bettercaptcha

Captcha

Captcha illustrator with Pillow and Numpy for Python 3.6 or higher

Installation

Install without distortion setting:

pip install captcha-MagicaFreak

Install with distortion setting:

pip install bettercaptcha[distortion]

or

pip install bettercaptcha
pip install numpy

Captcha Class

All the possile CAPTCHA variants

CaptchaImage

The class CaptchaImage generates an Image with given Text or generated Text. The look of the CAPTCHA can be changed through the default Arguments, setting Arguments and diffrent overwrites.

The most basic usage of this class looks like this:

from bettercaptcha import CaptchaImage

captcha = CaptchaImage()

image = captcha.create()
text = captcha.text

Here is the image a PIL.Image instance and text the text what is written in the CAPTCHA

Default Arguments:

Name	Type	Description	Default
font	str	path to the font file	arial.ttf
font_size	int	size of the font	24
color	tuple	text color of the CAPTCHA	(0, 0, 0) => black
text	str	text on the captcha	random generated string
char_amount	int	length of the random generated string	12

Setting Arguments:

Name	Type	Description	Default
background	Background	sets a background on the captcha (see more under Background)	None
distortion	bool	distorts the text (Needs the numpy package)	False
lines	int	adds the amount of lines between random Points	None
points	int	adds the amount of points on random Points	None
frame	bool	adds a frame around the CAPTCHA	False

Background

The class Background has all modes for the setting background.

The first mode is Background.RANDOM. It generates a random color in every Pixel of the background. The color is choosen from the list CaptchaImage.colors that can be overwritten.

The second mode is Background.COLOR. It colors the background with the text color with only 50% alpha.

Overwrites

Overwrites can be used in many diffrent cases, but the 2 biggest would be if the captcha is generated with some flexibiltiy, or you generate another captcha with diffrent look but same text.

The default argumnets can't be overwritten except color The overwrites for color would look like this:

captcha = CaptchaImage()
captcha.color = (52, 52, 52)

Important to notice that captcha.color can only be a rgb value!

All the setting arguments can be overwritten, this would look like this:

captcha = CaptchaImage()
captcha.settings['distortion'] = True
captcha.settings['background'] = Background.RANDOM

The third kind of overwrites are things that are normaly contstaned. Under this counts CaptchaImage.colors, CaptchaImage.distortion and CaptchaImage.distortion_factor.

CaptchaImage.colors is a list or RGB colors for the background setting Background.RANDOM, it can be overwritten with another list of RGB colors, like so:

captcha = CaptchaImage(background=Background.RANDOM)
captcha.colors = [(255, 0, 0), (0, 255, 0), (0, 0, 255)]

CaptchaImage.distortion is a mathematical function. To overwrite it you need to create another function with the values x, a and w. It can be overwritten like this:

from math import sin, pi

captcha = CaptchaImage(distrotion=True)


def distortion(x, a, w):
    return a * sin(2 * pi * x * w)
# or
distortion = lambda x, a, w: a * sin(2 * pi * x * w)

captcha.distortion = distortion

CaptchaImage.distortion_factor is the last overwrite it is a tuple of ints that are influencing the values a and w . The calculation of a and w looks like this:

a = image_height / distortion_factor[1]
w = distortion_factor[0] / image_width

To overwrite the complete distortion factor you would do this:

captcha = CaptchaImage(distrotion=True)
captcha.distortion_factor = (0.59, 17.0)

CaptchaImage.rand_chars(amount, chars)

This function generates a random string with the length of the amount, but it only usese every character one time out of the list. Important to notice that if you give a chars list smaller than the amount it will not work!

text = CaptchaImage.rand_chars(12)

Arguments:

Name	Type	Description	Default
amount	int	length of the string	12
chars	list	list of used characters	all ascii letters and digits