Skip to content

Reading Element State: Text, Attributes, Visibility

What was run

Real comparisons between .text, get_attribute('textContent'), and get_attribute('innerHTML') on the same real elements, plus real size/location measurements and is_displayed() checks across every tested element on the local test page — isolating exactly when .text and textContent genuinely diverge.

Real text properties, identical on a normal, visible element

label.text
label.get_attribute("textContent")
label.get_attribute("innerHTML")

REAL .text: 'I agree'
real get_attribute('textContent'): 'I agree'
real get_attribute('innerHTML'): 'I agree'

All three agree completely — for a normal, visible element with no nested markup, there's no real difference.

A real, measured divergence on an off-screen element

offscreen.text                              # ''
offscreen.get_attribute("textContent")      # bypasses visibility entirely

real offscreen.text (via .text property): ''
real offscreen text via textContent (bypasses visibility): 'Off-screen'
real offscreen.is_displayed(): False

A real, precise confirmation of what .text actually checks: the exact same element's real text content is genuinely still there (textContent returns 'Off-screen'), but .text returns an empty string, because .text specifically reflects rendered, visible text, not raw DOM content — and this element's real is_displayed() is False.

Real size, location, and visibility across every tested element

btn.size      # {'height': 21, 'width': 67}
btn.location  # {'x': 8, 'y': 198}

#text-input: is_displayed=True
#disabled-btn: is_displayed=True
#hidden-btn: is_displayed=False
#not-interactable: is_displayed=False
#action-btn: is_displayed=True

A real, informative result: disabled-btn is is_displayed=True — disabled is not the same real condition as hidden, consistent with Lesson 6's finding that disabled elements remain genuinely visible, just not clickable in a way that fires events.

How It Actually Works

.text is computed from the real, rendered page — it reflects what a human would actually see on screen, which is why an element positioned off-screen (position: absolute; left: -9999px in this lesson's real test case) reports empty text even though its real underlying HTML content hasn't changed at all. get_attribute('textContent') and innerHTML read the real DOM directly, completely bypassing any rendering or visibility consideration — which is exactly why they still returned the real 'Off-screen' text when .text could not.

This is a real, practical distinction worth internalizing: if a test needs to confirm something is actually visible to a real user, .text (combined with is_displayed()) is the right real check. If a test needs to confirm specific content exists in the DOM regardless of whether it's currently shown — useful for testing accessibility-only text, or content that's deliberately hidden until a later interaction — textContent is the real, correct tool.

Common mistakes

  • Using .text to verify that specific content exists in the DOM, then being confused when it returns empty for a real, legitimately-hidden-for-now element — this lesson's exact result is the direct explanation: the content is there, it's just not currently rendered.
  • Assuming is_displayed() and is_enabled() measure the same thing — this lesson's real disabled-btn: is_displayed=True result shows a disabled element is still fully visible; the two checks are independent.
  • Reading innerHTML when you actually want plain text — innerHTML includes real nested markup; for an element with child tags, it will differ from textContent/.text in ways this lesson's simple, tag-free label didn't happen to expose.

Exercise

Find or construct a real element on the test app with nested HTML inside it (for example, a <span> wrapped inside a <div>), and compare its real .text, textContent, and innerHTML values directly. Confirm innerHTML is the only one of the three that includes the real nested tag markup itself, not just its text.