Selenium

selenium Best Practices

This guide outlines the definitive best practices for using Selenium with Python. Adhere to these rules to ensure your automation scripts are robust, maintainable, and performant.

1. Code Organization: Page Object Model (POM)

Always structure your Selenium code using the Page Object Model (POM). This separates UI elements and interactions from your test or scraping logic, making your code highly maintainable and reusable.

❌ BAD: Logic and locators mixed

from selenium import webdriver
from selenium.webdriver.common.by import By

driver = webdriver.Chrome()
driver.get("https://example.com/login")
driver.find_element(By.ID, "username").send_keys("user")
driver.find_element(By.ID, "password").send_keys("pass")
driver.find_element(By.XPATH, "//button[text()='Login']").click()
# ... more logic

✅ GOOD: Page Object Model

# pages/base_page.py
from selenium.webdriver.remote.webdriver import WebDriver
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.common.by import By

class BasePage:
    def __init__(self, driver: WebDriver, timeout: int = 10):
        self.driver = driver
        self.wait = WebDriverWait(driver, timeout)

    def _find_element(self, by: By, value: str):
        return self.wait.until(EC.presence_of_element_located((by, value)))

# pages/login_page.py
from pages.base_page import BasePage
from selenium.webdriver.common.by import By

class LoginPage(BasePage):
    URL = "https://example.com/login"
    _USERNAME_INPUT = (By.ID, "username")
    _PASSWORD_INPUT = (By.ID, "password")
    _LOGIN_BUTTON = (By.XPATH, "//button[text()='Login']")

    def open(self):
        self.driver.get(self.URL)

    def login(self, username, password):
        self._find_element(*self._USERNAME_INPUT).send_keys(username)
        self._find_element(*self._PASSWORD_INPUT).send_keys(password)
        self._find_element(*self._LOGIN_BUTTON).click()

# tests/test_login.py (or main scraping script)
from selenium import webdriver
from pages.login_page import LoginPage

def test_successful_login():
    driver = webdriver.Chrome()
    try:
        login_page = LoginPage(driver)
        login_page.open()
        login_page.login("valid_user", "valid_pass")
        assert "dashboard" in driver.current_url # Example assertion
    finally:
        driver.quit()

2. Robustness: Explicit Waits

Never use time.sleep() or rely solely on implicit waits. Use WebDriverWait with expected_conditions to explicitly wait for elements to be in a specific state. This prevents flaky tests and ensures scripts are resilient to dynamic page loading.

❌ BAD: Arbitrary waits or implicit waits

import time
from selenium.webdriver.remote.webdriver import WebDriver
from selenium.webdriver.common.by import By

driver: WebDriver = ...
driver.implicitly_wait(10) # Still prone to issues
driver.get("https://example.com")
time.sleep(5) # NEVER DO THIS
element = driver.find_element(By.ID, "dynamic_content")

✅ GOOD: Explicit Waits with Expected Conditions

from selenium.webdriver.remote.webdriver import WebDriver
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.common.by import By

driver: WebDriver = ...
driver.get("https://example.com")
wait = WebDriverWait(driver, 10)

# Wait for element to be visible
dynamic_element = wait.until(EC.visibility_of_element_located((By.ID, "dynamic_content")))

# Wait for element to be clickable
button = wait.until(EC.element_to_be_clickable((By.CSS_SELECTOR, ".submit-btn")))
button.click()

3. Driver Management & Headless Execution

Always use Selenium Manager (built into Selenium 4.6+) or webdriver-manager to automatically handle browser driver binaries. Run browsers in headless mode for performance and CI/CD environments.

❌ BAD: Manual driver downloads, visible browser

from selenium import webdriver
# Requires chromedriver.exe in PATH or specified manually
driver = webdriver.Chrome(executable_path="/path/to/chromedriver")
driver.get("https://example.com") # Opens a visible browser window

✅ GOOD: Automated driver management, headless with optimized options

from selenium import webdriver
from selenium.webdriver.chrome.options import Options

def get_headless_chrome_driver() -> webdriver.Chrome:
    options = Options()
    options.add_argument("--headless=new")  # Modern headless mode
    options.add_argument("--no-sandbox")    # Essential for CI/Docker
    options.add_argument("--disable-dev-shm-usage") # Avoids /dev/shm issues
    options.add_argument("--window-size=1920,1080") # Ensure consistent rendering
    options.add_argument("--disable-gpu") # Recommended for headless
    options.add_argument("--log-level=3") # Suppress verbose logging
    return webdriver.Chrome(options=options)

driver = get_headless_chrome_driver()
try:
    driver.get("https://example.com")
    print(driver.title)
finally:
    driver.quit() # ALWAYS quit the driver

4. Locator Strategy Prioritization

Choose locators that are robust and least likely to change. Prioritize By.ID and By.CSS_SELECTOR. Avoid fragile By.XPATH unless absolutely necessary, and By.CLASS_NAME for elements with multiple classes.

❌ BAD: Fragile Locators

# Absolute XPath (breaks with any DOM change)
element = driver.find_element(By.XPATH, "/html/body/div[1]/div[2]/form/input[3]")
# Class name for multiple classes (ambiguous)
element = driver.find_element(By.CLASS_NAME, "btn btn-primary active")

✅ GOOD: Robust Locators

# Unique ID (most reliable)
element = driver.find_element(By.ID, "submitButton")
# CSS Selector (flexible, powerful, and stable)
element = driver.find_element(By.CSS_SELECTOR, "button.submit-btn[name='action']")
# Partial Link Text (for navigation)
element = driver.find_element(By.PARTIAL_LINK_TEXT, "Read More")

5. Error Handling & Teardown

Implement robust try...finally blocks to ensure driver.quit() is always called, preventing lingering browser processes. Catch common Selenium exceptions for graceful failure and better debugging.

from selenium import webdriver
from selenium.common.exceptions import NoSuchElementException, TimeoutException

driver = get_headless_chrome_driver() # Using the good practice from above
try:
    driver.get("https://example.com/nonexistent")
    # Attempt to find an element that might not exist
    element = driver.find_element(By.ID, "some_element")
    element.click()
except NoSuchElementException:
    print("Element not found. Check locator or page state.")
except TimeoutException:
    print("Page load or element wait timed out.")
except Exception as e:
    print(f"An unexpected error occurred: {e}")
finally:
    driver.quit() # Ensures browser process is terminated

6. Type Hints

Always use Python type hints. They improve code readability, enable static analysis, and reduce bugs, especially in larger projects and Page Object Models.

from selenium.webdriver.remote.webdriver import WebDriver
from selenium.webdriver.remote.webelement import WebElement
from selenium.webdriver.common.by import By
from typing import Tuple

class MyPage:
    _SEARCH_INPUT: Tuple[By, str] = (By.ID, "search")

    def __init__(self, driver: WebDriver):
        self.driver = driver

    def search(self, query: str) -> None:
        search_box: WebElement = self.driver.find_element(*self._SEARCH_INPUT)
        search_box.send_keys(query)
        search_box.submit()

7. Virtual Environments & Packaging

Always use a virtual environment (venv or uv) for dependency management. For modern projects, use uv and pyproject.toml for fast, reproducible builds.

# Recommended modern setup with uv
uv init my-selenium-project
cd my-selenium-project
uv add selenium pytest # Add your dependencies

# Or classic venv
python -m venv .venv
source .venv/bin/activate # Linux/macOS
.venv\Scripts\Activate.ps1 # Windows PowerShell
pip install selenium pytest