REST API Automated Testing


Keywords
rest, api, graphql, automated, continuous, test, flow, yaml, postman, continuous-testing, nodejs, rest-api, test-automation, testing, typescript, workflow
License
MIT
Install
npm install ply-ct@3.0.7-beta.2

Documentation

ply-logo

API Automated Testing

Installation

npm install @ply-ct/ply --save-dev

Or, to run anywhere:

npm install -g @ply-ct/ply

Usage

Ply API testing starts with a YAML file containing requests. Here's a GET request to retrieve topics for the ply-demo repository using GitHub API v3:

repositoryTopics:
  url: 'https://api.github.com/repos/ply-ct/ply-demo/topics'
  method: GET
  headers:
    Accept: application/vnd.github.mercy-preview+json

Submit a request

Suppose you save this in a file named "github.ply.yaml". Then you can submit the repositoryTopics request from a command line by typing:

ply -s github.ply.yaml

The -s argument tells Ply not to verify the response (-s is short for --submit, meaning submit an ad hoc request and don't bother with verification).

Verify response

If you run without -s you'll get an error saying, "Expected result file not found". Ply verification works by comparing expected vs actual. So a complete test requires an expected result file. Run again with --create, and the expected result file will be created from the actual response.

ply --create github.ply.yaml

Output looks like this:

Request 'repositoryTopics' submitted at 4/11/2022, 11:19:46:292
Creating expected result: ./results/expected/github.yaml
Request 'repositoryTopics' PASSED in 332 ms

Overall Results: {"Passed":1,"Failed":0,"Errored":0,"Pending":0,"Submitted":0}
Overall Time: 373 ms

During execution Ply submits the request and writes actual result file "./results/actual/github.yaml" based on the response. Because of --create, Ply then copies the actual result over expected result file "./results/expected/github.yaml" before comparing. This test naturally passes since the results are identical.

Expected results

Auto-creating an expected result provides a good starting point. But if you run the request again (without creating), it'll fail:

ply github.ply.yaml
Request 'repositoryTopics' submitted at 4/11/2022, 11:20:44:478
Request 'repositoryTopics' FAILED in 372 ms: Results differ from line 24
24
-       x-github-request-id: E8C2:3201:51B1B:D9917:62546332
+       x-github-request-id: E8C3:7857:386D8:7DDEE:6254636C
===
26
-       x-ratelimit-remaining: '56'
+       x-ratelimit-remaining: '55'
===
29
-       x-ratelimit-used: '4'
+       x-ratelimit-used: '5'
===

But looking at "./results/expected/github.yaml", you'll notice that it includes many response headers that are not of interest for testing purposes. Here's a cleaned-up version of similar expected results from ply-demo:

repositoryTopics:
  request:
    url: https://api.github.com/repos/${github.organization}/${github.repository}/topics
    method: GET
    headers:
      Accept: application/vnd.github.mercy-preview+json
  response:
    status:
      code: 200
      message: OK
    headers:
      content-type: application/json; charset=utf-8
    body: |-
      {
        "names": [
          "rest-api",
          "testing",
          "ply",
          "example-project",
          "graphql",
          "typescript",
          "workflow"
        ]
      }

The subset of response headers included in expected results YAML are those we care about for comparison. In this test, body content is our main concern.

Expressions

Something else about this example that may be noticed by sharp-eyed observers: our request URL contains placeholders like ${github.organization}. Ply supports JavaScript template literal syntax for substituting dynamic values in both requests and results. Values come from JSON files and/or environment variables, as described in the docs under Values.

Even more powerfully, your multi-request suites can embed expressions that reference runtime values from previous responses. For instance, the URL or body of a subsequent request in our github.ply.yaml file could have something like this:

${@repositoryTopics.response.body.names[0]}

which uses the special @ character to reference the first topic name from above (resolving to 'rest-api'). This enables you to string together sequential requests that each depend on response output from preceding ones. Check out the Results topic for details and examples.

Flows

If you have Visual Studio Code with the Ply extension, you can graphically chain multiple requests into a workflow. See the Ply flows documentation for details.

Flows can include custom TypeScript steps to perform complex interactions and update runtime values.

Running a flow from the command line is similar to running a request suite:

ply test/flows/movies-api.flow

GraphQL

Body content in request YAML can be any text payload (typically JSON). GraphQL syntax is also supported, as in this example which queries the GitHub GraphQL API for ply-demo repository topics:

repositoryTopicsQuery:
  url: 'https://api.github.com/graphql'
  method: POST
  headers:
    Authorization: Bearer ${githubToken}
    Content-Type: application/json
    User-Agent: ${github.organization}
  body: |-
    query {
      repository(owner: "${github.organization}", name: "${github.repository}") {
        repositoryTopics(first: 10) {
          edges {
            node {
              topic {
                name
              }
            }
          }
        }
      }
    }

Documentation

Guide

https://ply-ct.org/ply/topics/requests

CLI

https://ply-ct.org/ply/topics/cli

API

https://ply-ct.org/ply/api

Demo Project

https://github.com/ply-ct/ply-demo/

VS Code Extension

https://marketplace.visualstudio.com/items?itemName=ply-ct.vscode-ply