apologies for formatting, striving for better readability this time...
not repeating the source code, just the other part.
Reformatted:
-----------
From: kirby urner <kirby.ur...@gmail.com>
Date: Thu, Jul 21, 2016 at 2:02 PM
Subject: Re: teaching Python 3.5 in 2016: snapshot
To: "edu-sig@python.org" <edu-sig@python.org>
My "arc" (trajectory) for teaching Python is something like:
>>> dir()
>>> dir(__builtins__)
>>> type(3)
>>> type('A')
>>> import math
... using Python as a calculator and wordprocessor,
including file i/o (open, close, 'r' 'w' 'a')
>>> list, tuple, dict
>>> immutability vs. mutability, explaining how ([], []) is an immutable
"containing" mutables.
>>> import collections.namedtuple
... dot notation access to Atom(protons=, symbol=, long_name=, mass=)
talk about APIs as: car dashboard (driverless car
controversies); web APIs (point to API companies
like Apigee, Google, programmable web).
Linking Object.verb()
and
Object.noun to API concept also
(docstrings = owner's manual / user's manaul
-- but not a repair manual, which is for #comments
in source code and tests).
tell stories about Browser Wars, emergence of
browser as universal client, HTTP as API, RESTful
vs SOAP web services, play with existing APIs that
return JSON
>>> string formatting with new mini-language feature set
>>> def function() notation, arguments vs parameters
... *args, **kwargs as either
>>> save function objects to a list, show functions as
... top-level, lambda syntax
>>> class Element: that mimics namedtuple Atom.
.... Compare side-by-side
>>> composer.py (introduce operator overloading)
now show how "namedtuple" API may be duplicated
using __getitem__
>From here on, we're prepared to use special names
(or "__ribs__" as I sometimes
call them) as a way to talk about APIs.
E.g.
* the iterator protocol around __next__ and __iter__
may now be presented using classes.
* Context managers (__enter__, __exit__)
* Descriptors (__set__ and __get__)
We'll get to some of these below, after decorator syntax...
>>> inheritance, MRO (method resolution order)
>>> old fashioned menu-driven API based in input() to do
....CRUD against all_elements
# <--- this is where I'm at in the current 5-week course
I'm teaching
>>> annotations
What I call "grand unification" is up ahead where we show
how a context manager maybe be built with generator
as a decorator.
Haven't gotten there yet with this class. That's almost
black belt, like maybe brown.
====
_______________________________________________
Edu-sig mailing list
Edu-sig@python.org
https://mail.python.org/mailman/listinfo/edu-sig
