all
I tried to access an existing chrome instant.

First, I opened chrome from console:

chrome.exe --remote-debugging-port=9222 --user-data-dir="C:SeleniumChromeProfile"

Then I input url and directed to web page

In my code, I created dirver and attached to the existing chrome:

chrome_options = Options()
chrome_options.add_experimental_option("debuggerAddress", "127.0.0.1:9222")
service = Service("C:\chromedriver.exe")
driver = webdriver.Chrome(service=service, options=chrome_options)

Then I got shadow_root:

shadow_host = WebDriverWait(driver, 20).until(
    EC.presence_of_element_located((By.CSS_SELECTOR, "#shadow_host"))
)
shadow_root = driver.find_element(By.CSS_SELECTOR, "#shadow_host").shadow_root
print(shadow_root is None)

So far, everything is ok, the output is False, I got a not null shadow_root.
Then I tried to get element via shadow_root:

chartlist = shadow_root.find_element(By.CLASS_NAME, "chart_wrapper")
print(len(chartlist ))

I got error like below:

selenium.common.exceptions.InvalidSelectorException: Message: invalid selector
from javascript error: {"status":10,"value":"stale element not found in the current frame"}
  (Session info: chrome=127.0.6533.120); For documentation on this error, please visit: https://www.selenium.dev/documentation/webdriver/troubleshooting/errors#invalid-selector-exception

And I also tried to get element via javascript:

shadow_root = driver.execute_script("return arguments[0].shadowRoot", shadow_host)
chart_list = driver.execute_script('return arguments[0].querySelector('div[class="chart_wrapper"]')',shadow_root )

But got the same error!

But I tried to get innerHTML like below:

content = driver.execute_script("""
    var shadowHost = arguments[0];
    var shadowRoot = shadowHost.shadowRoot;
    return shadowRoot.innerHTML;
""", shadow_host)

print("Shadow root content:", content)

This time, I got the html text!

And I tried to get element again:

eles = driver.execute_script("""
    var shadowHost = arguments[0];
    var shadowRoot = shadowHost.shadowRoot;
    var eles = shadowRoot.querySelector('div[class="chart_wrapper"]')'
    return eles ;
""", shadow_host)

I got the same error again!

Now it seemed I can get the html text but can’t get element in shadowRoot!

Is there any way to access shadow_root element on an existing page via selenium?