google-search-playwright 4.0.2

Unofficial Python bindings to the Google search engine using Playwright. Not affiliated with Google.

Google Search (Playwright)

Google search from Python, powered by Playwright with automatic browser detection and interactive CAPTCHA solving.

Note: this project is not affiliated with Google in any way.

Why this fork?

Google's web search interface now requires JavaScript for most requests, which broke the classic urllib + BeautifulSoup scraping used by the original googlesearch library (pip install google). This package replaces that stack with a real headless browser driven by Playwright, so it keeps working against the current Google search experience.

It also auto-detects the browser already installed on your system (Chromium, Chrome, Edge, or Firefox) — no Playwright browser downloads required.

Installing

pip install google-search-playwright

Requirements

• Python 3.8+
• A Chromium, Chrome, Chrome-based (e.g. Microsoft Edge), or Firefox browser installed on your system.

The browser is found automatically. See find_browser() for the search order.

Usage example

from googlesearch import search

# Get the first 20 hits for: "Breaking Code" WordPress blog
for url in search('"Breaking Code" WordPress blog', stop=20):
    print(url)

from googlesearch import search, lucky

# "I'm feeling lucky"
first = lucky('python programming')
print(first)  # https://www.python.org/

CAPTCHA handling

If Google flags your IP and returns a CAPTCHA page, the library automatically opens a visible browser window so you can solve it interactively:

============================================================
Google requires a CAPTCHA verification. A visible browser
window has opened.

  1. Solve the captcha in the browser.
  2. When you see search results, return here and press Enter.

The search will then continue in the background using the
authenticated session.
============================================================
Press Enter after solving the captcha...

Once solved, the authenticated cookies are carried into the rest of the headless session so subsequent requests succeed. A CAPTCHA prompt only appears once per search; if it still fails afterward a ConnectionError is raised.

Command-line usage

A google console script is provided:

google "python programming" --stop 5
google "linux kernel" --browser chromium
google "breaking news" --no-stealth

Run google --help for all options.

Browser auto-detection

The library searches in this order:

1. Chromium/Chrome family via PATH: chromium, chromium-browser, google-chrome, google-chrome-stable, chrome, chrom, microsoft-edge, edge.
2. Chromium/Chrome family at platform-specific paths (macOS .app bundles, Windows Program Files).
3. Firefox via PATH: firefox, firefox-esr, firefox-nightly.
4. Firefox at platform-specific paths.

You can override this with search(..., browser_type="firefox", exe_path="/path").

API reference

• search(query, ...) — generator yielding result URLs
• lucky(query, ...) — returns the first result URL as a string
• get_tbs(from_date, to_date) — builds the tbs date-range parameter
• find_browser() — returns (browser_type, executable_path) or None

Full Sphinx docs: https://google-search-playwright.readthedocs.io/en/latest/

License

BSD 3-Clause. See LICENSE.