feat(stepfunctions-tasks): algorithmName validation for SageMakerCreateTrainingJob (#26877)

Referencing PR #26675, I have added validation for the `algorithmName` parameter in `SageMakerCreateTrainingJob`.
However, it was suggested that changes for validation should be separated.  So, I have created this PR.

Docs for `algorithmName`:
https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_AlgorithmSpecification.html#API_AlgorithmSpecification_Contents


Exemption Request:
This change does not alter the behavior.
I believe the unit test `create-training-job.test.ts` that I have added is sufficient to test this change.

----

*By submitting this pull request, I confirm that my contribution is made under the terms of the Apache-2.0 license*

Author: tmyoda

packages/aws-cdk-lib/aws-stepfunctions-tasks/lib/sagemaker/create-training-job.ts

@@ -4,7 +4,7 @@ import { renderEnvironment, renderTags } from './private/utils';
 import * as ec2 from '../../../aws-ec2';
 import * as iam from '../../../aws-iam';
 import * as sfn from '../../../aws-stepfunctions';
-import { Duration, Lazy, Size, Stack } from '../../../core';
+import { Duration, Lazy, Size, Stack, Token } from '../../../core';
 import { integrationResourceArn, validatePatternSupported } from '../private/task-utils';
 
 /**
@@ -163,6 +163,14 @@ export class SageMakerCreateTrainingJob extends sfn.TaskStateBase implements iam
       throw new Error('Must define either an algorithm name or training image URI in the algorithm specification');
     }
 
+    // check that both algorithm name and image are not defined
+    if (props.algorithmSpecification.algorithmName && props.algorithmSpecification.trainingImage) {
+      throw new Error('Cannot define both an algorithm name and training image URI in the algorithm specification');
+    }
+
+    // validate algorithm name
+    this.validateAlgorithmName(props.algorithmSpecification.algorithmName);
+
     // set the input mode to 'File' if not defined
     this.algorithmSpecification = props.algorithmSpecification.trainingInputMode
       ? props.algorithmSpecification
@@ -324,6 +332,21 @@ export class SageMakerCreateTrainingJob extends sfn.TaskStateBase implements iam
       : {};
   }
 
+  private validateAlgorithmName(algorithmName?: string): void {
+    if (algorithmName === undefined || Token.isUnresolved(algorithmName)) {
+      return;
+    }
+
+    if (algorithmName.length < 1 || 170 < algorithmName.length) {
+      throw new Error(`Algorithm name length must be between 1 and 170, but got ${algorithmName.length}`);
+    }
+
+    const regex = /^(arn:aws[a-z\-]*:sagemaker:[a-z0-9\-]*:[0-9]{12}:[a-z\-]*\/)?([a-zA-Z0-9]([a-zA-Z0-9-]){0,62})(?<!-)$/;
+    if (!regex.test(algorithmName)) {
+      throw new Error(`Expected algorithm name to match pattern ${regex.source}, but got ${algorithmName}`);
+    }
+  }
+
   private makePolicyStatements(): iam.PolicyStatement[] {
     // set the sagemaker role or create new one
     this._grantPrincipal = this._role =

packages/aws-cdk-lib/aws-stepfunctions-tasks/test/sagemaker/create-training-job.test.ts

@@ -408,3 +408,146 @@ test('Cannot create a SageMaker train task with both algorithm name and image na
   }))
     .toThrowError(/Must define either an algorithm name or training image URI in the algorithm specification/);
 });
+
+test('Cannot create a SageMaker train task with both algorithm name and image name defined', () => {
+
+  expect(() => new SageMakerCreateTrainingJob(stack, 'SageMakerTrainingTask', {
+    trainingJobName: 'myTrainJob',
+    algorithmSpecification: {
+      algorithmName: 'BlazingText',
+      trainingImage: tasks.DockerImage.fromJsonExpression(sfn.JsonPath.stringAt('$.Training.imageName')),
+    },
+    inputDataConfig: [
+      {
+        channelName: 'train',
+        dataSource: {
+          s3DataSource: {
+            s3DataType: tasks.S3DataType.S3_PREFIX,
+            s3Location: tasks.S3Location.fromJsonExpression('$.S3Bucket'),
+          },
+        },
+      },
+    ],
+    outputDataConfig: {
+      s3OutputLocation: tasks.S3Location.fromBucket(s3.Bucket.fromBucketName(stack, 'Bucket', 'mybucket'), 'myoutputpath/'),
+    },
+  }))
+    .toThrowError(/Cannot define both an algorithm name and training image URI in the algorithm specification/);
+});
+
+test('create a SageMaker train task with trainingImage', () => {
+
+  const task = new SageMakerCreateTrainingJob(stack, 'SageMakerTrainingTask', {
+    trainingJobName: 'myTrainJob',
+    algorithmSpecification: {
+      trainingImage: tasks.DockerImage.fromJsonExpression(sfn.JsonPath.stringAt('$.Training.imageName')),
+    },
+    inputDataConfig: [
+      {
+        channelName: 'train',
+        dataSource: {
+          s3DataSource: {
+            s3DataType: tasks.S3DataType.S3_PREFIX,
+            s3Location: tasks.S3Location.fromJsonExpression('$.S3Bucket'),
+          },
+        },
+      },
+    ],
+    outputDataConfig: {
+      s3OutputLocation: tasks.S3Location.fromBucket(s3.Bucket.fromBucketName(stack, 'Bucket', 'mybucket'), 'myoutputpath/'),
+    },
+  });
+
+  // THEN
+  expect(stack.resolve(task.toStateJson())).toMatchObject({
+    Parameters: {
+      AlgorithmSpecification: {
+        'TrainingImage.$': '$.Training.imageName',
+        'TrainingInputMode': 'File',
+      },
+    },
+  });
+});
+
+test('create a SageMaker train task with image URI algorithmName', () => {
+
+  const task = new SageMakerCreateTrainingJob(stack, 'SageMakerTrainingTask', {
+    trainingJobName: 'myTrainJob',
+    algorithmSpecification: {
+      algorithmName: 'arn:aws:sagemaker:us-east-1:123456789012:algorithm/scikit-decision-trees',
+    },
+    inputDataConfig: [
+      {
+        channelName: 'train',
+        dataSource: {
+          s3DataSource: {
+            s3DataType: tasks.S3DataType.S3_PREFIX,
+            s3Location: tasks.S3Location.fromJsonExpression('$.S3Bucket'),
+          },
+        },
+      },
+    ],
+    outputDataConfig: {
+      s3OutputLocation: tasks.S3Location.fromBucket(s3.Bucket.fromBucketName(stack, 'Bucket', 'mybucket'), 'myoutputpath/'),
+    },
+  });
+
+  // THEN
+  expect(stack.resolve(task.toStateJson())).toMatchObject({
+    Parameters: {
+      AlgorithmSpecification: {
+        AlgorithmName: 'arn:aws:sagemaker:us-east-1:123456789012:algorithm/scikit-decision-trees',
+      },
+    },
+  });
+});
+
+test('Cannot create a SageMaker train task when algorithmName length is 171 or more', () => {
+
+  expect(() => new SageMakerCreateTrainingJob(stack, 'SageMakerTrainingTask', {
+    trainingJobName: 'myTrainJob',
+    algorithmSpecification: {
+      algorithmName: 'a'.repeat(171), // maximum length is 170
+    },
+    inputDataConfig: [
+      {
+        channelName: 'train',
+        dataSource: {
+          s3DataSource: {
+            s3DataType: tasks.S3DataType.S3_PREFIX,
+            s3Location: tasks.S3Location.fromJsonExpression('$.S3Bucket'),
+          },
+        },
+      },
+    ],
+    outputDataConfig: {
+      s3OutputLocation: tasks.S3Location.fromBucket(s3.Bucket.fromBucketName(stack, 'Bucket', 'mybucket'), 'myoutputpath/'),
+    },
+  }))
+    .toThrowError(/Algorithm name length must be between 1 and 170, but got 171/);
+});
+
+test('Cannot create a SageMaker train task with incorrect algorithmName', () => {
+
+  expect(() => new SageMakerCreateTrainingJob(stack, 'SageMakerTrainingTask', {
+    trainingJobName: 'myTrainJob',
+    algorithmSpecification: {
+      algorithmName: 'Blazing_Text', // underscores are not allowed
+    },
+    inputDataConfig: [
+      {
+        channelName: 'train',
+        dataSource: {
+          s3DataSource: {
+            s3DataType: tasks.S3DataType.S3_PREFIX,
+            s3Location: tasks.S3Location.fromJsonExpression('$.S3Bucket'),
+          },
+        },
+      },
+    ],
+    outputDataConfig: {
+      s3OutputLocation: tasks.S3Location.fromBucket(s3.Bucket.fromBucketName(stack, 'Bucket', 'mybucket'), 'myoutputpath/'),
+    },
+  }))
+    .toThrowError(/Expected algorithm name to match pattern/);
+});