All-in-One Test Automation Framework

SeleniumBase makes it easy to run WebDriver automation with pytest and nosetests.

pypi Build Status Python version MIT License Join the chat at GitHub stars Get Started with SeleniumBase

Table of Contents / Navigation: Intro and Demo

Simple Python syntax makes coding easy:

(By default, CSS Selectors are used for finding page elements.)

Customize scripts from the command line:

nosetests --demo_mode --with-selenium --browser=chrome -s

pytest --demo_mode -s

Watch run in Demo Mode:

No more flaky tests:
SeleniumBase automatically waits for page elements to finish loading before interacting with them (up to a timeout limit). This means you no longer need random time.sleep() statements in your code.

SeleniumBase helps keep your code clean:
This long line of WebDriver code…


…becomes the following in SeleniumBase:

self.update_text("textarea", "text")

You can still use self.driver in your code.

Business Mindset:
SeleniumBase makes it easy to automate tedious business tasks. (To learn about businesses using SeleniumBase, Click Here.)

Hybrid Automation:
SeleniumBase includes an automated/manual hybrid solution called MasterQA, which speeds up manual testing by having automation perform all the web browser actions while the manual tester only needs to validate what is seen. Learn more about it here. Full Feature List Setup Instructions for Mac, Ubuntu, and Windows

(Debian Linux users: Run to setup your Debian Linux machine.)

(Docker users: See the Docker ReadMe to setup your Docker machine.)

Step 0a: Setup your Python version Python/pip environment:

Step 0b: Install web browsers to run automation on:

Step 0c: Get web drivers for each browser you intend to run automation on:

To run automation on various web browsers, you’ll need to download a driver file for each one and place it on your System PATH. On a Mac, /usr/local/bin is a good spot. On Windows, make sure you set the System Path under Environment Variables to include the location where you placed the driver files:

(NOTE: For older versions of Firefox such as 46.0 and earlier, you don’t need Geckodriver. The older driver comes prepackaged with Selenium.)

(NOTE: If you don’t have access rights to update system variables, you can use the Anaconda Version of Python 2. In that case, place web drivers in Anaconda_Installation_Path/Scripts/)


brew install chromedriver phantomjs Step 1: Clone SeleniumBase

git clone
cd SeleniumBase

(A Git GUI tool like SourceTree may make things easier.)

Step 2: Create a Virtual Environment

If you’re not sure how to create one, follow these instructions.
For an overview of virtual environments and why it’s good practice to use them, see

Step 3: Install SeleniumBase

To install SeleniumBase from the Python Package Index use:

pip install seleniumbase --upgrade

To install your local customized version of SeleniumBase use:

pip install -r requirements.txt --upgrade

python install

(NOTE: If you’re using Python 3.* instead of Python 2.7, use pip3 in place of pip and python3 in place of python in the above commands.) Step 4: Run the Example Script

Here’s what the example script looks like:

from seleniumbase import BaseCase

class MyTestClass(BaseCase):

    def test_basic(self):'')            # Navigate to the web page
        self.assert_element('img[alt="Python"]')       # Assert element on page'a[rel="license"]')                  # Click element on page
        self.assert_text('copy and reuse', 'div center')  # Assert element text'')
        image_object = self.find_element('#comic img')    # Returns the element
        caption = image_object.get_attribute('title')   # Get element attribute
        self.assertTrue('connections to the server' in caption)
        self.click_link_text('Blag')              # Click on link with the text
        self.assert_text('xkcd', '#site-title')
        header_text = self.get_text('header h2')  # Grab text from page element
        self.assertTrue('The blag of the webcomic' in header_text)
        self.update_text('input#s', 'Robots!\n')  # Fill in field with the text
        self.assert_text('Hooray robots!', '#content')'')
        self.assert_text('Automation', 'div#ctitle')

Here’s how to run the example script on various web browsers by using nosetests:

(NOTE: You can interchange nosetests with pytest as seen here.)

cd examples/

nosetests --with-selenium --browser=chrome -s

nosetests --with-selenium --browser=firefox -s

nosetests --with-selenium --browser=phantomjs -s

After the test completes, in the console output you’ll see a dot (.) on a new line, representing a passing test. (On test failures you’ll see an F instead, and on test errors you’ll see an E). It looks more like a moving progress bar when you’re running a ton of unit tests side by side. This is part of nosetests. After all tests complete (in this case there is only one), you’ll see the “Ran 1 test in ...” line, followed by an “OK” if all nosetests passed. The --with-selenium option is required for running GUI tests with nosetests (not needed when using pytest). If no browser is specified, Chrome will become the default. The -s option is optional, and that makes sure that any standard output is printed immediately on the command line when tests have print statements in them, which makes debugging much easier.

(NOTE: If you’re confused about how SeleniumBase works with Nosetests and Pytest, jump to the How SeleniumBase Works section to learn more about it.)

If the example test is moving too fast for your eyes to see what’s going on, you can run it in Demo Mode by adding --demo_mode on the command line, which pauses the browser briefly between actions, and highlights page elements being acted on:

nosetests --with-selenium --browser=chrome -s --demo_mode

You can override the default wait time by either updating or by using --demo_sleep={NUM} when using Demo Mode. (NOTE: If you use --demo_sleep={NUM} without using --demo_mode, nothing will happen.)

If you ever make any changes to your local copy of, you may need to run python install for those changes to take effect.

nosetests --with-selenium --browser=chrome -s --demo_mode --demo_sleep=1.2

You can also add either of the following to your scripts to slow down the tests:

import time; time.sleep(5)  # sleep for 5 seconds (add this after the line you want to pause on)
import ipdb; ipdb.set_trace()  # waits for your command. n = next line of current method, c = continue, s = step / next executed line (will jump)

(NOTE: If you’re using pytest instead of nosetests and you want to use ipdb in your script for debugging purposes, you’ll either need to add “–capture=no” on the command line, or use “import pytest; pytest.set_trace()” instead of using ipdb. More info on that here.)

You may also want to have your test sleep in other situations where you need to have your test wait for something. If you know what you’re waiting for, you should be specific by using a command that waits for something specific to happen.

If you need to debug things on the fly (in case of errors), use this line to run the code:

nosetests --browser=chrome --with-selenium --pdb --pdb-failures -s

The above code (with –pdb) will leave your browser window open in case there’s a failure, which is possible if the web pages from the example change the data that’s displayed on the page. (ipdb commands: ‘c’, ‘s’, ‘n’ => continue, step, next).

Here are some other useful nosetest arguments that you may want to append to your run commands:

--logging-level=INFO  # Hide DEBUG messages, which can be overwhelming.
-x  # Stop running the tests after the first failure is reached.
-v  # Prints the full test name rather than a dot for each test.
--with-id  # If -v is also used, will number the tests for easy counting.

Here’s how to run the example script with pytest:

cd examples/

pytest --with-testing_base --browser=firefox -s

pytest --with-testing_base --browser=chrome -s

pytest --with-testing_base --browser=phantomjs -s

(NOTE: The --with-testing_base plugin gives you full logging on test failures, which saves screenshots, page source, and basic test info into the logs folder.)

(NOTE: If you’re using pytest instead of nosetests for running tests outside of the SeleniumBase repo, you’ll need a copy of pytest.ini at the base of the new folder structure. How SeleniumBase Works:

At the core, SeleniumBase works by extending Nosetests and Pytest as a direct plugin to each one. This plugin is activated by using “--with-selenium” as a command line argument when running Nosetest/Pytest. When activated, Selenium-WebDriver automatically spins up web browsers for tests, and then gives those tests access to the SeleniumBase libraries through the base class.

Once you’ve activated the main SeleniumBase plugin with “--with-selenium”, you can use “--browser=chrome” to specify the web browser to use (Default = “Chrome”). You can also include additional plugins for additional features such as “--with-testing_base” (for logging data/screenshots on test failures) and “--demo_mode” (for highlighting elements & slowing test runs). There are also other plugins available such as “--with-db_reporting”, “--with-s3_logging”, and more.

(NOTE: Nosetests and Pytest work by automatically running any Python method that starts with “test” from the file that you specified on the command line. You can also run all tests from a specific class in a file, or even pick out an individual test to run.)

To use SeleniumBase calls you need the following:

from seleniumbase import BaseCase

And then have your test classes inherit BaseCase:

class MyTestClass(BaseCase):

(See the example test,, for reference.) Creating Visual Test Suite Reports:

(NOTE: The command line args are different for Nosetests vs Pytest)

Nosetest Reports:

The --report option gives you a fancy report after your test suite completes. (Requires --with-testing_base to also be set when --report is used because it’s part of that plugin.)

nosetests --with-selenium --with-testing_base --report --browser=chrome -s

(NOTE: You can add --show_report to immediately display the report after the test suite completes. You don’t want to use this when running tests remotely because otherwise the test run will hang indefinitely until someone manually exits the report.)

Pytest Reports:

Using --html=report.html gives you a fancy report of the name specified after your test suite completes.

pytest --with-selenium --html=report.html Using Production Environments & Integrations:

Here are some things you can do to setup a production environment for your testing:

nosetests [YOUR_TEST_FILE].py --browser=chrome --with-selenium --with-testing_base --with-db_reporting --with-s3_logging -s

(NOTE: If you haven’t configured your MySQL or S3 connections in, don’t use --with-db_reporting or --with-s3_logging.)

When the testing_base plugin is used, if there’s a test failure, the basic_test_info plugin records test logs, the page_source plugin records the page source of the last web page seen by the test, and the screen_shots plugin records the image of the last page seen by the test where the failure occurred. Make sure you always include testing_base whenever you include a plugin that logs test data. The db_reporting plugin records the status of all tests run into your MySQL DB. The s3_logging plugin uploads basic test info, screenshots, and page source into your S3 storage folder.

To simplify that long run command, you can create a *.cfg file, such as the one provided in the example, and enter your plugins there so that you can run everything by typing:

nosetests [YOUR_TEST_FILE].py --config=[MY_CONFIG_FILE].cfg -s

You can simplify that even more by using a setup.cfg file, such as the one provided for you in the examples folder. If you kick off a test run from within the folder that setup.cfg is location in, that file will automatically be used as your configuration, meaning that you wouldn’t have to type out all the plugins that you want to use (or include a config file) everytime you run tests.

If you tell nosetests to run an entire file, it will run every method in that python file that starts with “test”. You can be more specific on what to run by doing something like:

nosetests [YOUR_TEST_FILE].py:[SOME_CLASS_NAME].test_[SOME_TEST_NAME] --config=[MY_CONFIG_FILE].cfg -s

Let’s try an example of a test that fails. Copy the following into a file called

""" """
from seleniumbase import BaseCase

class MyTestClass(BaseCase):

    def test_find_army_of_robots_on_xkcd_desert_island(self):"")
        self.assert_element("div#ARMY_OF_ROBOTS", timeout=3)  # This should fail

Now run it:

nosetests --browser=chrome --with-selenium --with-testing_base -s

You’ll notice that a logs folder, “latest_logs”, was created to hold information about the failing test, and screenshots. Take a look at what you get. Remember, this data can be saved in your MySQL DB and in S3 if you include the necessary plugins in your run command (and if you set up the neccessary connections properly). For future test runs, past test results will get stored in the archived_logs folder if you have ARCHIVE_EXISTING_LOGS set to True in Detailed Method Specifications and Examples:"")  # This method opens the specified page.

self.go_back()  # This method navigates the browser to the previous page.

self.go_forward()  # This method navigates the browser forward in history.

self.refresh_page()  # This method reloads the current page.

self.get_current_url()  # This method returns the current page URL.

self.get_page_source()  # This method returns the current page source.

ProTip™: You may need to use the get_page_source() method along with Python’s find() command to parse through the source to find something that Selenium wouldn’t be able to. (You may want to brush up on your Python programming skills for that.) Ex:

source = self.get_page_source()
first_image_open_tag = source.find('<img>')
first_image_close_tag = source.find'</img>', first_image_open_tag)
everything_inside_first_image_tags = source[first_image_open_tag+len('<img>'):first_image_close_tag]


To click an element on the page:"div#my_id")

ProTip™: In most web browsers, you can right-click on a page and select Inspect Element to see the CSS selector details that you’ll need to create your own scripts.

Typing Text

self.update_text(selector, text) # updates the text from the specified element with the specified value. An exception is raised if the element is missing or if the text field is not editable. Example:

self.update_text("input#id_value", "2012")

You can also use self.add_text() or the WebDriver .send_keys() command, but those won’t clear the text box first if there’s already text inside. If you want to type in special keys, that’s easy too. Here’s an example:

from selenium.webdriver.common.keys import Keys
self.find_element("textarea").send_keys(Keys.SPACE + Keys.BACK_SPACE + '\n')  # the backspace should cancel out the space, leaving you with the newline

Getting the text from an element on a page

text = self.get_text("header h2")

Getting the attribute value from an element on a page

attribute = self.get_attribute("#comic img", "title")

Asserting existance of an element on a page within some number of seconds:

self.wait_for_element_present("div.my_class", timeout=10)

(NOTE: You can also use: self.assert_element_present(ELEMENT))

Asserting visibility of an element on a page within some number of seconds:

self.wait_for_element_visible("a.my_class", timeout=5)

(NOTE: The short versions of this are self.find_element(ELEMENT) and self.assert_element(ELEMENT). The find_element() version returns the element)

Since the line above returns the element, you can combine that with .click() as shown below:

self.find_element("a.my_class", timeout=5).click()

# But you're better off using the following statement, which does the same thing:"a.my_class")  # DO IT THIS WAY!

ProTip™: You can use dots to signify class names (Ex: div.class_name) as a simplified version of div[class="class_name"] within a CSS selector.

You can also use *= to search for any partial value in a CSS selector as shown below:'a[name*="partial_name"]')

Asserting visibility of text inside an element on a page within some number of seconds:

self.wait_for_text_visible("Make it so!", "div#trek div.picard div.quotes", timeout=3)
self.wait_for_text_visible("Tea. Earl Grey. Hot.", "div#trek div.picard div.quotes", timeout=1)

(NOTE: The short versions of this are self.find_text(TEXT, ELEMENT) and self.assert_text(TEXT, ELEMENT))

Asserting Anything

self.assertTrue(myvar1 == something)

self.assertEqual(var1, var2)

Useful Conditional Statements (with creative examples in action)

is_element_visible(selector) # is an element visible on a page

import logging
if self.is_element_visible('div#warning'):
    logging.debug("Red Alert: Something bad might be happening!")

is_element_present(selector) # is an element present on a page

if self.is_element_present('div#top_secret img.tracking_cookie'):
    self.contact_cookie_monster()  # Not a real method unless you define it somewhere
    current_url = self.get_current_url()
    self.contact_the_nsa(url=current_url, message="Dark Zone Found")  # Not a real method unless you define it somewhere

Another example:

def is_there_a_cloaked_klingon_ship_on_this_page():
    if self.is_element_present("div.ships div.klingon"):
        return not self.is_element_visible("div.ships div.klingon")
    return False

is_text_visible(text, selector) # is text visible on a page

def get_mirror_universe_captain_picard_superbowl_ad(superbowl_year):
    selector = "div.superbowl_%s div.commercials div.transcript div.picard" % superbowl_year
    if self.is_text_visible("For the Love of Marketing and Earl Grey Tea!", selector):
        return "Picard HubSpot Superbowl Ad 2015"
    elif self.is_text_visible("Delivery Drones... Engage", selector):
        return "Picard Amazon Superbowl Ad 2015"
    elif self.is_text_visible("Bing it on Screen!", selector):
        return "Picard Microsoft Superbowl Ad 2015"
    elif self.is_text_visible("OK Glass, Make it So!", selector):
        return "Picard Google Superbowl Ad 2015"
    elif self.is_text_visible("Number One, I've Never Seen Anything Like It.", selector):
        return "Picard Tesla Superbowl Ad 2015"
    elif self.is_text_visible("""With the first link, the chain is forged.
                              The first speech censored, the first thought forbidden,
                              the first freedom denied, chains us all irrevocably.""", selector):
        return "Picard Wikimedia Superbowl Ad 2015"
    elif self.is_text_visible("Let us make sure history never forgets the name ... Facebook", selector):
        return "Picard Facebook Superbowl Ad 2015"
        raise Exception("Reports of my assimilation are greatly exaggerated.")

Switching Tabs

What if your test opens up a new tab/window and now you have more than one page? No problem. You need to specify which one you currently want Selenium to use. Switching between tabs/windows is easy: Ex:

self.switch_to_window(1)  # this switches to the new tab (0 is the first one)

ProTip™: iFrames follow the same principle as new windows - you need to specify the iFrame if you want to take action on something in there Ex:

# Now you can act inside the iFrame
# .... Do something cool (here)
self.switch_to_default_content()  # exit the iFrame when you're done

Handle Pop-Up Alerts

What if your test makes an alert pop up in your browser? No problem. You need to switch to it and either accept it or dismiss it: Ex:



If you’re not sure whether there’s an alert before trying to accept or dismiss it, one way to handle that is to wrap your alert-handling code in a try/except block. Other methods such as .text and .send_keys() will also work with alerts.

Executing Custom jQuery Scripts:

jQuery is a powerful JavaScript library that allows you to perform advanced actions in a web browser. If the web page you’re on already has jQuery loaded, you can start executing jQuery scripts immediately. You’d know this because the web page would contain something like the following in the HTML:

<script src=""></script>

It’s OK if you want to use jQuery on a page that doesn’t have it loaded yet. To do so, run the following command first:


Here are some examples of using jQuery in your scripts:

self.execute_script('jQuery, window.scrollTo(0, 600)')  # Scrolling the page

self.execute_script("jQuery('#annoying-widget').hide()")  # Hiding elements on a page

self.execute_script("jQuery('#annoying-button a').remove()")  # Removing elements on a page

self.execute_script("jQuery('%s').mouseover()" % (mouse_over_item))  # Mouse-over elements on a page

self.execute_script("jQuery('input#the_id').val('my_text')")  # Fast text input on a page

self.execute_script("jQuery('div#dropdown').click()")  # Click elements on a page

self.execute_script("return jQuery('div#amazing')[0].text")  # Returns the css "text" of the element given

self.execute_script("return jQuery('textarea')[2].value")  # Returns the css "value" of the 3rd textarea element on the page

In the following example, javascript is used to plant code on a page that Selenium can then touch after that:
referral_link = '<a class="analytics test" href="%s">Free-Referral Button!</a>' % DESTINATION_URL
self.execute_script("document.body.innerHTML = \"%s\"" % referral_link)"")  # Clicks the generated button

Using non-terminating verifications:

Let’s say you want to verify multiple different elements on a web page in a single test, but you don’t want the test to fail until you verified several elements at once so that you don’t have to rerun the test to find more missing elements on the same page. That’s where page checks come in. Here’s the example:

from seleniumbase import BaseCase

class MyTestClass(BaseCase):

    def test_non_terminating_checks(self):'')
        self.check_assert_element('img[alt="Brand Identity"]')
        self.check_assert_element('img[alt="Rocket Ship"]')  # Will Fail
        self.check_assert_text('Fake Item', '#middleContainer')  # Will Fail
        self.check_assert_text('Random', '#middleContainer')
        self.check_assert_element('a[name="Super Fake !!!"]')  # Will Fail

check_assert_element() and check_assert_text() will save any exceptions that would be raised. To flush out all the failed checks into a single exception, make sure to call self.process_checks() at the end of your test method. If your test hits multiple pages, you can call self.process_checks() at the end of all your checks for a single page. This way, the screenshot from your log file will make the location where the checks were made.

Accessing raw WebDriver

If you need access to any commands that come with standard WebDriver, you can call them directly like this:

capabilities = self.driver.capabilities

(In general, you’ll want to use the SeleniumBase versions of methods when available.)

Checking Email:

Let’s say you have a test that sends an email, and now you want to check that the email was received:

from seleniumbase.fixtures.email_manager import EmailManager, EmailException
num_email_results = 0
email_subject = "This is the subject to search for (maybe include a timestamp)"
email_manager = EmailManager("[YOUR SELENIUM GMAIL EMAIL ADDRESS]")  # the password for this is elsewhere (in the library) because this is a default email account
    html_text ="%s" % email_subject, timeout=300)
    num_email_results = len(html_text)
except EmailException:
    num_email_results = 0
self.assertTrue(num_email_results)  # true if not zero

Now you can parse through the email if you’re looking for specific text or want to navigate to a link listed there.

Database Powers:

Let’s say you have a test that needs to access the database. First make sure you already have a table ready. Then try this example:

from seleniumbase.core.mysql import DatabaseManager
def write_data_to_db(self, theId, theValue, theUrl):
    db = DatabaseManager()
    query = """INSERT INTO myTable(theId,theValue,theUrl)
               VALUES (%(theId)s,%(theValue)s,%(theUrl)s)"""
    db.execute_query_and_close(query, {"theId":theId,

Access credentials are stored in for your convenience (you have to add them first).

The following example below (taken from the Delayed Data Manager) shows how data can be pulled from the database.

import logging
from seleniumbase.core.mysql import DatabaseManager

def get_delayed_test_data(self, testcase_address, done=0):
    """ Returns a list of rows """
    db = DatabaseManager()
    query = """SELECT guid,testcaseAddress,insertedAt,expectedResult,done
               FROM delayedTestData
               WHERE testcaseAddress=%(testcase_address)s
               AND done=%(done)s"""
    data = db.fetchall_query_and_close(query, {"testcase_address":testcase_address, "done":done})
    if data:
        return data
        logging.debug("Could not find any rows in delayedTestData.")
        logging.debug("DB Query = " + query % {"testcase_address":testcase_address, "done":done})
        return []

Now you know how to pull data from your MySQL DB.

Delayed Data Manager usage example: If you scheduled an email to go out 12 hours from now and you wanted to check that the email gets received (but you don’t want your test sitting idle for 12 hours) you can store the email credentials as a unique time-stamp for the email subject in the DB (along with a time for when it’s safe for the email to be searched for) and then a later-running test can do the checking after the right amount of time has passed. Wrap-Up

Congratulations! You now know how to Automate like a Pro!

Questions or Comments? Join the chat at

Here are some other exciting open source projects on GitHub by smart people I’ve worked with:

~ Michael Mintz License

MIT License (The MIT License)