https://github.com/rectangletangle/nlplib https://github.com/rectangletangle/nlplib "What is your project" A bunch of tools for natural language processing, w

https://github.com/rectangletangle/nlplib https://github.com/rectangletangle/nlplib "What is your project" A bunch of tools for natural language processing, with an emphasis on code composability. "How has the decision to use Python 3 affected the project?" Because my project has an entirely new code base, it made sense to aim for longevity by using the most recent version of the language. It seems like a lot of the FUD around 3 has been about third-party library support. However, I've found that everything I really needed was readily available for 3. Though there're probably some specific usecases that 3 just doesn't quite cover yet. Seeing as most of the major third-party libraries have made the switch, I think that it will take major Linux distros shipping 3 as the default for it really to catch on. Print being a function, has been a nice addition seeing as it's handy to pass it as a logging callback; it can also now be used within a lambda. There's a bunch of nifty new syntactic sugar like extended iterable unpacking, which make life just a little bit easier. The more generalized comprehension syntax is nice, seeing as it's an extremely powerful way to make sets, dicts, or whatever. The cleanup done to the standard library, specifically urllib has been nice. I haven't found a good use for the new function annotation syntax; however, I'm sure someone will come up with a neat way to utilize it. Subgenerators for recursive flattening are nice, as well as the generator syntax as a whole.