Robot Framework Browser library powered by Playwright. Moving browser automation to year 2023!
Aiming for 🚀 speed, ✅ reliability and 🔬 visibility.
See keyword documentation and web page for more details.
Only Python 3.9 or newer is supported. From Node side 18, 20 and 22 LTS versions are supported.
- Install node.js e.g. from https://nodejs.org/en/download/
- Update pip
pip install -U pipto ensure latest version is used - Install robotframework-browser from the commandline:
pip install robotframework-browser - Install the node dependencies: run
rfbrowser initin your shell
- if
rfbrowseris not found, trypython -m Browser.entry init
Please note that by default Chromium, Firefox and WebKit browser are installed, even those would be already
installed in the system. The installation size depends on the operating system, but usually is +700Mb.
It is possible to skip browser binaries installation with rfbrowser init --skip-browsers command, but then user
is responsible for browser binary installation. It is possible to install only selected browser binaries by adding
chromium, firefox or webkit as arguments to init command. Example rfbrowser init firefox would install
only Firefox binaries and rfbrowser init firefox chromium would install both Firefox and Chromium binaries.
Or use the docker images. Documented at docker/README.md.
Starting from release 18.3.0 Browser library has optional dependency with
Robotidy. Install library with Robotidy, run install with:
pip install robotframework-browser[tidy]. Starting from 18.3.0 release, library will provide external
Robotidy transformer. Transformer provided
by Browser library can be run with command: rfbrowser transform --transformer-name /path/to/tests. Example:
rfbrowser transform --wait-until-network-is-idle /path/to/tests would transform deprecated Wait Until Network Is Idle
keyword to Wait For Load State keyword. To see full list of transformers provided by Browser library, run
command: rfbrowser transform --help.
To upgrade your already installed robotframework-browser library
- Update from commandline:
pip install -U robotframework-browser - Clean old node side dependencies and browser binaries:
rfbrowser clean-node - Install the node dependencies for the newly installed version:
rfbrowser init
To completely uninstall library, including the browser binaries installed by Playwright, run following commands:
- Clean old node side dependencies and browser binaries:
rfbrowser clean-node - Uninstall with pip:
pip uninstall robotframework-browser
Testing with Robot Framework
*** Settings ***
Library Browser
*** Test Cases ***
Example Test
New Page https://playwright.dev
Get Text h1 contains Playwrightand testing with Python.
import Browser
browser = Browser.Browser()
browser.new_page("https://playwright.dev")
assert 'Playwright' in browser.get_text("h1")
browser.close_browser()async function myGoToKeyword(url, page, logger) {
logger("Going to " + url)
return await page.goto(url);
}
myGoToKeyword.rfdoc = "This is my own go to keyword";
exports.__esModule = true;
exports.myGoToKeyword = myGoToKeyword;*** Settings ***
Library Browser jsextension=${CURDIR}/mymodule.js
*** Test Cases ***
Example Test
New Page
myGoToKeyword https://www.robotframework.orgSee example. Ready made extensions and a place to share your own at robotframework-browser-extensions.
# Select element containing text "Login" with text selector strategy
# and select it's parent `input` element with xpath
Click "Login" >> xpath=../input
# Select element with CSS strategy and select button in it with text strategy
Click div.dialog >> "Ok"New Page ${LOGIN_URL}
${ref}= Get Element h1
Get Property ${ref} innerText == Login Page
Evaluate JavaScript ${ref} (elem) => elem.innerText = "abc"
Get Property ${ref} innerText == abc# The button with id `delayed_request` fires a delayed request. We use a promise to capture it.
${promise}= Promise To Wait For Response matcher= timeout=3s
Click \#delayed_request
${body}= Wait For ${promise}${device}= Get Device iPhone X
New Context &{device}
New Page
Get Viewport Size # returns { "width": 375, "height": 812 }${response}= HTTP /api/post POST {"name": "John"}
Should Be Equal ${response.status} ${200}You can let RF Browser spawn separate processes for every pabot process. This is very simple, just run the tests normally using pabot (see https://github.com/mkorpela/pabot#basic-use ). However if you have small tests do not use --testlevelsplit, it will cause lots of overhead because tests cannot share the browsers in any case.
You can share the node side RF Browser processes by using the ROBOT_FRAMEWORK_BROWSER_NODE_PORT environment variable, and from Browser.utils import spawn_node_process helper (see the docs for the helper ). This saves some overhead based on how many splits of tests you are running. Clean up the process afterwards.
- Figure out how the page is storing authentication
- If it is localstorage or cookies
Save Storage Stateshould work. See usage example: https://marketsquare.github.io/robotframework-browser/Browser.html#Save%20Storage%20State
See CONTRIBUTING.md for development instructions.
In order of appearance.
- Mikko Korpela
- Tatu Aalto
- Janne Härkönen (Alumnus)
- Kerkko Pelttari
- René Rohner
This project is community driven and becomes a reality only through the work of all the people who contribute. Supported by Robocorp through Robot Framework Foundation.
