playwright-performance v1.0.2

playwright-performance

To ensure that your application is responsive and performing optimally, it is important to monitor the apparent response time of key procedures. Apparent response time is defined as the time it takes for a procedure to complete and make the application available to the user. With this Playwright plugin, you can easily add performance analysis to any flow in your tests, whether it's a pure UI, API, or a combination of both. This plugin provides a simple and efficient way to measure the response times of various procedures and identify potential bottlenecks in your application. With this information, you can make informed decisions about optimizations and improvements to enhance the overall performance of your application. Read more here.

You can install this module as a dev-dependency using the following command:

npm install playwright-performance --save-dev

Import playwright-peformance in your test file as follows:

import type { PerformanceOptions, PlaywrightPerformance, PerformanceWorker } from "playwright-performance";
import { playwrightPerformance } from "playwright-performance";

To use playwright-performance, simply import the playwright-performance object and types, and then extend your test object using test.extend<>(). This will include the performance functionality in your test. No further setup is required. Here's an example:

import {test as base} from '@playwright/test';
import type { PerformanceOptions, PlaywrightPerformance, PerformanceWorker } from "playwright-performance";
import { playwrightPerformance } from "playwright-performance";

const test = base.extend<PlaywrightPerformance, PerformanceOptions & PerformanceWorker>({
  performance: playwrightPerformance.performance,
  performanceOptions: [{
  }, { scope: 'worker' }],
  worker: [playwrightPerformance.worker, { scope: 'worker', auto: true }]
});

test('startup performance', async ({ page, performance }) => {
    performance.sampleStart("GH-startup");
    await page.goto('http://github.com/');
    performance.sampleEnd("GH-startup");

    performance.sampleStart("SF-startup");
    await page.goto('https://sourceforge.net/');
    performance.sampleEnd("SF-startup");
  });

• It is advisable to define the extended test object in a separate, reusable test-base file.

You can also get the time span for a single sample inside a test:

it("should test github startup performance", () => {
            performance.sampleStart("Startup");
            browser.url("https://github.com/");
            performance.sampleEnd("Startup");

            expect(performance.getSampleTime("Startup")).to.be.at.most(1000);         
        });

You can override the default options values in the performanceOptions fixture object as follows:

const test = base.extend<PlaywrightPerformance, PerformanceOptions & PerformanceWorker>({
  performance: playwrightPerformance.performance,
  performanceOptions: [{
    disableAppendToExistingFile: false,
    performanceResultsFileName: "performance-results",
    dropResultsFromFailedTest: false,
    performanceResultsDirectory: "performance-results-dir",
    analyzeByBrowser: false,
    suppressConsoleResults: false,
  }, { scope: 'worker' }],
  worker: [playwrightPerformance.worker, { scope: 'worker', auto: true }]
});

When set to true, new test runs will start fresh and overwrite any existing performance data. When set to false (default), performance data will be added to the existing data.

You can set the default results file name (performance-results). A newly created results file normally overwrites the old file. If you want to keep old files, it is recommended to add a timestamp to the file name. For example:

...
performanceResultsFileName: `performance-results_${new Date().getHours()}`
...

Default is false. When the value is set to true, performance analysis from failed tests would be excluded.

...
performanceResultsFileName: "results-dir/performance-total-results"
...

Default is false. If true, the performance data would be grouped also by the browser type.

Default is false. If true, the performance results won't be printed to the terminal log.

A new directory named performance-results (or a different specified name) is created inside your project's root folder. Once all the tests are completed, two files are created inside the performance-results directory: performance-results.json and performance-results.csv. The analyzed data includes average time, standard error of mean (SEM), number of samples, minimum value, maximum value, earliest time, and latest time. The results table is also printed to the terminal log.

Typescript is supported for this plugin.

For any questions or suggestions contact me at: tzur.paldi@outlook.com