tl dr Is there a way to merge pact verification results before sending it to the pact broker? Hi all, I have a provider multi-modules maven project where I expect some modules to execute parts of the pact verification. The sum of all the module verification would cover all the pacts. I am planning use filters , to decide what states a module should be testing. My understanding is that the results are accumulated in memory, and not shared across junit/test jvms. Which means I am not sure how to assemble those results before sending it to the broker. I want each module to verify some pacts since the business logic is replicated across them. This service is an integration service, so for each external system we support we have a dedicated module for this that implements our service api.

hmm I don’t think so, at least a search of the docs suggests it’s not readily available

Well it would still boil down to the same problem, I would have more than one module implementing A endpoints

I would have more than one module implementing A endpoints

This is certainly unusual - although it is a leaky part of pact’s abstraction that the consumer needs to know which services it has that are different

I would have more than one module implementing A endpoints

OK I see the problem now. The “uses {B, C}” was confusing to me. It sounds like “implements part of {B, C}” is actually what it does.

If I had this situation, I would probably just do contracts between Consumer -> X, Consumer -> Y and Consumer -> Z.

Yes I can test this by spinning up the full provider, but then I would have to control which module receives which request using states. Which means I would be pushing this “implementation detail” to the consumer, and will have to replicate the tests for each Systems. This is something I was trying to avoid, but I guess that will be the way to go.