dash.testing
provides custom Dash pytest fixtures and a set of testing APIs for unit and end-to-end testing.
This tutorial shows how to write and run tests for a Dash app.
dash.testing
Install dash.testing
with:
python -m pip install dash[testing]
In some shells (for example, Zsh), you may need to escape the opening bracket, [
:
python -m pip install dash\[testing]
New in Dash 2.6
dash.testing
supports unit testing of callbacks. Use callback unit tests to confirm that a callback’s outputs are as expected.
In the following example, we write two tests for our app. We have an app, app.py
, that has two callbacks. The update
callback outputs to an html.Div
the number of times btn-1
and btn-2
have been clicked. The display
callback uses callback_context to determine which input triggered the callback, adds the id
to a string, and outputs it to an html.Div
.
app.py
from dash import Dash, callback, html, Input, Output, ctx, callback
app = Dash()
app.layout = html.Div([
html.Button('Button 1', id='btn-1'),
html.Button('Button 2', id='btn-2'),
html.Button('Button 3', id='btn-3'),
html.Div(id='container'),
html.Div(id='container-no-ctx')
])
@callback(
Output('container-no-ctx', 'children'),
Input('btn-1', 'n_clicks'),
Input('btn-2', 'n_clicks'))
def update(btn1, btn2):
return f'button 1: {btn1} & button 2: {btn2}'
@callback(Output('container','children'),
Input('btn-1', 'n_clicks'),
Input('btn-2', 'n_clicks'),
Input('btn-3', 'n_clicks'))
def display(btn1, btn2, btn3):
button_clicked = ctx.triggered_id
return f'You last clicked button with ID {button_clicked}'
if __name__ == '__main__':
app.run(debug=True)
In the test file, test_app_callbacks.py
, we create two test cases, test_update_callback
and test_display_callback
.
from contextvars import copy_context
from dash._callback_context import context_value
from dash._utils import AttributeDict
# Import the names of callback functions you want to test
from app import display, update
def test_update_callback():
output = update(1, 0)
assert output == 'button 1: 1 & button 2: 0'
def test_display_callback():
def run_callback():
context_value.set(AttributeDict(**{"triggered_inputs": [{"prop_id": "btn-1-ctx-example.n_clicks"}]}))
return display(1, 0, 0)
ctx = copy_context()
output = ctx.run(run_callback)
assert output == f'You last clicked button with ID btn-1-ctx-example'
Run the tests with the command:
pytest
Notes:
app.py
for our tests.test_app_callbacks.py
, we import the callback functions to test, withfrom app import display, update
.The first test case, test_update_callback
, calls the update
function with the values 1
and 0
and stores the result in the variable output
.
Looking at our app (app.py
), we can see these inputs are the values of n_clicks
for each of the buttons:
```python
@callback(
Output(‘container-no-ctx’, ‘children’),
Input(‘btn-1’, ‘n_clicks’),
Input(‘btn-2’, ‘n_clicks’))
def update(btn1, btn2):
return f’button 1: {btn1} & button 2: {btn2}’
```
The test function then uses assert
to confirm that the output
is as expected, that it is equal to 'button 1: 1 & button 2: 0'
.
The second test case uses callback_context
. To test the callback, we need to mock callback_context
.
To do this, we need the following additional imports:
from contextvars import copy_context
from dash._callback_context import context_value
from dash._utils import AttributeDict
In the example:
ctx
.ctx.run
to call our run_callback
function in that context and save it to the output
variable.run_callback
function, we use context_value.set
to define which inputs trigger the callback in our test, and call the display
callback. context_value.set
takes an AttributeDict
(from dash._utils
).output
equals 'You last clicked button with ID btn-1-ctx-example'
.dash.testing
also supports end-to-end tests. End-to-end tests run programmatically, start a real browser session, and click through the Dash app UI. They are slower to run than unit tests and more complex to set up but more closely mimic the end user’s experience.
If you’re running end-to-end tests,
you’ll also need to install a WebDriver, unless you’re planning on running your tests with a Selenium Grid. Your tests will use either your locally installed WebDriver or the remote WebDriver on your grid to interact with the browser. See the Running Tests section below for details on how to run tests using a Selenium Grid.
We recommend the ChromeDriver WebDriver, which we use for dash
end-to-end tests. Install ChromeDriver by following the ChromeDriver Getting Started Guide.
Ensure you install the correct version of ChromeDriver for the version of Chrome you have installed.
Note: Mozilla Firefox geckodriver is also supported. To run your tests with geckodriver, you’ll need to add a flag when running tests. See the Running Tests section below for details on running tests with geckodriver.
Note: The Gecko (Marionette) driver from Mozilla is not fully compatible with Selenium specifications. Some features may not work as expected.
Here we create a test case where the browser driver opens the app, waits for an element with the id
nully-wrapper
to be displayed, confirms its text equals "0"
, and that there are no errors in the browser console.
# 1. imports of your dash app
import dash
from dash import html
# 2. give each testcase a test case ID, and pass the fixture
# dash_duo as a function argument
def test_001_child_with_0(dash_duo):
# 3. define your app inside the test function
app = dash.Dash()
app.layout = html.Div(id="nully-wrapper", children=0)
# 4. host the app locally in a thread, all dash server configs could be
# passed after the first app argument
dash_duo.start_server(app)
# 5. use wait_for_* if your target element is the result of a callback,
# keep in mind even the initial rendering can trigger callbacks
dash_duo.wait_for_text_to_equal("#nully-wrapper", "0", timeout=4)
# 6. use this form if its present is expected at the action point
assert dash_duo.find_element("#nully-wrapper").text == "0"
# 7. to make the checkpoint more readable, you can describe the
# acceptance criterion as an assert message after the comma.
assert dash_duo.get_logs() == [], "browser console should contain no error"
# 8. visual testing with percy snapshot
dash_duo.percy_snapshot("test_001_child_with_0-layout")
Notes:
test_{tcid}_{test title}
. The tcid
(test case ID) ismmffddd => module + file + three digits
.tcid
facilitates the test selection by just runningpytest -k {tcid}
. Its naming convention also helps code navigation withstart_server
API from dash_duo
.threading.Thread
.server_url
.Div
component’s text is identical to children
. #5 willfind_element
API call has an implicit global timeout of twofind_element('#nully-wrapper')
is a shortcut to driver.find_element_by_css_selector('#nully-wrapper')
.unittest
, pytest
allows you to use the standard Pythonassert
for verifyingassert
behavior. It’sassert
inside another helper API, also todcc.Graph
component. We integrate the Percy service with a PERCY_TOKEN
Use the dash_br
fixture to test a deployed app.
Set dash_br.server_url
to the URL of the app:
def test_002_nav_bar(dash_br):
dash_br.server_url = "https://dash-example-index.herokuapp.com/"
dash_br.wait_for_text_to_equal(".navbar-brand", "Dash Example Index", timeout=20)
assert dash_br.find_element(".navbar-brand").text == "Dash Example Index"
There are many ways to run your tests, and you can change how your tests run by adding flags to
the pytest
command.
You can run all tests in the current working directory (and its subdirectories) with:
pytest
Any tests in .py files with names that start with test_ or end with _test are run.
In the Basic Test example above, we gave our test case a test case ID, 001
.
We can use this to run that specific test:
pytest -k 001
There are different ways to configure your WebDriver when running end-to-end tests.
ChromeDriver is the default WebDriver, but dash.testing
also supports geckodriver for Firefox.
Add the --webdriver Firefox
flag when running tests to use it:
pytest --webdriver Firefox -k 001
You can run tests in headless mode, if you don’t need to see
the tests in a browser UI:
pytest --headless -k 001
You can use Selenium Grid to run tests across multiple machines. To run tests with a local hosted grid at http://localhost:4444/wd/hu
:
pytest --remote -k 001
http://localhost:4444/wd/hu
is the default value. To add an different remote, use the --remote-url
flag:
pytest --webdriver Firefox --remote-url <a href="https://grid_provider_endpoints">https://grid_provider_endpoints</a>
Note: If you set --remote-url
, and the value is different to the default value, http://localhost:4444/wd/hu
, you don’t need to include the --remote
flag.
We can’t guarantee that the above examples work with every Selenium Grid. There may be limitations because of how your network is configured. For example, because of:
If you encounter issues, it may be because you need to do some auxiliary WebDriver options tuning to run the tests. Here are some things to try:
--log-cli-level DEBUG
.pytest_setup_options
hook defined in plugin.py
. The example below is to use the headless mode with Chrome WebDriver in Windows. There is a workaround by adding --disable-gpu
in the options.from selenium.webdriver.chrome.options import Options
def pytest_setup_options():
options = Options()
options.add_argument('--disable-gpu')
return options
To avoid accidental name collision with other pytest plugins, all Dash test
fixtures start with the prefix dash
.
dash_br
dash_duo
thread_server
and a WebDriver wrapped with high-level Dash testing APIs.dash_thread_server
threading.Thread
, which isdash_process_server
waitress
(by default if raw_command
is not provided) in a Pythonsubprocess
. You can control the process runner with two supplementalraw_command
argument; to extend the timeout if your application needsstart_timeout
PYTHONPATH
so that the DashBoth dash_duo
and dash_br
expose the Selenium WebDriver via the
property driver
, e.g. dash_duo.driver
, which gives you full access to
the Python Selenium API.
(Note that this is not the official Selenium documentation site, but has
somehow become the de facto Python community reference)
One of the core components of Selenium testing is finding the
web element with a locator
, and performing some actions like click
or send_keys
on it, and waiting to verify if the expected state is met
after those actions. The check is considered as an acceptance criterion,
for which you can write in a standard Python assert
statement.
There are several strategies to
locate elements;
CSS selector and XPATH are the two most versatile ways. We recommend using
the CSS Selector in most cases due to its
better performance and robustness across browsers.
If you are new to using CSS Selectors, these
SauceLab tips
are a great start. Also, remember that
Chrome Dev Tools Console
is always your good friend and playground.
This link covers
this topic nicely. For impatient readers, a quick take away is
quoted as follows:
The Selenium WebDriver provides two types of waits:
- explicit wait
Makes WebDriver wait for a certain condition to occur before
proceeding further with execution. All our APIs with wait_for_*
falls into this category.
- implicit wait
Makes WebDriver poll the DOM for a certain amount of time when trying
to locate an element. We set a global two-second timeout at the
driver
level.
Note all custom wait conditions are defined in dash.testing.wait
and there are two extra APIs until
and until_not
which are similar to
the explicit wait with WebDriver, but they are not binding to
WebDriver context, i.e. they abstract a more generic mechanism to
poll and wait for certain condition to happen
This section lists a minimal set of browser testing APIs. They are
convenient shortcuts to Selenium APIs and have been approved in
our daily integration tests.
The following table might grow as we start migrating more legacy tests in
the near future. But we have no intention to build a comprehensive list,
the goal is to make writing Dash tests concise and error-free.
Please feel free to submit a community PR to add any missing ingredient,
we would be happy to accept that if it’s adequate for Dash testing.
API | Description |
---|---|
find_element(selector) |
return the first found element by the CSS selector , shortcut to driver.find_element_by_css_selector . note that this API will raise exceptions if not found, the find_elements API returns an empty list instead |
find_elements(selector) |
return a list of all elements matching by the CSS selector , shortcut to driver.find_elements_by_css_selector |
multiple_click(selector, clicks) |
find the element with the CSS selector and clicks it with number of clicks |
wait_for_element(selector, timeout=None) |
shortcut to wait_for_element_by_css_selector the long version is kept for back compatibility. timeout if not set, equals to the fixture’s wait_timeout |
wait_for_element_by_css_selector(selector, timeout=None) |
explicit wait until the element is present, shortcut to WebDriverWait with EC.presence_of_element_located |
wait_for_element_by_id(element_id, timeout=None) |
explicit wait until the element is present, shortcut to WebDriverWait with EC.presence_of_element_located |
wait_for_style_to_equal(selector, style, value, timeout=None) |
explicit wait until the element’s style has expected value . shortcut to WebDriverWait with custom wait condition style_to_equal . timeout if not set, equals to the fixture’s wait_timeout |
wait_for_text_to_equal(selector, text, timeout=None) |
explicit wait until the element’s text equals the expected text . shortcut to WebDriverWait with custom wait condition text_to_equal . timeout if not set, equals to the fixture’s wait_timeout |
wait_for_contains_text(selector, text, timeout=None) |
explicit wait until the element’s text contains the expected text . shortcut to WebDriverWait with custom wait condition contains_text condition. timeout if not set, equals to the fixture’s wait_timeout |
wait_for_class_to_equal(selector, classname, timeout=None) |
explicit wait until the element’s class has expected value . timeout if not set, equals to the fixture’s wait_timeout . shortcut to WebDriverWait with custom class_to_equal condition. |
wait_for_contains_class(selector, classname, timeout=None) |
explicit wait until the element’s classes contains the expected classname . timeout if not set, equals to the fixture’s wait_timeout . shortcut to WebDriverWait with custom contains_class condition. |
wait_for_page(url=None, timeout=10) |
navigate to the url in webdriver and wait until the dash renderer is loaded in browser. use server_url if url is None |
toggle_window() |
switch between the current working window and the newly opened one. |
switch_window(idx) |
switch to window by window index. shortcut to driver.switch_to.window . raise BrowserError if no second window present in browser |
open_new_tab(url=None) |
open a new tab in browser with window name new window . url if not set, equals to server_url |
percy_snapshot(name, wait_for_callbacks=False) |
visual test API shortcut to percy_runner.snapshot . it also combines the snapshot name with the actual python versions. The wait_for_callbacks parameter controls whether the snapshot is taken only after all callbacks have fired; the default is False . |
visit_and_snapshot(resource_path, hook_id, wait_for_callbacks=True, assert_check=True) |
This method automates a common task during dash-docs testing: the URL described by resource_path is visited, and completion of page loading is assured by waiting until the element described by hook_id is fetched. Once hook_id is available, visit_and_snapshot acquires a snapshot of the page and returns to the main page. wait_for_callbacks controls if the snapshot is taken until all dash callbacks are fired, default True. assert_check is a switch to enable/disable an assertion that there is no devtools error alert icon. |
take_snapshot(name) |
hook method to take a snapshot while Selenium test fails. the snapshot is placed under /tmp/dash_artifacts in Linux or %TEMP in windows with a filename combining test case name and the running Selenium session id |
zoom_in_graph_by_ratio(elem_or_selector, start_fraction=0.5, zoom_box_fraction=0.2, compare=True) |
zoom out a graph (provided with either a Selenium WebElement or CSS selector) with a zoom box fraction of component dimension, default start at middle with a rectangle of 1/5 of the dimension use compare to control if we check the SVG get changed |
click_at_coord_fractions(elem_or_selector, fx, fy) |
Use ActionChains to click a Selenium WebElement at a location a given fraction of the way fx between its left (0) and right (1) edges, and fy between its top (0) and bottom (1) edges. |
get_logs() |
return a list of SEVERE level logs after last reset time stamps (default to 0, resettable by reset_log_timestamp . Chrome only |
clear_input() |
simulate key press to clear the input |
driver |
property exposes the Selenium WebDriver as fixture property |
session_id |
property returns the Selenium session_id, shortcut to driver.session_id |
server_url |
set the server_url as setter so the Selenium is aware of the local server port, it also implicitly calls wait_for_page . return the server_url as property |
download_path |
property returns the download_path, note that dash fixtures are initialized with a temporary path from pytest tmpdir |
This section enumerates a full list of Dash App related properties and APIs
apart from the previous browser ones.
API | Description |
---|---|
devtools_error_count_locator |
property returns the selector of the error count number in the devtool UI |
dash_entry_locator |
property returns the selector of react entry point, it can be used to verify if an Dash app is loaded |
dash_outerhtml_dom |
property returns the BeautifulSoup parsed Dash DOM from outerHTML |
dash_innerhtml_dom |
property returns the BeautifulSoup parsed Dash DOM from innerHTML |
redux_state_paths |
property returns the window.store.getState().paths |
redux_state_rqs |
property returns window.store.getState().requestQueue |
window_store |
property returns window.store |
get_local_storage(store_id="local") |
get the value of local storage item by the id, default is local |
get_session_storage(session_id="session") |
get the value of session storage item by the id, default is session |
clear_local_storage() |
shortcut to window.localStorage.clear() |
clear_session_storage() |
shortcut to window.sessionStorage.clear() |
clear_storage() |
clears both local and session storages |
If you run the integration in a virtual environment, make sure you are
getting the latest commit in the master branch from each component, and
that the installed pip
versions are correct.
Note: We have some enhancement initiatives tracking in this issue
The CircleCI Local CLI is a
handy tool to execute some jobs locally. It gives you an earlier warning
before even pushing your commits to remote. For example, it’s always
recommended to pass lint and unit tests job first on your local machine. So
we can make sure there are no simple mistakes in the commit.
# install the cli (first time only)
$ curl -fLSs <a href="https://circle.ci/cli">https://circle.ci/cli</a> | bash && circleci version
# run at least the lint & unit test job on both python 2 and 3
# note: the current config requires all tests pass on python 2.7, 3.6 and 3.7.
$ circleci local execute --job lint-unit-27 && $ circleci local execute --job lint-unit-37
pytest --log-cli-level DEBUG -k bsly001
You can get more logging information from Selenium WebDriver, Flask server,
and our test APIs.
14:05:41 | DEBUG | selenium.webdriver.remote.remote_connection:388 | DELETE <a href="http://127.0.0.1:53672/session/87b6f1ed3710173eff8037447e2b8f56">http://127.0.0.1:53672/session/87b6f1ed3710173eff8037447e2b8f56</a> {"sessionId": "87b6f1ed3710173eff8037447e2b8f56"}
14:05:41 | DEBUG | urllib3.connectionpool:393 | <a href="http://127.0.0.1:53672">http://127.0.0.1:53672</a> "DELETE /session/87b6f1ed3710173eff8037447e2b8f56 HTTP/1.1" 200 72
14:05:41 | DEBUG | selenium.webdriver.remote.remote_connection:440 | Finished Request
14:05:41 | INFO | dash.testing.application_runners:80 | killing the app runner
14:05:41 | DEBUG | urllib3.connectionpool:205 | Starting new HTTP connection (1): localhost:8050
14:05:41 | DEBUG | urllib3.connectionpool:393 | <a href="http://localhost:8050">http://localhost:8050</a> "GET /_stop-3ef0e64e8688436caced44e9f39d4263 HTTP/1.1" 200 29
If you run your tests with CircleCI dockers (locally with CircleCI CLI
and/or remotely with CircleCI
).
Inside a docker run or VM instance where there is no direct access to the
video card, there is a known limitation that you cannot see anything from
the Selenium browser on your screen. Automation developers use
Xvfb as
a workaround to solve this issue. It enables you to run graphical
applications without a display (e.g., browser tests on a CI server) while
also having the ability to take screenshots.
We implemented an automatic hook at the test report stage, it checks if a
test case failed with a Selenium test fixture. Before tearing down every
instance, it will take a snapshot at the moment where your assertion is
False
or having a runtime error. refer to Browser APIs
Note: you can also check the snapshot directly in CircleCI web page
under Artifacts
Tab
There are two customized pytest
arguments to tune Percy runner:
1. --nopercyfinalize
disables the Percy finalize in dash fixtures. This
is required if you run your tests in parallel, then you add an extra
percy finalize --all
step at the end. For more details, please visit
Percy Documents.
2. --percy-assets
lets Percy know where to collect additional assets
such as CSS files.