LambdaInvokerFactory and @LambdaFunction do not support multiple environments

[codan84]

If I create an interface:

interface MyLambdas {
    @LambdaFunction(functionName = "my-aws-lambda-name")
    public String helloWorld(String name);
}

And then build an implementation:

AWSLambdaClient lambda = new AWSLambdaClient(); lambda.configureRegion(Regions.EU_WEST_1);
MyLambdas lambdas = LambdaInvokerFactory.build(MyLambdas.class, lambda);

This is all fine. However, it doesn't support different deployment environments (say int/test/live), where a lambda can have different name (say created via CloudFormation).

To address this, I would propose to overload LambdaInvokerFactory.build to take an additional map of key->lambdaName. Key would store the value passed to @LambdaFunction, whilst value for this key would be the actual function name in AWS. This would allow to use different function names for different environments. An example of use:

interface MyLambdas {
    @LambdaFunction(functionName = "hello-world-lambda-key")
    public String helloWorld(String name);
}

Map<String, String> myLambdaNames = new HashMap<String, String>();
myLambdaNames.put("hello-world-lambda-key", "my-aws-lambda-name");

AWSLambdaClient lambda = new AWSLambdaClient();
lambda.configureRegion(Regions.EU_WEST_1);
MyLambdas lambdas = LambdaInvokerFactory.build(MyLambdas.class, lambda, myLambdaNames);

As you can see this let's us to pull the names at runtime, from places like config etc. I think this would be a beneficial addition to the API.
I have implemented this functionality in this fork

[shorea]

Made some comments in the PR.

[gigi888]

so cool! it solves our problem too! Many thanks