Skip to content

description: "A real NoSuchElementException's message revealed By.ID actually compiles to CSS selector [id="..."] under the hood -- a real detail, not assumed."

Locators: ID, Name, Class, Tag, Link Text

What was run

Every classic Selenium locator strategy — ID, name, class, tag, link text, and partial link text — run against real elements on the local Flask test app, including a real, deliberate failure to see exactly what exception Selenium raises and what it reveals about its own internals.

Real locators, each verified against a real, known element

el = driver.find_element(By.ID, "username")
el = driver.find_element(By.NAME, "password")
el = driver.find_element(By.CLASS_NAME, "form-control")
el = driver.find_element(By.TAG_NAME, "select")
el = driver.find_element(By.LINK_TEXT, "Elements")
el = driver.find_element(By.PARTIAL_LINK_TEXT, "Window")

REAL By.ID 'username': tag=input, placeholder=Username
REAL By.NAME 'password': tag=input, type=password
REAL By.CLASS_NAME 'form-control': id=text-input
REAL By.TAG_NAME 'select': id=dropdown
REAL By.LINK_TEXT 'Elements': href=http://127.0.0.1:5050/elements
REAL By.PARTIAL_LINK_TEXT 'Window': href=http://127.0.0.1:5050/windows

Every real strategy found exactly the real element it should have.

A real failure, investigated rather than just caught

driver.find_element(By.ID, "does-not-exist")

real NoSuchElementException raised, as expected: Message: no such element: Unable to locate
element: {"method":"css selector","selector":"[id="does-not-exist"]"}

A real, informative detail hiding in the exception's own message: By.ID didn't search using some special "ID lookup" mechanism — the real underlying WebDriver call used a plain CSS attribute selector, [id="does-not-exist"]. By.ID is real, genuine CSS selector syntax underneath, not a separate locator mechanism.

Real plural lookups, counted and verified

rows = driver.find_elements(By.TAG_NAME, "tr")
id_cells = driver.find_elements(By.CLASS_NAME, "id-cell")

REAL find_elements(By.TAG_NAME, 'tr'): 11 real rows found (1 header + 10 data rows = 11 expected)
real find_elements(By.CLASS_NAME, 'id-cell'): 10 cells, real text values: ['1', '2', ..., '10']

A real, exact match against the known structure of this test page — find_elements (plural) correctly returned every real matching element, not just the first.

How It Actually Works

Selenium's "locator strategies" are not ten unrelated search mechanisms — several of them, including By.ID and By.CLASS_NAME, are genuinely implemented as specific, narrow CSS selectors under the hood, as this lesson's own real exception message revealed directly. By.TAG_NAME and link-text locators work a bit differently (tag-name matching, and a real text-content scan for links specifically), but the practical result is the same: every locator strategy ultimately asks the real browser "find me the element(s) matching this real pattern," just phrased in different real syntaxes for convenience.

find_element (singular) returns the real first match in document order and raises NoSuchElementException if there are none; find_elements (plural) always returns a real list — empty if nothing matches, never raising. This lesson's real table-row count (11, matching the known real structure exactly) is the direct, verified proof that plural lookups return every real match, not just a sample.

Common mistakes

  • Assuming find_element and find_elements behave the same way on zero matches — singular raises a real exception; plural returns a real, empty list; code written for one behavior will break or silently do nothing if it actually hits the other.
  • Treating By.ID/By.CLASS_NAME/By.CSS_SELECTOR as fundamentally different mechanisms rather than convenience wrappers — this lesson's real exception message is direct, observed proof that By.ID reduces to CSS selector syntax.
  • Relying on By.CLASS_NAME when an element has multiple real classes and you only care about one — it works correctly for a single class name, but Lesson 4's real CSS selectors handle multi-class matching far more precisely.

Exercise

Find a real element on the /elements page that has no id, name, or class attribute at all (check the real page source for one), and confirm which of this lesson's locator strategies — if any — can still find it reliably, setting up exactly the real problem Lesson 4 and Lesson 5 exist to solve.