How to Integrate CB with a custom Python script?

Batman

I'm writing a custom script in python and would like to use CB for captcha solving. The captchas are easy and brute force is giving me 90+ success rate.

But I have no clue how to integrate CB with my script.
@s4nt0s advised me this:

If you go into your GSA CB installation directory and go to folder called source, there might be an example there to help you.

But that didn't really help. Calling out @Sven or anyone who can help in this regard.
Thanks

Sven

You can take any of the simulated captcha servic apis (preferable de-captcher.com one) and exchange their domain with 127.0.0.1 and it will work fine with our tool.

See here: http://forum.gsa-online.de/discussion/7322/how-to-work-via-php

Batman

Re: How to integrate GSA Captcha Breaker with Python 

Thanks @Sven
Okay, made SOME progress. But I'm getting "unknown" for every request. Here's my code:

import os
import requests
class decaptcher:
    def __init__(self, username, password, product_id=""):
        self.__action_url = "http://127.0.0.1"
        self.__username = username
        self.__password = password
        self.__product_id = product_id
    def solve_image(self, image_path):
    	from PIL import Image
    	img = Image.open('cap.jpg')
    	data = {"function": "picture2",
            "username": self.__username,
            "password": self.__password,
            "pict_to": "0",
            "pict_type": "0",
            "captcha_platform":"Guestbook",
            "pict": open(image_path, "rb").read()}
    	request = requests.post(self.__action_url, data)
    	answer = request.content.split("|")[-1]
    	return answer
d = decaptcher("a", "b")
print d.solve_image('cap.jpg')

I'm not sure if CB is even receiving the request as nothing shows up in the log (bottom right of CB window).
I've tried changing the captcha_platform to different things like "Article Script" etc.

Sven

you can check if CB gets your request if it appears in the log of it. Replying with Unknown is happening if you it was not able to solve it.

Also use "captcha_platform" with the same name as you call the captcha type in CB. Thats not required but helps to identify the captcha typebetter.

Batman

The decapthcher API method didn't work for me. 

import requests

session = requests.session()
r = session.get("http://127.0.0.1/gsa_test.gsa")
files = {"file": open("captcha.jpg", 'rb')}
r2 = session.post("http://127.0.0.1/gsa_test.gsa", files=files)
answer = r2.content[102:107]
return answer

The captcha has 5 letters hence r2.content[102:107].

It's not a super effective approach but it works.

Sven

OK but this works right?

Batman

Yes, it does for captchas that are 5 characters long. This line can be modified for captchas with different length:

answer = r2.content[102:107]