How to check URL for 404 using Selenium WebDriver?

Question

What is the most convenient way using Selenium WebDriver to check if an URL GET returns successfully (HTTP 200)?

In this particular case I'm most interested in verifying that no images of the current page are broken.

Answer 1

You could use the getEval command to verify the value returned from the following JavaScript for each image on the page.

@Test
public void checkForBrokenImages() {
    selenium.open("http://www.example.com/");
    int imageCount = selenium.getXpathCount("//img").intValue();
    for (int i = 0; i < imageCount; i++) {
        String currentImage = "this.browserbot.getUserWindow().document.images[" + i + "]";
        assertEquals(selenium.getEval("(!" + currentImage + ".complete) ? false : !(typeof " + currentImage + ".naturalWidth != \"undefined\" && " + currentImage + ".naturalWidth == 0);"), "true", "Broken image: " + selenium.getEval(currentImage + ".src"));
 }
}

Updated:

Added tested TestNG/Java example.

Answer 2

Try this:

List<WebElement> allImages = driver.findElements(By.tagName("img"));
for (WebElement image : allImages) {
  boolean loaded = ((JavaScriptExecutor) driver).executeScript(
      "return arguments[0].complete", image);
  if (!loaded) {
    // Your error handling here.
  }
}

Answer 3

I don't think that first response will work. When I src a misnamed image, it throws a 404 error as expected. However, when I check the javascript in firebug, that (broken) image has .complete set to true. So, it was a completed 404, but still a broken image.

The second response seems to be more accurate in that it checks that it's complete and then checks that there is some width to the image.

I made a python version of the second response that works for me. Could be cleaned up a bit, but hopefully it will help.

def checkForBrokenImages(self):
    sel = self.selenium
    imgCount = int(sel.get_xpath_count("//img"))
    for i in range(0,imgCount):
        isComplete = sel.get_eval("selenium.browserbot.getCurrentWindow().document.images[" + str(i) + "].complete")
        self.assertTrue(isComplete, "Bad Img (!complete): "+sel.get_eval("selenium.browserbot.getCurrentWindow().document.images[" + str(i) + "].src"))
        typeOf = sel.get_eval("typeof selenium.browserbot.getCurrentWindow().document.images[" + str(i) + "].naturalWidth")
        self.assertTrue(typeOf != 'undefined', "Bad Img (w=undef): "+sel.get_eval("selenium.browserbot.getCurrentWindow().document.images[" + str(i) + "].src"))
        natWidth = int(sel.get_eval("selenium.browserbot.getCurrentWindow().document.images[" + str(i) + "].naturalWidth"))
        self.assertTrue(natWidth > 0, "Bad Img (w=0): "+sel.get_eval("selenium.browserbot.getCurrentWindow().document.images[" + str(i) + "].src"))

Answer 4

Hi Dave,

Can you help me with the code to check for the broken links on the page please.

Regards

Answer 5

One of the alternative solutions is analyzing web server logs after test executing. This approach allows to catch not only missed images, but css, scripts and other resources.

Description of how to do it is here.

Answer 6

yes please,

Someone can help with the code to check for broken links on any page using selenium