Ask HN: Learning Advanced Python

[poplarstand]

I've accidentally assumed the role of local Python guru at my current company. Coworkers ask me about best practices, popular libraries, and deployment tooling. I find it very flattering, but I'm really just a tinkerer with a language preference.

Would anyone be able and willing to recommend me some sources to help in answering their questions? In particular, I'd be interested in hearing about:

* Publicly hosted examples of great Python code

* Concise, well-written books that discuss Python best practices and "gotchas"

* Tip sheets and style guides

Thank you in advance for any assistance you can offer.

[knopkop_]

I've seen many recommendations for the book 'Fluent Python'.

[erikschoster]

Reading this was really useful for me. It does a great job of going into depth on high level topics like the python data model and also goes deep into lots of more isolated practical topics like covering useful data structures buried in the standard library, a very good discussion of unicode, and concurrency topics, etc, etc.

Python is changing a lot and there is for example no in-depth discussion of the typing module in this book because PEP 484 had only just been written the year before its publication. There's still about four pages dedicated to the topic written as an aside to chapter 11 which discusses interfaces.

Even though it's 5 years old and there have been plenty of additions to the language in that time, I think the book is still plenty useful today -- I still pull it off the shelf as reference from time to time if I'm going into the weeds.

[existencebox]

It makes me chuckle a little bit that you feel this way, only because I recently took on similar roles in my own career, and, despite working directly with python luminaries for pretty much the last ~5 years, I still constantly question my expertise next to them. (I say this to perhaps lend solidarity/"it might not be as bad as we fear")

That said, let me give some actual answers to your questions.

1. Good hosted code. I actually asked this exact question to a python core dev a few years ago when having the same concerns you are, and I'll point you to where they pointed me: [0] (Obviously this is only one breadcrumb but it may serve to lead you to more)

2. This is a hard question since I personally shy away from books as opposed to code samples/going elbow deep. Many moons ago I used to hear "dive into python" recommended a lot but this advice may be very dated, so I'd leave this to other responders who may have more up-to-date advice.

3. There are a few here; [1] is kinda the "obvious" one if you weren't aware, we also produce a "targeted" version for my dev team (I'll caveat ahead of time, I work for the python folks at MS so I obviously have the bias of this coming to mind first, but this is our style/contribution guide. I'll also be open that we regularly find things we want to detail more effectively, so it's definitely a 'living document.' [2] In an effort to offset some of my bias :P I'd mention that the Google style guides are also notoriously good, although I am familiar with those in other spaces, such as the SRE handbook/C++ style guide.)

[0] https://github.com/brettcannon/gidgethub [1] https://pep8.org/ [2] https://azure.github.io/azure-sdk/python_documentation.html

Outside your bullets I'd mention the things that most helped me start overcoming my own concerns about being sufficiently experienced:

1. Force myself to broach ALL of the aspects I'd black boxed for years. Packaging, _proper_ async practices, keeping myself honest WRT style via black, AST, etc (obviously tune this list towards the things your team uses, the crux of this thought is more, "I used to be fine with a 'pretty good' grasp and some duct tape, I now need to go deeper to actually ensure CONFIDENT knowledge.")

2. Constructive post-mortems of my own code. Having individuals more experienced than you participate is invaluable, but not necessary. Things like, "what did I miss? What did I do wrong? What here is non-idiomatic or regrettable?" and using that to guide a list of reminders I give myself when doing work in the future; eventually these corrections become muscle memory. This will be very well served by the questions you've asked in your original post.

3. This may sound silly but in supplementing 2/1/all of the above, actually _keep up with what's going on in python._ Stuff I never did before or thought was irrelevant, reading PEPs, reading email discussion lists, reading debates about parts of the language I've never used or even style/semantic nitpicks all gives me "osmosis knowledge" from other devs, if for no other reason than exposing me to their thought processes, priorities, concepts and mental models.

Don't hesitate to shout if I can clarify any of this, and good luck!

[probinso]

David Beazley, Raymond Heddinger, Brandon Rhodes