Python

Footguns

difference between is and ==

Tips and Tricks

from time import sleep
Open your current file in interactive mode to inspect/debug it python -i file.py
String templating print("%s : %s" % (name, age))
Pipenv environments are stored in ~/.local/share/virtualenvs. Can be deleted as part of system cleanup

Lists

`a = [1, 2, 3]

Python now has tools inspired from haskell to operate with lists in the module itertools. Documentation is here

Use a list as stack

stack = [2,3,4]
stack.append(5)
stack.append(6)
stack.pop()
stack.pop()

Use a list as Queue

from collections import deque
queue = deque(["apple", "banana", "cat"])
queue.append("dog")
queue.append("elephant")
queue.popleft()
queue.popleft()

List Comprehensions a concise way to create list

A list comprehension consists of brackets containing an expression followed by a for clause, then zero or more for or if clauses. The result will be a new list resulting from evaluating the expression in the context of the for and if clauses which follow it.

List Display and Comprehension a = [x*x for x in range(10)]

Generators

If your function uses the yield command, its a generator.

def natural_numbers():
    i = 0
    while True:
        yield i
        i ++

You can pass generators to functions as paramters by prefixing * to the parameter name - def fun(*gen)

Multiple ways to use iterators

with a for loop

    for num in natural_numbers:
        print(num)

inline usage

  value = sum(next(t) for n in slice(natural_numbers, 10))

Use yield without a value

Simply writing the command yield, yields control to the calling code.

Convert a generator into a context manager

from contextlib import contextmanager

@contextmanager
def db_test(cur):
  cur.execute('create table points(x int, y int)')
  try:
    yield
  finally:
    cur.execute('drop table points ')

Class

Special functions in a Class

__init__ : constructor
__repr__ : print your class sensibly
__add__ : operator overload the + operator
__iter__, __next__ : make the class iterable
- raise StopIteration()
__enter__, __exit__ : make your class a ContextManger, with setup and teardown functions.

Internals

see what happens in the python bytecode

from dis import dis
dis(function_name)

inspect source code of a function in runtime

from inspect import getsource
getsource(add)

Ecosystem

Eventlet brings concurrency to python

References

So you want to be a python expert by James Powell
Transforming Code into Beautiful Idiomatic Python by Raymond Hettinger
Documentation for Collections
An opinionated guide to python

Starting a new Python project

Pipenv for package management and virtual environment

Notes from Idiomatic Python

ChainMap
Clarify function calls with keyword arguments
- You are sacrificing performance but improving developer time
named tuples

from collections import namedtuple


def named_tuples(a: int):
    Results = namedtuple("Result", ["status", "payload"])
    if a < 0:
        return Results("error", {})
    if a > 0:
        return Results("success", {"number": a})


print(named_tuples(3))
print(named_tuples(-1))

Updating Multiple state variables
Decorators and Context Managers : Help separate business logic from administrative ones

Async

Notes from Keynote on Concurrency, PyBay 2017 Goal : When to use threads, processes, asyncs. Advantages and disadvantages

Functional Programming

Creating partial functions - documenttion

This reminds me of currying in haskell

def user(name, age, role):
    return {"name":name, "age":age, "role":role}

manager = partial(user, role="manager")
admin   = partial(user, role="admin")

print(user(name="denny", age="32", role="user"))
print(manager(name="denny", age="38"))
print(admin(name="denny", age=42))

Pattern and Matching

https://peps.python.org/pep-0636/ How to implement delegation using pattern matching

Requires python 3.10+

action_pass = {"type": "pass", "payload": {"name": "denny"}}
action_keep = {"type": "keep", "payload": {}}

action = action_pass

match action:
    case {type, payload}:
        print(type)

Live Development Process with Reload

Uses the inotify syscall on linux to trigger reload of the program upon file changes

import inotify.adapters
import subprocess

def watch():
    i = inotify.adapters.Inotify()

    i.add_watch('./')

    for event in i.event_gen(yield_nones=False):
        (_, type_names, path, filename) = event

        #print("PATH=[{}] FILENAME=[{}] EVENT_TYPES=[{}]".format(path,filename,type_names))
        for type in type_names:
            #print("type: {} path: {} filename: {}".format(type, path, filename))
            if type == "IN_CLOSE_WRITE" and ".py" in filename:
                print('reload file')
                subprocess.run(["python", "hello.py"])

if __name__ == '__main__':
    watch()

Numerical and Scientific Computing

import numpy as np
import plotly.express as px

x = np.linspace(-np.pi, np.pi, 50)
y = np.array([np.sin(i) for i in x])
fig = px.scatter(x=x, y=y)
fig.show()

SciPy, Pandas and OpenCV use numpy array as the common format for data exchange

denntenna