Hi, given a scenario where I am targeting. signup endpoint, the consumer is frontend, would the consumer test live in the frontend repo?

Hi, I added below canIDeploy stage in my gitlab pipeline. I can see the installation log as "pact-ruby-standalone v2.0.7 installed to /builds/xrGFoCjz/0/mpl/jetdelivery/pact" But then, the pipeline fails due to to error -> "/builds/xrGFoCjz/0/mpl/jetdelivery/pact/lib/ruby/bin/ruby: exec: line 6: /builds/xrGFoCjz/0/mpl/jetdelivery/pact/lib/ruby/bin.real/ruby: not found"

canIDeploy:
    stage: canIDeploy
    before_script:
        - apk add bash
        - apk add curl
    needs:
        - test
    script:
        - curl -fsSL <https://raw.githubusercontent.com/pact-foundation/pact-ruby-standalone/master/install.sh> | PACT_CLI_VERSION=v2.0.7 bash
        - ./pact/bin/pact-broker can-i-deploy --pacticipant Foo --latest
    tags:
        - k8s

Hi, If I add a field to the endpoint, what do I need to make sure consumer fails for such changes?

Hi, I have a question that always troubles me from using pact. 1. Pact is a great tool for contract/integration test, which is usually at the border of the system. 2. Both parties (consumer/producer) of the test needs to include Pact in their test suite so that contract testing really takes effect. 3. However, since it’s at the border, it usually means they belong to two different teams. It’s hard to impact the other team to accept the contract test and implement it. Is my understanding true? or do I miss something? Thanks

Good morning, I’m trying to pull some reports from the PactFlow DB, and I wanted to figure out whether a participant is doing CDC or BDC. Is there a way to check for this on the DB?

Hi, I want to turn my consumer Pact test response body objects into mocks for unit tests. The problem I have is that these mocks will have matcher objects nested in the response dictionary (python) that my unit tests won't know how to handle. Other than avoiding using matchers like

Like

, how can I "clean" these dictionaries or otherwise use them for unit tests?

Hi, it seems your latest release of pact-python 2.1.0 didn't get published with pypy? Was that a mistake or intentional? https://pypi.org/project/pact-python/

Hi, Did anyone approach Pact+DAPR pub/sub contract testing? Everything in this setup is twisted and I am confused how to approach it in the best way. The consumer is invoked through REST API by DAPR sidecar, while producer is invoking REST API of Dapr sidecar. Anyone can share experience?

Hello everyone! In my company, we are trying to understand the boundaries of the contract tests written in pacts. Of course, we've read this article (and watched the awesome video placed there), but we still have some doubts. Currently, our consumer messaging tests are very naive, we just return a JSON created from a DTO object created by hand -> so in fact, any of the layers of our application is not called. • Is there any pattern that could we use to make our tests a little more intelligent to verify whether the message is correctly handled by our service? • Is it a good idea to mock repositories and verify if an object that we get asynchronously is correctly written to a repository? • If not, why :)? Could you please share any articles/presentations related to such topic?

Hi, anyone using webhook with event contract_requiring_verification_published? There are lots of webhook triggered by this event type. It doesn't look right to me. As I expected this to trigger only provider main branch pipeline with latest version. Somehow it trigger all version under provider main branch, and I end up getting a bigger list of trigger pipeline day by day. Something definitely wrong here, anyone can help point me to the right direction?

Hi. Is it considered best practice to use a dynamic date for includeWipPactsSince? The documentation only shows examples with static dates like "2020-10-31". But will it be a problem say a year from now? That it could fetch WIP pacts since that date. Rather, is it better to just configure it as something like new Date().minusDays(7) for WIP pacts within a week.

<!here> Catch @Alejandro Pena & José Maciñeiras live now on YouTube, wrapping up Pactober with their introduction of a new tool for automating the body generation for Pact interactions in Java https://pact.io/pactober/2023/pact-jvm-dsl

Good morning! When I run the verifier locally via the Python API, what does the variable

PACT_URL

in this snippet represent?

Another question - hope this is the right place to ask. When I attempt to run pact tests locally (local pact file, local provider, pact mock host), I get an error that my request headers are different than previously. How does it come to this conclusion? Am I actually pointing it to my real broker somehow?

Hi. I am trying to upgrade our flow from tags to contract selectors and I stumbled upon chicken or an egg problem that was already there but I hoped that contract selector will solve it. I have working process for https://docs.pact.io/pact_nirvana/step_7 but how should I handle the first time when I publish new contract between two services? I have two repositories with branch

feature/abc

containing first contract between two services. I cannot merge consumer as it will fail

can-i-deploy

(provider supporting contract is not yet deployed on target environment). If I merge provider it won't record deployment as contract selector looks for main branch and release/deployed pacts. Pending pacts and WIP pacts won't help as this contract was already verified from branch so selector returns no contracts when run from main branch after merge, it won't fail build just because the annotation

@IgnoreNoPactsToVerify

. It seems that I have to manually verify against branch

feature/abc

, rerun deployment pipeline to register deployment and then I can start merging consumer. Is there any valid way to do that without the need to manually running verification or hardcoding branch

feature/abc

in the selector? Did I miss something?

hi all, I'm afraid my searching skills were not enough and I have to ask here. I'm considering Pact for Messages with the diagram I attached as our example of event-driven architecture. Imagine that we want to introduce a change in the

OrderRequested

message. In Pact, the code of the providers has to be merged first, and the code of the consumer after it. This makes sense to me but I'm struggling to see if it also makes sense for event-driven architecture. If we merged

Basket Service

first, there would be errors (or at least a queue of messages) because

Payment Service

, the consumer, doesn't understand the new changes introduced by the provider. For me, in this case, we would introduce some code to the consumer to be able to handle both the old and new versions of the

OrderRequested

message, and after that, we'd merge the code to

Basket Service

. How would you handle this scenario, please? Was it clear enough? Thanks!

Question related to the pactfile generated from tools like wiremock or mountebank : What is the usecase for such pactfile where matchers are not present at the source itself ? Is such pactfile only targeted for bi-directional contract verification supported by pactflow ? as they are not so useful to run as tests for traditional pact provider verification. If yes, then what is the difference in validating the stubs/mocks against the

OpenAPISpecification

for the consumers and validating the tests request/response against the

OpenAPISpecification

using simple library like this ? doesn't it gives the same schema-based validation same as the bidirectional contract verification ?

Hey - Question regarding the

pactfoundation/pact-cli:latest

How to run a pact message provider test by using the docker image ? There is a verifier command but no mention about how to run message provider pact test. Any pointers please ?

Hi, i just had a meeting where i was trying to explain that contract testing should only test the contract; ideally stub out (aka mock) the downstream software components and services as thats the concern of other tests. Of course there may be exceptions to this ideal (very small service/laziness/some other reason). The issue I was advising against, was trying to drive your apps berhavioural testing et al (say in the provider - starting at the top layer) as giant 'integration' tests through pact - thats not the responsibility of PACT and you can get very complicated flakey pacts and harder to maintain code (i.e. you can get yourself into trouble). I do recall a video by @Beth (pactflow.io/Pact Broker/pact-ruby) on youtube, that i believe covered this subject, not sure if anyone can link me it (i cant find the right one) ? or comment; am I wrong - is my advice poor ? Thanks!! 😉

hi @Matt (pactflow.io / pact-js / pact-go) do we have a query all providers by a consumer from the broker?

Hello, I have written a POC for Java Pactflow. One service is sending a message to another service and I would like to compare the messages. I am trying to do the same in Python. Is that possible in Python? here is my Java code: My Consumer class:

@ExtendWith(PactConsumerTestExt.class)
@PactTestFor(providerName = "ProviderCDCT",
        providerType = ProviderType.ASYNCH,
        pactVersion = PactSpecVersion.V3)
public class ConsumerContractTest {

    @Pact(consumer = "ConsumerCDCT")
    public MessagePact createMyPact(MessagePactBuilder builder) {
        Map<String, String> metadata = new HashMap<>();
        metadata.put("content-type", "application/json");

        DslPart sqsBody = new PactDslJsonBody()
                .stringType("messageId", "1")
                .stringValue("receiptHandle", "testRadek")
                .stringType("eventSource", "aws:sqs")
                .stringType("awsRegion", "us-east-1")
                .object("messageAttributes")
                .closeObject();

        return builder
                .expectsToReceive("a message with information")
                .withMetadata(metadata)
                .withContent(sqsBody)
                .toPact();

    }

    //This method is needed to create the pact files.
    @Test
    @PactTestFor(pactMethod = "createMyPact")
    public void testMyMethod(List<Message> messages) {
    }
}

My Provider Class:

@PactBroker(scheme = "https", host = "",
        authentication = @PactBrokerAuth(token = ""))
@Provider("ProviderCDCT")
@Consumer("ConsumerCDCT")
public class ProviderContractTest {
    private final ObjectMapper objectMapper = new ObjectMapper();

    @BeforeEach
    public void setUp(PactVerificationContext context) {
        context.setTarget(new MessageTestTarget(List.of("Provider")));
        System.setProperty("pact.verifier.publishResults",
                System.getenv("PACT_BROKER_PUBLISH_VERIFICATION_RESULTS") == null ? "true" : "true");
    }

    @TestTemplate
    @ExtendWith(PactVerificationInvocationContextProvider.class)
    void testTemplate(PactVerificationContext context) {
        context.verifyInteraction();
    }

    @PactVerifyProvider("a message with information")
    public MessageAndMetadata providerMessage() throws Exception {
        Map<String, String> metadata = new HashMap<>();
        metadata.put("content-type", "application/json");

        Map<String, Object> mainMessage = new HashMap<>();
        mainMessage.put("messageId", "");
        mainMessage.put("receiptHandle", "testRadek");
        mainMessage.put("eventSource", "aws:sqs");
        mainMessage.put("awsRegion", "us-east-1");
        mainMessage.put("messageAttributes", new HashMap<>());

        return new MessageAndMetadata(objectMapper.writeValueAsBytes(mainMessage), metadata);
    }
}

Hi all, I'm working on contract testing for a message broker scenario. My consumer has a modular structure with a handler->transformer (simplified). I've realized that running the test with a mocked transformer dependency won't cover some cases from the transformer. Alternatively, I'm thinking about extracting necessary assertions directly into the contract test. This way, I can manually check the contract's compliance with specific aspects of the application's functionality, including those from the transformer and other modules. However, this approach might lead to a disconnection from the actual app logic and could be prone to errors if the logic changes. Which approach is more aligned with the best practices of contract testing using Pact?

Hey everyone! I have a consumer contract test where I want to reflect that we expect an array with at least one element containing a boolean with value true. From the forums and docs, I understand I can use "*arrayContaining*" to reflect this. I have two questions regarding this: • I actually want to say we want only one element to have this boolean true, all other elements must have it as false. But I believe this is not possible to reflect? • The bigger problem, is that I don't know how to use this "arrayContaining" in the DSL, while it seems to only be possible on arrays contained within an object. The event structure we are consuming is however containing an array on top level, so no parent object. And it is this array which has this requirement. I hope my problem is understandable, and that someone could help out 🙂 I am using Pact JVM btw

I have a question. I have heard that contract testing focuses on testing the arrows in your microservice architecture diagram (the communication between your services). I can see how it works for direct invocation API calls. But can it work for pub/sub? I have been searching for anything explaining how that would work, and I have found nothing. Is it correct to say that pub/sub is a form of service communication that cannot be tested as a contract?

blobwave Hey all, Want to know whats been happening in the world of Pact this month? If so, check out the latest Pact blog https://docs.pact.io/blog/2023/11/22/pact-open-source-update-nov-2023 You can find out what @Tien Vo and @Joshua Ellis have been up to in Pact PHP and Pact Python, and how working with the Pact compatibility suite is leading to improvements in the Pact Core. You can read about #pact-nestjs author @Omer Morad latest project Automock and if you are a Bazel fan, you might want to check out @Olivier Picaud’s contributions. As always thanks to everyone for their hard work, and if you fancy getting featured in a Pact blog, why not reach out. Have a great rest of your day (or start, depending on where you are in the world!)

WE removed some fields from provider aPI , I changed my consumer tests and that passes at consumer side but provider is still failing

Anybody know much about DAPR? https://pact-foundation.slack.com/archives/C9UTHV2AD/p1701146304630789

Hey folks, a somewhat newbie question: What if the same value is used in multiple places across request and response? An example: we send a request to create a resource with id x123. The response can be smth like:

{
  "id": "x123",
  "url": "<https://example.com/api/v1/entity/x123>",
  "fullName": "entity-x123"
}

How can this be modeled in pact? If each field just specifies a regex like

x\d+

or smth, and

x123

as an example, is this enough?

webinar Webinar webinar Show and Tell: Discover the Benefits SmartBear’s Open Source Tools Bring to the SDLC Tuesday, 5th December 2023 at 3:00PM GMT / 10:00AM EDT Join Yousaf and Jo from the DevRel team at SmartBear to hear about the Open Source tools and communities at SmartBear! Expect to learn how to harness Swagger, SoapUI, Pact, and more to ace API development. 🛠️💻 👉🏻 Register to join