Python has grown tremendously this year, and there’s been a lot of new stuff. Here are five libraries, tools, and developers to keep an eye on in 2018. See if there are any you haven’t heard of?
Five Python libraries to watch
Web field: Sanic
channelcat/sanic github.com 
The name of the library is a reference to a popular YouTube meme in which someone drew Sonic the little blue man and ended up with a serious drawing of the following, Sanic, with the caption “Gotta Go Faster.”
This library is similar to Flask, but much faster, reaching 36,000 requests per second in tests. Star growth nearly doubled in 2017. Gotta go faster!
Environment and package management: Pen V
pypa/pipenv github.com 
The library is a combination of environment management and package management, written by Kenneth Reitz (author of Requests) and now handed over to the Python authorities for maintenance, providing better package management than the PIP experience. Slogon is Python Development Workflow for Humans, which solves the problem of inconsistent environments and package installation.
Reptile: Requestium
tryolabs/requestium github.com 
Now the anti-crawler technology is more and more powerful, the common request + parse crawler is not so effective with the increasing frequency and the complexity of front-end development. The benefit of Requestium is that it incorporates Chrome 59’s headless feature (instead of PhantomJS). Although slightly lower efficiency, but stable, front-end rendering of the web page can also be climbed and resolved. It’s a combination of Requests, Selenium, and Parsel.
Deep learning: Caffe2
caffe2/caffe2 github.com 
Is a deep learning library with the advantage that it is maintained by the Facebook team. Tensorflow is probably the one that most people hear about in deep learning, but Caffe2 is much simpler and lighter to learn, and its background allows it to have a robust development.
Text Processing: FlashText
vi3k6i5/flashtext github.com 
The developers of the library posted an article on Medium with the headline: Regex was taking 5 days to run. So I built a tool that did it in 15 minutes. This library can quickly search and replace text in large corpora. When the number of keywords is >500, FlashText searches start to outpace regex.
Five tools to watch
The strongest terminal: Upterm
railsware/upterm github.com 
In fact, I would have recommended fish or ZSH, but in fact, I mainly like the feature of autocompletion. The Upterm I am using recently is actually very simple and easy to use. It is a full-platform terminal, which can be said to be an IDE in the terminal, with powerful auto-completion function. His previous name was BlackWindow, and someone told him that this name was not good for community promotion. After he changed his name to Upterm, it now has 17000+ Star.
Interactive interpreter: Ptpython
jonathanslenders/ptpython github.com 
An interactive Python interpreter. Syntax highlighting, hints and even vim and Emacs typing modes are supported. Ptpython is also built into the online terminals we provide in the course.
Package management essential: Anaconda
anaconda.org/anaconda.org
Anaconda is really highly recommended as it helps you install a lot of cumbersome things including: Python environments, PIP package management tools, common libraries, setting up environment paths, and more. These things to do one by one, easy to encounter all kinds of problems, bring frustration.
If you want to do data things with Python, just install it. It even developed Numba, a JIT interpreter. So with the JIT, Anaconda can also do things that require higher efficiency of online scientific computation.
Editor: Sublime3
A sophisticated text editor for code, markup and prosewww.sublimetext.com
If you don’t want to start an editor war, of course you should start with PyCharm, but sometimes when you’re writing a lightweight script, you’ll want to use a lighter tool.
Using various editors this year, Sublime has been updated with the True, Matchless, Triple zero version, and in many areas (especially in the aspect area of concern) is much easier to use than before. With the installation of Anaconda or CodeIntel plug-ins, Sublime can have an almost IDE experience.
In addition, I think that Sublime font color rendering is very important for OBSESSIVE compulsive disorder. Like VSCode Atom editor, I always feel that the color is a little dark and not bright, just like the image made in PS will be different in the browser, which will make me uncomfortable and even nauseating if I look at it for a long time.
Front-end online Editor: CodeSandbox
codesandbox.io/codesandbox.io
It’s not a Python development tool, but it’s a handy online editor for back-end engineers who want to write front ends. Instead of installing thousands of NPM packages, it’s already done in the cloud, allowing you to write code and see what it looks like. React, Vue and other mainstream front-end frameworks are supported.
Five developers to watch
If you are having trouble with other Python tutorials, try muggle Programming, an introductory course for beginners who are addicted to it