October Meet Minutes

// under Meeting Minutes

Python Test Automation

The first talk of the evening was given by Krishnan in Python Test Automation. He explained how Python could be used for all types of testing - web testing, GUI testing, functional testing, unit testing, embedded testing etc. He explained how Python was a good choice for test automation, because it allowed you to use the same language for all the different types of testing, in contrast to specific automation test software, each specialised in one type of testing, and each requiring a different tool language to be learnt.

He then went on to list useful python modules for testing, some of which are

• shutil - for copying files quickly
• subprocess - for running another process from your program, and monitoring its status, getting and using its output
• pexpect - for interacting with console applications
• logging - for logging, but where u can specify different levels of errors, like debug, error, warning and critical
• unittest - for writing unit tests
• regex -
• pep8 and pylint - for checking if your source code confirms to python coding standards
• traceback and pdb - standard python debugging libraries to step through your code.
• pywinauto - for automating windows tasks, including installations!

Two other interesting things mentioned were image based automation and the difference between unittests and integration tests.

In image based automation you automate windows tasks like clicking a shortcut, emptying the recycle bin by getting the object you want to automate by searching for it through its image, rather than through the windows API or some other dictionary/list.

Unit tests are where you check each and every part of the code separately for its functionality, error conditions etc. Integration testing is where you check one module of the code for its ability to work properly with other outside objects. Each module then acts like a black box, and you are more concerned about its working with other modules rather than how the module itself works ie: whats inside the black box

Docopt

The second talk was given by Rengaraj on Docopt. Docopt is a python module used to generate an interface for your command line application. it also automatically generates a parser for it. I had some confusion on where this module could be used, so I went back home and checked it out. Basically, you can use it to easily create a command line program where you can give options like -v, -x -u and -uv following your command. The great thing about Docopt is that it will then create a dictionary with your options as keys and the options the user put as values to those keys, so that you can easily find out the user input commands were. Its much much easier than other similar functional modules like optparse. The only thing to remember with Docopt is that the usage format has to be in the POSIX format for it to work

Building Embedded Systems with Twisted

The last talk was given by Vijay Kumar on a problem he solved at his company using Python and some embedded hardware. He explained how at his company they were faced with the problem of keeping track of who had borrowed common resources and of getting the person who had borrowed it to return the resource after he/she had used it. After trying simple solutions like keeping a notebook, reminding people, and finding that they did not work, they finally decided to use a tech solution. They used an RFID reader and put an RFID tag on each resource and linked it to a server so that they would know who had borrowed which resource. Python was used to interface and talk to the hardware, which included a reed switch, RFID reader, and the server. The Python framework used was Twisted, and one interesting thing was the idea of an event driven framework, instead of a blocking read/write framework. It was nice to see the actual hardware, software and also a video of the system in action! Vijay also showed us the state machine of the system which showed the various states of the system and how it changed depending on the users actions.

Credits

• Venue Coordinator: Shrinivasan
• Networking over Tea, Sponsor: Clay Labs
• Marketing: Krishnan Shankar, Prasanna Venkadesh, Chintu Philips, Shrinivasan
• Photographs: Krishnan Shankar, Kenash Kanagaraj
• Minutes: Harsha

Go Top