director of PythonLabs at Zope Corporation Guido van Rossum director of PythonLabs at Zope Corporation mailto:guido@python.org http://www.python.org
What’s in a name? Snake logos and mascot notwithstanding, it’s named after Monty Python’s Flying Circus Humor-impaired can safely ignore the spam references :-) Nobody expects the Spanish inquisition
What is Python? O-O rapid prototyping language Not just a scripting language Not just another Perl Easy to learn, read, use Extensible (add new modules) C/C++/Fortran/whatever Java (through Jython) Embeddable in applications
Touchy-feely properties Open Source (OSI Certified) copyrighted but use not restricted no "viral" license owned by independent non-profit, PSF Mature (13 years old) Supportive user community plenty of good books, too Simple design, easy to learn reads like “pseudo-code” Suitable as first language Suitable as last language :-)
High-level properties Extremely portable Unix/Linux, Windows, Mac, PalmOS, WindowsCE, RiscOS, VxWorks, QNX, OS/2, OS/390, AS/400, PlayStation, Sharp Zaurus, BeOS, VMS… Compiles to interpreted byte code compilation is implicit and automatic Memory management automatic reference counting for most situations GC added for cycle detection “Safe”: no core dumps due to your bugs
What is it used for? rapid prototyping web programming (client and server side) ad hoc programming ("scripting") steering scientific applications extension language XML processing database applications GUI applications education
Who is using it? Google (various projects) NASA (several projects) NYSE (one of only three languages "on the floor") Industrial Light & Magic (everything) Yahoo! (Yahoo mail & groups) RealNetworks (function and load testing) RedHat (Linux installation tools) LLNL, Fermilab (steering scientific applications) Zope Corporation (content management) ObjectDomain (embedded Jython in UML tool) Alice project at CMU (accessible 3D graphics) More success stories at www.pythonology.com
Language properties Everything is an object Packages, modules, classes, functions Exception handling Dynamic typing, polymorphism Static scoping Operator overloading Indentation for block structure Otherwise conventional syntax
High-level data types Numbers: int, long, float, complex Strings, Unicode: immutable Lists and dictionaries: containers Other types for e.g. binary data, regular expressions, introspection Extension modules can define new “built-in” data types
Interfaces to... XML Relational databases Java (via Jython) DOM, expat XMLRPC, SOAP, Web Services Relational databases MySQL, PostgreSQL, Oracle , ODBC, Sybase, Informix Java (via Jython) Objective C COM, DCOM (.NET too) Many GUI libraries cross-platform Tk, wxWindows, GTK, Qt platform-specific MFC, Mac (classic, Cocoa), X11
Compared to Perl Easier to learn More readable code very important for infrequent users More readable code More maintainable code Fewer “magical” side effects More “safety” guarantees Better Java integration
Compared to Java Code up to 5 times shorter Dynamic typing and more readable Dynamic typing Multiple inheritance, operator overloading Quicker development no compilation phase less typing Yes, it may run a bit slower but development is much faster and Python uses less memory (studies show) Similar (but more so) for C/C++
Jython Seamless integration with Java Separate implementation Implements the same language Different set of standard modules differences in “gray areas” e.g. some different introspection calls different command line options, etc.
Jython's Java integration Interactive Compiles directly to Java bytecode Import Java classes directly Subclass Java classes pass instances back to Java Java beans integration Can compile into Java class files
Example function def gcd(a, b): "greatest common divisor" while a != 0: a, b = b%a, a # parallel assignment return b
Example class class Stack: "A well-known data structure" # doc string def __init__(self): # constructor self.items = [] def push(self, x): self.items.append(x) # the sky is the limit def pop(self): x = self.items[-1] # what happens if it’s empty? del self.items[-1] return x def empty(self): return len(self.items) == 0
References and plugs References: www.python.org - Python home site documentation, downloads, community, PSF www.pythonology.org - success stories www.artima.com/intv - interview with GvR www.zope.org - Zope community site www.zope.com - Zope corporate site Python Conferences (see www.python.org): PyCon DC March 26-28 Washington, DC on-line registration ends today ($200) Python UK April 2-3 Oxford, England EuroPython June 25-27 Charleroi, Belgium Python11 at OSCON July 7-11 Portland, OR