ST Selenium Interview Questions & Answers

---

1️⃣ Using Robust and Dynamic Locators

I avoid absolute XPath and static IDs as much as possible.

What I use instead:

• Relative XPath
• contains(), starts-with()
• Stable attributes like data-testid, aria-label
• Parent–child relationships

Real‑time example:
In one Angular app, button IDs were generated dynamically on every refresh. I used:

//button[contains(text(),’Submit’)]

instead of ID‑based locators.

---

2️⃣ Explicit Waits (Most Important Technique)

Dynamic elements often appear after an API call or JS execution, so timing is critical.

I use:

• WebDriverWait
• ExpectedConditions like:
  • visibilityOfElementLocated
  • elementToBeClickable
  • presenceOfElementLocated

Example:
A search result list appeared only after backend response. Without wait, scripts failed intermittently.

✅ Replacing Thread.sleep() with explicit waits reduced flaky failures significantly.

---

3️⃣ Handling Stale Element Reference Exception

Dynamic pages often refresh the DOM, making previously located elements invalid.

How I handle it:

• Re‑locate the element before interacting
• Wrap actions in retry logic
• Use ExpectedConditions.refreshed()

Real‑time scenario:
After clicking “Save”, the page refreshed partially and old references broke. Re‑fetching elements solved it.

---

4️⃣ Handling Dynamic Dropdowns and Auto‑Suggestions

For AJAX‑based dropdowns:

• Wait until options count > 1
• Fetch list dynamically
• Iterate and select based on visible text

Example:
State → City dropdown where cities load dynamically after state selection.

---

5️⃣ Using JavaScriptExecutor (When Selenium Fails)

Some dynamic elements are:

• Hidden
• Overlapped
• Not interactable via normal click

In such cases, I use:

js.executeScript(“arguments[0].click();”, element);

Used mainly as a last resort, not default.

---

6️⃣ Page Object Model (POM) for Stability

I always implement POM, so when a dynamic locator changes:

• Fix happens in one place
• No impact on test logic

This is critical in large regression suites.