Hello! My team would love to create reusable

ComponentResources

that encode our best practices while allowing the calling code to override some aspects of the underlying AWS resources. We want to expose the available overrides as Pulumi’s

Args

objects. E.g.,:

class SecureS3Bucket(pulumi.ComponentResource):
    def __init__(
        self,
        name: str,
        bucket_overrides: aws.s3.BucketArgs = aws.s3.BucketArgs(),

Is there a convenient way to manipulate/merge

BucketArgs

objects? E.g., we might overlay required attribute values on top of the supplied

bucket_overrides

argument before passing the merged object into the

Bucket(BucketArgs)

constructor. One way we’ve found is using a combination of

pulumi._types.input_type_to_dict(bucket_overrides)

to turn

BucketArgs

into a

dict

and

pulumi.set( args_object, "attribute", "value" )

to apply

dict

entries to a

BucketArgs

object.

This is one of the reasons I recommend TypeScript and interfaces. 😞

Hey @nice-father-44210 How did you structure your schema.json for this component resource? I’m trying to do something similar and am looking to unwrap the S3 BucketArgs and use them as inputs to my component resource

@kind-jelly-61624 for now, we’re using a mono Git repo that contains both, the reusable components and the calling code. Haven’t yet had to interact with

schema.json

. But.. please do let me know if you solve it! I don’t know how much longer we’ll be able to get away with the mono repo approach 🙂

ah gotcha! will do!