1.0.1 • Published 2 years ago

clet v1.0.1

Weekly downloads
-
License
MIT
Repository
github
Last release
2 years ago

CLET - Command Line E2E Testing

CLET - Command Line E2E Testing

NPM Version NPM Download NPM Quality GitHub Actions CI Coverage

CLET aims to make end-to-end testing for command-line apps as simple as possible.

  • Powerful, stop writing util functions yourself.
  • Simplified, every API is chainable.
  • Modern, ESM first, but not leaving commonjs behind.

Inspired by coffee and nixt.

How it looks

Boilerplate && prompts

import { runner, KEYS } from 'clet';

it('should works with boilerplate', async () => {
  await runner()
    .cwd(tmpDir, { init: true })
    .spawn('npm init')
    .stdin(/name:/, 'example') // wait for stdout, then respond
    .stdin(/version:/, new Array(9).fill(KEYS.ENTER))
    .stdout(/"name": "example"/) // validate stdout
    .notStderr(/npm ERR/)
    .file('package.json', { name: 'example', version: '1.0.0' }) // validate file
});

Command line apps

import { runner } from 'clet';

it('should works with command-line apps', async () => {
  const baseDir = path.resolve('test/fixtures/example');
  await runner()
    .cwd(baseDir)
    .fork('bin/cli.js', [ '--name=test' ], { execArgv: [ '--no-deprecation' ] })
    .stdout('this is example bin')
    .stdout(`cwd=${baseDir}`)
    .stdout(/argv=\["--name=\w+"\]/)
    .stderr(/this is a warning/);
});

Build tools && Long-run servers

import { runner } from 'clet';
import request from 'supertest';

it('should works with long-run apps', async () => {
  await runner()
    .cwd('test/fixtures/server')
    .fork('bin/cli.js')
    .wait('stdout', /server started/)
    .expect(async () => {
      // using supertest
      return request('http://localhost:3000')
        .get('/')
        .query({ name: 'tz' })
        .expect(200)
        .expect('hi, tz');
    })
    .kill(); // long-run server will not auto exit, so kill it manually after test
});

Work with CommonJS

describe('test/commonjs.test.cjs', () => {
  it('should support spawn', async () => {
    const { runner } = await import('clet');
    await runner()
      .spawn('npm -v')
      .log('result.stdout')
      .stdout(/\d+\.\d+\.\d+/);
  });
});

Installation

$ npm i --save clet

Command

fork(cmd, args, opts)

Execute a Node.js script as a child process.

it('should fork', async () => {
  await runner()
    .cwd(fixtures)
    .fork('example.js', [ '--name=test' ], { execArgv: [ '--no-deprecation' ] })
    .stdout('this is example bin')
    .stdout(/argv=\["--name=\w+"\]/)
    .stdout(/execArgv=\["--no-deprecation"\]/)
    .stderr(/this is a warning/);
});

Options:

  • timeout: {Number} - will kill after timeout.
  • execArgv: {Array} - pass to child process's execArgv, default to process.execArgv.
  • cwd: {String} - working directory, prefer to use .cwd() instead of this.
  • env: {Object} - prefer to use .env() instead of this.
  • extendEnv: {Boolean} - whether extend process.env, default to true.
  • more detail: https://github.com/sindresorhus/execa#options

spawn(cmd, args, opts)

Execute a shell script as a child process.

it('should support spawn', async () => {
  await runner()
    .spawn('node -v')
    .stdout(/v\d+\.\d+\.\d+/);
});

cwd(dir, opts)

Change the current working directory.

Notice: it affects the relative path in fork(), file(), mkdir(), etc.

it('support cwd()', async () => {
  await runner()
    .cwd(targetDir)
    .fork(cliPath);
});

Support options:

  • init: delete and create the directory before tests.
  • clean: delete the directory after tests.

Use trash instead of fs.rm to prevent misoperation.

it('support cwd() with opts', async () => {
  await runner()
    .cwd(targetDir, { init: true, clean: true })
    .fork(cliPath)
    .notFile('should-delete.md')
    .file('test.md', /# test/);
});

env(key, value)

Set environment variables.

Notice: if you don't want to extend the environment variables, set opts.extendEnv to false.

it('support env', async () => {
  await runner()
    .env('DEBUG', 'CLI')
    .fork('./example.js', [], { extendEnv: false });
});

timeout(ms)

Set a timeout. Your application would receive SIGTERM and SIGKILL in sequent order.

it('support timeout', async () => {
  await runner()
    .timeout(5000)
    .fork('./example.js');
});

wait(type, expected)

Wait for your expectations to pass. It's useful for testing long-run apps such as build tools or http servers.

  • type: {String} - support message / stdout / stderr / close
  • expected: {String|RegExp|Object|Function}
    • {String}: check whether the specified string is included
    • {RegExp}: check whether it matches the specified regexp
    • {Object}: check whether it partially includes the specified JSON
    • {Function}: check whether it passes the specified function

Notice: don't forgot to wait('end') or kill() later.

it('should wait', async () => {
  await runner()
    .fork('./wait.js')
    .wait('stdout', /server started/)
    // .wait('message', { action: 'egg-ready' }) // ipc message
    .file('logs/web.log')
    .kill();
});

kill()

Kill the child process. It's useful for manually ending long-run apps after validation.

Notice: when kill, exit code may be undefined if the command doesn't hook on signal event.

it('should kill() manually after test server', async () => {
  await runner()
    .cwd(fixtures)
    .fork('server.js')
    .wait('stdout', /server started/)
    .kill();
});

stdin(expected, respond)

Responde to a prompt input.

  • expected: {String|RegExp} - test if stdout includes a string or matches regexp.
  • respond: {String|Array} - content to respond. CLET would write each with a delay if an array is set.

You could use KEYS.UP / KEYS.DOWN to respond to a prompt that has multiple choices.

import { runner, KEYS } from 'clet';

it('should support stdin respond', async () => {
  await runner()
    .cwd(fixtures)
    .fork('./prompt.js')
    .stdin(/Name:/, 'tz')
    .stdin(/Email:/, 'tz@eggjs.com')
    .stdin(/Gender:/, [ KEYS.DOWN + KEYS.DOWN ])
    .stdout(/Author: tz <tz@eggjs.com>/)
    .stdout(/Gender: unknown/)
    .code(0);
});

Tips: type ENTER repeatedly if needed

it('should works with boilerplate', async () => {
  await runner()
    .cwd(tmpDir, { init: true })
    .spawn('npm init')
    .stdin(/name:/, 'example')
    .stdin(/version:/, new Array(9).fill(KEYS.ENTER)) // don't care about others, just enter
    .stdout(/"name": "example"/)
    .notStderr(/npm ERR/)
    .file('package.json', { name: 'example', version: '1.0.0' })
});

Validator

stdout(expected)

Validate stdout, support regexp and string.includes.

it('should support stdout()', async () => {
  await runner()
    .spawn('node -v')
    .stdout(/v\d+\.\d+\.\d+/) // regexp match
    .stdout(process.version)  // string includes;
});

notStdout(unexpected)

The opposite of stdout().

stderr(expected)

Validate stdout, support regexp and string.includes.

it('should support stderr()', async () => {
  await runner()
    .cwd(fixtures)
    .fork('example.js')
    .stderr(/a warning/)
    .stderr('this is a warning');
});

notStderr(unexpected)

The opposite of stderr().

code(n)

Validate child process exit code.

No need to explicitly check if the process exits successfully, use code(n) only if you want to check other exit codes.

Notice: when a process is killed, exit code may be undefined if you don't hook on signal events.

it('should support code()', async () => {
  await runner()
    .spawn('node --unknown-argv')
    .code(1);
});

file(filePath, expected)

Validate the file.

  • file(filePath): check whether the file exists
  • file(filePath, 'some string'): check whether the file content includes the specified string
  • file(filePath, /some regexp/): checke whether the file content matches regexp
  • file(filePath, {}): check whether the file content partially includes the specified JSON
it('should support file()', async () => {
  await runner()
    .cwd(tmpDir, { init: true })
    .spawn('npm init -y')
    .file('package.json')
    .file('package.json', /"name":/)
    .file('package.json', { name: 'example', config: { port: 8080 } });
});

notFile(filePath, unexpected)

The opposite of file().

Notice: .notFile('not-exist.md', 'abc') will throw because the file is not existing.

expect(fn)

Validate with a custom function.

it('should support expect()', async () => {
  await runner()
    .spawn('node -v')
    .expect(ctx => {
      const { assert, result } = ctx;
      assert.match(result.stdout, /v\d+\.\d+\.\d+/);
    });
});

Operation

log(format, ...keys)

Print log for debugging. key supports dot path such as result.stdout.

it('should support log()', async () => {
  await runner()
    .spawn('node -v')
    .log('result: %j', 'result')
    .log('result.stdout')
    .stdout(/v\d+\.\d+\.\d+/);
});

tap(fn)

Tap a method to the chain sequence.

it('should support tap()', async () => {
  await runner()
    .spawn('node -v')
    .tap(async ({ result, assert}) => {
      assert(result.stdout, /v\d+\.\d+\.\d+/);
    });
});

sleep(ms)

it('should support sleep()', async () => {
  await runner()
    .fork(cliPath)
    .sleep(2000)
    .log('result.stdout');
});

shell(cmd, args, opts)

Run a shell script. For example, run npm install after boilerplate init.

it('should support shell', async () => {
  await runner()
    .cwd(tmpDir, { init: true })
    .spawn('npm init -y')
    .file('package.json', { name: 'shell', version: '1.0.0' })
    .shell('npm version minor --no-git-tag-version', { reject: false })
    .file('package.json', { version: '1.1.0' });
});

The output log could validate by stdout() and stderr() by default, if you don't want this, just pass { collectLog: false }.

mkdir(path)

Act like mkdir -p.

it('should support mkdir', async () => {
  await runner()
    .cwd(tmpDir, { init: true })
    .mkdir('a/b')
    .file('a/b')
    .spawn('npm -v');
});

rm(path)

Move a file or a folder to trash (instead of permanently delete it). It doesn't throw if the file or the folder doesn't exist.

it('should support rm', async () => {
  await runner()
    .cwd(tmpDir, { init: true })
    .mkdir('a/b')
    .rm('a/b')
    .notFile('a/b')
    .spawn('npm -v');
});

writeFile(filePath, content)

Write content to a file, support JSON and PlainText.

it('should support writeFile', async () => {
  await runner()
    .cwd(tmpDir, { init: true })
    .writeFile('test.json', { name: 'writeFile' })
    .writeFile('test.md', 'this is a test')
    .file('test.json', /"name": "writeFile"/)
    .file('test.md', /this is a test/)
    .spawn('npm -v');
});

Context

/**
 * @typedef Context
 *
 * @property {Object} result - child process execute result
 * @property {String} result.stdout - child process stdout
 * @property {String} result.stderr - child process stderr
 * @property {Number} result.code - child process exit code
 *
 * @property {execa.ExecaChildProcess} proc - child process instance
 * @property {TestRunner} instance - runner instance
 * @property {String} cwd - child process current workspace directory
 *
 * @property {Object} assert - assert helper
 * @property {Object} utils -  utils helper
 * @property {Object} logger - built-in logger
 */

assert

Extend Node.js built-in assert with some powerful assertions.

/**
 * assert `actual` matches `expected`
 *  - when `expected` is regexp, assert by `RegExp.test`
 *  - when `expected` is json, assert by `lodash.isMatch`
 *  - when `expected` is string, assert by `String.includes`
 *
 * @param {String|Object} actual - actual string
 * @param {String|RegExp|Object} expected - rule to validate
 */
function matchRule(actual, expected) {}

/**
 * assert `actual` does not match `expected`
 *  - when `expected` is regexp, assert by `RegExp.test`
 *  - when `expected` is json, assert by `lodash.isMatch`
 *  - when `expected` is string, assert by `String.includes`
 *
 * @param {String|Object} actual - actual string
 * @param {String|RegExp|Object} expected - rule to validate
 */
function doesNotMatchRule(actual, expected) {}

/**
 * validate file
 *
 *  - `matchFile('/path/to/file')`: check whether the file exists
 *  - `matchFile('/path/to/file', /\w+/)`: check whether the file content matches regexp
 *  - `matchFile('/path/to/file', 'usage')`: check whether the file content includes the specified string
 *  - `matchFile('/path/to/file', { version: '1.0.0' })`: checke whether the file content partially includes the specified JSON
 *
 * @param {String} filePath - target path to validate, could be relative path
 * @param {String|RegExp|Object} [expected] - rule to validate
 * @throws {AssertionError}
 */
async function matchFile(filePath, expected) {}

/**
 * validate file with opposite rule
 *
 *  - `doesNotMatchFile('/path/to/file')`: check whether the file exists
 *  - `doesNotMatchFile('/path/to/file', /\w+/)`: check whether the file content does not match regex
 *  - `doesNotMatchFile('/path/to/file', 'usage')`: check whether the file content does not include the specified string
 *  - `doesNotMatchFile('/path/to/file', { version: '1.0.0' })`: checke whether the file content does not partially include the specified JSON
 *
 * @param {String} filePath - target path to validate, could be relative path
 * @param {String|RegExp|Object} [expected] - rule to validate
 * @throws {AssertionError}
 */
async function doesNotMatchFile(filePath, expected) {}

debug(level)

Set level of logger.

import { runner, LogLevel } from 'clet';

it('should debug(level)', async () => {
  await runner()
    .debug(LogLevel.DEBUG)
    // .debug('DEBUG')
    .spawn('npm -v');
});

Extendable

use(fn)

Middleware, always run before child process chains.

// middleware.pre -> before -> fork -> running -> after -> end -> middleware.post -> cleanup

it('should support middleware', async () => {
  await runner()
    .use(async (ctx, next) => {
      // pre
      await utils.rm(dir);
      await utils.mkdir(dir);

      await next();

      // post
      await utils.rm(dir);
    })
    .spawn('npm -v');
});

register(Function|Object)

Register your custom APIs.

it('should register(fn)', async () => {
  await runner()
    .register(({ ctx }) => {
      ctx.cache = {};
      cache = function(key, value) {
        this.ctx.cache[key] = value;
        return this;
      };
    })
    .cache('a', 'b')
    .tap(ctx => {
      console.log(ctx.cache);
    })
    .spawn('node', [ '-v' ]);
});

Known Issues

Help Wanted

  • when answer prompt with inquirer or enquirer, stdout will recieve duplicate output.
  • when print child error log with .error(), the log order maybe in disorder.

License

MIT