WebDriver UDF - Help & Support (II)

[Siwa]

It did not work, and I added _WD_GetFrameCount($sSession) which returned 0.

How can I check for frames or iframes inside he html ?  

[AttilaP]

Hello!

With this addon I can start Chrome headless (by passing --headless to args in $sDesiredCapabilities).

Is there any way to show the window later on different conditions? Or once it started without the UI, it will not be visible anymore?

[Siwa]

1 hour ago, AttilaP said:

Hello!

With this addon I can start Chrome headless (by passing --headless to args in $sDesiredCapabilities).

Is there any way to show the window later on different conditions? Or once it started without the UI, it will not be visible anymore?

I guess you can do that by using WInsetState, but you have to get the handle of the windows first.

[Danp2]

@AttilaP Haven't tried, but I always assumed that a headless session will always be headless.

[Danp2]

@Siwa Have you checked out the ChroPath extension?

[AttilaP]

14 minutes ago, Danp2 said:

@AttilaP Haven't tried, but I always assumed that a headless session will always be headless.

And it is possible to start "normally" and then hide the Window?🤔

[Danp2]

@AttilaP I imagine it's possible using standard Autoit commands, but haven't tested to see what effect it will have on the webdriver functionality, if any.

[Nine]

1 hour ago, AttilaP said:

and then hide the Window?

Here what I do :

_WD_WINDOW($sSession, "rect", '{"x":-1000,"y":-1000,"width":500,"height":500}')

 

[Siwa]

I installed it and the result is the same. I tried the element by id, by class, by name .... but with no success in clicking the button, or reading the ui.

If I get the position of the element, can I send a mouse click to the location ? because the ui dialog is placed middle no matter of my scroll.

and I want to be headless, so MouseCLick() is not an option for me, I want the webdriver to the clicking. 

[Danp2]

@Siwa Can you provide a link to the website? If not, then how about finding another site with the same type of dialog?

FWIW, I suspect that you are dealing with a jQuery Dialog like this, but can't tell for sure based on info provided.

[Danp2]

@Siwa The link you sent me via PM won't load for me. Not sure why... maybe limited access geographically. <shrug>

Maybe this will help you --

#include <wd_core.au3>

Local $sDesiredCapabilities

SetupChrome()
_WD_Startup()

Local $sSession = _WD_CreateSession($sDesiredCapabilities)
_WD_Navigate($sSession, "https://cdn.rawgit.com/salmanarshad2000/demos/v1.0.4/jquery-ui-dialog/modal-dialog-with-buttons.html")

Local $oElement1 = _WD_FindElement($sSession,$_WD_LOCATOR_ByXPath,"//div[@class='ui-dialog-buttonset']/button[1]")
Local $oElement2 = _WD_FindElement($sSession,$_WD_LOCATOR_ByXPath,"//div[@class='ui-dialog-buttonset']/button[2]")

_WD_ElementAction($sSession, $oElement1, 'click')


Func SetupChrome()
_WD_Option('Driver', 'chromedriver.exe')
_WD_Option('Port', 9515)
_WD_Option('DriverParams', '--verbose --log-path="' & @ScriptDir & '\chrome.log"')

$sDesiredCapabilities = '{"capabilities": {"alwaysMatch": {"goog:chromeOptions": {"w3c": true, "excludeSwitches": [ "enable-automation"], "useAutomationExtension": false }}}}'
EndFunc

 

[Siwa]

It did not work, and this is the console :

 

Spoiler

__WD_Post: URL=HTTP://127.0.0.1:9515/session/03801389c6069e2bb080dae7d55d802f/element; $sData={"using":"xpath","value":"//*[@id='j_idt250']"}
__WD_Post: StatusCode=200; ResponseText={"value":{"element-6066-11e4-a52e-4f735466cecf":"1a414e59-c9f0-430a-a98c-4ec010ce78db"}}...
_WD_FindElement: {"value":{"element-6066-11e4-a52e-4f735466cecf":"1a414e59-c9f0-430a-a98c-4ec010ce78db"}}
__WD_Post: URL=HTTP://127.0.0.1:9515/session/03801389c6069e2bb080dae7d55d802f/element/1a414e59-c9f0-430a-a98c-4ec010ce78db/click; $sData={"id":"1a414e59-c9f0-430a-a98c-4ec010ce78db"}
__WD_Post: StatusCode=400; ResponseText={"value":{"error":"element not interactable","message":"element not interactable\n  (Session info: c...
__WD_Post ==> Element interaction issue: {"value":{"error":"element not interactable","message":"element not interactable\n  (Session info: chrome=87.0.4280.66)","stacktrace":"Backtrace:\n\tOrdinal0 [0x0085ECE3+3337443]\n\tOrdinal0 [0x0073F041+2158657]\n\tOrdinal0 [0x005C6E50+618064]\n\tOrdinal0 [0x0054CC0B+117771]\n\tOrdinal0 [0x00545B67+88935]\n\tOrdinal0 [0x00561973+203123]\n\tOrdinal0 [0x00545996+88470]\n\tOrdinal0 [0x00561A5A+203354]\n\tOrdinal0 [0x0056E4A9+255145]\n\tOrdinal0 [0x0056181B+202779]\n\tOrdinal0 [0x00543D74+81268]\n\tOrdinal0 [0x00544D6E+85358]\n\tOrdinal0 [0x00544CF9+85241]\n\tOrdinal0 [0x00756F2C+2256684]\n\tGetHandleVerifier [0x009E2FAC+1478332]\n\tGetHandleVerifier [0x009E2AFD+1477133]\n\tGetHandleVerifier [0x009EAAD8+1509864]\n\tGetHandleVerifier [0x009E369E+1480110]\n\tOrdinal0 [0x0074D6BD+2217661]\n\tOrdinal0 [0x007589AB+2263467]\n\tOrdinal0 [0x00758AEF+2263791]\n\tOrdinal0 [0x0076D363+2347875]\n\tBaseThreadInitThunk [0x75D06359+25]\n\tRtlGetAppContainerNamedObjectPath [0x776F7B74+228]\n\tRtlGetAppContainerNamedObjectPath [0x776F7B44+180]\n"}}
_WD_ElementAction: {"value":{"error":"element not interactable","message":"element not interactable\n  (Session info: c...
_WD_ElementAction ==> Element interaction issue: {"value":{"error":"element not interactable","message":"element not interactable\n  (Session info: chrome=87.0.4280.66)","stacktrace":"Backtrace:\n\tOrdinal0 [0x0085ECE3+3337443]\n\tOrdinal0 [0x0073F041+2158657]\n\tOrdinal0 [0x005C6E50+618064]\n\tOrdinal0 [0x0054CC0B+117771]\n\tOrdinal0 [0x00545B67+88935]\n\tOrdinal0 [0x00561973+203123]\n\tOrdinal0 [0x00545996+88470]\n\tOrdinal0 [0x00561A5A+203354]\n\tOrdinal0 [0x0056E4A9+255145]\n\tOrdinal0 [0x0056181B+202779]\n\tOrdinal0 [0x00543D74+81268]\n\tOrdinal0 [0x00544D6E+85358]\n\tOrdinal0 [0x00544CF9+85241]\n\tOrdinal0 [0x00756F2C+2256684]\n\tGetHandleVerifier [0x009E2FAC+1478332]\n\tGetHandleVerifier [0x009E2AFD+1477133]\n\tGetHandleVerifier [0x009EAAD8+1509864]\n\tGetHandleVerifier [0x009E369E+1480110]\n\tOrdinal0 [0x0074D6BD+2217661]\n\tOrdinal0 [0x007589AB+2263467]\n\tOrdinal0 [0x00758AEF+2263791]\n\tOrdinal0 [0x0076D363+2347875]\n\tBaseThreadInitThunk [0x75D06359+25]\n\tRtlGetAppContainerNamedObjectPath [0x776F7B74+228]\n\tRtlGetAppContainerNamedObjectPath [0x776F7B44+180]\n"}}
__WD_Post: URL=HTTP://127.0.0.1:9515/session/03801389c6069e2bb080dae7d55d802f/execute/sync; $sData={"script":"return document.readyState", "args":[]}
__WD_Post: StatusCode=200; ResponseText={"value":"complete"}...
_WD_ExecuteScript: {"value":"complete"}...

Edit :

Finally I got it working by :

_WD_ExecuteScript($sSession,"document.getElementById('j_idt250').click();")

Glad to solve it finally, and learned how to use JS scripts the right way 😉

(I guess it will be a huge step forward for new learners like me, if you could include this as a example in your demo )

Edited November 22, 2020 by Siwa

[Danp2]

@Siwa Not sure why the javascript worked where the standard click didn't, but glad you found a solution. FWIW, the control is being found successfully, but the click attempt is failing with "element not interactable". From the webdriver specs --

Quote

Error Code	HTTP Status	JSON Error Code	Description
element not interactable	400	element not interactable	A command could not be completed because the element is not pointer- or keyboard interactable.

 

[Danp2]

@Siwa It's possible that the element wasn't visible / enabled, which would cause the issue you encountered. Does it work when you do this?

$Confirm_Button = _WD_WaitElement($sSession,$_WD_LOCATOR_ByXPath,"//*[@id='j_idt250']", Default, Default, True, True, True)
_WD_ElementAction($sSession, $Confirm_Button, 'click')

 

[Siwa]

@Danp2 No, it did not work. 

[AttilaP]

Hi!

I am pretty new to this UDF and I try to fill a login form with it, using ChromeDriver. However, it always replace the @ sign in the username to the text what I have on my clipboard.

ChromeDriver 86.0.4240.22

AutoIt 3.3.14.5

A minimalistic code to reproduce this behavior:

#include "wd_core.au3"
#include "wd_helper.au3"

$_WD_DEBUG = $_WD_DEBUG_None
$_WD_DEBUG = $_WD_DEBUG_Info
_WD_Option('Driver', 'chromedriver.exe')
_WD_Option('Port', 9515)
_WD_Option('DriverParams', '--verbose --log-path="' & @ScriptDir & '\chrome.log"')

$sDesiredCapabilities = '{"capabilities": {"alwaysMatch": {"goog:chromeOptions": {"w3c": true, "excludeSwitches": [ "enable-automation"], "useAutomationExtension": false }}}}'
_WD_Startup()
If @error <> $_WD_ERROR_Success Then
    Exit -1
EndIf

$sSession = _WD_CreateSession($sDesiredCapabilities)

$username = "user@test.com"
_WD_Navigate($sSession, "https://www.w3schools.com/html/html_forms.asp")
$sElement = _WD_FindElement($sSession, $_WD_LOCATOR_ByXPath, '//input[@id="fname"]');
_WD_ElementAction($sSession, $sElement, 'value', $username);
Sleep(15000)
_WD_DeleteSession($sSession)
_WD_Shutdown()
Exit

 

First I copied the text 'CLIPBOARD' to the clipboard with Ctrl+C, then run the above - the result is attached.

I see the following in the chrome.log:

[1606145142.741][INFO]: Cannot switch to US keyboard layout - some keys may be interpreted incorrectly

And below the following, after the letter 'r' in user:
 

Spoiler

[1606145148.091][DEBUG]: DevTools WebSocket Command: Input.dispatchKeyEvent (id=61) BCA32ABAF10344BBC3E00EFCD9DC77B9 {
   "code": "ControlLeft",
   "key": "Control",
   "modifiers": 0,
   "text": "",
   "type": "rawKeyDown",
   "unmodifiedText": "",
   "windowsVirtualKeyCode": 17
}
[1606145148.092][DEBUG]: DevTools WebSocket Command: Input.dispatchKeyEvent (id=62) BCA32ABAF10344BBC3E00EFCD9DC77B9 {
   "code": "AltLeft",
   "key": "Alt",
   "modifiers": 0,
   "text": "",
   "type": "rawKeyDown",
   "unmodifiedText": "",
   "windowsVirtualKeyCode": 18
}
[1606145148.092][DEBUG]: DevTools WebSocket Command: Input.dispatchKeyEvent (id=63) BCA32ABAF10344BBC3E00EFCD9DC77B9 {
   "code": "KeyV",
   "commands": [ "Paste" ],
   "key": "v",
   "modifiers": 3,
   "text": "@",
   "type": "rawKeyDown",
   "unmodifiedText": "v",
   "windowsVirtualKeyCode": 86
}
[1606145148.092][DEBUG]: DevTools WebSocket Command: Input.dispatchKeyEvent (id=64) BCA32ABAF10344BBC3E00EFCD9DC77B9 {
   "code": "KeyV",
   "commands": [ "Paste" ],
   "key": "v",
   "modifiers": 3,
   "text": "@",
   "type": "char",
   "unmodifiedText": "v",
   "windowsVirtualKeyCode": 86
}
[1606145148.093][DEBUG]: DevTools WebSocket Command: Input.dispatchKeyEvent (id=65) BCA32ABAF10344BBC3E00EFCD9DC77B9 {
   "code": "KeyV",
   "commands": [ "Paste" ],
   "key": "v",
   "modifiers": 3,
   "text": "@",
   "type": "keyUp",
   "unmodifiedText": "v",
   "windowsVirtualKeyCode": 86
}
[1606145148.093][DEBUG]: DevTools WebSocket Command: Input.dispatchKeyEvent (id=66) BCA32ABAF10344BBC3E00EFCD9DC77B9 {
   "code": "AltLeft",
   "key": "Alt",
   "modifiers": 0,
   "text": "",
   "type": "keyUp",
   "unmodifiedText": "",
   "windowsVirtualKeyCode": 18
}
[1606145148.094][DEBUG]: DevTools WebSocket Command: Input.dispatchKeyEvent (id=67) BCA32ABAF10344BBC3E00EFCD9DC77B9 {
   "code": "ControlLeft",
   "key": "Control",
   "modifiers": 0,
   "text": "",
   "type": "keyUp",
   "unmodifiedText": "",
   "windowsVirtualKeyCode": 17

}

For me it seems it tries to write the @ sign with sending 17, 18 down, then v, finally 17, 18 up. Which is Alt + Ctrl  + v I assume, and that should be correct, that's how I write it on my hungarian keyboard. Nevertheless it is interpreted as simple Ctrl + V.

Is there any other method I can use? Also, what shall I do if I want to overwrite the actual form element? Like in the example, I do not want to keep the text John there.

Thank you in advance!

[Danp2]

@AttilaP Thanks for the concise reproducer. You could have coded the clipboard piece with ClipPut. 😉

Your code runs fine for me, so I think your diagnosis that this is a keyboard layout or language issue is on target.

Edit: Read this -- https://chromedriver.chromium.org/help/keyboard-support

Edited November 23, 2020 by Danp2

[AttilaP]

Thanks Danp2, I installed the US language pack which solved the problem.

Now I face an another one: I try to download a file from a URL which is not a direct link to the file. I tried to use _WD_DownloadFile but that saved the HTML source instead. When I check the request in the browser, I see that the response header has this:

content-disposition: attachment;filename="=?ISO-8859-1?Q?=D6n=E9letrajz.pdf?="
content-length: 36855
content-type: application/pdf;charset=UTF-8

And the file is downloaded.

I can not find a link to mimic this behavior, unfortunately. Is there any way to grab these files and save them into a specific folder? I found this, but could not get it working.

Related part (Python source):

def enable_download_headless(browser,download_dir):
    browser.command_executor._commands["send_command"] = ("POST", '/session/$sessionId/chromium/send_command')
    params = {'cmd':'Page.setDownloadBehavior', 'params': {'behavior': 'allow', 'downloadPath': download_dir}}
    browser.execute("send_command", params)

[Danp2]

@AttilaP I don't know how to solve this for you, but if I come up with any ideas I'll let you know.

A couple of questions --

• Are you running in headless mode?
• Did you notice where they added some prefs related to downloads to their Chrome Options?

[Danp2]

@AttilaP I wonder if @TheDcoder's _WD_ExecuteCdpCommand function is what you need here.