#21 Python has a new star framework for RESTful APIs

This episode has been sponsored by Rollbar. Get a special offer via http://rollbar.com/pythonbytes

#1 Brian: profile and pstats — Performance Analysis

• Doug Hellman is working on the Python 3 MOTW series that was so successful for Python 2.
• Recent edition is profile and pstats, for profiling parts of your code you may have concerns with and finding out where the slow bits are.

#2 Michael: API Star by Tom Christie

• A smart Web API framework, designed for Python 3.
• A few things to try right away:

    $ pip3 install apistar
    $ apistar new --template minimal
    $ apistar run
    $ apistar test

• API Star allows you to dynamically inject various information about the incoming request into your views using type annotation.
  • e.g.

    def show_query_params(query_params: http.QueryParams):
        return {
            'params': dict(query_params)
        }

• You can instead set the status code or headers by annotating the view as returning a Response

def create_project() -> Response: ...

• Parameters are automatically passed into views from routes (annotations!):

    def echo_username(user_id: int):
        return {'message': f'Welcome, user {user_id}!'}

• Performance: Faster than sanic!

#3 Brian: Yes, Python is Slow, and I Don’t Care

• Optimize for your most expensive resource. That’s YOU, not the computer.
• Choose a language/framework/architecture that helps you develop quickly (such as Python). Do not choose technologies simply because they are fast.
• When you do have performance issues: find your bottleneck
• Your bottleneck is most likely not CPU or Python itself.
• If Python is your bottleneck (you’ve already optimized algorithms/etc.), then move the hot-spot to Cython/C
• Go back to enjoying getting things done quickly

#4 Michael: A Quick Introduction: Hashing

• Article by Gerald Nash
• Hashing is a method of determining the equivalence of two chunks of data.
• A cryptographic hash function is an irreversible function that generates a unique string for any set of data.
• Example

    import hashlib as hash
    sha = hash.sha256()
    # Insert the string we want to hash
    sha.update('Hello World!')
    # Print the hexadecimal format of the binary hash we just created
    print(sha.hexdigest())
    # 4d3cf15aa67c88742e63918825f3c80f203f2bd59f399c81be4705a095c9fa0e

• Know when to choose “weak” hashes vs. strong ones
• Straight hashes are not enough for security (e.g. passwords). Use passlib and be done.

#5 Brian: Wedding at Scale: How I Used Twilio, Python and Google to Automate My Wedding

• gspread to access a google spreadsheet of guests and phone numbers
• SMS guests with twilio
• replies handled by a flask app
• gathered accept/decline/didn't reply statistics
• reminder texts
• food selections and replies and reminders, all handled by Python

# 6 Michael: python-alexa: A Python framework for Alexa Development

• by Neil Stewart
• Ordered an amazon assistant.
• Before it arrived, I had challenged myself to develop something for it
• Project: VoiceOps, interact with an AWS account, such as telling me how many running and stopped instances there is or what RDS databases are in an account
• Wanted a framework that would make Alexa development super easy.
• Decided a new framework was needed: python-alexa
• python-alexa on github
• echo shim for testing without hardware

Our news:

Michael: Just added full text search (including within videos) to Talk Python courses.

Brian: Netflix chaos engineering interview on Test & Code