Also shared here: https://github.com/tomakehurst/wiremock/issues/625
I'm writing an integration test to verify that my application that interacts with a REST API handles unsuccessful requests appropriately. To do this, I'm wanting to simulate a scenario where a GET requests is made twice to a HTTP endpoint. First time, request is not successful with a response status code of 500; second time, request is successful with a response status code of 200.
Consider the example below:
@Rule
public WireMockRule wireMockRule
= new WireMockRule(wireMockConfig().dynamicPort().dynamicHttpsPort());
@Test
public void testRetryScenario(){
// First StubMapping
stubFor(get(urlEqualTo("/my/resource"))
.withHeader("Accept", equalTo("text/xml"))
.willReturn(aResponse()
.withStatus(500) // request unsuccessful with status code 500
.withHeader("Content-Type", "text/xml")
.withBody("<response>Some content</response>")));
// Second StubMapping
stubFor(get(urlEqualTo("/my/resource"))
.withHeader("Accept", equalTo("text/xml"))
.willReturn(aResponse()
.withStatus(200) // request successful with status code 200
.withHeader("Content-Type", "text/xml")
.withBody("<response>Some content</response>")));
//Method under test that makes calls to endpoint
doSomething();
Thread.sleep(5000);
//Verify GET request was made again after first attempt
verify(exactly(2), getRequestedFor(urlEqualTo("/my/resource")));
}
Is there a way to avoid the second StubMapping from overriding the first -- to make sure that the first time doSomething()
makes a request, a response with status code 500 is returned, and the second time, a different response with status code 200 is returned?
This is what the Scenarios feature is for.
You'll need to put both stubs into a scenario (i.e. same scenario name), make the first stub trigger a transition to a new state, then make the second stub contingent on the scenario being in the second state and the first stub contingent on the scenario being in the STARTED
state.