How to maintain legacy Python codebases?

Key Takeaways

how can you maintain a legacy Python code basis great question by a reader in this video I'm gonna give some thoughts around the topic so let's define what a legacy Python code base is so generally when people refer to legacy code they mean code that has aged a little bit you know we're talking about an older code base maybe this is some code written in python 2.5 2.6 Python 2 in general or maybe it's you know it's even older it's a system that's kind of hobbling along on Python 1 and it's never been upgraded to the latest version of the language and so now you've been assigned the task to maintain a code base like that in general I think you have two choices there either keep it on life support keep it running as is don't upgrade to a newer version of Python or option 2 you're gonna upgraded to a newer version of Python your going to take the code you're gonna make some refactoring and you're gonna make it work you're gonna take the existing code base and gonna make it work on a more recent version of Python or on the latest version of Python even so these are kind of the two options that you have really right and there are a couple of strategies you can you can apply to to each of them so if you want to maintain the existing code base which in a lot of cases will make the most sense or will make a lot of sense like generally it is very hard to rewrite an existing code base without introducing bugs and so upgrading a Python one code base to Python 3 if it's a lot of code that could be a very very stressful a very very tough project not saying it doesn't make sense but it is often very hard to make a real good business case for doing that you know if you're asking me like what would I do as a sort of idealistic developer totally I would upgrade it and rewrite it in Python 3 and make it super nice and you know add linters and everything but sometimes that is not the smartest business decision you know if this thing generates value maybe you don't want to touch it when you want to focus on something else or your employer wants you to focus on something else so you know if you have to maintain a legacy code base then that's that's just a reality of it and you can do some things to make your life a little bit easier if you're maintaining a legacy Python code base so maybe you can package the whole thing up into virtual machines or at least you don't have to worry about future OS updates breaking your application you know maybe maybe you can do things like that that basically seal this thing off and kind of preserve it and keep it hobbling along and hopefully you don't have to make too many changes to it now if you're going the other route and you want to take an existing code base and upgrade it to the latest version of Python then that's an entirely different game right then you're likely rewriting at least parts of the application to make it work with newer versions of Python you don't always have to go all the way and let's say upgrade from from python 2.2 to python 3.7 or 3.6 whatever the latest version is at the time sometimes it can just make sense to go to the latest version of python 2 for example you know it's a lot more work to go from python 2.2 to python 3 compared to going from python 2.2 to python 2.7 and python 2 is going to be around for a long time so if you're on 2.7 then you know that gives you some more mileage out of the existing code base without having to make more drastic changes and really those I think are the two main options that you have for maintaining a legacy Python code base either keep it around make it bearable or upgraded to the latest version or a newer version of Python all right I hope this was helpful if you've got any tips to share and how to work with legacy Python code bases then please leave a comment below if you've got a request for a video that you'll like me to make on Python or Safa development in future and please also leave a comment and I'll try and cover that in a future video all right thanks so much for listening and happy Python eek