Developed by Jon Winsley
- numpy
- pillow
- opencv
- keyboard
Lackey is a Python implementation of Sikuli script, allowing you to run automation scripts developed in the Sikuli editor with pure Python. If you're trying to run Sikuli scripts in a Java-free environment or integrate them into an existing Python testing structure, this is the library for you.
Installation is easy:
pip install Lackey
Then you can just import Lackey at the head of your Sikuli-script python file:
from lackey import *
WARNING Be aware that this will create global methods that will overwrite certain Python functions, such as type(). For more information, see the Sikuli Patching section below.
The Lackey library is divided up into classes for finding and interacting with particular regions of the screen. Patterns are provided as bitmap files (supported formats include .bmp, .pbm, .ras, .jpg, .tiff, and .png). These patterns are compared to a Region of the screen, and, if they exist, can target a mouse move/click action.
If you've used Sikuli, you'll feel right at home. Lackey is designed to be a drop-in shim for Sikuli.
Sample code (note that you'll need to provide your own PNGs):
from lackey import *
click("Start_Button.png")
wait("Control_Panel.png", 5) # Maybe the Start menu is slow
click("Notepad.png")
In most cases, you won't need to run Lackey with elevated privileges. However, Windows will not let a non-elevated script send mouse/keyboard events to a program with elevated privileges (an installer running as administrator, for example). If you run into this problem, running Lackey as administrator (for example, by calling it from an Administrator-level Powershell instance) should solve your issue.
Full API documentation can be found at ReadTheDocs.
In my line of work, I have a lot of tasks walking through line-of-business applications to do boring things that any computer could do. Laziness being the mother of invention, I decided to script what I could. I found SikuliX to be a tremendously valuable tool for the job, but its Java dependencies and limited Python coupling posed problems in several cases. So, I decided to create a pure Python implementation of Sikuli script.
There are some existing libraries for this purpose, like pywinauto and autopy, but they didn't work for me for one reason or another. I wasn't doing a lot of Windows GUI interaction with these particular applications, so pywinauto's approach wouldn't help. I needed something that could search for and use images on screen. autopy was closer, but it had quite a few outstanding issues and hadn't been updated in a while.
Most of my automation is in Windows, so I've begun this library with only Windows support. As of version 0.7.0, it also includes Mac OS X support,and it's designed to eventually be extended with support for Linux by implementing an additional "PlatformManager" class. I'll get around to this at some point, but if you'd like to contribute one sooner, please feel free!
My goal with this project is to be able to reuse my existing library of Sikuli scripts with minimal modifications. To that end, Lackey will map certain functions of the screen region (find(), click(), etc.) to the global scope. This means you can use the Sikuli IDE for development, and run the final product with pure Python! Add the following line to your Sikuli python script, and you should be able to run it in Python largely without issue:
from lackey import *
Note that I have had to adjust some of my image search similarity settings in a couple cases. Your mileage may vary. Please report any issues that you encounter and I'll try to get them patched.
Be aware that some Sikuli-script methods actually overwrite Python-native functions, namely type() and input(). Where this is the case, I've remapped the native functions by adding a trailing underscore. They can be accessed as follows:
from lackey import *
username = input_("Enter your username: ") # built-in Python input
Each platform (Windows/OSX/Linux) needs its own PlatformManager (see documentation above) to abstract OS-level functionality, like simulating mouse clicks or key presses. Ideally, these should be implemented with as few 3rd-party library dependencies as possible. If you'd like to contribute a PlatformManager for your OS, feel free to submit a pull request!
Don't forget to update the unit tests and verify that they still run.
This library is currently under development, and may have some bugs. Check the Issues list to find features/bugs you can help with!
To build the wheel from source, cd to the project directory and run:
python setup.py bdist_wheel
To link directly to the repository (if you want to work with the develop ring, for example), cd to the project directory and run:
pip install -e ./
(Note that you may need to install the wheel package.)
Debugging contributions: