How to parse a CloudFormation template programmatically and resolve field? - amazon-web-services

Question

I want to extract a particular field from a CloudFormation template. Specifically, I am creating a stack using CDK and within it there is a StepFunction resource whose definition inside the template looks like this:
"MyStateMachineE7CD0EAE": {
"Type": "AWS::StepFunctions::StateMachine",
"Properties": {
"RoleArn": {
"Fn::GetAtt": [
"MyStateMachineRoleEC8990B2",
"Arn"
]
},
"DefinitionString": {
"Fn::Join": [
"",
[
"{\"StartAt\": ...part of state machine definition here... ,\"Resource\":\"arn:",
{
"Ref": "AWS::Partition"
},
":states:::lambda:invoke\",\"Parameters\":{\"FunctionName\":\"",
{
"Fn::GetAtt": [
"SingletonLambda2a860ff582bf4d828a504282814af94c7CEF2D65",
"Arn"
]
},
"...remainder of state machine definition here...}"
]
]
},
"LoggingConfiguration": {
"Destinations": [
{
"CloudWatchLogsLogGroup": {
"LogGroupArn": {
"Fn::GetAtt": [
"MyStatemachineLogGroup2DEF8C9E",
"Arn"
]
}
}
}
],
"IncludeExecutionData": true,
"Level": "ALL"
},
"StateMachineType": "STANDARD"
},
"DependsOn": [
"MyStateMachineRoleDefaultPolicy4C064A65",
"MyStateMachineRoleEC8990B2"
],
"Metadata": {
"aws:cdk:path": "some/path/here"
}
}
The DefinitionString is made up of a call to Fn:Join, when there is an array of strings, some of which are references to other CloudFormation resources. There can be many such references, depending on how complicated the StateMachine definition is.
I want to be able to use StepFunctions Local in order to test the flow of data through the StepFunctions state machine, and in particular to check the InputPath and ResultSelector data input and output as part of some integration tests of the state machine as a whole. The StepFunction Local CLI tool requires as input a single string, and that string is the evaluation of the DefinitionString field.
Is there a way of extracting the field from the CloudFormation template, perhaps by a CloudFormation client API? I'd rather not have to try and parse the JSON directly unless I absolutely have to.