Scripts for Scaleway CLI development

Script	Description
`build.sh`	Build the CLI binary for all supported platform (Linux, Darwin, Windows)
`install.sh`	Build and install the CLI binary (MacOSX support only)
`lint.sh`	Run the linter golangci-lint on the codebase
`run-tests.sh`	Execute the test suite for the CLI v2

Requirements

Some of those scripts have external requirements such as:

build.sh

$ ./scripts/build.sh

install.sh

./scripts/install.sh

lint.sh

$ ./scripts/lint.sh -h
Usage:
  ./scripts/lint.sh [OPTIONS]

Options:
  -w, --write
    Fix found issues (if it's supported by the linter).
  --list
    List current linters configuration.
  -h, --help
    Display this help.

run-tests.sh

CLI testing uses cassettes and golden files to run a part of its test suite.

Cassettes

Cassettes record a set of interactions between the CLI and the Scaleway API. A given test would define a scenario that would be described as a YAML file that contains all those interactions. For instance:

version: 1
interactions:
- request:
    body: ""
    form: {}
    headers:
      User-Agent:
      - scaleway-sdk-go/v1.0.0-beta.5+dev (go1.13.6; darwin; amd64) cli-e2e-test
    url: https://api.scaleway.com/
    method: GET
  response:
    body: '{"Name":"Scaleway Api","Description":"Welcome to the Scaleway public API!","Version":"v0.0.140","ProtobufVersion":"a0be3c28","DocumentationUrl":"https://www.scaleway.com/en/developers/"}

Golden

A golden file stores the output of a command and is used as a reference by the test as the expected output. Those files should be updated any time the output changes for reasons such as: API changes, new format, etc. The output is stored as a separate file rather than as a string literal inside the test code. So when the test is executed, it will read in the file and compare it to the output produced by the system under test.

How should I write a test?

Let's suppose you want to write the test for the list server in the bare-metal namespace. You would do the following:

Write the test, here we will put it in internal/namespaces/baremetal/v1alpha1/baremetal_cli_test.go

func Test_ListServer(t *testing.T) {
    t.Run("Simple", core.Test(&core.TestConfig{
        Commands: GetCommands(),
        Cmd:      "scw baremetal server list",
        Check: core.TestCheckCombine(
            core.TestCheckGolden(),
            core.TestCheckExitCode(0),
        ),
    }))
}

Update the cassette and golden for this particular test:

./scripts/run-tests.sh --update --run Test_ListServer

Beware! Updating the cassette and golden launch real calls on the API! Be sure your tests are not modifying your cloud resources. Also, a valid Scaleway token is required in your environment in order to record cassettes.

Help command

$ ./scripts/run-tests.sh -h
Usage:
  ./scripts/run-tests.sh [OPTIONS]

Options:
  -r, --run <regex>
    Run a specific test or set of tests matching the given regex. Similar to the '-run' Go test flag.
  -u, --update
    Update goldens and record cassettes during integration tests.
  -g, --update-goldens
    Update goldens during integration tests.
  -c, --update-cassettes
    Record cassettes during integration tests. Warning: a valid Scaleway token is required in your environment in order to record cassettes.
  -D, --debug
    Enable CLI debug mode.
  -h, --help
    Display this help.