如何在 MechanicalSoup 中识别 reCAPTCHA 并提交表单

# Python 3.10 or later; MechanicalSoup brings requests, bs4 and lxml
pip install capskip mechanicalsoup

# pip install capskip mechanicalsoup
import mechanicalsoup

PAGE_URL = "https://example.com/login"

# One StatefulBrowser per session: it keeps cookies, follows redirects
# and sends the current page as Referer when it submits a form.
browser = mechanicalsoup.StatefulBrowser(raise_on_404=True)
browser.open(PAGE_URL)

# The checkbox div, or the button on invisible reCAPTCHA, holds the key.
widget = browser.page.select_one(".g-recaptcha[data-sitekey]")
if widget is None:
    raise RuntimeError("No reCAPTCHA widget in the HTML; see the v3 section.")
sitekey = widget["data-sitekey"]

# The form around the widget; a page often has a search form first.
browser.select_form(widget.find_parent("form") or 'form[method="post" i]')
print(sitekey, browser.url)

from capskip import CapSkip

solver = CapSkip(host="127.0.0.1", port=8080)

# enterprise.js in place of api.js means reCAPTCHA Enterprise.
enterprise = int("recaptcha/enterprise.js" in str(browser.page))
if widget.name == "button" and widget.get("data-action"):
    # Google's v3 button binding: a g-recaptcha button with an action is v3.
    result = solver.recaptcha(sitekey=sitekey, url=browser.url, version="v3",
                              action=widget["data-action"], enterprise=enterprise)
else:
    # A g-recaptcha button, or data-size="invisible", is the invisible widget.
    invisible = widget.name == "button" or widget.get("data-size") == "invisible"
    result = solver.recaptcha(sitekey=sitekey, url=browser.url,
                              invisible=int(invisible), enterprise=enterprise)

token = result["code"]
print(token[:40])   # what the widget would have written into the form

# This line raises LinkNotFoundError on most pages: the widget script
# creates the field in a real browser, so the parsed HTML lacks it.
# browser["g-recaptcha-response"] = token

# Add the field the widget would have written, replacing any copy.
browser.new_control("hidden", "g-recaptcha-response", token)
browser["email"] = "YOUR_EMAIL"
browser["password"] = "YOUR_PASSWORD"

response = browser.submit_selected()
print(response.status_code, browser.url)

import re

# A v3 page has no widget element; the key sits in the script tags.
html = str(browser.page)
key = re.search(r"recaptcha/(api|enterprise)\.js\?[^\"'<>\s]*?\brender=(?!explicit\b)([\w-]+)", html)
action = re.search(r"execute\(\s*[^,()]+,\s*\{\s*[\"']?action[\"']?\s*:\s*[\"']([\w/]+)", html)
if key is None:
    raise RuntimeError("No v3 render key in the HTML; check DevTools.")

v3 = solver.recaptcha(sitekey=key.group(2), url=browser.url, version="v3",
                      enterprise=int(key.group(1) == "enterprise"),
                      **({"action": action.group(1)} if action else {}))

browser.select_form('form[method="post" i]')
# The field the page's own script fills in; read its name in DevTools.
browser.new_control("hidden", "g-recaptcha-response", v3["code"])

# pip install capskip mechanicalsoup
# login.py: run with python login.py
import os
import re

import mechanicalsoup
from capskip import CapSkip, CapSkipError, TimeoutException

PAGE_URL = "https://example.com/login"

# render= carries a v3 key on api.js or enterprise.js; render=explicit is a v2 page.
V3_KEY = re.compile(r"recaptcha/(api|enterprise)\.js\?[^\"'<>\s]*?\brender=(?!explicit\b)([\w-]+)")
# The action passed to execute, whether the key is a string or a variable.
V3_ACTION = re.compile(r"execute\(\s*[^,()]+,\s*\{\s*[\"']?action[\"']?\s*:\s*[\"']([\w/]+)")

solver = CapSkip(
    apiKey=os.getenv("CAPSKIP_API_KEY", "capskip"),
    host=os.getenv("CAPSKIP_HOST", "127.0.0.1"),
    port=int(os.getenv("CAPSKIP_PORT", "8080")),
)


def solve_page(browser):
    """Solve the page's reCAPTCHA and select its form. Returns the token."""
    html = str(browser.page)
    enterprise = int("recaptcha/enterprise.js" in html)
    widget = browser.page.select_one(".g-recaptcha[data-sitekey]")
    if widget is not None:
        browser.select_form(widget.find_parent("form") or 'form[method="post" i]')
        if widget.name == "button" and widget.get("data-action"):
            # Google's v3 button binding: plain api.js, the action on the button.
            return solver.recaptcha(sitekey=widget["data-sitekey"], url=browser.url, version="v3",
                                    action=widget["data-action"], enterprise=enterprise)["code"]
        invisible = widget.name == "button" or widget.get("data-size") == "invisible"
        return solver.recaptcha(sitekey=widget["data-sitekey"], url=browser.url,
                                invisible=int(invisible), enterprise=enterprise)["code"]
    key = V3_KEY.search(html)
    if key is None:
        raise RuntimeError(f"No reCAPTCHA found on {browser.url}")
    action = V3_ACTION.search(html)
    browser.select_form('form[method="post" i]')
    return solver.recaptcha(sitekey=key.group(2), url=browser.url, version="v3",
                            enterprise=int(key.group(1) == "enterprise"),
                            **({"action": action.group(1)} if action else {}))["code"]


def login(email, password, page_url=PAGE_URL):
    # A new browser per login, so sessions never share cookies.
    with mechanicalsoup.StatefulBrowser(raise_on_404=True) as browser:
        browser.open(page_url)
        token = solve_page(browser)
        browser.new_control("hidden", "g-recaptcha-response", token)
        browser["email"] = email
        browser["password"] = password
        response = browser.submit_selected()
        return response.status_code, browser.url


if __name__ == "__main__":
    try:
        print(login("YOUR_EMAIL", "YOUR_PASSWORD"))
    except TimeoutException:
        raise SystemExit("gave up after recaptchaTimeout")
    except CapSkipError as exc:
        raise SystemExit(f"solve failed: {type(exc).__name__}: {exc}")
    except mechanicalsoup.LinkNotFoundError as exc:
        # A 404, or a form or field the page does not have.
        raise SystemExit(f"not found: {str(exc) or PAGE_URL}")
    except RuntimeError as exc:
        raise SystemExit(str(exc))

# pool.py: save next to login.py, the full example, and run this file
from concurrent.futures import ThreadPoolExecutor

from login import login

accounts = [("YOUR_EMAIL_1", "YOUR_PASSWORD_1"), ("YOUR_EMAIL_2", "YOUR_PASSWORD_2")]


def attempt(account):
    # One failed login must not stop the others.
    try:
        return login(*account)
    except Exception as exc:
        return f"{type(exc).__name__}: {exc}"


# Ten at a time, to match CapSkip's default Max. Threads for reCAPTCHA.
with ThreadPoolExecutor(max_workers=10) as pool:
    for account, outcome in zip(accounts, pool.map(attempt, accounts)):
        print(account[0], outcome)