Interview Corner

A: In Playwright, I use page.frameLocator() to scope into an iframe by its CSS selector. Then I chain regular locators like getByRole() and getByLabel(). For nested iframes, I chain multiple frameLocator() calls. The key difference from Selenium: there is no switchTo().frame(). In Selenium, you switch the entire driver context to the iframe and then back to the parent. In Playwright, frameLocator() creates a scoped locator context. I can interact with both the main page and the iframe in the same test without any switching.

A: Each tab is a separate Page object. When a link opens a new tab, I use Promise.all with context.waitForEvent('page') and the click. For window.open popups, I use page.waitForEvent('popup'). I create manual tabs with context.newPage(). There is no switchTo().window() like Selenium -- I just use the Page variable directly. The critical rule: set up the event listener BEFORE the click that opens the tab.

A: I use page.once('dialog', handler) to register a handler before the action that triggers the dialog. The handler gets a Dialog object with type() (alert/confirm/prompt), message(), accept(), and dismiss() methods. For prompts, I pass text to accept('my text'). The critical mistake is registering the handler after the click -- Playwright auto-dismisses unhandled dialogs, so the handler never runs. I use once() for single dialogs and on() for multiple.

A: When the file input is hidden behind a custom button, I use the fileChooser event. I set up page.waitForEvent('filechooser') in Promise.all with the click on the custom upload button. The fileChooser object has a setFiles() method where I pass the file path. For visible file inputs, I use the simpler setInputFiles() directly on the input element. I always verify the upload by checking for the filename or a success message.

A: Two approaches. First, locator.dragTo(target) -- one line, works for standard HTML5 drag-and-drop. Second, manual mouse steps for custom drag libraries like react-dnd: get bounding boxes, mouse.move() to source center, mouse.down(), mouse.move() to target center with { steps: 10 } for smooth motion, then mouse.up(). The steps parameter is critical because some libraries need intermediate mousemove events. I always try dragTo() first and fall back to manual only if it fails.

A: Playwright auto-waits for most actions, which eliminates most timing issues. For explicit waits, I use waitForResponse() for API calls, waitForURL() for navigation, waitForSelector({ state: 'hidden' }) for spinners to disappear, and expect assertions that auto-retry. For complex conditions, I use expect.poll(). I never use waitForTimeout() -- that is a hard sleep. When a test is flaky, I find the right condition to wait for instead of adding delays. The right wait is always a condition, never a duration.

A: First, I inspect the DOM to check if it is a native <select> or a custom component. For native, I use selectOption(). For custom dropdowns (React Select, Material UI), I click the trigger to open the list, wait for the listbox to be visible, then click the desired option using getByRole('option'). For searchable dropdowns, I fill() the search text first, wait for filtered options, then click. I always verify the selected value after selection. The most common mistake is trying selectOption() on a custom dropdown -- it throws an error.