensure that running the check_crd_equivalence MCP on the CRD should return EQUIVALENT.
   • Changes to the schema (fields added/removed) are NOT acceptable.

5. Copyright Headers:
   Ensure that new files have the correct copyright header:

   // Copyright 2026 Google LLC

   Please do not change the copyright on existing files.

6. Labels:
   Ensure the controller-runtime annotations match the existing CRD labels, including:

   // +kubebuilder:metadata:labels="cnrm.cloud.google.com/managed-by-kcc=true"
   // +kubebuilder:metadata:labels="cnrm.cloud.google.com/system=true"
   // +kubebuilder:metadata:labels="cnrm.cloud.google.com/stability-level=stable"
   // +kubebuilder:metadata:labels="cnrm.cloud.google.com/tf2crd=true"

   The goal is to maintain these annotations, not add an annotation if it is missing.

7. Status:
   status.observedGeneration should be an *int64.

8. Generate Mappers:

   • Running dev/tasks/generate-types-and-mappers will generate the mapper code once the apis/bigquery/v1beta1/datasetaccess_types.go file is generating an equivalent CRD.
   • Run make all-binary to ensure the generated mapper code compiles. Please fix any issue discovered by this compilation.

This issue is part of Epic #5954.