[Question] How to use this with selenium-python on my current page? #23

M-Zubair10 · 2022-03-27T09:38:08Z

How to use this with selenium-python on my current page?

QIN2DIM · 2022-03-27T09:41:25Z

Imitate and modify.

hcaptcha-challenger/src/apis/scaffold/challenge.py

Lines 18 to 75 in 7fc13d9

    
           def runner( 
        
               sample_site: str, 
        
               lang: Optional[str] = "zh", 
        
               silence: Optional[bool] = False, 
        
               onnx_prefix: Optional[str] = None, 
        
           ): 
        
               """人机挑战演示 顶级接口""" 
        
               logger.info("Starting demo project...") 
        
               # Instantiating embedded models 
        
               yolo = YOLO(DIR_MODEL, onnx_prefix=onnx_prefix) 
        
               # Instantiating Challenger Components 
        
               challenger = ArmorCaptcha(dir_workspace=DIR_CHALLENGE, lang=lang, debug=True) 
        
               challenger_utils = ArmorUtils() 
        
               # Instantiating the Challenger Drive 
        
               ctx = get_challenge_ctx(silence=silence, lang=lang) 
        
               try: 
        
                   for _ in range(5): 
        
                       try: 
        
                           # Read the hCaptcha challenge test site 
        
                           ctx.get(sample_site) 
        
                           # Necessary waiting time 
        
                           time.sleep(3) 
        
                           # Detects if a clickable `hcaptcha checkbox` appears on the current page. 
        
                           # The `sample site` must pop up the `checkbox`, where the flexible wait time defaults to 5s. 
        
                           # If the `checkbox` does not load in 5s, your network is in a bad state. 
        
                           if challenger_utils.face_the_checkbox(ctx): 
        
                               start = time.time() 
        
                               # Enter iframe-checkbox --> Process hcaptcha checkbox --> Exit iframe-checkbox 
        
                               challenger.anti_checkbox(ctx) 
        
                               # Enter iframe-content --> process hcaptcha challenge --> exit iframe-content 
        
                               result = challenger.anti_hcaptcha(ctx, model=yolo) 
        
                               if not result: 
        
                                   continue 
        
                               challenger.log( 
        
                                   f"End of demo - total: {round(time.time() - start, 2)}s" 
        
                               ) 
        
                           break 
        
                       # Do not capture the `ChallengeReset` signal in the outermost layer. 
        
                       # In the demo project, we wanted the human challenge to pop up, not pass after processing the checkbox. 
        
                       # So when this happens, we reload the page to activate hcaptcha repeatedly. 
        
                       # But in your project, if you've passed the challenge by just handling the checkbox, 
        
                       # there's no need to refresh the page! 
        
                       except (WebDriverException, ChallengeReset): 
        
                           continue 
        
               finally: 
        
                   input("[EXIT] Press any key to exit...") 
        
                   ctx.quit()

M-Zubair10 · 2022-03-27T11:32:36Z

Can you make standalone class like class.solve() to solve captcha on current page

…

On Sun, Mar 27, 2022, 2:41 PM QIN2DIM ***@***.***> wrote: Imitate and modify. https://github.com/QIN2DIM/hcaptcha-challenger/blob/7fc13d9e193ea8cb527af91d3bbf1737d1ac10c4/src/apis/scaffold/challenge.py#L18-L75 — Reply to this email directly, view it on GitHub <#23 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALL7Y5C5OZLDJJPRN3VDWALVCAUM7ANCNFSM5RYRAYHQ> . You are receiving this because you authored the thread.Message ID: ***@***.***>

QIN2DIM · 2022-03-27T14:18:13Z

@M-Zubair10 in progress🐒

M-Zubair10 · 2022-03-27T14:19:30Z

Thanks a lot.

…

On Sun, Mar 27, 2022, 7:18 PM QIN2DIM ***@***.***> wrote: @M-Zubair10 <https://github.com/M-Zubair10> in process🐒 — Reply to this email directly, view it on GitHub <#23 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALL7Y5DVNVR4H5L4ZVIFEM3VCBU27ANCNFSM5RYRAYHQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

Revadike · 2022-03-28T03:53:52Z

maybe u can also make a standalone proxy server exe/linux

M-Zubair10 · 2022-03-28T06:54:14Z

Nah! I only need to use it during my webscraping in selenium So needed a pypi release with function solve to automatically solve captcha on current selenium page I actually done this with yolov3 using image ai but the result ratio is 2 out of 5 which is bad, your model is quite good but i can't load it in imageai since i don't have a knowledge of tensorflow, keras for now.

…

On Mon, Mar 28, 2022, 8:54 AM Revadike ***@***.***> wrote: maybe u can also make a standalone proxy server exe/linux — Reply to this email directly, view it on GitHub <#23 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALL7Y5HURWPAG3SPYSZUBBTVCEUNVANCNFSM5RYRAYHQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

QIN2DIM · 2022-03-28T10:58:20Z

@M-Zubair10 The solution for this project uses the ONNX compression model, and you can call it directly using the opencv-python interface without caring about the details of deep learning. If you look at requirement.txt, you will see that torch or tensorflow is missing.

@M-Zubair10 @Revadike btw I don't have a good grasp of pypi and container server, which will take some time to learn. and then, I've been busy with school lately, so I may not be able to push it anytime soon.🤦‍♂️

QIN2DIM · 2022-03-29T13:48:04Z

After a round of research I found that you guys are actually asking the same question...

M-Zubair10 · 2022-03-29T13:50:51Z

I don't know, we just want ease to use it with our program right? Can you tell me how you detect object in image, so that i will make a class with pypi release

…

On Tue, Mar 29, 2022, 6:48 PM QIN2DIM ***@***.***> wrote: After a round of research I found that you guys are actually asking the same question... — Reply to this email directly, view it on GitHub <#23 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALL7Y5HFAQUYOCMJIG2HHGTVCMCZ5ANCNFSM5RYRAYHQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

QIN2DIM · 2022-03-29T14:02:28Z

这个项目名叫 hcaptcha-challenger 而非 hcaptcha-solver。我的本意是搭建一个服务接口用以演示 YOLOv5(ONNX) 嵌入式解决方案 带来的提升，我只在乎识别的速度以及挑战的通过效率。

对于 @Revadike 这位老哥的问题，用什么方法对抗 hCaptcha 并不重要，重要的是获取挑战结束后返回的 TOKEN 。

对于 @M-Zubair10 的问题，如果你只是想在你编写的 selenium 程序里使用这个挑战方法，你可以按照我上面说的，根据 demo 改一下你的上下文业务既可。如果你要把它打包成 pypi，编码难度极高。

M-Zubair10 · 2022-03-29T14:05:15Z

Can you again send this in english?

…

On Tue, Mar 29, 2022, 7:02 PM QIN2DIM ***@***.***> wrote: 这个项目名叫 hcaptcha-challenger 而非 hcaptcha-solver。我的本意是搭建一个服务接口用以演示 YOLOv5(ONNX) 嵌入式解决方案带来的提升，我只在乎识别的速度以及挑战的通过效率。对于 @Revadike <https://github.com/Revadike> 这位老哥的问题，用什么方法对抗 hCaptcha 并不重要，重要的是获取挑战结束后返回的 TOKEN 。对于 @M-Zubair10 <https://github.com/M-Zubair10> 的问题，如果你只是想在你编写的 selenium 程序里使用这个挑战方法，你可以按照我上面说的，根据 demo 改一下上下文业务既可。如果你要把它打包成 pypi，编码难度极高。 — Reply to this email directly, view it on GitHub <#23 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALL7Y5BLT7NOZYEJBV2VX73VCMEP5ANCNFSM5RYRAYHQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

QIN2DIM · 2022-03-29T14:07:06Z

The project is called hcaptcha-challenger, not hcaptcha-solver, and my intention was to build a service interface to demonstrate the improvements made by the YOLOv5 (ONNX) embedded solution, I only cared about the speed of recognition and the efficiency of passing challenges.

To @Revadike， it doesn't matter what method is used against hCaptcha, what matters is getting the TOKEN returned at the end of the challenge.

To @M-Zubair10， if you just want to use this challenge method in your selenium application, you can change your contextual business according to the demo as I mentioned above. If you want to package it as a pypi, it's extremely difficult to code.

Revadike · 2022-03-29T14:09:16Z

I meant like an API server

QIN2DIM · 2022-03-29T14:10:48Z

ah~ @Revadike I know what you mean, you need a cross programming language solution.

QIN2DIM · 2022-03-30T05:07:10Z

@M-Zubair10 ah - - I think I didn't express my meaning accurately.

The coding difficulty I mentioned yesterday is mainly due to the fact that hcaptcha-challenger cannot be started in various selenium contexts. Because the challenge label is multilingual,

when you use selenium on a PC to trigger the challenge, the lang of label depends on the --lang parameter of the WebDriver Options;
and when you use selenium on linux, the lang of label depends on the value of the process environment variable LANGUAGE.

Obviously all such recognition operations involve label matching, and if the label text is translated into other languages because of multilingual issues, not only will the model fail, but the whole challenge logic will be seriously flawed.

I would like to make this solution available to developers using a variety of languages with a single set of code, however, the two factors I just mentioned that determine the lang of label are set before the WebDriver starts. To reference hcaptcha-challenger's method in the process for a challenge, you would have to do redundant transcoding, which is unnecessary.

M-Zubair10 · 2022-03-30T08:28:55Z

Thank you for the clarification.

…

On Wed, Mar 30, 2022, 10:07 AM QIN2DIM ***@***.***> wrote: @M-Zubair10 <https://github.com/M-Zubair10> ah - - I think I didn't express my meaning accurately. The coding difficulty I mentioned yesterday is mainly due to the fact that hcaptcha-challenger cannot be started in various selenium contexts. Because the challenge label is multilingual, when you use selenium on a PC to trigger the challenge, the lang of label depends on the --lang parameter of the WebDriver Options, and when you use selenium on linux, the lang of label depends on the value of the process environment variable LANGUAGE. Obviously all such recognition operations involve label matching, and if the label text is translated into other languages because of multilingual issues, not only will the model fail, but the whole challenge logic will be seriously flawed. I would like to make this solution available to developers using a variety of languages with a single set of code, however, the two factors I just mentioned that determine the lang of label are set before the WebDriver starts. To reference hcaptcha-challenger's method in the process for a challenge, you would have to do redundant transcoding, which is unnecessary. — Reply to this email directly, view it on GitHub <#23 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALL7Y5H4TZCUJ6GPUPHPIWLVCPOQRANCNFSM5RYRAYHQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

izoomrud · 2022-04-01T10:35:23Z

For @Revadike，it doesn't matter what method is used against hCaptcha, it's important to get the TOKEN returned at the end of the challenge.

How can I get TOKEN from there?

I tried putting token = ctx.find_element_by_tag_name('iframe').get_attribute("data-hcaptcha-response") after self.log("Challenge Success") in core.py but it doesn't work.

QIN2DIM · 2022-04-01T16:05:12Z

but it doesn't work.

@izoomrud sure

izoomrud · 2022-04-02T03:01:22Z

but it doesn't work.

@izoomrud sure

ctx.switch_to.default_content()
time.sleep(3)
token = ctx.find_elements(By.XPATH, "//iframe[@title='widget containing checkbox for hCaptcha security challenge']")
for value in token:
    file = open('token.txt', 'a')
    file.write(value.get_attribute('data-hcaptcha-response'))

.|.

QIN2DIM · 2022-04-02T03:14:23Z

@izoomrud 有被可爱到，路漫漫其修远兮哈哈哈哈

3281448091 · 2022-04-04T00:26:21Z

execute javascript to get the token.
just use getAttribute("data-hcaptcha-response") and then set the token to the webpage's title and do webdriver.title
and that will work

QIN2DIM · 2022-04-04T13:39:08Z

execute javascript to get the token. just use getAttribute("data-hcaptcha-response") and then set the token to the webpage's title and do webdriver.title and that will work

yeah

shahzain345 · 2022-04-04T14:57:59Z

Well, what u can do is execute javascript in your webdriver. You can use the hcaptcha.getResponse() method to get the response token, which you can then use in your script. The code should look like this.

token = ctx.execute_script("return hcaptcha.getResponse();")

yeshenshuijiao · 2022-04-21T13:09:48Z

可以使用playwright吗

QIN2DIM · 2022-04-21T13:14:26Z

可以使用playwright吗

一样的，思路都是在运行时判断是否遇到 challenge，然后用相应的接口处理。但现在这个项目我做成了 demo了，耦合性非常强，如果你要自己的项目中使用的话，目前只能照葫芦画瓢敲一个出来了，因为 selenium 和 playwright 的 api 完全不一样。

QIN2DIM changed the title ~~[Question]~~ [Question] How to use this with selenium-python on my current page? Mar 27, 2022

QIN2DIM added the fixed BUG 已修复或问题已解决 label Mar 27, 2022

QIN2DIM mentioned this issue Apr 4, 2022

[Question]Selenium #28

Closed

QIN2DIM closed this as completed Jul 23, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Question] How to use this with selenium-python on my current page? #23

[Question] How to use this with selenium-python on my current page? #23

M-Zubair10 commented Mar 27, 2022

QIN2DIM commented Mar 27, 2022

M-Zubair10 commented Mar 27, 2022 via email

QIN2DIM commented Mar 27, 2022 •

edited

Loading

M-Zubair10 commented Mar 27, 2022 via email

Revadike commented Mar 28, 2022

M-Zubair10 commented Mar 28, 2022 via email

QIN2DIM commented Mar 28, 2022 •

edited

Loading

QIN2DIM commented Mar 29, 2022

M-Zubair10 commented Mar 29, 2022 via email

QIN2DIM commented Mar 29, 2022 •

edited

Loading

M-Zubair10 commented Mar 29, 2022 via email

QIN2DIM commented Mar 29, 2022

Revadike commented Mar 29, 2022

QIN2DIM commented Mar 29, 2022

QIN2DIM commented Mar 30, 2022 •

edited

Loading

M-Zubair10 commented Mar 30, 2022 via email

izoomrud commented Apr 1, 2022 •

edited

Loading

QIN2DIM commented Apr 1, 2022 •

edited

Loading

izoomrud commented Apr 2, 2022 •

edited by QIN2DIM

Loading

QIN2DIM commented Apr 2, 2022

3281448091 commented Apr 4, 2022

QIN2DIM commented Apr 4, 2022

shahzain345 commented Apr 4, 2022

yeshenshuijiao commented Apr 21, 2022

QIN2DIM commented Apr 21, 2022 •

edited

Loading

[Question] How to use this with selenium-python on my current page? #23

[Question] How to use this with selenium-python on my current page? #23

Comments

M-Zubair10 commented Mar 27, 2022

QIN2DIM commented Mar 27, 2022

M-Zubair10 commented Mar 27, 2022 via email

QIN2DIM commented Mar 27, 2022 • edited Loading

M-Zubair10 commented Mar 27, 2022 via email

Revadike commented Mar 28, 2022

M-Zubair10 commented Mar 28, 2022 via email

QIN2DIM commented Mar 28, 2022 • edited Loading

QIN2DIM commented Mar 29, 2022

M-Zubair10 commented Mar 29, 2022 via email

QIN2DIM commented Mar 29, 2022 • edited Loading

M-Zubair10 commented Mar 29, 2022 via email

QIN2DIM commented Mar 29, 2022

Revadike commented Mar 29, 2022

QIN2DIM commented Mar 29, 2022

QIN2DIM commented Mar 30, 2022 • edited Loading

M-Zubair10 commented Mar 30, 2022 via email

izoomrud commented Apr 1, 2022 • edited Loading

QIN2DIM commented Apr 1, 2022 • edited Loading

izoomrud commented Apr 2, 2022 • edited by QIN2DIM Loading

QIN2DIM commented Apr 2, 2022

3281448091 commented Apr 4, 2022

QIN2DIM commented Apr 4, 2022

shahzain345 commented Apr 4, 2022

yeshenshuijiao commented Apr 21, 2022

QIN2DIM commented Apr 21, 2022 • edited Loading

QIN2DIM commented Mar 27, 2022 •

edited

Loading

QIN2DIM commented Mar 28, 2022 •

edited

Loading

QIN2DIM commented Mar 29, 2022 •

edited

Loading

QIN2DIM commented Mar 30, 2022 •

edited

Loading

izoomrud commented Apr 1, 2022 •

edited

Loading

QIN2DIM commented Apr 1, 2022 •

edited

Loading

izoomrud commented Apr 2, 2022 •

edited by QIN2DIM

Loading

QIN2DIM commented Apr 21, 2022 •

edited

Loading