path: root/Doc/whatsnew/whatsnew24.tex

\documentclass{howto}
\usepackage{distutils}
% $Id$

% Don't write extensive text for new sections; I'll do that.  
% Feel free to add commented-out reminders of things that need
% to be covered.  --amk

% XXX pydoc can display links to module docs -- but when?
% 

\title{What's New in Python 2.4}
\release{0.3}
\author{A.M.\ Kuchling}
\authoraddress{
	\strong{Python Software Foundation}\\
	Email: \email{amk@amk.ca}
}

\begin{document}
\maketitle
\tableofcontents

This article explains the new features in Python 2.4 alpha3, scheduled
for release in early September.  The final version of Python 2.4 is
expected to be released around December 2004.

Python 2.4 is a medium-sized release.  It doesn't introduce as many
changes as the radical Python 2.2, but introduces more features than
the conservative 2.3 release did.  The most significant new language
features (as of this writing) are function decorators and generator
expressions; most other changes are to the standard library.

This article doesn't attempt to provide a complete specification of
every single new feature, but instead provides a convenient overview.
For full details, you should refer to the documentation for Python
2.4, such as the \citetitle[../lib/lib.html]{Python Library Reference}
and the \citetitle[../ref/ref.html]{Python Reference Manual}.  If you
want to understand the complete implementation and design rationale,
refer to the PEP for a particular new feature or to the module
documentation.


%======================================================================
\section{PEP 218: Built-In Set Objects}

Python 2.3 introduced the \module{sets} module.  C implementations of
set data types have now been added to the Python core as two new
built-in types, \function{set(\var{iterable})} and
\function{frozenset(\var{iterable})}.  They provide high speed
operations for membership testing, for eliminating duplicates from
sequences, and for mathematical operations like unions, intersections,
differences, and symmetric differences.

\begin{verbatim}
>>> a = set('abracadabra')              # form a set from a string
>>> 'z' in a                            # fast membership testing
False
>>> a                                   # unique letters in a
set(['a', 'r', 'b', 'c', 'd'])
>>> ''.join(a)                          # convert back into a string
'arbcd'

>>> b = set('alacazam')                 # form a second set
>>> a - b                               # letters in a but not in b
set(['r', 'd', 'b'])
>>> a | b                               # letters in either a or b
set(['a', 'c', 'r', 'd', 'b', 'm', 'z', 'l'])
>>> a & b                               # letters in both a and b
set(['a', 'c'])
>>> a ^ b                               # letters in a or b but not both
set(['r', 'd', 'b', 'm', 'z', 'l'])

>>> a.add('z')                          # add a new element
>>> a.update('wxy')                     # add multiple new elements
>>> a
set(['a', 'c', 'b', 'd', 'r', 'w', 'y', 'x', 'z'])       
>>> a.remove('x')                       # take one element out
>>> a
set(['a', 'c', 'b', 'd', 'r', 'w', 'y', 'z'])       
\end{verbatim}

The \function{frozenset} type is an immutable version of \function{set}.
Since it is immutable and hashable, it may be used as a dictionary key or
as a member of another set.  

The \module{sets} module remains in the standard library, and may be
useful if you wish to subclass the \class{Set} or \class{ImmutableSet}
classes.  There are currently no plans to deprecate the module.

\begin{seealso}
\seepep{218}{Adding a Built-In Set Object Type}{Originally proposed by
Greg Wilson and ultimately implemented by Raymond Hettinger.}
\end{seealso}

%======================================================================
\section{PEP 237: Unifying Long Integers and Integers}

The lengthy transition process for this PEP, begun in Python 2.2,
takes another step forward in Python 2.4.  In 2.3, certain integer
operations that would behave differently after int/long unification
triggered \exception{FutureWarning} warnings and returned values
limited to 32 or 64 bits (depending on your platform).  In 2.4, these
expressions no longer produce a warning and instead produce a
different result that's usually a long integer.

The problematic expressions are primarily left shifts and lengthy
hexadecimal and octal constants.  For example,
\code{2 \textless{}\textless{} 32} results
in a warning in 2.3, evaluating to 0 on 32-bit platforms.  In Python
2.4, this expression now returns the correct answer, 8589934592.

\begin{seealso}
\seepep{237}{Unifying Long Integers and Integers}{Original PEP
written by Moshe Zadka and GvR.  The changes for 2.4 were implemented by 
Kalle Svensson.}
\end{seealso}

%======================================================================
\section{PEP 289: Generator Expressions}

The iterator feature introduced in Python 2.2 and the
\module{itertools} module make it easier to write programs that loop
through large data sets without having the entire data set in memory
at one time.  List comprehensions don't fit into this picture very
well because they produce a Python list object containing all of the
items, unavoidably pulling them all into memory.  When trying to write
a functionally-styled program, it would be natural to write something
like:

\begin{verbatim}
links = [link for link in get_all_links() if not link.followed]
for link in links:
    ...
\end{verbatim}

instead of 

\begin{verbatim}
for link in get_all_links():
    if link.followed:
        continue
    ...
\end{verbatim}

The first form is more concise and perhaps more readable, but if
you're dealing with a large number of link objects the second form
would have to be used to avoid having all link objects in memory at
the same time.

Generator expressions work similarly to list comprehensions but don't
materialize the entire list; instead they create a generator that will
return elements one by one.  The above example could be written as:

\begin{verbatim}
links = (link for link in get_all_links() if not link.followed)
for link in links:
    ...
\end{verbatim}

Generator expressions always have to be written inside parentheses, as
in the above example.  The parentheses signalling a function call also
count, so if you want to create a iterator that will be immediately
passed to a function you could write:

\begin{verbatim}
print sum(obj.count for obj in list_all_objects())
\end{verbatim}

Generator expressions differ from list comprehensions in various small
ways.  Most notably, the loop variable (\var{obj} in the above
example) is not accessible outside of the generator expression.  List
comprehensions leave the variable assigned to its last value; future
versions of Python will change this, making list comprehensions match
generator expressions in this respect.

\begin{seealso}
\seepep{289}{Generator Expressions}{Proposed by Raymond Hettinger and
implemented by Jiwon Seo with early efforts steered by Hye-Shik Chang.}
\end{seealso}


%======================================================================
\section{PEP 292: Simpler String Substitutions}

Some new classes in the standard library provide a
alternative mechanism for substituting variables into strings that's 
better-suited for applications where untrained users need to edit templates.

The usual way of substituting variables by name is the \code{\%}
operator:

\begin{verbatim}
>>> '%(page)i: %(title)s' % {'page':2, 'title': 'The Best of Times'}
'2: The Best of Times'
\end{verbatim}

When writing the template string, it can be easy to forget the
\samp{i} or \samp{s} after the closing parenthesis.  This isn't a big
problem if the template is in a Python module, because you run the
code, get an ``Unsupported format character'' \exception{ValueError},
and fix the problem.  However, consider an application such as Mailman
where template strings or translations are being edited by users who
aren't aware of the Python language; the syntax is complicated to
explain to such users, and if they make a mistake, it's difficult to
provide helpful feedback to them.

PEP 292 adds a \class{Template} class to the \module{string} module
that uses \samp{\$} to indicate a substitution.  \class{Template} is a
subclass of the built-in Unicode type, so the result is always a
Unicode string:

\begin{verbatim}
>>> import string
>>> t = string.Template('$page: $title')
>>> t % {'page':2, 'title': 'The Best of Times'}
u'2: The Best of Times'
>>> t2 % {'cost':42.50, 'action':'polish'}
u'$ 42.5: polishing'
\end{verbatim}

% $ Terminate $-mode for Emacs

If a key is missing from the dictionary, the \class{Template} class 
will raise a \exception{KeyError}.  There's also a \class{SafeTemplate}
class that ignores missing keys:

\begin{verbatim}
>>> t = string.SafeTemplate('$page: $title')
>>> t % {'page':3}
u'3: $title'
\end{verbatim}

Because templates are Unicode strings, you can use a template with the
\module{gettext} module to look up translated versions of a message.

\begin{seealso}
\seepep{292}{Simpler String Substitutions}{Written and implemented 
by Barry Warsaw.}
\end{seealso}


%======================================================================
\section{PEP 318: Decorators for Functions, Methods and Classes}

Python 2.2 extended Python's object model by adding static methods and
class methods, but it didn't extend Python's syntax to provide any new
way of defining static or class methods.  Instead, you had to write a
\keyword{def} statement in the usual way, and pass the resulting
method to a \function{staticmethod()} or \function{classmethod()}
function that would wrap up the function as a method of the new type.
Your code would look like this:

\begin{verbatim}
class C:
   def meth (cls):
       ...
   
   meth = classmethod(meth)   # Rebind name to wrapped-up class method
\end{verbatim}

If the method was very long, it would be easy to miss or forget the
\function{classmethod()} invocation after the function body.  

The intention was always to add some syntax to make such definitions
more readable, but at the time of 2.2's release a good syntax was not
obvious.  Years later, when Python 2.4 is coming out, a good syntax
\emph{still} isn't obvious but users are asking for easier access to
the feature, so a new syntactic feature has been added.

The feature is called ``function decorators''.  The name comes from
the idea that \function{classmethod}, \function{staticmethod}, and
friends are storing additional information on a function object; they're
\emph{decorating} functions with more details.  

The notation borrows from Java and uses the \character{@} character as an
indicator.  Using the new syntax, the example above would be written:

\begin{verbatim}
class C:

   @classmethod
   def meth (cls):
       ...
   
\end{verbatim}

The \code{@classmethod} is shorthand for the
\code{meth=classmethod(meth)} assignment.  More generally, if you have
the following:

\begin{verbatim}
@A @B @C
def f ():
    ...
\end{verbatim}

It's equivalent to:

\begin{verbatim}
def f(): ...
f = C(B(A(f)))
\end{verbatim}

Decorators must come on the line before a function definition, and
can't be on the same line, meaning that \code{@A def f(): ...} is
illegal.  You can only decorate function definitions, either at the
module-level or inside a class; you can't decorate class definitions.

A decorator is just a function that takes the function to be decorated
as an argument and returns either the same function or some new
callable thing.  It's easy to write your own decorators.  The
following simple example just sets an attribute on the function
object:

\begin{verbatim}
>>> def deco(func):
...    func.attr = 'decorated'
...    return func
...
>>> @deco
... def f(): pass
...
>>> f
<function f at 0x402ef0d4>
>>> f.attr
'decorated'
>>>
\end{verbatim}

As a slightly more realistic example, the following decorator checks
that the supplied argument is an integer:

\begin{verbatim}
def require_int (func):
    def wrapper (arg):
        assert isinstance(arg, int)
        return func(arg)

    return wrapper

@require_int
def p1 (arg):
    print arg

@require_int
def p2(arg):
    print arg*2
\end{verbatim}

An example in \pep{318} contains a fancier version of this idea that
lets you specify the required type and check the returned type as
well.

Decorator functions can take arguments.  If arguments are supplied,
the decorator function is called with only those arguments and must
return a new decorator function; this new function must take a single
function and return a function, as previously described.  In other
words, \code{@A @B @C(args)} becomes:

\begin{verbatim}
def f(): ...
_deco = C(args)
f = _deco(B(A(f)))
\end{verbatim}

Getting this right can be slightly brain-bending, but it's not too
difficult.

A small related change makes the \member{func_name} attribute of
functions writable.  This attribute is used to display function names
in tracebacks, so decorators should change the name of any new
function that's constructed and returned.

The new syntax was provisionally added in 2.4alpha2, and is subject to
change during the 2.4alpha release cycle depending on the Python
community's reaction.  Post-2.4 versions of Python will preserve
compatibility with whatever syntax is used in 2.4final.

\begin{seealso}
\seepep{318}{Decorators for Functions, Methods and Classes}{Written 
by Kevin D. Smith, Jim Jewett, and Skip Montanaro.  Several people
wrote patches implementing function decorators, but the one that was
actually checked in was patch \#979728, written by Mark Russell.}
\end{seealso}

%======================================================================
\section{PEP 322: Reverse Iteration}

A new built-in function, \function{reversed(\var{seq})}, takes a sequence
and returns an iterator that loops over the elements of the sequence 
in reverse order.  

\begin{verbatim}
>>> for i in reversed(xrange(1,4)):
...    print i
... 
3
2
1
\end{verbatim}

Compared to extended slicing, such as \code{range(1,4)[::-1]},
\function{reversed()} is easier to read, runs faster, and uses
substantially less memory.

Note that \function{reversed()} only accepts sequences, not arbitrary
iterators.  If you want to reverse an iterator, first convert it to 
a list with \function{list()}.

\begin{verbatim}
>>> input= open('/etc/passwd', 'r')
>>> for line in reversed(list(input)):
...   print line
... 
root:*:0:0:System Administrator:/var/root:/bin/tcsh
  ...
\end{verbatim}

\begin{seealso}
\seepep{322}{Reverse Iteration}{Written and implemented by Raymond Hettinger.}

\end{seealso}


%======================================================================
\section{PEP 327: Decimal Data Type}

Python has always supported floating-point (FP) numbers as a data
type, based on the underlying C \ctype{double} type.  However, while
most programming languages provide a floating-point type, most people
(even programmers) are unaware that computing with floating-point
numbers entails certain unavoidable inaccuracies.  The new decimal
type provides a way to avoid these inaccuracies.

\subsection{Why is Decimal needed?}

The limitations arise from the representation used for floating-point numbers.
FP numbers are made up of three components:

\begin{itemize}
\item The sign, which is positive or negative.
\item The mantissa, which is a single-digit binary number  
followed by a fractional part.  For example, \code{1.01} in base-2 notation
is \code{1 + 0/2 + 1/4}, or 1.25 in decimal notation.
\item The exponent, which tells where the decimal point is located in the number represented.  
\end{itemize}

For example, the number 1.25 has positive sign, a mantissa value of
1.01 (in binary), and an exponent of 0 (the decimal point doesn't need
to be shifted).  The number 5 has the same sign and mantissa, but the
exponent is 2 because the mantissa is multiplied by 4 (2 to the power
of the exponent 2).

Modern systems usually provide floating-point support that conforms to
a relevant standard called IEEE 754.  C's \ctype{double} type is
usually implemented as a 64-bit IEEE 754 number, which uses 52 bits of
space for the mantissa.  This means that numbers can only be specified
to 52 bits of precision.  If you're trying to represent numbers whose
expansion repeats endlessly, the expansion is cut off after 52 bits.
Unfortunately, most software needs to produce output in base 10, and
base 10 often gives rise to such repeating decimals in the binary
expansion.  For example, 1.1 decimal is binary \code{1.0001100110011
...}; .1 = 1/16 + 1/32 + 1/256 plus an infinite number of additional
terms.  IEEE 754 has to chop off that infinitely repeated decimal
after 52 digits, so the representation is slightly inaccurate.

Sometimes you can see this inaccuracy when the number is printed:
\begin{verbatim}
>>> 1.1
1.1000000000000001
\end{verbatim}

The inaccuracy isn't always visible when you print the number because
the FP-to-decimal-string conversion is provided by the C library, and
most C libraries try to produce sensible output.  Even if it's not
displayed, however, the inaccuracy is still there and subsequent
operations can magnify the error.

For many applications this doesn't matter.  If I'm plotting points and
displaying them on my monitor, the difference between 1.1 and
1.1000000000000001 is too small to be visible.  Reports often limit
output to a certain number of decimal places, and if you round the
number to two or three or even eight decimal places, the error is
never apparent.  However, for applications where it does matter, 
it's a lot of work to implement your own custom arithmetic routines.

Hence, the \class{Decimal} type was created.

\subsection{The \class{Decimal} type}

A new module, \module{decimal}, was added to Python's standard library.
It contains two classes, \class{Decimal} and \class{Context}.
\class{Decimal} instances represent numbers, and
\class{Context} instances are used to wrap up various settings such as the precision and default rounding mode.

\class{Decimal} instances, like regular Python integers and FP
numbers, are immutable; once they've been created, you can't change
the value it represents.  \class{Decimal} instances can be created
from integers or strings: 

\begin{verbatim}
>>> import decimal
>>> decimal.Decimal(1972)
Decimal("1972")
>>> decimal.Decimal("1.1")
Decimal("1.1")
\end{verbatim}

You can also provide tuples containing the sign, the mantissa represented 
as a tuple of decimal digits, and the exponent:

\begin{verbatim}
>>> decimal.Decimal((1, (1, 4, 7, 5), -2))
Decimal("-14.75")
\end{verbatim}

Cautionary note: the sign bit is a Boolean value, so 0 is positive and
1 is negative. 

Converting from floating-point numbers poses a bit of a problem:
should the FP number representing 1.1 turn into the decimal number for
exactly 1.1, or for 1.1 plus whatever inaccuracies are introduced?
The decision was to leave such a conversion out of the API.  Instead,
you should convert the floating-point number into a string using the
desired precision and pass the string to the \class{Decimal}
constructor:

\begin{verbatim}
>>> f = 1.1
>>> decimal.Decimal(str(f))
Decimal("1.1")
>>> decimal.Decimal('%.12f' % f)
Decimal("1.100000000000")
\end{verbatim}

Once you have \class{Decimal} instances, you can perform the usual
mathematical operations on them.  One limitation: exponentiation
requires an integer exponent:

\begin{verbatim}
>>> a = decimal.Decimal('35.72')
>>> b = decimal.Decimal('1.73')
>>> a+b
Decimal("37.45")
>>> a-b
Decimal("33.99")
>>> a*b
Decimal("61.7956")
>>> a/b
Decimal("20.64739884393063583815028902")
>>> a ** 2
Decimal("1275.9184")
>>> a**b
Traceback (most recent call last):
  ...
decimal.InvalidOperation: x ** (non-integer)
\end{verbatim}

You can combine \class{Decimal} instances with integers, but not with
floating-point numbers:

\begin{verbatim}
>>> a + 4
Decimal("39.72")
>>> a + 4.5
Traceback (most recent call last):
  ...
TypeError: You can interact Decimal only with int, long or Decimal data types.
>>>
\end{verbatim}

\class{Decimal} numbers can be used with the \module{math} and
\module{cmath} modules, but note that they'll be immediately converted to 
floating-point numbers before the operation is performed, resulting in
a possible loss of precision and accuracy.  You'll also get back a
regular floating-point number and not a \class{Decimal}.  

\begin{verbatim}
>>> import math, cmath
>>> d = decimal.Decimal('123456789012.345')
>>> math.sqrt(d)
351364.18288201344
>>> cmath.sqrt(-d)
351364.18288201344j
\end{verbatim}

Instances also have a \method{sqrt()} method that returns a
\class{Decimal}, but if you need other things such as trigonometric
functions you'll have to implement them.

\begin{verbatim}
>>> d.sqrt()
Decimal("351364.1828820134592177245001")
\end{verbatim}


\subsection{The \class{Context} type}

Instances of the \class{Context} class encapsulate several settings for 
decimal operations:

\begin{itemize}
 \item \member{prec} is the precision, the number of decimal places.
 \item \member{rounding} specifies the rounding mode.  The \module{decimal}
       module has constants for the various possibilities:
       \constant{ROUND_DOWN}, \constant{ROUND_CEILING}, \constant{ROUND_HALF_EVEN}, and various others.
 \item \member{traps} is a dictionary specifying what happens on
encountering certain error conditions: either  an exception is raised or 
a value is returned.  Some examples of error conditions are
division by zero, loss of precision, and overflow.
\end{itemize}

There's a thread-local default context available by calling
\function{getcontext()}; you can change the properties of this context
to alter the default precision, rounding, or trap handling.

\begin{verbatim}
>>> decimal.getcontext().prec
28
>>> decimal.Decimal(1) / decimal.Decimal(7)
Decimal("0.1428571428571428571428571429")
>>> decimal.getcontext().prec = 9 
>>> decimal.Decimal(1) / decimal.Decimal(7)
Decimal("0.142857143")
\end{verbatim}

The default action for error conditions is selectable; the module can
either return a special value such as infinity or not-a-number, or
exceptions can be raised:

\begin{verbatim}
>>> decimal.Decimal(1) / decimal.Decimal(0)
Traceback (most recent call last):
  ...
decimal.DivisionByZero: x / 0
>>> decimal.getcontext().traps[decimal.DivisionByZero] = False
>>> decimal.Decimal(1) / decimal.Decimal(0)
Decimal("Infinity")
>>> 
\end{verbatim}

The \class{Context} instance also has various methods for formatting 
numbers such as \method{to_eng_string()} and \method{to_sci_string()}.

For more information, see the documentation for the \module{decimal}
module, which includes a quick-start tutorial and a reference.

\begin{seealso}
\seepep{327}{Decimal Data Type}{Written by Facundo Batista and implemented
  by Facundo Batista, Eric Price, Raymond Hettinger, Aahz, and Tim Peters.}

\seeurl{http://research.microsoft.com/\textasciitilde hollasch/cgindex/coding/ieeefloat.html}
{A more detailed overview of the IEEE-754 representation.}

\seeurl{http://www.lahey.com/float.htm}
{The article uses Fortran code to illustrate many of the problems
that floating-point inaccuracy can cause.}

\seeurl{http://www2.hursley.ibm.com/decimal/}
{A description of a decimal-based representation.  This representation
is being proposed as a standard, and underlies the new Python decimal
type.  Much of this material was written by Mike Cowlishaw, designer of the
Rexx language.}

\end{seealso}      


%======================================================================
\section{PEP 328: Multi-line Imports}

One language change is a small syntactic tweak aimed at making it
easier to import many names from a module.  In a
\code{from \var{module} import \var{names}} statement, 
\var{names} is a sequence of names separated by commas.  If the sequence is 
very long, you can either write multiple imports from the same module,
or you can use backslashes to escape the line endings:

\begin{verbatim}
from SimpleXMLRPCServer import SimpleXMLRPCServer,\
            SimpleXMLRPCRequestHandler,\
            CGIXMLRPCRequestHandler,\
            resolve_dotted_attribute
\end{verbatim}

The syntactic change simply allows putting the names within
parentheses.  Python ignores newlines within a parenthesized
expression, so the backslashes are no longer needed:

\begin{verbatim}
from SimpleXMLRPCServer import (SimpleXMLRPCServer,
            SimpleXMLRPCRequestHandler,
            CGIXMLRPCRequestHandler,
            resolve_dotted_attribute)
\end{verbatim}

The PEP also proposes that all \keyword{import} statements be
absolute imports, with a leading \samp{.} character to indicate a
relative import.  This part of the PEP is not yet implemented.

\begin{seealso}
\seepep{328}{Imports: Multi-Line and Absolute/Relative}{Written by Aahz.  Multi-line imports were implemented by Dima Dorfman.}


%======================================================================
\section{PEP 331: Locale-Independent Float/String Conversions}

The \module{locale} modules lets Python software select various
conversions and display conventions that are localized to a particular
country or language.  However, the module was careful to not change
the numeric locale because various functions in Python's
implementation required that the numeric locale remain set to the
\code{'C'} locale.  Often this was because the code was using the C library's
\cfunction{atof()} function.  

Not setting the numeric locale caused trouble for extensions that used
third-party C libraries, however, because they wouldn't have the
correct locale set.  The motivating example was GTK+, whose user
interface widgets weren't displaying numbers in the current locale.

The solution described in the PEP is to add three new functions to the
Python API that perform ASCII-only conversions, ignoring the locale
setting:

\begin{itemize}
 \item \cfunction{PyOS_ascii_strtod(\var{str}, \var{ptr})} 
and \cfunction{PyOS_ascii_atof(\var{str}, \var{ptr})} 
both convert a string to a C \ctype{double}.
 \item \cfunction{PyOS_ascii_formatd(\var{buffer}, \var{buf_len}, \var{format}, \var{d})} converts a \ctype{double} to an ASCII string.
\end{itemize}

The code for these functions came from the GLib library
(\url{http://developer.gnome.org/arch/gtk/glib.html}), whose
developers kindly relicensed the relevant functions and donated them
to the Python Software Foundation.  The \module{locale} module 
can now change the numeric locale, letting extensions such as GTK+ 
produce the correct results.

\begin{seealso}
\seepep{331}{Locale-Independent Float/String Conversions}{Written by Christian R. Reis, and implemented by Gustavo Carneiro.}
\end{seealso}      

%======================================================================
\section{Other Language Changes}

Here are all of the changes that Python 2.4 makes to the core Python
language.

\begin{itemize}

\item The \method{dict.update()} method now accepts the same
argument forms as the \class{dict} constructor.  This includes any
mapping, any iterable of key/value pairs, and keyword arguments.       

\item The string methods \method{ljust()}, \method{rjust()}, and
\method{center()} now take an optional argument for specifying a
fill character other than a space.

\item Strings also gained an \method{rsplit()} method that
works like the \method{split()} method but splits from the end of
the string.  

\begin{verbatim}
>>> 'www.python.org'.split('.', 1)
['www', 'python.org']
'www.python.org'.rsplit('.', 1)
['www.python', 'org']        
\end{verbatim}      

\item The \method{sort()} method of lists gained three keyword
arguments: \var{cmp}, \var{key}, and \var{reverse}.  These arguments
make some common usages of \method{sort()} simpler.  All are optional.

\var{cmp} is the same as the previous single argument to
\method{sort()}; if provided, the value should be a comparison
function that takes two arguments and returns -1, 0, or +1 depending
on how the arguments compare.  

\var{key} should be a single-argument function that takes a list
element and returns a comparison key for the element.  The list is
then sorted using the comparison keys.  The following example sorts a
list case-insensitively:

\begin{verbatim}
>>> L = ['A', 'b', 'c', 'D']
>>> L.sort()                 # Case-sensitive sort
>>> L
['A', 'D', 'b', 'c']
>>> L.sort(key=lambda x: x.lower())
>>> L
['A', 'b', 'c', 'D']
>>> L.sort(cmp=lambda x,y: cmp(x.lower(), y.lower()))
>>> L
['A', 'b', 'c', 'D']
\end{verbatim}

The last example, which uses the \var{cmp} parameter, is the old way
to perform a case-insensitive sort.  It works but is slower than
using a \var{key} parameter.  Using \var{key} results in calling the
\method{lower()} method once for each element in the list while using
\var{cmp} will call it twice for each comparison.

For simple key functions and comparison functions, it is often
possible to avoid a \keyword{lambda} expression by using an unbound
method instead.  For example, the above case-insensitive sort is best
coded as:

\begin{verbatim}
>>> L.sort(key=str.lower)
>>> L
['A', 'b', 'c', 'D']
\end{verbatim}       

The \var{reverse} parameter should have a Boolean value.  If the value
is \constant{True}, the list will be sorted into reverse order.
Instead of \code{L.sort(lambda x,y: cmp(x.score, y.score)) ;
L.reverse()}, you can now write: \code{L.sort(key = lambda x: x.score,
reverse=True)}.

The results of sorting are now guaranteed to be stable.  This means
that two entries with equal keys will be returned in the same order as
they were input.  For example, you can sort a list of people by name,
and then sort the list by age, resulting in a list sorted by age where
people with the same age are in name-sorted order.

\item There is a new built-in function
\function{sorted(\var{iterable})} that works like the in-place
\method{list.sort()} method but can be used in
expressions.  The differences are:
  \begin{itemize}
  \item the input may be any iterable;
  \item a newly formed copy is sorted, leaving the original intact; and
  \item the expression returns the new sorted copy
  \end{itemize}

\begin{verbatim}
>>> L = [9,7,8,3,2,4,1,6,5]
>>> [10+i for i in sorted(L)]       # usable in a list comprehension
[11, 12, 13, 14, 15, 16, 17, 18, 19]
>>> L                               # original is left unchanged
[9,7,8,3,2,4,1,6,5]
>>> sorted('Monty Python')          # any iterable may be an input
[' ', 'M', 'P', 'h', 'n', 'n', 'o', 'o', 't', 't', 'y', 'y']

>>> # List the contents of a dict sorted by key values
>>> colormap = dict(red=1, blue=2, green=3, black=4, yellow=5)
>>> for k, v in sorted(colormap.iteritems()):
...     print k, v
...
black 4
blue 2
green 3
red 1
yellow 5
\end{verbatim}

\item Integer operations will no longer trigger an \exception{OverflowWarning}.
The \exception{OverflowWarning} warning will disappear in Python 2.5.

\item The \function{eval(\var{expr}, \var{globals}, \var{locals})}
and \function{execfile(\var{filename}, \var{globals}, \var{locals})}
functions and the \keyword{exec} statement now accept any mapping type
for the \var{locals} argument.  Previously this had to be a regular
Python dictionary.  (Contributed by Raymond Hettinger.)

\item The \function{zip()} built-in function and \function{itertools.izip()}
  now return an empty list if called with no arguments.
  Previously they raised a \exception{TypeError}
  exception.  This makes them more
  suitable for use with variable length argument lists:

\begin{verbatim}
>>> def transpose(array):
...    return zip(*array)
...
>>> transpose([(1,2,3), (4,5,6)])
[(1, 4), (2, 5), (3, 6)]
>>> transpose([])
[]
\end{verbatim}

\item Encountering a failure while importing a module no longer leaves
a partially-initialized module object in \code{sys.modules}.  The
incomplete module object left behind would fool further imports of the
same module into succeeding, leading to confusing errors.  

\item \constant{None} is now a constant; code that binds a new value to 
the name \samp{None} is now a syntax error.

\end{itemize}


%======================================================================
\subsection{Optimizations}

\begin{itemize}

\item The inner loops for list and tuple slicing
 were optimized and now run about one-third faster.  The inner loops
 were also optimized for dictionaries, resulting in performance boosts for
 \method{keys()}, \method{values()}, \method{items()},
 \method{iterkeys()}, \method{itervalues()}, and \method{iteritems()}.

\item The machinery for growing and shrinking lists was optimized for
 speed and for space efficiency.  Appending and popping from lists now
 runs faster due to more efficient code paths and less frequent use of
 the underlying system \cfunction{realloc()}.  List comprehensions
 also benefit.   \method{list.extend()} was also optimized and no
 longer converts its argument into a temporary list before extending
 the base list.

\item \function{list()}, \function{tuple()}, \function{map()},
  \function{filter()}, and \function{zip()} now run several times
  faster with non-sequence arguments that supply a \method{__len__()}
  method.  

\item The methods \method{list.__getitem__()},
  \method{dict.__getitem__()}, and \method{dict.__contains__()} are
  are now implemented as \class{method_descriptor} objects rather
  than \class{wrapper_descriptor} objects.  This form of optimized
  access doubles their performance and makes them more suitable for
  use as arguments to functionals:
  \samp{map(mydict.__getitem__, keylist)}.

\item Added a new opcode, \code{LIST_APPEND}, that simplifies
  the generated bytecode for list comprehensions and speeds them up
  by about a third.

\item String concatenations in statements of the form \code{s = s +
"abc"} and \code{s += "abc"} are now performed more efficiently in
certain circumstances.  This optimization won't be present in other
Python implementations such as Jython, so you shouldn't rely on it;
using the \method{join()} method of strings is still recommended when
you want to efficiently glue a large number of strings together.

\end{itemize}

The net result of the 2.4 optimizations is that Python 2.4 runs the
pystone benchmark around XX\% faster than Python 2.3 and YY\% faster
than Python 2.2.


%======================================================================
\section{New, Improved, and Deprecated Modules}

As usual, Python's standard library received a number of enhancements and
bug fixes.  Here's a partial list of the most notable changes, sorted
alphabetically by module name. Consult the
\file{Misc/NEWS} file in the source tree for a more
complete list of changes, or look through the CVS logs for all the
details.

\begin{itemize}

% XXX new email parser

\item The \module{asyncore} module's \function{loop()} now has a 
   \var{count} parameter that lets you perform a limited number 
   of passes through the polling loop.  The default is still to loop
   forever.

\item The \module{curses} modules now supports the ncurses extension 
   \function{use_default_colors()}.  On platforms where the terminal
   supports transparency, this makes it possible to use a transparent
   background.  (Contributed by J\"org Lehmann.)

\item The \module{bisect} module now has an underlying C implementation
   for improved performance.
   (Contributed by Dmitry Vasiliev.)

\item The CJKCodecs collections of East Asian codecs, maintained
by Hye-Shik Chang, was integrated into 2.4.  
The new encodings are:

\begin{itemize}
 \item Chinese (PRC): gb2312, gbk, gb18030, big5hkscs, hz
 \item Chinese (ROC): big5, cp950
 \item Japanese: cp932, euc-jis-2004, euc-jp,
euc-jisx0213, iso-2022-jp, iso-2022-jp-1, iso-2022-jp-2,
 iso-2022-jp-3, iso-2022-jp-ext, iso-2022-jp-2004,
 shift-jis, shift-jisx0213, shift-jis-2004
 \item Korean: cp949, euc-kr, johab, iso-2022-kr
\end{itemize} 

\item Some other new encodings were added: HP Roman8, 
ISO_8859-11, ISO_8859-16, PCTP-154, 
and TIS-620.

\item There is a new \module{collections} module for 
   various specialized collection datatypes.  
   Currently it contains just one type, \class{deque}, 
   a double-ended queue that supports efficiently adding and removing
   elements from either end.

\begin{verbatim}
>>> from collections import deque
>>> d = deque('ghi')        # make a new deque with three items
>>> d.append('j')           # add a new entry to the right side
>>> d.appendleft('f')       # add a new entry to the left side
>>> d                       # show the representation of the deque
deque(['f', 'g', 'h', 'i', 'j'])
>>> d.pop()                 # return and remove the rightmost item
'j'
>>> d.popleft()             # return and remove the leftmost item
'f'
>>> list(d)                 # list the contents of the deque
['g', 'h', 'i']
>>> 'h' in d                # search the deque
True  
\end{verbatim}

Several modules now take advantage of \class{collections.deque} for
improved performance, such as the \module{Queue} and
\module{threading} modules.

\item The \module{ConfigParser} classes have been enhanced slightly.
   The \method{read()} method now returns a list of the files that
   were successfully parsed, and the \method{set()} method raises
   \exception{TypeError} if passed a \var{value} argument that isn't a
   string.

\item The \module{heapq} module has been converted to C.  The resulting
   tenfold improvement in speed makes the module suitable for handling
   high volumes of data.  In addition, the module has two new functions
   \function{nlargest()} and \function{nsmallest()} that use heaps to
   find the N largest or smallest values in a dataset without the
   expense of a full sort.

\item The \module{imaplib} module now supports IMAP's THREAD command
(contributed by Yves Dionne) and new \method{deleteacl()} and
\method{myrights()} methods (contributed by Arnaud Mazin).

\item The \module{itertools} module gained a
  \function{groupby(\var{iterable}\optional{, \var{func}})} function.
  \var{iterable} returns a succession of elements, and the optional
  \var{func} is a function that takes an element and returns a key
  value; if omitted, the key is simply the element itself.
  \function{groupby()} then groups the elements into subsequences
  which have matching values of the key, and returns a series of 2-tuples
  containing the key value and an iterator over the subsequence.
 
Here's an example.  The \var{key} function simply returns whether a
number is even or odd, so the result of \function{groupby()} is to
return consecutive runs of odd or even numbers.

\begin{verbatim}
>>> import itertools
>>> L = [2,4,6, 7,8,9,11, 12, 14]
>>> for key_val, it in itertools.groupby(L, lambda x: x % 2):
...    print key_val, list(it)
... 
0 [2, 4, 6]
1 [7]
0 [8]
1 [9, 11]
0 [12, 14]
>>> 
\end{verbatim}

\function{groupby()} is typically used with sorted input.  The logic
for \function{groupby()} is similar to the \UNIX{} \code{uniq} filter
which makes it handy for eliminating, counting, or identifying
duplicate elements:

\begin{verbatim}
>>> word = 'abracadabra'
>>> letters = sorted(word)   # Turn string into a sorted list of letters
>>> letters 
['a', 'a', 'a', 'a', 'a', 'b', 'b', 'c', 'd', 'r', 'r']
>>> for k, g in itertools.groupby(letters):
...    print k, list(g)
... 
a ['a', 'a', 'a', 'a', 'a']
b ['b', 'b']
c ['c']
d ['d']
r ['r', 'r']
>>> # List unique letters
>>> [k for k, g in groupby(letters)]                     
['a', 'b', 'c', 'd', 'r']
>>> # Count letter occurences
>>> [(k, len(list(g))) for k, g in groupby(letters)]     
[('a', 5), ('b', 2), ('c', 1), ('d', 1), ('r', 2)]
\end{verbatim}

\item \module{itertools} also gained a function named
\function{tee(\var{iterator}, \var{N})} that returns \var{N} independent
iterators that replicate \var{iterator}.  If \var{N} is omitted, the
default is 2.

\begin{verbatim}
>>> L = [1,2,3]
>>> i1, i2 = itertools.tee(L)
>>> i1,i2
(<itertools.tee object at 0x402c2080>, <itertools.tee object at 0x402c2090>)
>>> list(i1)               # Run the first iterator to exhaustion
[1, 2, 3]
>>> list(i2)               # Run the second iterator to exhaustion
[1, 2, 3]
>\end{verbatim}

Note that \function{tee()} has to keep copies of the values returned 
by the iterator; in the worst case, it may need to keep all of them.  
This should therefore be used carefully if the leading iterator
can run far ahead of the trailing iterator in a long stream of inputs.
If the separation is large, then you might as well use 
\function{list()} instead.  When the iterators track closely with one
another, \function{tee()} is ideal.  Possible applications include
bookmarking, windowing, or lookahead iterators.

\item  A number of functions were added to the \module{locale} 
module, such as \function{bind_textdomain_codeset()} to specify a
particular encoding, and a family of \function{l*gettext()} functions
that return messages in the chosen encoding.
(Contributed by Gustavo Niemeyer.)

\item The \module{logging} package's \function{basicConfig} function 
gained some keyword arguments to simplify log configuration.  The
default behavior is to log messages to standard error, but
various keyword arguments can be specified to log to a particular file,
change the logging format, or set the logging level.  For example:

\begin{verbatim}
import logging
logging.basicConfig(filename = '/var/log/application.log',
    level=0,  # Log all messages, including debugging,
    format='%(levelname):%(process):%(thread):%(message)')	            
\end{verbatim}

Another addition to \module{logging} is a
\class{TimedRotatingFileHandler} class which rotates its log files at
a timed interval.  The module already had \class{RotatingFileHandler},
which rotated logs once the file exceeded a certain size.  Both
classes derive from a new \class{BaseRotatingHandler} class that can
be used to implement other rotating handlers.

\item The \module{nntplib} module's \class{NNTP} class gained
\method{description()} and \method{descriptions()} methods to retrieve 
newsgroup descriptions for a single group or for a range of groups.
(Contributed by J\"urgen A. Erhard.)

\item The \module{operator} module gained two new functions, 
\function{attrgetter(\var{attr})} and \function{itemgetter(\var{index})}.
Both functions return callables that take a single argument and return
the corresponding attribute or item; these callables make excellent
data extractors when used with \function{map()} or
\function{sorted()}.  For example:

\begin{verbatim}
>>> L = [('c', 2), ('d', 1), ('a', 4), ('b', 3)]
>>> map(operator.itemgetter(0), L)
['c', 'd', 'a', 'b']
>>> map(operator.itemgetter(1), L)
[2, 1, 4, 3]
>>> sorted(L, key=operator.itemgetter(1)) # Sort list by second tuple item
[('d', 1), ('c', 2), ('b', 3), ('a', 4)]
\end{verbatim}

\item The \module{optparse} module was updated.  The module now passes
its messages through \function{gettext.gettext()}, making it possible
to internationalize Optik's help and error messages.  Help messages
for options can now include the string \code{'\%default'}, which will
be replaced by the option's default value.

\item A new \function{urandom(\var{n})} function 
was added to the \module{os} module, providing access to
platform-specific sources of randomness such as
\file{/dev/urandom} on Linux or the Windows CryptoAPI.  The
function returns a string containing \var{n} bytes of random data.
(Contributed by Trevor Perrin.)

\item Another new function: \function{os.path.lexists(\var{path})} 
returns true if the file specified by \var{path} exists, whether or
not it's a symbolic link.  This differs from the existing
\function{os.path.exists(\var{path})} function, which returns false if 
\var{path} is a symlink that points to a destination that doesn't exist.
(Contributed by Beni Cherniavsky.)

\item A new \function{getsid()} function was added to the
\module{posix} module that underlies the \module{os} module.
(Contributed by J. Raynor.)

\item The \module{poplib} module now supports POP over SSL.

\item The \module{profile} module can now profile C extension functions.
% XXX more to say about this? 

\item The \module{random} module has a new method called \method{getrandbits(N)} 
   which returns an N-bit long integer.  This method supports the existing
   \method{randrange()} method, making it possible to efficiently generate
   arbitrarily large random numbers.

\item The regular expression language accepted by the \module{re} module
   was extended with simple conditional expressions, written as
   \regexp{(?(\var{group})\var{A}|\var{B})}.  \var{group} is either a
   numeric group ID or a group name defined with \regexp{(?P<group>...)} 
   earlier in the expression.  If the specified group matched, the
   regular expression pattern \var{A} will be tested against the string; if
   the group didn't match, the pattern \var{B} will be used instead.

\item The \module{re} module is also no longer recursive, thanks 
to a massive amount of work by Gustavo Niemeyer.  In a recursive
regular expression engine, certain patterns result in a large amount
of C stack space being consumed, and it was possible to overflow the
stack.  For example, if you matched a 30000-byte string of \samp{a}
characters against the expression \regexp{(a|b)+}, one stack frame was
consumed per character.  Python 2.3 tried to check for stack overflow
and raise a \exception{RuntimeError} exception, but if you were
unlucky Python could dump core.  Python 2.4's regular expression
engine can match this pattern without problems.

\item A new \function{socketpair()} function was added to the
\module{socket} module, returning a pair of connected sockets. 
(Contributed by Dave Cole.)

\item The \function{sys.exitfunc()} function has been deprecated.  Code
should be using the existing \module{atexit} module, which correctly
handles calling multiple exit functions.  Eventually
\function{sys.exitfunc()} will become a purely internal interface,
accessed only by \module{atexit}.

\item The \module{tarfile} module now generates GNU-format tar files
by default.

\item The \module{threading} module now has an elegantly simple way to support 
thread-local data.  The module contains a \class{local} class whose
attribute values are local to different threads.

\begin{verbatim}
import threading

data = threading.local()
data.number = 42
data.url = ('www.python.org', 80)
\end{verbatim}

Other threads can assign and retrieve their own values for the
\member{number} and \member{url} attributes.  You can subclass
\class{local} to initialize attributes or to add methods.
(Contributed by Jim Fulton.)

\item The \module{weakref} module now supports a wider variety of objects
   including Python functions, class instances, sets, frozensets, deques,
   arrays, files, sockets, and regular expression pattern objects.

\item The \module{xmlrpclib} module now supports a multi-call extension for 
transmitting multiple XML-RPC calls in a single HTTP operation.
   
\end{itemize}


%======================================================================
% whole new modules get described in subsections here

\subsection{cookielib}

The \module{cookielib} library supports client-side handling for HTTP
cookies, just as the \module{Cookie} provides server-side cookie
support in CGI scripts. Cookies are stored in cookie jars; the library
transparently stores cookies offered by the web server in the cookie
jar, and fetches the cookie from the jar when connecting to the
server. Similar to web browsers, policy objects control whether
cookies are accepted or not.

In order to store cookies across sessions, two implementations of
cookie jars are provided: one that stores cookies in the Netscape
format, so applications can use the Mozilla or Lynx cookie jars, and
one that stores cookies in the same format as the Perl libwww libary.

\module{urllib2} has been changed to interact with \module{cookielib}:
\class{HTTPCookieProcessor} manages a cookie jar that is used when
accessing URLs.

\subsection{doctest}

The \module{doctest} module underwent considerable refactoring thanks
to Edward Loper and Tim Peters.

% XXX describe this

% ======================================================================
\section{Build and C API Changes}

Changes to Python's build process and to the C API include:

\begin{itemize}

  \item Three new convenience macros were added for common return
  values from extension functions: \csimplemacro{Py_RETURN_NONE},
  \csimplemacro{Py_RETURN_TRUE}, and \csimplemacro{Py_RETURN_FALSE}.

  \item Another new macro, \csimplemacro{Py_CLEAR(\var{obj})}, 
  decreases the reference count of \var{obj} and sets \var{obj} to the
  null pointer.

  \item A new function, \cfunction{PyTuple_Pack(\var{N}, \var{obj1},
  \var{obj2}, ..., \var{objN})}, constructs tuples from a variable
  length argument list of Python objects.

  \item A new function, \cfunction{PyDict_Contains(\var{d}, \var{k})},
  implements fast dictionary lookups without masking exceptions raised
  during the look-up process.

  \item A new function, \cfunction{PyArg_VaParseTupleAndKeywords()},
  is the same as \cfunction{PyArg_ParseTupleAndKeywords()} but takes a 
  \ctype{va_list} instead of a number of arguments.
  (Contributed by Greg Chapman.)

  \item A new method flag, \constant{METH_COEXISTS}, allows a function
  defined in slots to co-exist with a \ctype{PyCFunction} having the
  same name.  This can halve the access time for a method such as
  \method{set.__contains__()}.

  \item Python can now be built with additional profiling for the
  interpreter itself.  This is intended for people developing on the
  Python core.  Providing \longprogramopt{--enable-profiling} to the
  \program{configure} script will let you profile the interpreter with
  \program{gprof}, and providing the \longprogramopt{--with-tsc}
  switch enables profiling using the Pentium's Time-Stamp-Counter
  register.  The switch is slightly misnamed, because the profiling
  feature also works on the PowerPC platform, though that processor
  architecture doesn't called that register the TSC.
   
  \item The \ctype{tracebackobject} type has been renamed to \ctype{PyTracebackObject}.

\end{itemize}


%======================================================================
\subsection{Port-Specific Changes}

\begin{itemize}

\item The Windows port now builds under MSVC++ 7.1 as well as version 6.

\end{itemize}


%======================================================================
\section{Other Changes and Fixes \label{section-other}}

% XXX update these figures as we go
As usual, there were a bunch of other improvements and bugfixes
scattered throughout the source tree.  A search through the CVS change
logs finds there were 421 patches applied and 413 bugs fixed between
Python 2.3 and 2.4.  Both figures are likely to be underestimates.

Some of the more notable changes are:

\begin{itemize}

\item The \module{timeit} module now automatically disables periodic
  garbarge collection during the timing loop.  This change makes
  consecutive timings more comparable.

\item The \module{base64} module now has more complete RFC 3548 support
  for Base64, Base32, and Base16 encoding and decoding, including
  optional case folding and optional alternative alphabets.
  (Contributed by Barry Warsaw.)

\end{itemize}


%======================================================================
\section{Porting to Python 2.4}

This section lists previously described changes that may require
changes to your code:

\begin{itemize}

\item The \function{zip()} built-in function and \function{itertools.izip()}
  now return  an empty list instead of raising a \exception{TypeError}
  exception if called with no arguments.

\item \function{dircache.listdir()} now passes exceptions to the caller
      instead of returning empty lists.

\item \function{LexicalHandler.startDTD()} used to receive the public and
  system IDs in the wrong order.  This has been corrected; applications
  relying on the wrong order need to be fixed.

\item \function{fcntl.ioctl} now warns if the \var{mutate} 
 argument is omitted and relevant.

\item The \module{tarfile} module now generates GNU-format tar files
by default.

\end{itemize}


%======================================================================
\section{Acknowledgements \label{acks}}

The author would like to thank the following people for offering
suggestions, corrections and assistance with various drafts of this
article: Hye-Shik Chang, Michael Dyck, Raymond Hettinger.

\end{document}