利用 Python + Selenium 实现对页面的指定元素截图(可截长图元素)

对WebElement截图

WebDriver.Chrome自带的方法只能对当前窗口截屏，且不能指定特定元素。若是需要截取特定元素或是窗口超过了一屏，就只能另辟蹊径了。
WebDriver.PhantomJS自带的方法支持对整个网页截屏。
下面提供几种思路。

方式一

针对WebDriver.Chrome

通过WebDriver的js脚本注入功能，曲线救国。
1. 注入第三方html转canvas的js库(见下方推荐)
2. 获取元素html源码
3. 将html转换为canvas
4. 下载canvas

优点: 截取长图容易实现
缺点: 加载第三方库耗费时间，转换原理请参考这篇文章:
将 DOM 对象绘制到 canvas 中

方式二

针对WebDriver.Chrome

截取全图，自行裁剪、拼接
1. 获取元素位置、大小
2. 获取窗口大小
3. 截取包含元素的窗口
4. 进行相应的裁剪和拼接。

具体算法思路很清晰，但需要注意的细节较多。这里就不在赘述。示例代码请移步:
[Github]PythonSpiderLibs

优点: 不需太多js工作，python+少量js代码即可完成
缺点: 拼接等工作会被WebDriver的实现差异、图片加载速度等因素影响，需多加注意。 在保证截图质量的情况下，速度较慢

方式三

针对WebDriver.PhantomJS
由于接口实现的差异，PhantomJS相比于Chrome，可以截取到整个网页。所以获取指定元素的截图也就简单很多

1. 截取网页全图
2. 裁剪指定元素

Python

driver = webdriver.Chrome() driver.<span class="hljs-keyword">get</span>(<span class="hljs-string">'<a class="vglnk" href="http://stackoverflow.com/" rel="nofollow">http://stackoverflow.com/</a>'</span>) driver.save_screenshot(<span class="hljs-string">'screenshot.png'</span>) left = element.location[<span class="hljs-string">'x'</span>] top = element.location[<span class="hljs-string">'y'</span>] right = element.location[<span class="hljs-string">'x'</span>] + element.size[<span class="hljs-string">'width'</span>] bottom = element.location[<span class="hljs-string">'y'</span>] + element.size[<span class="hljs-string">'height'</span>] im = Image.open(<span class="hljs-string">'screenshot.png'</span>) im = im.crop((left, top, right, bottom)) im.save(<span class="hljs-string">'screenshot.png'</span>)

1 2 3 4 5 6 7 8 9 10 11 12 13

driver = webdriver . Chrome ( ) driver . < span class = "hljs-keyword" > get < / span > ( < span class = "hljs-string" > '<a class="vglnk" href="http://stackoverflow.com/" rel="nofollow">http://stackoverflow.com/</a>' < / span > ) driver . save_screenshot ( < span class = "hljs-string" > 'screenshot.png' < / span > )   left = element . location [ < span class = "hljs-string" > 'x' < / span > ] top = element . location [ < span class = "hljs-string" > 'y' < / span > ] right = element . location [ < span class = "hljs-string" > 'x' < / span > ] + element . size [ < span class = "hljs-string" > 'width' < / span > ] bottom = element . location [ < span class = "hljs-string" > 'y' < / span > ] + element . size [ < span class = "hljs-string" > 'height' < / span > ]   im = Image . open ( < span class = "hljs-string" > 'screenshot.png' < / span > ) im = im . crop ( ( left , top , right , bottom ) ) im . save ( < span class = "hljs-string" > 'screenshot.png' < / span > )

优点: 实现简单
缺点: 对于高度太高的页面会导致文件过大，处理会有问题，我测试的最大图片尺寸是12.8M。

解决图片加载不完整的问题

参考: 利用 Python + Selenium 自动化快速截图

我们先在首页上执行一段 JavaScript 脚本，将页面的滚动条拖到最下方，然后再拖回顶部，最后才截图。这样可以解决像上面那种按需加载图片的情况。

Python

<span class="hljs-keyword">from</span> <span class="wp_keywordlink"><a href="http://www.168seo.cn/selenium/" title="selenium">selenium</a></span> <span class="hljs-keyword">import</span> webdriver <span class="hljs-keyword">import</span> time <span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">take_screenshot</span><span class="hljs-params">(url, save_fn=<span class="hljs-string">"capture.png"</span>)</span>:</span> browser = webdriver.Firefox() <span class="hljs-comment"># Get local session of firefox</span> browser.set_window_size(<span class="hljs-number">1200</span>, <span class="hljs-number">900</span>) browser.get(url) <span class="hljs-comment"># Load page</span> browser.execute_script(<span class="hljs-string">""" (function () { var y = 0; var step = 100; window.scroll(0, 0); function f() { if (y < document.body.scrollHeight) { y += step; window.scroll(0, y); setTimeout(f, 100); } else { window.scroll(0, 0); document.title += "scroll-done"; } } setTimeout(f, 1000); })(); """</span>) <span class="hljs-keyword">for</span> i <span class="hljs-keyword">in</span> xrange(<span class="hljs-number">30</span>): <span class="hljs-keyword">if</span> <span class="hljs-string">"scroll-done"</span> <span class="hljs-keyword">in</span> browser.title: <span class="hljs-keyword">break</span> time.sleep(<span class="hljs-number">10</span>) browser.save_screenshot(save_fn) browser.close() <span class="hljs-keyword">if</span> __name__ == <span class="hljs-string">"__main__"</span>: take_screenshot(<span class="hljs-string">"<a class="vglnk" href="http://codingpy.com/" rel="nofollow">http://codingpy.com</a>"</span>)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

< span class = "hljs-keyword" > from < / span > selenium < span class = "hljs-keyword" > import < / span > webdriver < span class = "hljs-keyword" > import < / span > time     < span class = "hljs-function" > < span class = "hljs-keyword" > def < / span > < span class = "hljs-title" > take_screenshot < / span > < span class = "hljs-params" > ( url , save_fn = < span class = "hljs-string" > "capture.png" < / span > ) < / span > : < / span >      browser = webdriver . Firefox ( ) < span class = "hljs-comment" > # Get local session of firefox</span>      browser . set_window_size ( < span class = "hljs-number" > 1200 < / span > , < span class = "hljs-number" > 900 < / span > )      browser . get ( url ) < span class = "hljs-comment" > # Load page</span>      browser . execute_script ( < span class = "hljs-string" > """         (function () {             var y = 0;             var step = 100;             window.scroll(0, 0);               function f() {                 if (y < document.body.scrollHeight) {                     y += step;                     window.scroll(0, y);                     setTimeout(f, 100);                 } else {                     window.scroll(0, 0);                     document.title += "scroll-done";                 }             }               setTimeout(f, 1000);         })();     """ < / span > )        < span class = "hljs-keyword" > for < / span > i < span class = "hljs-keyword" > in < / span > xrange ( < span class = "hljs-number" > 30 < / span > ) :          < span class = "hljs-keyword" > if < / span > < span class = "hljs-string" > "scroll-done" < / span > < span class = "hljs-keyword" > in < / span > browser . title :              < span class = "hljs-keyword" > break < / span >          time . sleep ( < span class = "hljs-number" > 10 < / span > )        browser . save_screenshot ( save_fn )      browser . close ( )     < span class = "hljs-keyword" > if < / span > __name__ == < span class = "hljs-string" > "__main__" < / span > :        take_screenshot ( < span class = "hljs-string" > "<a class=" vglnk " href=" http : / / codingpy . com / " rel=" nofollow ">http://codingpy.com</a>" < / span > )

不同wewbdriver对某些方法的实现不同

Chrome和PhantomJS 的接口差异
抓知乎时的坑，
1. Chrome用WebElement.text可以正常得到值，用PhantomJS只能用 WebElement.get_attribute('innerHTML')
2. WebDriver.Chrome截图只能截当前屏幕区域。WebDriver.PhantomJS截图可以获取整个页面的长图。

其它还有一些坑等待发现

• zeropython 微信公众号 5868037 QQ号 5868037@qq.com QQ邮箱