Brought to you by:

#193
Get More Refcardz! Visit refcardz.com

CONTENTS INCLUDE:

Core Python
❱ Python 2.x vs. 3.x
❱ Branching, Looping, and Exceptions
❱ The Zen of Python
❱ Popular Python Libraries
❱ Python Operators By: Naomi Ceder and Mike Driscoll
❱ Instantiating Classes... and More!

Python is an interpreted dynamically typed Language. Python uses Comments and docstrings
indentation to create readable, even beautiful, code. Python comes with To mark a comment from the current location to the end of the line, use a
so many libraries that you can handle many jobs with no further libraries. pound sign, ‘#’.
Python fits in your head and tries not to surprise you, which means you
can write useful code almost immediately.
# this is a comment on a line by itself
x = 3 # this is a partial line comment after some code
Python was created in 1990 by Guido van Rossum. While the snake is
used as totem for the language and community, the name actually derives
from Monty Python and references to Monty Python skits are common For longer comments and more complete documentation, especially at the
in code examples and library names. There are several other popular beginning of a module or of a function or class, use a triple quoted string.
implementations of Python, including PyPy (JIT compiler), Jython (JVM You can use 3 single or 3 double quotes. Triple quoted strings can cover
integration) and IronPython (.NET CLR integration). multiple lines and any unassigned string in a Python program is ignored.
Such strings are often used for documentation of modules, functions,
classes and methods. By convention, the “docstring” is the first state-
Python 2.x vs. Python 3.x ment in its enclosing scope. Following this convention allows automated
Python comes in two basic flavors these days – Python 2.x (currently production of documentation using the pydoc module.
2.7) and Python 3.x (currently 3.3). This is an important difference –
some code written for one won’t run on the other. However, most code is In general, you use one line comments for commenting code from the
interchangeable. Here are some of the key differences: point of view of a developer trying to understand the code itself. Docstrings
are more properly used to document what the code does, more from the
Python 2.x Python 3.x point of view of someone who is going to be using the code.

Python is the sort of language that you can just dive into, so let’s dive in
print “hello” (print print(“hello”) (print is a function)
with this example Python script:
is a keyword)

except Exception, e: # OR except Exception as e: # ONLY #! /usr/bin/env python

except Exception as e “”” An example Python script
Note that triple quotes allow multiline strings
Naming of Libraries Improved (but still imperfect) consistency with “””
and APIs are frequently PEP 8 guidelines
inconsistent with PEP 8 # single line comments are indicated with a “#”

Strings and unicode Strings are all unicode and bytes type is for import sys # loads the sys (system) library
unencoded 8 bit values def main_function(parameter):
“”” This is the docstring for the function “””
There is a utility called 2to3.py that you can use to convert Python 2.x print “ here is where we do stuff with the parameter”
print parameter
code to 3.x, while the ‘-3’ command line switch in 2.x enables additional
deprecation warnings for cases the automated converter cannot handle. return a_result # this could also be multiples
Third party tools like python-modernize and the ‘six’ support package
if __name__ == “__main__”:
make it easy to target the large common subset of the two variants for “”” this will only be true if the script is called
libraries and applications which support both 2.x and 3.x. as the main program “””

language features

Programming as Guido indented it...

Indentation rules in Python. There are no curly braces, no begin and end
keywords, no need for semicolons at the ends of lines - the only thing that
organizes code into blocks, functions, or classes is indentation. If some-
thing is indented, it forms a block with everything indented at the same
level until the end of the file or a line with less indentation.

While there are several options for indentation, the common standard is 4
spaces per level:

def function_block():
# first block
# second block within first block
Core Python

stuff
for x in an_iterator:
# this is the block for the for loop
print x
# back out to this level ends the for loop
# and the second block...
more first block stuff
def another_function_block()

DZone, Inc. | www.dzone.com

 Core Python

Keyword Usage
# command line parameters are numbered from 0
# sys.argv[0] is the script name if <expression>: Conditional expression that only executes if True
param = sys.argv[1] # first param after script name
# the line below calls the main_function and else: Used primarily as a catchall. If <expression> is False,
# puts the result into function_result then we fall into the else
function_result = main_function(param)
elif: Use elif to test multiple conditions.
while <expression>: The while loop only loops while an expression
evaluates to True.

Branching, looping, and exceptions break Breaks out of a loop

continue Ends current iteration of loop and goes back to top
Branching of loop
Python has a very straightforward set of if/else statements: try: Begins a block to check for exceptions
except <exception>: Followed by Exception type being checked for,
if something_is_true: begins block of code to handle exception
do this
elif something_else_is_true: finally Code that will be executed whether exception
do that occurs or not
else:
do the other thing

DaTA OBJECTS
The expressions that are part of if and elif statements can be comparisons
(==, <, >, <=, >=, etc) or they can be any python object. In general, zero and Variables and Types
empty sequences are False, and everything else is True. Python does not Python is a dynamically typed language, but it is also a fairly strongly typed
have a switch statement. language. So a variable could end up referring to different types of objects,
but the object that it’s referring to at any given moment is strongly typed.
Loops For example:
Python has two loops. The for loop iterates over a sequence, such as a list, a
file, or some other series:
x = 1 # x points to an integer object
y = 2 # y also points to an integer object
for item in [‘spam’, ‘spam’, ‘spam’, ‘spam’]:
print item z = x + y # z points to an integer object – 3
a = y # a points to the same int object as y

The code above will print “spam” four times. The while loop executes while a y = “2” # y now points to a different object, a string
condition is true:
z = x + y # throws a type mismatch (TypeError) exception since an
integer and a string are different types and can’t be added.
counter = 5
while counter > 0: z = x + a # z now points to an int (3), since a is pointing to
counter -= 1 an int

With each iteration, the counter variable is reduced by one. This code Duck typing - if it quacks like a ...
executes until the expression is False, which in this case is when “counter” While Python objects themselves are strongly typed there is a large amount
reaches zero. of flexibility in how they are used. In many languages there is a pattern of
checking to be sure an object is of the correct type before attempting an
Handling Exceptions operation. This approach limits flexibility and code reuse – even slightly dif-
Python is different from languages like C or Java in how it thinks about er-
ferent objects (say, a tuple vs. a list) will require different explicit checking.
rors. Languages like Java are “look before you leap” (LBYL) languages. That
is, there is a tendency to check types and values to make sure that they are
In Python, things are different. Because the exception handling is strong
legal before they are used. Python, on the other hand, thinks of things more in
we can just go ahead and try an operation. If the object we are operating
a “easier to ask for forgiveness than permission”(EAFP) style. In other words,
on has the methods or data members we need, the operation succeeds. If
Pythonic style would be more likely to go ahead and try the operation and
not, the operation raises an exception. In other words, in the Python world
then handle any problems if they occur:
if something walks like a duck and quacks like a duck, we can treat it like a
duck. This is called “duck typing”.
try:
item = x[0]
except TypeError:
#this will print only on a TypeError exception
Python Data Types
print “x isn’t a list!” Python has several data types. The most commonly found ones are shown
else: in the following table:
# executes if the code in the “try” does NOT
# raise an exception
print “You didn’t raise an exception!” Type Description
finally:
#this will always print An integer of the same size as a long in C on the
int
print “processing complete”
current platform.
In this case, a list or sequence operation is attempted and if it fails because An integer of unlimited precision (In Python 3.x this
long
it’s the wrong type, the except clause just deals with it. Otherwise the becomes an int).
exception will be raised normally. Then, whether an exception happens or
not the finally clause will be executed, usually to clean up after the opera- float A floating point number , usually a double in C on the
tion in either case. current platform.

complex Complex numbers have a real and an imaginary

component, each is a float.

boolean True or False.

DZone, Inc. | www.dzone.com

 Core Python

Python built-in object types

def a_function():
Python also has built-in object types that are closely related to the data “”” document function here”””
types mentioned above. Once you are familiar with these two sets of tables, print “something”
you will know how to code almost anything!

Type Description Parameters

Parameters can be passed in several ways:
list Mutable sequence, always in square brackets: [1,
Default parameters:
2, 3]

tuple Immutable sequence, always in parentheses: (a, b, c) def foo(a=2, b=3):

print a
dict Dictionary - key, value storage. Uses curly braces: foo()
{key:value}
Collection of unique elements unordered, no By position:
set
duplicates
foo(1, 2)
str String - sequence of characters, immutable

unicode Sequence of Unicode encoded characters By name:

foo(b=4)
Python operators
The following table lists Python’s common operators:
As a list:
Operator Action Example def bar(*args):
print args
+ Adds items together; for 1 + 1 -> 2 bar(1, 2, 3)
strings and sequences "one" + "one" -> "oneone"
concatenates As a dictionary:
- subtraction 1 - 1 -> 0
def foo(a, b=2, c= 3):
* multiplication, with strings, 2 * 3 -> 6 print a, b, c
repeats string "one" * 2 -> "oneone" d = {a:5, b:6, c:7}
foo(**d)

/ (//) division, division of integers 3/4 -> 0 (2.x)

results in an integer with 3/4 -> 0.75 (3.x) See also keyword arguments (i.e. **kwargs), which allows you to take an
truncation in Python 2.x, 3//4 -> 0 (3.x) arbitrary number of keyword arguments. You can read more about it here:
a float in Python 3.x (// is http://docs.python.org/2/tutorial/controlflow.html#keyword-arguments.
integer division in Python
3.x)
Returning values
** Exponent - raises a number You can return any Python object from a function – ints, floats, lists,
to the given exponent dictionaries, anything.

Sequence indexes and slicing return dict(“color”: “blue”)

There are several Python types that are all sequential collections of items
that you access by using numeric indexes, like lists, tuples, and strings.
Accessing a single item from one of these sequences is straightforward Thanks to tuple packing and unpacking you can also return more than one
– just use the index, or a negative index to count back from the end of the item a time. Items separated by commas are automatically ‘packed’ into a
sequences. E.g., my_list[-1] will return the last item in my_list, my_list[-2] will tuple and can be ‘unpacked’ on the receiving end:
return the second to last, and so on.
a, b, c = (1, 2, 3)
Notation Returns Examples - if x =
[0,1,2,3]
Expression will return

x[0] First element of a sequence 0

Classes

Defining classes
x[1] Second element of a sequence 1 You define a class with the class keyword:

x[-1] Last element of a sequence 3

class MyClass(object):
x[1:] Second element through last [1,2,3] def __init__(self, par):
element # initialize some stuff
self.foo = “bar”
x[:-1] First element up to (but NOT [0,1,2] def a_method(self):
including last element # do something
def another_method(self, parameter):
x[:] All elements - returns a copy of list [0,1,2,3] # do something with parameter

x[0::2] Start at first element, then every 2nd [0,2]

element Note: In Python 3.x, you can create classes without inheriting from “object”
because that’s the default. Also don’t write getters/setters up front, use
the @ property instead which lets you add them transparently later.

FUNCTIONS Instantiating classes

Classes are instantiated using the class name:
Function definitions
Functions are defined with the def keyword and parenthesis after the
function name:

DZone, Inc. | www.dzone.com

 Core Python

my_class_object = my_class()
The Zen of Python, by Tim Peters
When a class object is instantiated, the classe’s __init__(self) method is Beautiful is better than ugly.
called on the instance, usually doing any set up that is needed: initializing Explicit is better than implicit.
variables and the like.
Simple is better than complex.
If the class __init__( ) method accepts a parameter, it can be passed in: Complex is better than complicated.
Flat is better than nested.
my_class_object = my_class(param) Sparse is better than dense.
Readability counts.
Special cases aren’t special enough to break
Inheritance and mixins
Python supports multiple inheritance. This does provide you with more the rules.
ways to shoot yourself in the foot, but a common pattern for multiple Although practicality beats purity.
inheritance is to use “mixin” classes.
Errors should never pass silently.
Unless explicitly silenced.
Abstract Base Classes, Metaclasses
Abstract base classes are defined in PEP 3119. You can create abstract In the face of ambiguity, refuse the temptation
base classes via the abc module, which was added in Python 2.6. to guess.
There should be one -and preferably only
A metaclass is a class for creating classes. You can see examples of this in
Python built-ins, such as int, str or type. All of these are metaclasses. You one -obvious way to do it.
can create a class using a specific metaclass via __metaclass__. If that is Although that way may not be obvious at first
not specified, then type will be used.
unless you’re Dutch.
Comprehensions Now is better than never.
Python comes with a concept known as comprehensions. There are 3 Although never is often better than *right* now.
types: list comprehensions, dict comprehensions and set comprehensions. If the implementation is hard to explain, it’s a bad
Following is an example of a list comprehension:
idea.
If the implementation is easy to explain, it may be
new_list = [x for x in range(5)]
a good idea.
Namespaces are one honking great idea -- let’s do
This will create a list from 0-5. It is the equivalent of the following for loop: more of those!

new_list = []
for x in range(5):
new_list.append(x)
PEP-8 - the Python style guide
A dict comprehension is similar. It looks like this: Python has its own style guide known as PEP8 that outlines various
guidelines that are good to follow. In fact, you must follow them if you plan
new_dict = {key: str(key) for key in range(5)} to contribute to Python Core. PEP 8 specifies such things as indentation
amount, maximum line length, docstrings, whitespace, naming
conventions, etc.
A set comprehension will create a Python set, which means you will end
up with an unordered collection with no duplicates. The syntax for a set
comprehension is as follows:
USing the shell

new_set = {x for x in ‘mississippi’} Python's default shell

Python is one of several languages that has an interactive shell which is
a read-eval-print-loop (REPL). The shell can be enormously helpful for
experimenting with new libraries or unfamiliar features and for accessing
Style Tips documentation.

What does 'Pythonic' mean? BATTERIES INCLUDED: USING LIBRARIES

‘Pythonic’ is the term that Pythonistas use when they are talking about
code that uses the language well and the way that it’s creators intended.
Pythonic is a very good thing. Using Java-esque camel cased variable Importing and using modules and libraries
names is not Pythonic, but using it for class names is. Writing for loops Using external modules and libraries is as simple as using the import key-
in the style of C/C++ is considered un-Pythonic. On the other hand, using word at the top of your code.
Python data structures intelligently and following the Python style guide
makes your code Pythonic.
Import Explanation

The Zen of Python from lib import x Imports single element x from lib, no dot prefix
from lib import x as y needed
PEP(Python Enhancement Proposal)-20 is the Zen of Python. Written by
x()
long time Python developer Tim Peters, the Zen is acknowledged as the
y()
core philosophy of Python. In fact, it is always accessible in any Python
environment by using import this: import lib Imports all of lib, dot prefix needed

lib.x()

from lib import * Imports all of lib, no dot prefix needed "NOT FOR"
PRODUCTION CODE - POSSIBLE VARIABLE NAME
CLASHES!

DZone, Inc. | www.dzone.com

 Core Python

Of the three styles of import the second (import lib) has the advantage To install the packages, you can use pip or easy_install, both of which
that it is always clear what library an imported element comes from and you’ll need to download from PyPI. For full instructions on bootstrapping
the chances for namespace collision and pollution are low. If you are only with these tools, see http://www.pip-installer.org/en/latest/installing.html
using one or two components of a library the first style (from lib import Sometimes those utilities won’t work and you’ll have to use the package’s
x) makes typing the element name a bit easier. The last style (from lib included setup.py to do the installation, which normally goes something like
import *) is NOT for production code – namespace collisions are very this:
likely and you can break module reloading. There is one major exception to
this rule that you will see in many examples and that concerns the include python setup.py install
Tkinter GUI toolkit. Most Tkinter tutorials import it as follow: from Tkinter
import *. The reason is that Tkinter has been designed so that it is unlikely You will see a lot of information output to your screen when you execute
to cause namespace collisions. the above. In some cases, the module has C headers and will require a C/
C++ compiler installed on your machine to complete installation correctly.
Creating modules and libraries
POPULAR PYTHON LIBRARIES
def my_module(foo, bar):
print foo
print bar
if __name__ == “__main__”: numpy and scipy
my_module(1, 2) Numpy and scipy are extensive mathematical libraries written to make op-
erating on large data collections easier. As Python’s presence in scientific
communities has grown, so has the popularity of numpy and scipy. Cur-
Any Python script file can be treated like a module and imported. However
rently there are conferences devoted to them and to scientific computing.
be aware that when a module is imported, its code is executed – that’s the
For graphing, you might want to try matplotlib.
reason for the if __name__ == “__main__”: structure in the example above.
In other words, to be safely used as a module, a script should be organized
IPython - the shell and more
into functions (or classes), with the if statement at the very end.
The default Python shell has some annoying limitations – it’s inconvenient
Here is an example module:
to access the host operating system, there is no good way to save and
recover sessions, and it’s not easy to export the commands of a session
The Python standard library - selected library groups to an ordinary script file. This is particularly irksome for scientists and
Python comes with a standard library of modules that can do much of what researchers who may want to spend extensive time exploring their data
you need to get done. The standard library is quite extensive – it would take using an interactive shell.
weeks to become familiar with everything in it.
To address these issues IPython answers these and other problems.
Whenever you feel the need to go looking for an additional external library,
you should first look carefully in the standard library – more often than not,
a perfectly good implementation of what you need is already there.

Library Group Contains Libraries for

File and Directory File paths, tempfiles, file comparisons (see the os and
Access tempfile modules)
Numeric and Math Math, decimal, fractions, random numbers/sequences,
iterators (see math, decimal, and collections)
Data Types Math, decimal, fractions, random numbers/sequences,
iterators (see math, decimal, and collections)
Data Persistence Object serialization (pickle), sqlite, database access
File Formats Csv files, config files - see ConfigParser
Generic OS Operating system functions, time, command line
Services arguments, logging (see os, logging, time, argparse)
To get IPython, go to http://ipython.org/ and download the version best
Interprocess Communication with other processes, low-level suited to your operating system.
sockets (see subprocess and the socket module)
Interned Data Handling Internet data, including json, email and Web libraries
Handling mailboxes, mime encoding (see json, email, smtplib One of the main uses for Python these days is for web programming. There
and mimetools) are several popular web frameworks as described below, as well as other
Structured Markup Parsing HTML and XML (see xml.minidom and libraries for dealing with web content.
ElementTree)
Django
Internet Protocols HTTP, FTP, CGI, URL parsing, SMTP, POP, IMAP, Arguably the most popular web framework, django has taken the Python
Telnet, simple servers (see httplib, urllib, smtplib,
world by storm in the past few years. It has its own ORM, which makes it
imaplib)
very easy to interact with databases.
Development Documentation, test, Python 2 to Python 3 conversion
(see doctest and 2to3) Pyramid
Debugging Debugging, profiling (see pdb and profile) A Python framework originally based on Pylons, but is now a rebranding
of repoze.bfg. Pyramid supports single file applications, decorator-base
Runtime System parameters and settings, builtins, warnings, config, URL generation, etc.
contexts (see the dir command and the inspect
module) Flask
GUI Tkinter GUI libraries, turtle graphics Flask is also a micro web framework for Python, but it is based on
Werkzeug and Jinja2.

Getting other libraries Requests

If you find yourself needing additional functionality, you should go take a Requests is an HTTP library that provides a more Pythonic API to HTTP
look in the Python Package Index (PyPI). There you will find thousands of Requests. In other words, it makes it easier to download files and work with
packages that cover a vast array of topics. HTTP requests than the standard library.

DZone, Inc. | www.dzone.com

 Core Python

Beautifulsoup Learn Python the Hard Way - http://learnpythonthehardway.org/book/

A great HTML parser that allows the developer to nagivate, search and
modify a parse tree as well as dissecting and extracting data from a Python Anywhere - https://www.pythonanywhere.com/
web page.
Books
Other Libraries
Python has many other libraries ranging over all kinds of topics. Here is a Python 3 Object Oriented Programming by Dusty Phillips
sampling:
Python Cookbook (2nd Edition) (Python 2.x) by Alex Martelli, Anna
Twisted – Networking Ravenscroft, David Ascher

Natural Language Tool Kit (NLTK) - Language Processing Python Cookbook (3rd Edition) (Python 3.x) by David Beazley, Brian K.
Jones
Pygame - Games in Python
Python Standard Library by Example by Doug Hellmann
SQLAlchemy - Database Toolkit
Python in Practice by Mark Summerfield

Dive Into Python by Mark Pilgrim

RESOURCES

Python Documentation

Python 3 - http://docs.python.org/3/

Python 2 - http://docs.python.org/2.7/

Tutorials

Official - http://docs.python.org/2/tutorial/

Learn Python - http://www.learnpython.org/

A B O U T the A uthor R ecommended B ook

Naomi Ceder has been Mike Driscoll has been The Quick Python Guide introduces Python’s
involved in teaching programming in Python syntax, control flow, and basic data structures, then
and promoting Python since 2006. He enjoys
walks through creating, testing, and deploying full
for over a decade. She writing about Python
applications and larger code libraries. Includes survey
has taught Python on his blog at www.
blog.pythonlibrary.org/, of GUI programming, testing, database access, and
to everyone from 6th
graders to adults and is occasionally writes for web frameworks.
a member of the Python the Python Software
Software Foundation and started both the Foundation. Mike has also been a
poster session and the education summit technical reviewer for Packt Publishing’s
Python books, such as Python 3 Object
BUY NOW!
at PyCon. She is also the author of The
Quick Python Book, 2nd ed. from Manning Oriented Programming, and Python 2.6
Publishers. Graphics Cookbook and more.

Browse our collection of over 150 Free Cheat Sheets

Upcoming Refcardz

Free PDF
Ruby on Rails
Regex
Clean Code
HTML5 IndexedDB

DZone, Inc.
150 Preston Executive Dr.
Suite 201
Cary, NC 27513
DZone communities deliver over 6 million pages each month to 888.678.0399
more than 3.3 million software developers, architects and decision 919.678.0300
makers. DZone offers something for everyone, including news,
Refcardz Feedback Welcome
$7.95

tutorials, cheat sheets, blogs, feature articles, source code and more.
"DZone is a developer's dream.", says PC Magazine. [email protected]

Copyright © 2013 DZone, Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval Sponsorship Opportunities
system, or transmitted, in any form or by means electronic, mechanical, photocopying, or otherwise, without prior
written permission of the publisher. [email protected]