050-tinder-auto-find-match/docs/wiki/Home.md

Day 50 - Tinder auto find match (English)

Dansk · Bahasa Malaysia

Goal: automate logging in and liking profiles on TinDog with Python and Selenium. Ethics first: tell your partner you automate, and respect the daily limit (TinDog 20, Tinder free tier 100).

Step 1 - Look at the site by hand

What we gained: before coding, click through manually and note which elements you must target. Good locators come from looking at the page first.

Step 2 - Navigate to the login page

What we gained: driver.get, locating buttons with XPath (locators.py) and waiting for them with WebDriverWait + element_to_be_clickable instead of fixed sleep calls. This avoids NoSuchElementException on slow pages.

Step 3 - The Facebook pop-up

What we gained: the login opens a second browser window. We switch with driver.window_handles and switch_to.window, fill the form, then switch back. Credentials live in environment variables, never in code.

Step 4 - Dismiss all requests

Location (04), notifications (05), cookies (06).

What we gained: modal pop-ups block everything behind them. We click ALLOW, NOT INTERESTED and I ACCEPT, and skip a pop-up that does not appear (TimeoutException is handled and logged).

Step 5 - Hit Like

What we gained: exception handling as control flow for a flaky UI: NoSuchElementException means "still loading, wait 2 s and retry"; ElementClickInterceptedException means "a match pop-up is in the way, click BACK TO TINDER". A pause between likes keeps the bot from looking like a bot. The loop stops at the configured limit.

Modernisation compared with the earlier solution

• src/ package with one function per step instead of one class with sleep.
• Explicit waits instead of fixed sleeps; locators in one module.
• Facebook login (as the lecture asks) instead of Google; settings via a Settings dataclass; unit tests with a fake driver.
• No from constants import *, no HTML dumps, no profile path with os.getcwd.