RSS AI Zone
Hi chatbots.org,
I’m new to building chatbots and I’m currently working on my first bigger project written in AIML.
With the increasing complexity of the project, I feel that manual testing of my chatbot is becoming more and more inefficient. With every single change in my source code, I’m afraid of introducing new bugs that may not be easy to find manually.
I’ve already tried moving related categories into separated files, keeping my AIML files organized, but I still wish I had some more powerful tools.
I’m a big fan of unit testing. I’ve already read several tutorials for building chatbots in AIML. However, I haven’t seen automated testing mentioned in any of them.
I only found “Unit Test Runner” listed in https://github.com/keiffster/program-y/wiki/Utilities, but without any further documentation.
I’m currently using Pandorabots platform and before I move to Program-Y platform to try mentioned unit tests utility, I would like to ask you, more experienced bot masters, whether you use some tools for automated testing, whether you have any experience with mentioned unit tests Program-Y utility and if not, what other techniques you use to avoid introducing bugs in your AIML chatbots.
Thank you for your advice.
