What’s New In Python 3.0¶

Author:Guido van Rossum

This article explains the new features in Python 3.0, compared to 2.6.Python 3.0, also known as “Python 3000” or “Py3K”, is the first everintentionally backwards incompatible Python release. There are morechanges than in a typical release, and more that are important for allPython users. Nevertheless, after digesting the changes, you’ll findthat Python really hasn’t changed all that much – by and large, we’remostly fixing well-known annoyances and warts, and removing a lot ofold cruft.

This article doesn’t attempt to provide a complete specification ofall new features, but instead tries to give a convenient overview.For full details, you should refer to the documentation for Python3.0, and/or the many PEPs referenced in the text. If you want tounderstand the complete implementation and design rationale for aparticular feature, PEPs usually have more details than the regulardocumentation; but note that PEPs usually are not kept up-to-date oncea feature has been fully implemented.

Due to time constraints this document is not as complete as it shouldhave been. As always for a new release, the Misc/NEWS file in thesource distribution contains a wealth of detailed information aboutevery small thing that was changed.

Common Stumbling Blocks¶

This section lists those few changes that are most likely to trip youup if you’re used to Python 2.5.

Print Is A Function¶

The print statement has been replaced with a print()function, with keyword arguments to replace most of the special syntaxof the old print statement (PEP 3105). Examples:

Old: print "The answer is", 2*2New: print("The answer is", 2*2)Old: print x,           # Trailing comma suppresses newlineNew: print(x, end=" ")  # Appends a space instead of a newlineOld: print              # Prints a newlineNew: print()            # You must call the function!Old: print >>sys.stderr, "fatal error"New: print("fatal error", file=sys.stderr)Old: print (x, y)       # prints repr((x, y))New: print((x, y))      # Not the same as print(x, y)!

You can also customize the separator between items, e.g.:

print("There are <", 2**32, "> possibilities!", sep="")

which produces:

There are <4294967296> possibilities!

Note:

• The print() function doesn’t support the “softspace” feature ofthe old print statement. For example, in Python 2.x,print "A\n", "B" would write "A\nB\n"; but in Python 3.0,print("A\n", "B") writes "A\n B\n".
• Initially, you’ll be finding yourself typing the old print xa lot in interactive mode. Time to retrain your fingers to typeprint(x) instead!
• When using the 2to3 source-to-source conversion tool, allprint statements are automatically converted toprint() function calls, so this is mostly a non-issue forlarger projects.

Views And Iterators Instead Of Lists¶

Some well-known APIs no longer return lists:

• dict methods dict.keys(), dict.items() anddict.values() return “views” instead of lists. For example,this no longer works: k = d.keys(); k.sort(). Use k =sorted(d) instead (this works in Python 2.5 too and is justas efficient).

• Also, the dict.iterkeys(), dict.iteritems() anddict.itervalues() methods are no longer supported.

• map() and filter() return iterators. If you really needa list and the input sequences are all of equal length, a quickfix is to wrap map() in list(), e.g. list(map(...)),but a better fix isoften to use a list comprehension (especially when the original codeuses lambda), or rewriting the code so it doesn’t need alist at all. Particularly tricky is map() invoked for theside effects of the function; the correct transformation is to use aregular for loop (since creating a list would just bewasteful).

  If the input sequences are not of equal length, map() willstop at the termination of the shortest of the sequences. For fullcompatibility with map() from Python 2.x, also wrap the sequences initertools.zip_longest(), e.g. map(func, *sequences) becomeslist(map(func, itertools.zip_longest(*sequences))).

• range() now behaves like xrange() used to behave, exceptit works with values of arbitrary size. The latter no longerexists.

• zip() now returns an iterator.

Ordering Comparisons¶

Python 3.0 has simplified the rules for ordering comparisons:

• The ordering comparison operators (<, <=, >=, >)raise a TypeError exception when the operands don’t have ameaningful natural ordering. Thus, expressions like 1 < '', 0> None or len <= len are no longer valid, and e.g. None <None raises TypeError instead of returningFalse. A corollary is that sorting a heterogeneous listno longer makes sense – all the elements must be comparable to eachother. Note that this does not apply to the == and !=operators: objects of different incomparable types always compareunequal to each other.
• builtin.sorted() and list.sort() no longer accept thecmp argument providing a comparison function. Use the keyargument instead. N.B. the key and reverse arguments are now“keyword-only”.
• The cmp() function should be treated as gone, and the __cmp__()special method is no longer supported. Use __lt__() for sorting,__eq__() with __hash__(), and other rich comparisons as needed.(If you really need the cmp() functionality, you could use theexpression (a > b) - (a < b) as the equivalent for cmp(a, b).)

Integers¶

• PEP 237: Essentially, long renamed to int.That is, there is only one built-in integral type, namedint; but it behaves mostly like the old long type.
• PEP 238: An expression like 1/2 returns a float. Use1//2 to get the truncating behavior. (The latter syntax hasexisted for years, at least since Python 2.2.)
• The sys.maxint constant was removed, since there is nolonger a limit to the value of integers. However, sys.maxsizecan be used as an integer larger than any practical list or stringindex. It conforms to the implementation’s “natural” integer sizeand is typically the same as sys.maxint in previous releaseson the same platform (assuming the same build options).
• The repr() of a long integer doesn’t include the trailing Lanymore, so code that unconditionally strips that character willchop off the last digit instead. (Use str() instead.)
• Octal literals are no longer of the form 0720; use 0o720instead.

Text Vs. Data Instead Of Unicode Vs. 8-bit¶

Everything you thought you knew about binary data and Unicode haschanged.

• Python 3.0 uses the concepts of text and (binary) data insteadof Unicode strings and 8-bit strings. All text is Unicode; howeverencoded Unicode is represented as binary data. The type used tohold text is str, the type used to hold data isbytes. The biggest difference with the 2.x situation isthat any attempt to mix text and data in Python 3.0 raisesTypeError, whereas if you were to mix Unicode and 8-bitstrings in Python 2.x, it would work if the 8-bit string happened tocontain only 7-bit (ASCII) bytes, but you would getUnicodeDecodeError if it contained non-ASCII values. Thisvalue-specific behavior has caused numerous sad faces over theyears.
• As a consequence of this change in philosophy, pretty much all codethat uses Unicode, encodings or binary data most likely has tochange. The change is for the better, as in the 2.x world therewere numerous bugs having to do with mixing encoded and unencodedtext. To be prepared in Python 2.x, start using unicodefor all unencoded text, and str for binary or encoded dataonly. Then the 2to3 tool will do most of the work for you.
• You can no longer use u"..." literals for Unicode text.However, you must use b"..." literals for binary data.
• As the str and bytes types cannot be mixed, youmust always explicitly convert between them. Use str.encode()to go from str to bytes, and bytes.decode()to go from bytes to str. You can also usebytes(s, encoding=...) and str(b, encoding=...),respectively.
• Like str, the bytes type is immutable. There is aseparate mutable type to hold buffered binary data,bytearray. Nearly all APIs that accept bytes alsoaccept bytearray. The mutable API is based oncollections.MutableSequence.
• All backslashes in raw string literals are interpreted literally.This means that '\U' and '\u' escapes in raw strings are nottreated specially. For example, r'\u20ac' is a string of 6characters in Python 3.0, whereas in 2.6, ur'\u20ac' was thesingle “euro” character. (Of course, this change only affects rawstring literals; the euro character is '\u20ac' in Python 3.0.)
• The built-in basestring abstract type was removed. Usestr instead. The str and bytes typesdon’t have functionality enough in common to warrant a shared baseclass. The 2to3 tool (see below) replaces every occurrence ofbasestring with str.
• Files opened as text files (still the default mode for open())always use an encoding to map between strings (in memory) and bytes(on disk). Binary files (opened with a b in the mode argument)always use bytes in memory. This means that if a file is openedusing an incorrect mode or encoding, I/O will likely fail loudly,instead of silently producing incorrect data. It also means thateven Unix users will have to specify the correct mode (text orbinary) when opening a file. There is a platform-dependent defaultencoding, which on Unixy platforms can be set with the LANGenvironment variable (and sometimes also with some otherplatform-specific locale-related environment variables). In manycases, but not all, the system default is UTF-8; you should nevercount on this default. Any application reading or writing more thanpure ASCII text should probably have a way to override the encoding.There is no longer any need for using the encoding-aware streamsin the codecs module.
• The initial values of sys.stdin, sys.stdout andsys.stderr are now unicode-only text files (i.e., they areinstances of io.TextIOBase). To read and write bytes datawith these streams, you need to use their io.TextIOBase.bufferattribute.
• Filenames are passed to and returned from APIs as (Unicode) strings.This can present platform-specific problems because on someplatforms filenames are arbitrary byte strings. (On the other hand,on Windows filenames are natively stored as Unicode.) As awork-around, most APIs (e.g. open() and many functions in theos module) that take filenames accept bytes objectsas well as strings, and a few APIs have a way to ask for abytes return value. Thus, os.listdir() returns alist of bytes instances if the argument is a bytesinstance, and os.getcwdb() returns the current workingdirectory as a bytes instance. Note that whenos.listdir() returns a list of strings, filenames thatcannot be decoded properly are omitted rather than raisingUnicodeError.
• Some system APIs like os.environ and sys.argv canalso present problems when the bytes made available by the system isnot interpretable using the default encoding. Setting the LANGvariable and rerunning the program is probably the best approach.
• PEP 3138: The repr() of a string no longer escapesnon-ASCII characters. It still escapes control characters and codepoints with non-printable status in the Unicode standard, however.
• PEP 3120: The default source encoding is now UTF-8.
• PEP 3131: Non-ASCII letters are now allowed in identifiers.(However, the standard library remains ASCII-only with the exceptionof contributor names in comments.)
• The StringIO and cStringIO modules are gone. Instead,import the io module and use io.StringIO orio.BytesIO for text and data respectively.
• See also the Unicode HOWTO, which was updated for Python 3.0.

Overview Of Syntax Changes¶

This section gives a brief overview of every syntactic change inPython 3.0.

New Syntax¶

• PEP 3107: Function argument and return value annotations. Thisprovides a standardized way of annotating a function’s parametersand return value. There are no semantics attached to suchannotations except that they can be introspected at runtime usingthe __annotations__ attribute. The intent is to encourageexperimentation through metaclasses, decorators or frameworks.

• PEP 3102: Keyword-only arguments. Named parameters occurringafter *args in the parameter list must be specified usingkeyword syntax in the call. You can also use a bare * in theparameter list to indicate that you don’t accept a variable-lengthargument list, but you do have keyword-only arguments.

• Keyword arguments are allowed after the list of base classes in aclass definition. This is used by the new convention for specifyinga metaclass (see next section), but can be used for other purposesas well, as long as the metaclass supports it.

• PEP 3104: nonlocal statement. Using nonlocal xyou can now assign directly to a variable in an outer (butnon-global) scope. nonlocal is a new reserved word.

• PEP 3132: Extended Iterable Unpacking. You can now write thingslike a, b, *rest = some_sequence. And even *rest, a =stuff. The rest object is always a (possibly empty) list; theright-hand side may be any iterable. Example:

  (a, *rest, b) = range(5)

  This sets a to 0, b to 4, and rest to [1, 2, 3].

• Dictionary comprehensions: {k: v for k, v in stuff} means the
  same thing as dict(stuff) but is more flexible. (This is
  PEP 274 vindicated. :-)

  
  

• Set literals, e.g. {1, 2}. Note that {} is an empty
  dictionary; use set() for an empty set. Set comprehensions are
  also supported; e.g., {x for x in stuff} means the same thing as
  set(stuff) but is more flexible.

  
  

• New octal literals, e.g. 0o720 (already in 2.6). The old octal
  literals (0720) are gone.

  
  

• New binary literals, e.g. 0b1010 (already in 2.6), and
  there is a new corresponding built-in function, bin().

  
  

• Bytes literals are introduced with a leading b or B, and
  there is a new corresponding built-in function, bytes().

  
  

Changed Syntax¶

• PEP 3109 and PEP 3134: new raise statement syntax:raise [expr [from expr]]. See below.

• as and with are now reserved words. (Since2.6, actually.)

• True, False, and None are reserved words. (2.6 partially enforcedthe restrictions on None already.)

• Change from except exc, var toexcept exc as var. See PEP 3110.

• PEP 3115: New Metaclass Syntax. Instead of:

  class C:    __metaclass__ = M    ...

  you must now use:

  class C(metaclass=M):    ...

  The module-global __metaclass__ variable is no longersupported. (It was a crutch to make it easier to default tonew-style classes without deriving every class fromobject.)

• List comprehensions no longer support the syntactic form
  [... for var in item1, item2, ...]. Use
  [... for var in (item1, item2, ...)] instead.
  Also note that list comprehensions have different semantics: they
  are closer to syntactic sugar for a generator expression inside a
  list() constructor, and in particular the loop control
  variables are no longer leaked into the surrounding scope.

  
  

• The ellipsis (...) can be used as an atomic expression
  anywhere. (Previously it was only allowed in slices.) Also, it
  must now be spelled as .... (Previously it could also be
  spelled as . . ., by a mere accident of the grammar.)

  
  

Removed Syntax¶

• PEP 3113: Tuple parameter unpacking removed. You can no longerwrite def foo(a, (b, c)): ....Use def foo(a, b_c): b, c = b_c instead.
• Removed backticks (use repr() instead).
• Removed <> (use != instead).
• Removed keyword: exec() is no longer a keyword; it remains asa function. (Fortunately the function syntax was also accepted in2.x.) Also note that exec() no longer takes a stream argument;instead of exec(f) you can use exec(f.read()).
• Integer literals no longer support a trailing l or L.
• String literals no longer support a leading u or U.
• The from module import * syntax is onlyallowed at the module level, no longer inside functions.
• The only acceptable syntax for relative imports is from .[module]import name. All import forms not starting with . areinterpreted as absolute imports. (PEP 328)
• Classic classes are gone.

Changes Already Present In Python 2.6¶

Since many users presumably make the jump straight from Python 2.5 toPython 3.0, this section reminds the reader of new features that wereoriginally designed for Python 3.0 but that were back-ported to Python2.6. The corresponding sections in What’s New in Python 2.6 should beconsulted for longer descriptions.

• PEP 343: The ‘with’ statement. The with statement is now a standardfeature and no longer needs to be imported from the __future__.Also check out Writing Context Managers andThe contextlib module.
• PEP 366: Explicit Relative Imports From a Main Module. This enhances the usefulness of the -moption when the referenced module lives in a package.
• PEP 370: Per-user site-packages Directory.
• PEP 371: The multiprocessing Package.
• PEP 3101: Advanced String Formatting. Note: the 2.6 description mentions theformat() method for both 8-bit and Unicode strings. In 3.0,only the str type (text strings with Unicode support)supports this method; the bytes type does not. The plan isto eventually make this the only API for string formatting, and tostart deprecating the % operator in Python 3.1.
• PEP 3105: print As a Function. This is now a standard feature and no longer needsto be imported from __future__. More details were given above.
• PEP 3110: Exception-Handling Changes. The except exc as varsyntax is now standard and except exc, var is nolonger supported. (Of course, the as var part is stilloptional.)
• PEP 3112: Byte Literals. The b"..." string literal notation (and itsvariants like b'...', b"""...""", and br"...") nowproduces a literal of type bytes.
• PEP 3116: New I/O Library. The io module is now the standard way ofdoing file I/O. The built-in open() function is now analias for io.open() and has additional keyword argumentsencoding, errors, newline and closefd. Also note that aninvalid mode argument now raises ValueError, notIOError. The binary file object underlying a text fileobject can be accessed as f.buffer (but beware that thetext object maintains a buffer of itself in order to speed upthe encoding and decoding operations).
• PEP 3118: Revised Buffer Protocol. The old builtin buffer() is now really gone;the new builtin memoryview() provides (mostly) similarfunctionality.
• PEP 3119: Abstract Base Classes. The abc module and the ABCs defined in thecollections module plays a somewhat more prominent role inthe language now, and built-in collection types like dictand list conform to the collections.MutableMappingand collections.MutableSequence ABCs, respectively.
• PEP 3127: Integer Literal Support and Syntax. As mentioned above, the new octal literalnotation is the only one supported, and binary literals have beenadded.
• PEP 3129: Class Decorators.
• PEP 3141: A Type Hierarchy for Numbers. The numbers module is another new use ofABCs, defining Python’s “numeric tower”. Also note the newfractions module which implements numbers.Rational.

Library Changes¶

Due to time constraints, this document does not exhaustively cover thevery extensive changes to the standard library. PEP 3108 is thereference for the major changes to the library. Here’s a capsulereview:

• Many old modules were removed. Some, like gopherlib (nolonger used) and md5 (replaced by hashlib), werealready deprecated by PEP 4. Others were removed as a resultof the removal of support for various platforms such as Irix, BeOSand Mac OS 9 (see PEP 11). Some modules were also selected forremoval in Python 3.0 due to lack of use or because a betterreplacement exists. See PEP 3108 for an exhaustive list.

• The bsddb3 package was removed because its presence in thecore standard library has proved over time to be a particular burdenfor the core developers due to testing instability and Berkeley DB’srelease schedule. However, the package is alive and well,externally maintained at https://www.jcea.es/programacion/pybsddb.htm.

• Some modules were renamed because their old name disobeyedPEP 8, or for various other reasons. Here’s the list:

  Old NameNew Name_winregwinregConfigParserconfigparsercopy_regcopyregQueuequeueSocketServersocketservermarkupbase_markupbasereprreprlibtest.test_supporttest.support

• A common pattern in Python 2.x is to have one version of a moduleimplemented in pure Python, with an optional accelerated versionimplemented as a C extension; for example, pickle andcPickle. This places the burden of importing the acceleratedversion and falling back on the pure Python version on each user ofthese modules. In Python 3.0, the accelerated versions areconsidered implementation details of the pure Python versions.Users should always import the standard version, which attempts toimport the accelerated version and falls back to the pure Pythonversion. The pickle / cPickle pair received thistreatment. The profile module is on the list for 3.1. TheStringIO module has been turned into a class in the iomodule.

• Some related modules have been grouped into packages, and usuallythe submodule names have been simplified. The resulting newpackages are:

  • dbm (anydbm, dbhash, dbm,dumbdbm, gdbm, whichdb).
  • html (HTMLParser, htmlentitydefs).
  • http (httplib, BaseHTTPServer,CGIHTTPServer, SimpleHTTPServer, Cookie,cookielib).
  • tkinter (all Tkinter-related modules exceptturtle). The target audience of turtle doesn’treally care about tkinter. Also note that as of Python2.6, the functionality of turtle has been greatly enhanced.
  • urllib (urllib, urllib2, urlparse,robotparse).
  • xmlrpc (xmlrpclib, DocXMLRPCServer,SimpleXMLRPCServer).

Some other changes to standard library modules, not covered byPEP 3108:

• Killed sets. Use the built-in set() class.
• Cleanup of the sys module: removed sys.exitfunc(),sys.exc_clear(), sys.exc_type, sys.exc_value,sys.exc_traceback. (Note that sys.last_typeetc. remain.)
• Cleanup of the array.array type: the read() andwrite() methods are gone; use fromfile() andtofile() instead. Also, the 'c' typecode for array isgone – use either 'b' for bytes or 'u' for Unicodecharacters.
• Cleanup of the operator module: removedsequenceIncludes() and isCallable().
• Cleanup of the thread module: acquire_lock() andrelease_lock() are gone; use acquire() andrelease() instead.
• Cleanup of the random module: removed the jumpahead() API.
• The new module is gone.
• The functions os.tmpnam(), os.tempnam() andos.tmpfile() have been removed in favor of the tempfilemodule.
• The tokenize module has been changed to work with bytes. Themain entry point is now tokenize.tokenize(), instead ofgenerate_tokens.
• string.letters and its friends (string.lowercase andstring.uppercase) are gone. Usestring.ascii_letters etc. instead. (The reason for theremoval is that string.letters and friends hadlocale-specific behavior, which is a bad idea for suchattractively-named global “constants”.)
• Renamed module __builtin__ to builtins (removing theunderscores, adding an ‘s’). The __builtins__ variablefound in most global namespaces is unchanged. To modify a builtin,you should use builtins, not __builtins__!

PEP 3101: A New Approach To String Formatting¶

• A new system for built-in string formatting operations replaces the% string formatting operator. (However, the % operator isstill supported; it will be deprecated in Python 3.1 and removedfrom the language at some later time.) Read PEP 3101 for the fullscoop.

Changes To Exceptions¶

The APIs for raising and catching exception have been cleaned up andnew powerful features added:

• PEP 352: All exceptions must be derived (directly or indirectly)from BaseException. This is the root of the exceptionhierarchy. This is not new as a recommendation, but therequirement to inherit from BaseException is new. (Python2.6 still allowed classic classes to be raised, and placed norestriction on what you can catch.) As a consequence, stringexceptions are finally truly and utterly dead.

• Almost all exceptions should actually derive from Exception;BaseException should only be used as a base class forexceptions that should only be handled at the top level, such asSystemExit or KeyboardInterrupt. The recommendedidiom for handling all exceptions except for this latter category isto use except Exception.

• StandardError was removed.

• Exceptions no longer behave as sequences. Use the argsattribute instead.

• PEP 3109: Raising exceptions. You must now use raiseException(args) instead of raise Exception, args.Additionally, you can no longer explicitly specify a traceback;instead, if you have to do this, you can assign directly to the__traceback__ attribute (see below).

• PEP 3110: Catching exceptions. You must now useexcept SomeException as variable insteadof except SomeException, variable. Moreover, thevariable is explicitly deleted when the except blockis left.

• PEP 3134: Exception chaining. There are two cases: implicitchaining and explicit chaining. Implicit chaining happens when anexception is raised in an except or finallyhandler block. This usually happens due to a bug in the handlerblock; we call this a secondary exception. In this case, theoriginal exception (that was being handled) is saved as the__context__ attribute of the secondary exception.Explicit chaining is invoked with this syntax:

  raise SecondaryException() from primary_exception

  (where primary_exception is any expression that produces anexception object, probably an exception that was previously caught).In this case, the primary exception is stored on the__cause__ attribute of the secondary exception. Thetraceback printed when an unhandled exception occurs walks the chainof __cause__ and __context__ attributes and prints aseparate traceback for each component of the chain, with the primaryexception at the top. (Java users may recognize this behavior.)

• PEP 3134: Exception objects now store their traceback as the
  traceback attribute. This means that an exception
  object now contains all the information pertaining to an exception,
  and there are fewer reasons to use sys.exc_info() (though the
  latter is not removed).

  
  

• A few exception messages are improved when Windows fails to load an
  extension module. For example, error code 193 is now %1 is
not a valid Win32 application. Strings now deal with non-English
  locales.

  
  

Miscellaneous Other Changes¶

Operators And Special Methods¶

• != now returns the opposite of ==, unless == returnsNotImplemented.
• The concept of “unbound methods” has been removed from the language.When referencing a method as a class attribute, you now get a plainfunction object.
• __getslice__(), __setslice__() and __delslice__()were killed. The syntax a[i:j] now translates toa.__getitem__(slice(i, j)) (or __setitem__() or__delitem__(), when used as an assignment or deletion target,respectively).
• PEP 3114: the standard next() method has been renamed to__next__().
• The __oct__() and __hex__() special methods are removed– oct() and hex() use __index__() now to convertthe argument to an integer.
• Removed support for __members__ and __methods__.
• The function attributes named func_X have been renamed touse the __X__ form, freeing up these names in the functionattribute namespace for user-defined attributes. To wit,func_closure, func_code, func_defaults,func_dict, func_doc, func_globals,func_name were renamed to __closure__,__code__, __defaults__, __dict__,__doc__, __globals__, __name__,respectively.
• __nonzero__() is now __bool__().

Builtins¶

• PEP 3135: New super(). You can now invoke super()without arguments and (assuming this is in a regular instance methoddefined inside a class statement) the right class andinstance will automatically be chosen. With arguments, the behaviorof super() is unchanged.
• PEP 3111: raw_input() was renamed to input(). Thatis, the new input() function reads a line fromsys.stdin and returns it with the trailing newline stripped.It raises EOFError if the input is terminated prematurely.To get the old behavior of input(), use eval(input()).
• A new built-in function next() was added to call the__next__() method on an object.
• The round() function rounding strategy and return type havechanged. Exact halfway cases are now rounded to the nearest evenresult instead of away from zero. (For example, round(2.5) nowreturns 2 rather than 3.) round(x[, n]) nowdelegates to x.__round__([n]) instead of always returning afloat. It generally returns an integer when called with a singleargument and a value of the same type as x when called with twoarguments.
• Moved intern() to sys.intern().
• Removed: apply(). Instead of apply(f, args) usef(*args).
• Removed callable(). Instead of callable(f) you can useisinstance(f, collections.Callable). The operator.isCallable()function is also gone.
• Removed coerce(). This function no longer serves a purposenow that classic classes are gone.
• Removed execfile(). Instead of execfile(fn) useexec(open(fn).read()).
• Removed the file type. Use open(). There are now severaldifferent kinds of streams that open can return in the io module.
• Removed reduce(). Use functools.reduce() if you reallyneed it; however, 99 percent of the time an explicit forloop is more readable.
• Removed reload(). Use imp.reload().
• Removed. dict.has_key() – use the in operatorinstead.

Build and C API Changes¶

Due to time constraints, here is a very incomplete list of changesto the C API.

• Support for several platforms was dropped, including but not limitedto Mac OS 9, BeOS, RISCOS, Irix, and Tru64.
• PEP 3118: New Buffer API.
• PEP 3121: Extension Module Initialization & Finalization.
• PEP 3123: Making PyObject_HEAD conform to standard C.
• No more C API support for restricted execution.
• PyNumber_Coerce(), PyNumber_CoerceEx(),PyMember_Get(), and PyMember_Set() C APIs are removed.
• New C API PyImport_ImportModuleNoBlock(), works likePyImport_ImportModule() but won’t block on the import lock(returning an error instead).
• Renamed the boolean conversion C-level slot and method:nb_nonzero is now nb_bool.
• Removed METH_OLDARGS and WITH_CYCLE_GC from the C API.

Performance¶

The net result of the 3.0 generalizations is that Python 3.0 runs thepystone benchmark around 10% slower than Python 2.5. Most likely thebiggest cause is the removal of special-casing for small integers.There’s room for improvement, but it will happen after 3.0 isreleased!

Porting To Python 3.0¶

For porting existing Python 2.5 or 2.6 source code to Python 3.0, thebest strategy is the following:

0. (Prerequisite:) Start with excellent test coverage.
1. Port to Python 2.6. This should be no more work than the averageport from Python 2.x to Python 2.(x+1). Make sure all your testspass.
2. (Still using 2.6:) Turn on the -3 command line switch.This enables warnings about features that will be removed (orchange) in 3.0. Run your test suite again, and fix code that youget warnings about until there are no warnings left, and all yourtests still pass.
3. Run the 2to3 source-to-source translator over your source codetree. (See 2to3 - Automated Python 2 to 3 code translation for more on this tool.) Run theresult of the translation under Python 3.0. Manually fix up anyremaining issues, fixing problems until all tests pass again.

It is not recommended to try to write source code that runs unchangedunder both Python 2.6 and 3.0; you’d have to use a very contortedcoding style, e.g. avoiding print statements, metaclasses,and much more. If you are maintaining a library that needs to supportboth Python 2.6 and Python 3.0, the best approach is to modify step 3above by editing the 2.6 version of the source code and running the2to3 translator again, rather than editing the 3.0 version of thesource code.

For porting C extensions to Python 3.0, please see Porting Extension Modules to Python 3.