Add pdf2htmlEX workaround to screenshotter in html_render_diff

opendocument-app · Sep 4, 2024 · f975773 · f975773
1 parent 637105e
commit f975773
Showing 1 changed file with 24 additions and 2 deletions.
diff --git a/test/scripts/html_render_diff.py b/test/scripts/html_render_diff.py
@@ -5,8 +5,13 @@
 import sys
 import argparse
 import io
+import time
+
 from PIL import Image, ImageChops
 from selenium import webdriver
+from selenium.webdriver.common.by import By
+from selenium.webdriver.support import expected_conditions
+from selenium.webdriver.support.ui import WebDriverWait
 import pathlib
 
 
@@ -19,8 +24,25 @@ def to_url(something):
 def screenshot(browser, url):
     browser.get(url)
 
-    body = browser.find_element('tag name', 'body')
-    png = body.screenshot_as_png
+    target_find_by = By.TAG_NAME
+    target = 'body'
+    loaded_page_settling_time = 0
+
+    # Selenium doesn't like when we try to screenshot <body> element of documents generated by pdf2htmlEX
+    if 'output-pdf2htmlEX' in url:
+        target_find_by = By.ID
+        target = 'page-container'
+        loaded_page_settling_time = 1
+
+    web_driver_wait = WebDriverWait(browser, 5)
+    web_driver_wait.until(expected_conditions.presence_of_element_located((target_find_by, target)))
+    web_driver_wait.until(lambda driver: driver.execute_script("return document.readyState") == "complete")
+    if loaded_page_settling_time != 0:
+        time.sleep(loaded_page_settling_time)
+
+    target_element = browser.find_element(target_find_by, target)
+
+    png = target_element.screenshot_as_png
     return Image.open(io.BytesIO(png))