[BUG] Dash Testing: wait_for_text_to_equal may incorrectly succeed when used with text "None"

[emilykl]

Describe your context

• replace the result of pip list | grep dash below

dash                            2.16.1         
dash-core-components            2.0.0
dash-dangerously-set-inner-html 0.0.2
dash-flow-example               0.0.5
dash-html-components            2.0.0
dash-table                      5.0.0
dash-testing-stub               0.0.2

Describe the bug

When wait_for_text_to_equal is used to wait for the text "None", the function will often succeed even when you would reasonably expect it to fail.

I think this is part of the reason why the regression in #2733 wasn't caught by the tests.

This behavior is demonstrated by the following test case:

import dash
from dash import html


def test_wait_for_text_to_equal_none(dash_duo):
    app = dash.Dash(__name__)
    app.layout = html.Div(id="my-div", children="Hello world")
    dash_duo.start_server(app)
    dash_duo.wait_for_text_to_equal("#my-div", "None", timeout=4)

Expected behavior

The test should fail because the contents of the #my-div div are never equal to None or "None".

Actual behavior

The test passes.

Explanation

This happens because wait_for_text_to_equal checks not only the text content of the element, but also the value of the value attribute. (see here).

If value is not defined we get a value of None, which is then converted to a string and therefore matches the string "None".

So dash_duo.wait_for_text_to_equal("#my-div", "None") always succeeds unless the target element has a defined value.

Proposed solutions

IMO the cleanest solution would be to modify wait_for_text_to_equal to check only the element's text, and add a new function wait_for_value_to_equal which checks the value (or a generalized wait_for_attr_to_equal function). This would break backwards compatibility.

Alternatively we could have wait_for_text_to_equal ignore value if value is not defined, or issue a warning when used with the text "None".

[TillerBurr]

I was just coming to look into this further, but looks like I was beaten to it.

Looking at the function, I think there's a another solution. My thought would be to change

dash/dash/testing/wait.py

Lines 106 to 113 in f7f8fb4

	def __call__(self, driver):
	try:
	elem = self._get_element(driver)
	logger.debug("text to equal {%s} => expected %s", elem.text, self.text)
	return (
	str(elem.text) == self.text
	or str(elem.get_attribute("value")) == self.text
	)

to

def __call__(self, driver): 
    try:  
        elem = self._get_element(driver) 
        logger.debug("text to equal {%s} => expected %s", elem.text, self.text) 
        value = elem.get_attribute("value")
        value = str(value) if value is not None else None
        return ( 
            str(elem.text) == self.text 
            or value == self.text 
        ) 

That way, the overall functionality of wait_for_text_to_equal doesn't change (except for the "None" case). The only downside is that it would pass if you use dash_duo.wait_for_text_to_equal("#my-div", None, timeout=4), but I don't know how often that would come up.

There is also an issue with wait_for_contains_text. The test

from dash import html
import dash


def test_wait_for_contains_text_none(dash_duo):

    app = dash.Dash(__name__)
    app.layout = html.Div(id="my-div", children="Hello world")
    dash_duo.start_server(app)
    dash_duo.wait_for_contains_text("#my-div", "No", timeout=4)

also passes, but similarly, there is no value and "No" is not in "Hello world".

There are potential issues with wait_for_contains_class and wait_for_class_to_equal, but I imagine those would be far less common.