Performance Logs to Driver()

[Kamran-ov]

Hi,

Can you add an argument for Chrome Driver creation that will add the below capability to Driver?

chrome_options.set_capability('goog:loggingPrefs', {'performance': 'ALL'})

I am trying to do the below

from seleniumbase import Driver
# Create Driver and request Page
driver = Driver(uc_cdp=True, incognito=True, headless=False)
driver.get("https://nowsecure.nl/#relax")
# Get Logs to retrieve AJAX Requests
logs = driver.get_log('performance')

[mdmintz]

That option is already there for uc_cdp mode, which you already appear to be using:

SeleniumBase/seleniumbase/undetected/__init__.py

Line 168 in eea2441

"goog:loggingPrefs", {"performance": "ALL", "browser": "ALL"}

[MovingW]

Hi Michael @mdmintz ,

I've worked out a temporary solution for my issue. I set uc_cdp=True and commented out the listen method in Class Reactor. Then it worked. I could execute driver.get_log("performance") any time I need.

Though it's not perfect, it does fix my issue.
Waiting for your perfect upgrade.

[mdmintz]

Resolved in 4.20.8 - #2220

[Kamran-ov]

@mdmintz, thank you very much for this update. it works perfectly fine with uc=True or uc=None.
It will help a lot to people working with the logs.

[Kamran-ov]

Hi @MovingW and @mdmintz, ,

I just want to put this here for you and for somebody else who is working with logs, and had an issue with the amount of logs returned from the driver.get().

If you use driver.get(url) on the site with antibot systems like Cloudflare etc,
driver.get_logs('performance')
will return fewer logs than what is actually happening in the network activity tab of the browser.

However, if you use driver.default_get(url),
it will have all the logs without deleting any.

I believe it is because of tab changes etc. when .get() is done with SeleniumBase since it is not doing the default selenium .get() method.

Obviously, you risk getting blocked by using driver.default_get(), since you don't get all the functionalities of Seleniumbase bot bypass features.
But if you don't have a problem with getting blocked after the initial opening page (in my case, checks are only in the first request of the page), then I would suggest using default_get to get all logs.

[MovingW]

Hi @Kamran-ov,

I've never met with similar issues so far. I used seleniumbase to access websites and got logs successfully while I failed using undetected-chromedriver directly.

From my understanding, if you could access the website and all pages were loaded normally, the issues were not caused by the antibot mechanism. If your automation was detected by the antibot system, the server would not respond correctly, so the pages could not be loaded as you expected. And if the pages were loaded as you expected, that meant the server had gave you everything.