Python crash course

2010-01-15 v0.07
Philippe Lagadec
http://www.decalage.info

This document published on the site www.decalage.info is licensed by Philippe Lagadec under a Creative Commons
Attribution-Noncommercial-Share Alike 3.0 Unported License.
Introduction
 “Python is an easy to learn, powerful
programming language. It has efficient high-
level data structures and a simple but effective
approach to object-oriented programming.
Python's elegant syntax and dynamic typing,
together with its interpreted nature, make it an
ideal language for scripting and rapid
application development in many areas on
most platforms.”
 The Python Tutorial, http://docs.python.org/tutorial/
Why “Python” ?
 “Guido van Rossum, the creator of the
Python language, named the language
after the BBC show "Monty Python's
Flying Circus". He doesn't particularly
like snakes that kill animals for food by
winding their long bodies around them
and crushing them.” 
Swaroop C H.
Download a Python interpreter
 Python official website:
http://www.python.org
 Download:
http://www.python.org/download
 Warning: for now (Jan 2010), Python 2.6
or 2.5 are still recommended.
 Many third-party modules have issues with
Python 3.x, which is very different from v2.x.
Install a Python-aware editor
 On Windows, I recommend PyScripter to
get an effective IDE with syntax
highlighting, code completion, integrated
Python shell and debugger: http://mmm-
experts.com/Products.aspx?ProductId=4
or http://code.google.com/p/pyscripter/ for
latest versions.
 On other OSes such as Linux or
MacOSX, try Eclipse + PyDev.
Python Documentation
 On Windows, the manual provided with the
interpreter (Start menu / All Programs / Python /
Python Manuals) is usually the most convenient
way to access the Python documentation.
 Use the index tab !
 Online official documentation startpoint:
http://www.python.org/doc
 (pretty long) official tutorial:
http://docs.python.org/tut/tut.html
Python shell
 On Windows, use Start Menu / Python / Python
command line.
 Alternatively you may run "python" from a CMD
window.
 but since python.exe is not in the PATH environment
variable by default you may want to add it, or type its
complete path such as "C:\python25\python.exe".
 Quit with Ctrl+Z or simply close the window
when finished.
Python basics
Variables and Constants
 Variables are simply names which point to any
value or object:
a_string = "hello, world"
an_integer = 12
a_float = 3.14
a_boolean = True
nothing = None
 Dynamic typing: the value or type of a variable
may be changed at any time.
Print
 To print a constant or variable:
print "hello, world"
print a_string
print 12
print (5+3)/2
 To print several items, use commas
(items will be separated by spaces):
print "abc", 12, a_float
Long lines
 A long statement may be split using a
backslash:
my_very_long_variable = something + \
something_else

 It is not necessary if there are

parentheses or square brackets:
average_value = (some_value1 + some_value2 +
some_value3) / 3
Strings
 There are 3 syntaxes for string constants:
string_single_quotes = 'abc'

string_double_quotes = "abc"

string_triple_quotes = """this is
a multiline
string."""
Strings
 It's useful when we need to include
quotes in a string:
string1 = 'hello "world"'
string2 = "don't"
 otherwise we have to use backslashes:
string2 = 'don\'t'
 Be careful about backslashes in paths:
Win_path = 'C:\\Windows\\System32'
String operations and methods
 Strings are objects which support many operations:
strings = string1 + " : " + string2
 Get the length of a string:
len(strings)
 Convert to uppercase:
strings_uppercase = strings.upper()
 Strip spaces at beginning and end of a string:
stripped = a_string.strip()
 Replace a substring inside a string:
newstring = a_string.replace('abc', 'def')
 All string methods: http://docs.python.org/lib/string-methods.html
 Note: a string is immutable, all operations create a new string in
memory.
Conversions
 Convert a string to an integer and vice-
versa:
i = 12
s = str(i)

s = '17'
i = int(s)
Building strings
 Several ways to include a variable into a string:
 by concatenation:
string1 = 'the value is ' + str(an_integer) + '.'
 by "printf-like" formatting:
string2 = 'the value is %d.' % an_integer
 With several variables, we need to use
parentheses:
a = 17
b = 3
string3 = '%d + %d = %d' % (a, b, a+b)
Building strings
 To include strings into another string:
stringa = '17'
stringb = '3'
stringc = 'a = ' + stringa + ', b = ' + stringb
stringd = 'a = %s, b= %s' % (stringa, stringb)
 Everything about string formatting:
http://docs.python.org/library/stdtypes.html#strin
g-formatting-operations
Lists
 A list is a dynamic array of any objects.
It is declared with square brackets:
mylist = [1, 2, 3, 'abc', 'def']

 Access a specific element by index (index

starts at zero):
third_item = mylist[2]
List operations
 Operations on a list:
mylist[2] = „new value‟
mylist.append(„another value‟)
mylist.remove(„abc‟)
length = len(mylist)
longlist = mylist + [4,5,6]
 All list operations:
http://docs.python.org/lib/typesseq.html
Slicing
 Slicing is extracting a sublist from a list:
first_two_items = mylist[0:2]
third_and_fourth = mylist[2:4]
fourth_up_to_end = mylist[3:]
last_two_items = mylist[-2:]
first_two_items2 = mylist[:2]
More complex lists
 Lists may contain lists:
mylist2 = [mylist, 'abc', mylist,
[1, 2, 3]]
 It works like a two-dimensions array:
item1 = mylist2[3][2]
item2 = mylist2[0][4]
Tuples
 A tuple is similar to a list but it is a fixed-size,
immutable array: once a tuple has been
created, its elements may not be changed,
removed, appended or inserted.
 It is declared using parentheses and comma-
separated values:
a_tuple = (1, 2, 3, 'abc', 'def')
 But parentheses are optional:
a_tuple = 1, 2, 3, 'abc', 'def'
 Tuples may be seen as “complex constants”.
Dictionaries
 A dictionary is a mapping between
indexes to values.
 Indexes may be almost any type:
integers, strings, tuples, objects…
 Example:
countries = {„us‟:„USA‟, „fr‟:‟France‟,
„uk‟:‟United Kingdom‟}
print countries[„uk‟]
countries[„de‟] = „Germany‟
Dictionary operations
 Get a list of all indexes:
country_codes = countries.keys()
 Get a list of (index, value) tuples:
Countries_list = countries.items()
 Test if a specific index is there:
is_uk_there = „uk‟ in countries
 More info: http://docs.python.org/lib/typesmapping.html
 And
http://docs.python.org/tutorial/datastructures.html#dictio
naries
Blocks and Indentation
(control flow)
 Blocks of code are delimited using indentation,
either spaces or tabs at the beginning of lines.
 This is one of the main differences of Python over
other languages, and usually the main reason why
people love it or hate it. ;-)

 Tip: NEVER mix tabs and spaces in a script, it

may result in tricky bugs.
 From my experience, the safest solution is to always
use 4-spaces indents, never tabs. (because each
editor may convert tabs either to 2, 4 or 8 spaces)
if / else
if a == 3:
print 'The value of a is:'
print 'a=3'

if a != 'test':
print 'a is not "test"'
test_mode = False
else:
print 'The value of a is:'
print 'a="test"'
test_mode = True
if / elif / else
if choice == 1:
print “First choice.”
elif choice == 2:
print “Second choice.”
else:
print “Wrong choice.”
While loop
a=1
while a<10:
print a
a += 1
For loop
for a in range(10):
print a

my_list = [2, 4, 8, 16, 32]

for a in my_list:
print a
Sorting
 Use sorted() to get a sorted version of a
list/dict or any other iterable.
 Example:
my_list = [32, 4, 16, 8, 2]
for a in sorted(my_list):
print a
 See help for more advanced sorting:
http://docs.python.org/library/functions.ht
ml?highlight=sorted#sorted
Functions
 See tut1a_functions.py
 And tut1b_default_args.py
Exceptions
 TODO…

 Any error raises an exception:

def average (items):
return sum(items) / len(items)
print average([1, 3, 7])
print average([])
Classes and Objects
Object-oriented
programming
Classes and objects
 With Python it is possible to use either
procedural or object-oriented
programming, but most applications mix
both.
 You may use objects to model complex
data structures in an organized, self-
contained way.
A simple class - Attributes
 Attributes are defined in the constructor
method called __init__().
 “self” is a variable pointing to the object
itself.
 Each attribute is stored in self.attribute.
class Person:
def __init__(self, lastname, firstname, age):
self.lname = lastname
self.fname = firstname
self.age = age
john = Person(„Doe', „John„, 45)
print john.fname, “is”, john.age
john.age = 46
Methods
 Like a function defined inside the class.
 “self” must always be the first argument of
a method. It represents the object itself.
class Person:
def __init__(self, lastname, firstname, age):
self.lname = lastname
self.fname = firstname
self.age = age
def get_fullname(self):
return '%s %s' % (self.fname, self.lname)
def add_years(self, years):
self.age += years
john = Person(„Doe', „John„, 45)
print john.get_fullname()
john.add_years(2)
Common error: missing self
 When adding a new method, it‟s a common
mistake to forget “self”.
class Person:
[…]
def display():
print self.get_fullname()
john = Person(„Doe', „John')
john.display()

 Here is the cryptic error message you get in this

case:
 TypeError: display() takes no arguments
(1 given)
 Self is a hidden argument, hence “1 given”.
Attributes are flexible
 Contrary to other languages like Java or C++,
attributes are not protected or private.
 Attributes may be modified from anywhere
outside of the class.
 Attributes may even be created or deleted by
any method, or from outside.
 Simple naming convention: an attribute starting
with underscore is expected to be private. But
it‟s only a convention.
Initializing list and dict attributes
 Common pitfall: never use an empty list or
dict as default value for an attribute.
 All objects would then a get a pointer to
the same list/dict in memory.
 Tip: use “None” instead:
class Person:
def __init__(self, lastname, firstname, age, children=None):
self.lname = lastname
self.fname = firstname
self.age = age
self.children = children
if children == None:
self.children = []
The special __str__() method
 Used to return a custom string
representation of an object.
 Will be called when Python needs to
convert an object to a string, such as:
 print my_object
 s = str(my_object)
 s = “the object is %s.” % my_object
 See tut3a_class.py
Inheritance
 See tut3b_class_inherit.py
Class attributes
 TODO
Class methods
 TODO
Standard Library
Overview: a few useful
modules
Tip
 The CHM Python manual installed with
the Python interpreter is very handy:
 Start Menu / All Programs / Python / Python
Manuals

 Use the index tab to quickly find anything

Script arguments
 sys.argv is a list containing all arguments
that were specified when launching a
script.

import sys
print “all arguments:”, sys.argv
print “number of args:”, len(sys.argv)
print “first arg:”, sys.argv[1]
Launching a process/command
 Simplest way:
import os
os.system(“net accounts”)
 More effective and flexible:
 See subprocess module
• Capture output
• Control keyboard input
• Execution in parallel
Exercises
 1) Write a script which computes the
SHA1 hash of your name.

 2) Compute the SHA1 hash of a file given

as argument.
Solution
 TODO…
Network protocols
 Example 1: download a web page (HTTP
client)
 See urllib2

 Example 2: send an e-mail (SMTP client)

 See smtplib
Simple XML parsing
using ElementTree
XML parsers
 There are several XML parsers for
Python, usually with DOM or SAX API.
 Quite complex
 ElementTree provides a simpler, more
pythonic API to handle XML
 Included in Python standard library since
v2.5
ElementTree: base concepts
 An ElementTree object is a representation in
memory of a complete XML file (tree with a
root)
 An Element object is a single XML node. It has:
 A tag name, with optional namespace
 Attributes (optional) with values
 Text (optional)
 Children (optional sub-elements)
 Tail (optional)
 Element / ElementTree example
ElementTree object
Root Element, tag=„FOO‟
Element, child of root
<FOO
xmlns:ns="http://namespace.org/" Attribute (name+value)
ns:version=“2.0">
<ns:Body> Text
<BAR name=“value”>
<BAZ>hello</BAZ>
</BAR>
<BAR name=“value2”/>
</ns:Body> blabla Tail
</FOO>
 First, import the “ET” module
 Tip to support all Python versions:
try:
# ElementTree in Python 2.5
import xml.etree.ElementTree as ET
except:
try:
# external ElemenTree for Python <=2.4
import elementtree.ElementTree as ET
except:
raise ImportError, \
"the ElementTree module is not installed: "+\
"see http://effbot.org/zone/element-index.htm"
Parsing an XML file
 ET.parse returns an ElementTree object:
 tree = ET.parse(„myfile.xml‟)
 To obtain the root Element, use getroot:
 elem = tree.getroot()
Parsing a string containing XML
 Use ET.fromstring:
xmlstring = “<FOO>…</FOO>”
root = ET.fromstring(xmlstring)
Element data
 Tag name:
 print elem.tag
 Text and tail:
 print elem.text
 print elem.tail
 Attributes and values (dictionary):
 print elem.attrib
 List of children (sub-elements):
 print list(elem)
Element Attributes
 Iterating over the attributes (dictionary):
for name in elem.attrib:
value = elem.attrib[name]
print „%s=“%s”‟ % (name, value)
 Or simply:
for name, value in elem.attrib.items():
print „%s=“%s”‟ % (name, value)
Finding a tag
 Find the first child with a given tag:
elem.find(“FOO”)
 Obtain a list of children with that tag:
elem.findall(“FOO”)
 Look for a tag in the whole sub-tree:
elem.getiterator(“FOO”)
 More info:
http://effbot.org/zone/element.htm#searching-
for-subelements
Finding a tag by its path
 Knowing all tags on the path:
elem.find(“FOO/BAR/BAZ”)
 To look for that path anywhere:
elem.find(“.//FOO/BAR/BAZ”)
 Paths may contain “*”
elem.find(“FOO/*/BAZ”)
 More info: http://effbot.org/zone/element-
xpath.htm
Namespaces
 Modern XML formats use more and more
namespaces.
 When XML data contains namespaces, each
tag name MUST contain its full namespace
URL in ElementTree as “{namespace URL}tag”.
 Sample XML data:
<soap:Envelope
xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/">…

 Corresponding ElementTree code:

envelope = tree.find(
"{http://schemas.xmlsoap.org/soap/envelope/}Envelope")
Namespaces
 Tip: to ease development and clarify
code, prepare all useful tags in global
variables before using them.
NS_SOAP="http://schemas.xmlsoap.org/soap/envelope/“
ENVELOPE = “{%s}Envelope” % NS_SOAP
BODY = “{%s}Body” % NS_SOAP

root = ET.fromstring(xml_soap_message)
elem_env = root.find(ENVELOPE)
elem_body = elem_env.find(BODY)
More about ElementTree
 More complete tutorials about parsing,
generating and editing XML with
ElementTree:
 http://effbot.org/zone/element-index.htm
 http://effbot.org/zone/element.htm

 http://www.decalage.info/en/python/etree
Simple web applications
Using CherryPy
Simplest web application
import cherrypy

class HelloWorld:
def index(self):
return "Hello world!“
index.exposed = True

cherrypy.quickstart(HelloWorld())
Two pages
import cherrypy

class TwoPages:
def index(self):
return „<a href=“page2”>Go to page 2</a>‟
index.exposed = True

def page2(self):
return “This is page 2!“
page2.exposed = True

cherrypy.quickstart(TwoPages())
CherryPy tutorial
 Tutorial:
 http://www.cherrypy.org/wiki/CherryPyTutorial

 Sample scripts
 See also local samples in your Python directory,
such as:
• C:\Python25\Lib\site-packages\cherrypy\tutorial\

 Documentation:
 http://www.cherrypy.org/wiki/TableOfContents