Full Screenshot function not taking images for a page bigger than the window

[andrewfgarrison]

Using selenium-screenshot in a selenium script produces no errors, but when using the full-screenshot function on a page with a larger height than the viewing window, it only takes an image of the top portion despite fully scrolling down. The browser is Chrome and all modules are run through Conda.

[sam4u3]

Can you share snippet pf code which you are using?

[andrewfgarrison]

This is the function that includes the full screenshot call:

def getData():
    date = datetime.datetime.today()
    new_dir_path = date.now().strftime('%Y%m%d')
    if not os.path.exists(new_dir_path):
        os.makedirs(new_dir_path)
    with open(new_dir_path+'.csv', mode='w') as infor_file:
        infor_writer = csv.writer(
            infor_file, delimiter=',', quotechar='"', quoting=csv.QUOTE_MINIMAL)
        infor_writer.writerow(
            ['案件', '詳細', '一時/確定URL', '注文日', 'デザイン提出日', 'オプション'])
        for obj in list:
            driver.get(obj.detail_link)
            driver.implicitly_wait(1)
            btn = driver.find_element(By.XPATH, 
                "*//div[@id='GoodsInfo']/div/div/div[contains(., <text>)][1]")
            btn.click()
            driver.implicitly_wait(1)
            order_date = driver.find_element(By.XPATH, 
                "*//div[@id='GoodsInfo']/div/div/div[contains(., <text>)][1]/parent::div/div/table/tbody/tr[3]/th[contains(.,<text>)]/following::td").text
            option = driver.find_element(By.XPATH, 
                "*//div[@id='GoodsInfo']/div/div/div[contains(., <text>)][1]/parent::div/div/table/tbody/tr[7]/th[contains(.,'AI_OP')]/following::td").text
            try:
                design_due = driver.find_element(By.XPATH, 
                    "*//td[contains(.,<text>)]/parent::tr/td[2]").get_attribute("innerText")
            except (TimeoutException, NoSuchElementException):
                design_due = <text>
            try:
                url = driver.find_element(By.XPATH, 
                    "*//div[@id='GoodsInfo']/div/div/div[contains(., <text>)][1]/parent::div/div/table/tbody/tr[2]/th[contains(.,<text>)]/following::td/p/a")
                href = url.get_attribute('href')
                m = href.replace('://', <text>)
            except (TimeoutException, NoSuchElementException):
                url = driver.find_element(By.XPATH, 
                    "*//div[@id='GoodsInfo']/div/div/div[contains(., <text>)][1]/parent::div/div/table/tbody/tr[1]/th[contains(.,<text>)]/following::td/p/a")
                href = url.get_attribute('href')
                m = href.replace('http://', <text>)
            finally:
                driver.implicitly_wait(5)
                driver.execute_script('browserstack_executor: {"action": "sendBasicAuth", "arguments": {"username": <text>, "password": <text>, "timeout": "5000"}\}')
            infor_writer.writerow(
                [obj.name, obj.detail_link, href, order_date, design_due, option])
            print(obj.name, obj.detail_link, href,
                 <text> + order_date, <text> + design_due, option)
            driver.get(m)
            ob = Screenshot.Screenshot()
            ob.full_Screenshot(driver, save_path=new_dir_path, image_name=obj.name+'.png')

Sensitive information has been removed and replaced with <text>

[arturgeller6]

The same problem for me. After updating selenium and Selenium_Screenshot it does not capture webpages with large height anymore. It used to work before update

[arturgeller6]

tried older version of Selenium_Screenshot and it worked fine. so it is Selenium_Screenshot update and not selenium update that caused the problems.

[leandroeidi]

I have the same problem. Chrome, Firefox and Edge, all screenshots are only for the top part of the page...

[Fer-Bar]

I got the same issue. Why would be the reason?

[jakubstephan]

The same on my side with Python 3.8, ChromeDriver 107.0.5304.62 :(

Page is scrolling to the bottom, but screenshot contains only first part of webpage.

[adamws]

with multi_images=True, for example:

ob = Screenshot.Screenshot()
selenium.get("https://github.com/PyWizards/Selenium_Screenshot")
img_url = ob.full_Screenshot(selenium, save_path=r'.', image_name='Myimage.png', multi_images=True)

it is kind of ok, although there are gaps between screenshots and page bottom is missing (at least using firefox, didn't check others):

for me, quick & dirty fix looks like this:

M Screenshot/Screenshot.py
@@ -100,59 +100,30 @@ class Screenshot:
                     rectangles.append((ii, i, top_width, top_height))
                     ii = ii + viewport_width
                 i = i + viewport_height
-            stitched_image = None
-            previous = None
             part = 0
 
-            # This value indicates if is the first part of the screenshot, or it is the rest of the screenshot
-            init_canvas = True
-            # Get the screenshot and save the dimensions of the image, this will be useful for create a
-            # correct canvas with correct dimensions
-            # and not show images with a wrong size or cut
-            size_screenshot = self.__get_screen_size(driver)
-            # Get temporary sum of height of the total pages
-            height_canvas = size_screenshot['height'] * len(rectangles)
-
-            # Compare if the screenshot it's only one part and assign the correct dimensions of the screenshot size
-            # Or assign the height of the total screenshot
-            if multi_images:
-                # Assign the height of the total screenshot at canvas even if not used at the moment
-                stitched_image = Image.new('RGB', (size_screenshot['width'], height_canvas))
-            else:
-                # Assign the dimensions corresponding to the size of the uniq screenshot
-                stitched_image = Image.new('RGB', (size_screenshot['width'], size_screenshot['height']))
-
-            # This constant is used top offset the image in the canvas
-            jump = round(size_screenshot['height'] / 2)
-            # With take multiple screenshots this value is used to adjust the position of the image in the canvas
-            # This value know update every time the screenshot is taken
-            top_offset = jump
+            stitched_image = Image.new('RGB', (total_width, total_height))
 
+            height_remaining = total_height
             for rectangle in rectangles:
-                if previous is not None:
-                    driver.execute_script("window.scrollTo({0}, {1})".format(rectangle[0], rectangle[1]))
-                    time.sleep(1)
-                    self.hide_elements(driver, elements)
+                driver.execute_script("window.scrollTo({0}, {1})".format(rectangle[0], rectangle[1]))
+                time.sleep(1)
+                self.hide_elements(driver, elements)
 
                 file_name = "part_{0}.png".format(part)
                 driver.get_screenshot_as_file(file_name)
                 screenshot = Image.open(file_name)
+                if screenshot.height > height_remaining:
+                    screenshot = screenshot.crop((0, screenshot.height - height_remaining, screenshot.width, screenshot.height))
 
-                if rectangle[1] + viewport_height > total_height:
-                    offset = (rectangle[0], rectangle[1] + top_offset)
-                else:
-                    if init_canvas:
-                        offset = (rectangle[0], rectangle[1])
-                        init_canvas = False
-                    else:
-                        offset = (rectangle[0], rectangle[1] + top_offset)
-                        top_offset = jump + top_offset
+                offset = (rectangle[0], rectangle[1])
 
                 stitched_image.paste(screenshot, offset)
+                height_remaining = height_remaining - screenshot.height
                 del screenshot
                 os.remove(file_name)
                 part = part + 1
-                previous = rectangle
+
             save_path = os.path.abspath(os.path.join(save_path, image_name))
             stitched_image.save(save_path)
             return save_path

which results in:

but since full_Screenshot is used also in get_element this may not be proper solution. This may also be broken for other browsers.

[sam4u3]

I think this is long pending issue. I dont get much time to fix this issue. But someone has to contribute and fix.

[nirtal85]

saw this issue also, it's not working.

[elyamurza]

Hi, any updates?

[sam4u3]

updated code by @feelgood-interface . just check and re-open the issue if problem persists. Thank you @feelgood-interface