Add end to end integration tests for bundle JSON Schema

.github/workflows/push.yml

@@ -107,11 +107,13 @@ jobs:
         run: npm install -g ajv-cli@5.0.0
 
       # Assert that the generated bundle schema is a valid JSON schema by using
-      # ajv-cli to validate it against a sample configuration file.
+      # ajv-cli to validate it against bundle configuration files.
       # By default the ajv-cli runs in strict mode which will fail if the schema
       # itself is not valid. Strict mode is more strict than the JSON schema
       # specification. See for details: https://ajv.js.org/options.html#strict-mode-options
       - name: Validate bundle schema
         run: |
           go run main.go bundle schema > schema.json
-          ajv -s schema.json -d ./bundle/tests/basic/databricks.yml
+          for file in ./bundle/tests/schema/pass/*.yml; do
+            ajv -s schema.json -d $file
+          done

bundle/tests/schema/pass/basic.yml

@@ -0,0 +1,2 @@
+bundle:
+  name: basic

bundle/tests/schema/pass/job.yml

@@ -0,0 +1,72 @@
+bundle:
+  name: a job
+
+workspace:
+  host: "https://myworkspace.com"
+  root_path: /abc
+
+presets:
+  name_prefix: "[DEV]"
+  jobs_max_concurrent_runs: 10
+
+variables:
+  simplevar:
+    default: true
+    description: "simplevar description"
+
+  complexvar:
+    default:
+      key1: value1
+      key2: value2
+      key3:
+        - value3
+        - value4
+    description: "complexvar description"
+
+# TODO: Include enums for fields that are a part of the bundle config in the generate
+# JSON schema.
+# TODO: Are all references allowed for "resources" or just ids?
+# TODO: Ideas for fails:
+# - missing required fields
+# - values of the wrong type
+# - enums with the wrong value
+# - invalid variable references, like not the right format.
+
+run_as:
+  service_principal_name: myserviceprincipal
+
+resources:
+  jobs:
+    myjob:
+      name: myjob
+      continuous:
+        pause_status: PAUSED
+      edit_mode: EDITABLE
+      max_concurrent_runs: 10
+      description: "my job description"
+      email_notifications:
+        no_alert_for_skipped_runs: true
+      environments:
+        - environment_key: venv
+          spec:
+            dependencies:
+              - python=3.7
+            client: "myclient"
+      format: MULTI_TASK
+      tags:
+        foo: bar
+        bar: baz
+      tasks:
+        - task_key: mytask
+          notebook_task:
+            notebook_path: ${var.simplevar}
+          existing_cluster_id: abcd
+        - task_key: mytask2
+          for_each_task:
+            inputs: av
+            concurrency: 10
+            task:
+              task_key: inside_for_each
+              notebook_task:
+                notebook_path: ${var.complexvar.key3[0]}
+        - ${var.complexvar}

bundle/tests/schema/pass/ml.yml

@@ -0,0 +1,72 @@
+bundle:
+  name: ML
+
+workspace:
+  host: "https://myworkspace.com"
+  root_path: /abc
+
+presets:
+  name_prefix: "[DEV]"
+  jobs_max_concurrent_runs: 10
+
+variables:
+  simplevar:
+    default: "true"
+    description: "simplevar description"
+
+  complexvar:
+    default:
+      key1: value1
+      key2: value2
+      key3:
+        - value3
+        - value4
+    description: "complexvar description"
+
+resources:
+  models:
+    mymodel:
+      creation_timestamp: 123
+      description: "my model"
+      latest_versions:
+        - creation_timestamp: 123
+          tags: ${var.complexvar.key1}
+          status: READY
+      permissions:
+        - service_principal_name: myserviceprincipal
+          level: CAN_MANAGE
+
+  experiments:
+    myexperiment:
+      artifact_location: /dbfs/myexperiment
+      last_update_time: 123
+      lifecycle_stage: ${var.simplevar}
+      permissions:
+        - service_principal_name: myserviceprincipal
+          level: CAN_MANAGE
+
+  model_serving_endpoints:
+    myendpoint:
+      config:
+        served_models:
+          - model_name: ${resources.models.mymodel.name}
+            model_version: abc
+            scale_to_zero_enabled: true
+            workload_size: Large
+      name: myendpoint
+
+  schemas:
+    myschema:
+      catalog_name: mycatalog
+      name: myschema
+
+  registered_models:
+    myregisteredmodel:
+      catalog_name: mycatalog
+      name: myregisteredmodel
+      schema_name: ${resources.schemas.myschema.name}
+      grants:
+        - principal: abcd
+          privileges:
+            - SELECT
+            - INSERT

bundle/tests/schema/pass/pipeline.yml

@@ -0,0 +1,58 @@
+bundle:
+  name: a pipeline
+
+workspace:
+  host: "https://myworkspace.com"
+  root_path: /abc
+
+presets:
+  name_prefix: "[DEV]"
+  jobs_max_concurrent_runs: 10
+
+variables:
+  simplevar:
+    default: true
+    description: "simplevar description"
+
+  complexvar:
+    default:
+      key1: value1
+      key2: value2
+      key3:
+        - value3
+        - value4
+    description: "complexvar description"
+
+# TODO: Include enums for fields that are a part of the bundle config in the generate
+# JSON schema.
+# TODO: No need to include the regex patterns for string fields. Fix upstream.
+# TODO: Clean up yaml-ls referneces in the tests, and the spec file as well.
+artifacts:
+  mywheel:
+    path: ./mywheel.whl
+    type: WHEEL
+
+run_as:
+  service_principal_name: myserviceprincipal
+
+resources:
+  jobs:
+    myjob:
+      name: myjob
+      tasks:
+        - task_key: ${bundle.name} pipeline trigger
+          pipeline_task:
+            pipeline_id: ${resources.mypipeline.id}
+
+  pipelines:
+    mypipeline:
+      name: mypipeline
+      libraries:
+        - whl: ./mywheel.whl
+      catalog: 3{var.complexvar.key2}
+      development: true
+      clusters:
+        - autoscale:
+            mode: ENHANCED
+            max_workers: 10
+            min_workers: 1

bundle/tests/schema/pass/quality_monitor.yml

@@ -0,0 +1,16 @@
+bundle:
+  name: quality_monitor
+
+resources:
+  quality_monitors:
+    myqualitymonitor:
+      inference_log:
+        granularities:
+          - a
+          - b
+        model_id_col: a
+        prediction_col: b
+        timestamp_col: c
+        problem_type: PROBLEM_TYPE_CLASSIFICATION
+      assets_dir: /dbfs/mnt/abc
+      output_schema_name: default

bundle/tests/schema/pass/run_job_task.yml

@@ -0,0 +1,56 @@
+bundle:
+  name: a run job task
+  databricks_cli_version: 0.200.0
+  compute_id: "mycompute"
+
+
+variables:
+  simplevar:
+    default: 5678
+    description: "simplevar description"
+
+  complexvar:
+    default:
+      key1: 1234
+      key2: value2
+      key3:
+        - value3
+        - 9999
+    description: "complexvar description"
+
+resources:
+  jobs:
+    inner:
+      permissions:
+        - user_name: user1
+          level: CAN_MANAGE
+
+      name: inner job
+      tasks:
+        - task_key: inner notebook task
+          notebook_task:
+            notebook_path: /Users/abc/notebooks/inner
+          existing_cluster_id: abcd
+
+    outer:
+      name: outer job
+      tasks:
+        - task_key: run job task 1
+          run_job_task:
+            job_id: 1234
+
+        - task_key: run job task 2
+          run_job_task:
+            job_id: ${var.complexvar.key1}
+
+        - task_key: run job task 3
+          run_job_task:
+            job_id: ${var.simplevar}
+
+        - task_key: run job task 4
+          run_job_task:
+            job_id: ${resources.inner.id}
+
+        - task_key: run job task 5
+          run_job_task:
+            job_id: ${var.complexvar.key3[1]}

bundle/tests/schema/pass/schema.yml

@@ -0,0 +1,24 @@
+bundle:
+  name: basic
+
+variables:
+  complexvar:
+    default:
+      key1: 1234
+      key2: value2
+      key3:
+        - value3
+        - 9999
+    description: complexvar description
+
+resources:
+  schemas:
+    myschema:
+      name: myschema
+      catalog_name: main
+      grants:
+        - ${var.complexvar}
+        - principal: ${workspace.current_user.me}
+          privileges:
+            - ${var.complexvar.key3[0]}
+            - ${var.complexvar.key2}