[nodriver]How to select the frame or click 'Verify you are human'

[TonyLiooo]

url: https://masiro.me
I tried the following code:
checkbox = await page.wait_for(text="cf-chl-widget-", timeout=10) print(checkbox) await checkbox.mouse_move() await checkbox.mouse_click() print(checkbox)
The element information is:
<input type="hidden" name="cf-turnstile-response" id="cf-chl-widget-805o7_response"></input>
However, checkbox.mouse_move() reported the following error:
could not find position for <input type="hidden" name="cf-turnstile-response" id="cf-chl-widget-805o7_response"></input>
also:
iframes = await page.select_all('iframe')
but error:
time ran out while waiting for iframe

[sohaib17]

That is most probably because the iframe is hidden under a shadow root.

[marioeivissa]

That is most probably because the iframe is hidden under a shadow root.

great!!!any posibility to pass this shadow root????thanks

[pythonlw]

[pythonlw]

`
import json
from nodriver import start, cdp, loop
import nodriver as uc

async def switch_to_frame(browser, frame):
"""
change iframe
let iframe = document.querySelector("YOUR_IFRAME_SELECTOR")
let iframe_tab = iframe.contentWindow.document.body;
"""
iframe_tab: uc.Tab = next(
filter(
lambda x: str(x.target.target_id) == str(frame.frame_id), browser.targets
)
)
return iframe_tab

async def main():
browser_args = ['--disable-web-security']
browser = await uc.start(browser_args=browser_args)
tab = await browser.get("http://www.yescaptcha.cn/auth/login")
for _ in range(10):
await tab.scroll_down(50)
await tab
await tab.sleep(15)
# solve Ordinary iframe
# query_selector = await tab.select_all('button[class="widgetLabel moveFromRightLabel-enter-done"]', include_frames=True)
# print('query_selector:', query_selector)
# if len(query_selector) == 1:
# await query_selector[0].click()

# solve  Cross-origin iframe
recaptcha0 = await tab.select('iframe[title="reCAPTCHA"]')
print('recaptcha0:', recaptcha0.frame_id)
# # for tar in browser.targets:
# #     print('target_id:', tar.target.target_id)
iframe_tab = await switch_to_frame(browser, recaptcha0)
print('iframe_tabwebsocket_url:', iframe_tab.websocket_url)
iframe_tab.websocket_url = iframe_tab.websocket_url.replace("iframe", "page")
button = await iframe_tab.select("span#recaptcha-anchor")
await button.click()
input('stop')

if name == "main":
loop().run_until_complete(main())`

[today2004]

url: https://masiro.me I tried the following code: checkbox = await page.wait_for(text="cf-chl-widget-", timeout=10) print(checkbox) await checkbox.mouse_move() await checkbox.mouse_click() print(checkbox) The element information is: <input type="hidden" name="cf-turnstile-response" id="cf-chl-widget-805o7_response"></input> However, checkbox.mouse_move() reported the following error: could not find position for <input type="hidden" name="cf-turnstile-response" id="cf-chl-widget-805o7_response"></input> also: iframes = await page.select_all('iframe') but error: time ran out while waiting for iframe

May I ask if you have resolved it？

[ultrafunkamsterdam]

For many elements a position could not be determined since they are either nested or in a different origin

[TonyLiooo]

For many elements a position could not be determined since they are either nested or in a different origin

Thank you, I've found a solution and can now determine the element's position.

[ultrafunkamsterdam]

For many elements a position could not be determined since they are either nested or in a different origin

Thank you, I've found a solution and can now determine the element's position.

Good you found it, better to share your solution :)

[YogaSakti]

For many elements a position could not be determined since they are either nested or in a different origin

Thank you, I've found a solution and can now determine the element's position.

Glad to hear that the problem is solved. please share the solution :)

[TonyLiooo]

For many elements a position could not be determined since they are either nested or in a different origin

Thank you, I've found a solution and can now determine the element's position.

Good you found it, better to share your solution :)

Sorry, I haven't had much time lately. I'll provide the code I wrote two months ago; I just tried it, and it still works.

def xpath_to_css(xpath: str) -> str:
    """
    Convert an XPath expression to a CSS selector.
    
    Args:
        xpath (str): The XPath expression to convert.
    
    Returns:
        str: The equivalent CSS selector.
    """
    
    # Convert XPath axis and node tests to CSS selectors
    css = xpath
    
    # Convert predicate expressions (e.g., [1], [@attr='value']) to CSS attribute selectors
    css = re.sub(r'\[@([^\]]+)=["\']([^"\']+)["\']\]', r'[\1="\2"]', css)
    
    # Convert XPath predicates (e.g., [1]) to nth-child CSS selectors
    css = re.sub(r'\[(\d+)\]', r':nth-child(\1)', css)
    
    # Remove the XPath axis from the beginning of the XPath expression
    css = re.sub(r'^//', '', css)
    
    # Replace double slashes with a single slash (XPath to CSS path)
    css = re.sub(r'//', ' ', css)
    
    # Remove unnecessary leading and trailing spaces
    css = css.strip()

    # Ensure CSS selector is properly formatted
    css = re.sub(r'(\s+)', ' ', css)  # Replace multiple spaces with a single space

    # Clean up any residual syntax errors or unnecessary parts
    css = css.replace('[1]', '')

    # Replace common XPath functions and expressions
    css = re.sub(r'\[contains\(@class,["\']([^"\']+)["\']\)\]', r'.\1', css)
    css = re.sub(r'\[contains\(@id,["\']([^"\']+)["\']\)\]', r'#\1', css)
    css = re.sub(r'\[contains\(@name,["\']([^"\']+)["\']\)\]', r'[name="\1"]', css)
    css = re.sub(r'\[@id=["\']([^"\']+)["\']\]', r'#\1', css)
    css = re.sub(r'\[@class=["\']([^"\']+)["\']\]', r'.\1', css)

    return css

async def switch_to_frame(browser:uc.Browser, frame_id):
    """
    change iframe
    let iframe = document.querySelector("YOUR_IFRAME_SELECTOR")
    let iframe_tab = iframe.contentWindow.document.body;
    """
    iframe_tab: Tab = next(
        filter(
        lambda x: str(x.target.target_id) == str(frame_id), browser.targets
        )
    )
    return iframe_tab

def cdp_generator(method, params):
    cmd_dict = {"method": method, "params": params}
    json = yield cmd_dict
    return json

async def find_and_click_element(tab, selector):
    async def describe_node(_tab, node_id):
        return await _tab.send(cdp_generator("DOM.describeNode", {
            "nodeId": node_id,
            # "depth": -1,
            "pierce": True
        }))
    
    async def find_element_in_node(_tab, node_id, selector):
        result = await _tab.send(cdp_generator("DOM.querySelector", {
            "nodeId": node_id,
            "selector": xpath_to_css(selector)
        }))
        return result

    async def process_node(_tab, node):
        node_id = node['nodeId'] if 'nodeId' in node else None
        if not node_id:
            return _tab, None

        result = await find_element_in_node(_tab, node_id, selector)
        if result and result.get('nodeId'):
            return _tab, result

        if 'shadowRoots' in node:
            for shadow_root in node['shadowRoots']:
                process_tab, result = await process_node(_tab, shadow_root)
                if result and result.get('nodeId'):
                    return process_tab, result

        iframe_result = await find_element_in_node(_tab, node_id, 'iframe')
        if iframe_result and iframe_result.get('nodeId'):
            iframe_node_id = iframe_result['nodeId']
            process_tab, result = await process_iframe(_tab, iframe_node_id)
            if result and result.get('nodeId'):
                return process_tab, result

        if 'children' in node:
            process_tab, result = await process_child(_tab, node)
            if result and result.get('nodeId'):
                return process_tab, result
        
        return _tab, None

    async def process_iframe(_tab, node_id):
        iframe_response = await describe_node(_tab, node_id)
        frame_id = iframe_response['node']['frameId']
        iframe_tab = await switch_to_frame(_tab.browser, frame_id)
        
        if iframe_tab:
            iframe_document = await iframe_tab.send(cdp_generator("DOM.getDocument", {"depth": -1, "pierce": True}))
        process_tab, result = await process_node(iframe_tab, iframe_document['root'])
        if result and result.get('nodeId'):
            return process_tab, result
        return _tab, None

    async def process_child(_tab, node):
        if 'children' in node:
            for child in node.get('children'):
                if 'shadowRoots' in child:
                    for shadow_root in child['shadowRoots']:
                        process_tab, result = await process_node(_tab, shadow_root)
                        if result and result.get('nodeId'):
                            return process_tab, result
                if 'children' in child:
                    process_tab, result = await process_child(_tab, child)
                    if result and result.get('nodeId'):
                        return process_tab, result
        return _tab, None
    
    document = await tab.send(cdp_generator("DOM.getDocument", {"depth": -1, "pierce": True}))
    process_tab, result = await process_node(tab, document['root'])

    if result is None or not result.get('nodeId'):
        raise Exception(f"Element with selector '{selector}' not found.")

    node_id = result['nodeId']
    box_model = await process_tab.send(cdp_generator('DOM.getBoxModel', {
        'nodeId': node_id
    }))
    
    if 'model' in box_model and 'content' in box_model['model']:
        content = box_model['model']['content']
        x_min, y_min = content[0], content[1]
        x_max, y_max = content[4], content[5]
        x_center = (x_min + x_max) / 2
        y_center = (y_min + y_max) / 2

        await process_tab.send(cdp_generator('DOM.scrollIntoViewIfNeeded', {
            'nodeId': node_id
        }))

        await tab.send(cdp_generator('Input.dispatchMouseEvent', {
            'type': 'mouseMoved',
            'x': x_center,
            'y': y_center,
            'button': 'none'
        }))

        await process_tab.send(cdp_generator('Input.dispatchMouseEvent', {
            'type': 'mousePressed',
            'x': x_center,
            'y': y_center,
            'button': 'left',
            'clickCount': 1
        }))
        await process_tab.send(cdp_generator('Input.dispatchMouseEvent', {
            'type': 'mouseReleased',
            'x': x_center,
            'y': y_center,
            'button': 'left',
            'clickCount': 1
        }))
    else:
        raise Exception(f"Failed to get box model for element with selector '{selector}'.")

You can use the following code to click 'Verify you are human'.
await find_and_click_element(page, 'input[type="checkbox"]')