使用 Python 进行验证码抓取：实用指南

Python 的 requests 库可以有效地处理 HTTP，但验证码需要外部求解器。本指南展示了如何将 CaptchaAI 集成到 Python 抓取脚本中 - 大多数网站不需要浏览器。

要求

CaptchaAI 帮助程序类

为您的 Python 项目构建可重用的求解器类：

import requests
import time

class CaptchaSolver:
    def __init__(self, api_key):
        self.api_key = api_key
        self.base = "https://ocr.captchaai.com"

    def _submit(self, params):
        params["key"] = self.api_key
        resp = requests.get(f"{self.base}/in.php", params=params)
        if not resp.text.startswith("OK|"):
            raise Exception(f"Submit error: {resp.text}")
        return resp.text.split("|")[1]

    def _poll(self, task_id, timeout=300):
        deadline = time.time() + timeout
        while time.time() < deadline:
            time.sleep(5)
            resp = requests.get(f"{self.base}/res.php", params={
                "key": self.api_key,
                "action": "get",
                "id": task_id
            })
            if resp.text == "CAPCHA_NOT_READY":
                continue
            if resp.text.startswith("OK|"):
                return resp.text.split("|")[1]
            raise Exception(f"Solve error: {resp.text}")
        raise TimeoutError("Solve timed out")

    def solve_recaptcha_v2(self, site_key, page_url):
        task_id = self._submit({
            "method": "userrecaptcha",
            "googlekey": site_key,
            "pageurl": page_url
        })
        return self._poll(task_id)

    def solve_recaptcha_v3(self, site_key, page_url, action="verify"):
        task_id = self._submit({
            "method": "userrecaptcha",
            "googlekey": site_key,
            "pageurl": page_url,
            "version": "v3",
            "action": action
        })
        return self._poll(task_id)

    def solve_turnstile(self, site_key, page_url):
        task_id = self._submit({
            "method": "turnstile",
            "sitekey": site_key,
            "pageurl": page_url
        })
        return self._poll(task_id)

    def solve_image(self, image_base64):
        task_id = self._submit({
            "method": "base64",
            "body": image_base64
        })
        return self._poll(task_id)

抓取受 reCAPTCHA 保护的表单

from bs4 import BeautifulSoup
import requests

solver = CaptchaSolver("YOUR_API_KEY")
session = requests.Session()
session.headers.update({
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"
})

# Step 1: Load the page
url = "https://example.com/search"
page = session.get(url)
soup = BeautifulSoup(page.text, "html.parser")

# Step 2: Extract the site key
recaptcha_div = soup.find("div", class_="g-recaptcha")
site_key = recaptcha_div["data-sitekey"]

# Step 3: Solve the CAPTCHA
token = solver.solve_recaptcha_v2(site_key, url)

# Step 4: Submit the form with the token
form_data = {
    "q": "search term",
    "g-recaptcha-response": token
}
result = session.post(url, data=form_data)

# Step 5: Parse the results
result_soup = BeautifulSoup(result.text, "html.parser")
items = result_soup.find_all("div", class_="result-item")
for item in items:
    print(item.text.strip())

抓取多个页面

对于验证码后面的分页结果：

def scrape_all_pages(base_url, site_key, max_pages=10):
    solver = CaptchaSolver("YOUR_API_KEY")
    session = requests.Session()
    session.headers.update({
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"
    })
    all_results = []

    for page_num in range(1, max_pages + 1):
        page_url = f"{base_url}?page={page_num}"

        # Solve CAPTCHA for each page if needed
        token = solver.solve_recaptcha_v2(site_key, page_url)

        resp = session.get(page_url, params={
            "g-recaptcha-response": token,
            "page": page_num
        })

        soup = BeautifulSoup(resp.text, "html.parser")
        items = soup.find_all("div", class_="item")

        if not items:
            break

        all_results.extend([item.text.strip() for item in items])
        print(f"Page {page_num}: {len(items)} items")

        time.sleep(2)  # Polite delay

    return all_results

处理图像验证码

对于具有基于图像的文本验证码的网站：

import base64

def scrape_with_image_captcha(url):
    solver = CaptchaSolver("YOUR_API_KEY")
    session = requests.Session()

    page = session.get(url)
    soup = BeautifulSoup(page.text, "html.parser")

    # Find the CAPTCHA image
    captcha_img = soup.find("img", {"id": "captcha-image"})
    captcha_url = captcha_img["src"]

    # Download and encode the image
    img_resp = session.get(captcha_url)
    img_base64 = base64.b64encode(img_resp.content).decode()

    # Solve
    captcha_text = solver.solve_image(img_base64)

    # Submit
    form_data = {
        "captcha": captcha_text,
        "username": "user"
    }
    result = session.post(url, data=form_data)
    return result.text

错误处理和重试

为生产刮刀添加重试逻辑：

def solve_with_retry(solver, site_key, page_url, max_retries=3):
    for attempt in range(max_retries):
        try:
            return solver.solve_recaptcha_v2(site_key, page_url)
        except Exception as e:
            if attempt == max_retries - 1:
                raise
            print(f"Attempt {attempt + 1} failed: {e}. Retrying...")
            time.sleep(2)

故障排除

常问问题

我需要 Selenium 在 Python 中进行验证码抓取吗？

并非总是如此。如果站点的表单适用于标准 HTTP POST 请求，则 requests + CaptchaAI 比 Selenium 更快、更轻。仅当站点需要 JavaScript 渲染时才使用 Selenium。

我可以异步解决验证码吗？

是的。将 aiohttp 与 CaptchaAI 的 API 结合使用以实现异步工作流程。看aiohttp + CaptchaAI 集成。

如何处理速率限制？

在请求之间添加延迟 (time.sleep(2-5))、QA 测试会话并使用真实的标头。看用于验证码抓取的代理轮换。

相关指南

• 使用 Python 处理 Selenium 验证码
• 使用 Node.js 进行验证码抓取
• 抓取而不被阻止