Abstract

This PEP describes the changes currently envisioned in Python 3.0 (also called Python 3000), a hypothetical future release of Python that can break backwards compatibility with the existing body of Python code.

The list of features included in this document is subject to change and isn't binding on the Python development community; features may be added, removed, and modified at any time. The purpose of this list is to focus our language development effort on changes that are steps to 3.0, and to encourage people to invent ways to smooth the transition.

This document is not a wish-list that anyone can extend. While there are two authors of this PEP, we're just supplying the text; the decisions for which changes are listed in this document are made by Guido van Rossum, who has chosen them as goals for Python 3.0.

General goals

A general goal is to reduce feature duplication by removing old ways of doing things. A general principle of the design will be that one obvious way of doing something is enough. [1]

Core language

• Remove distinction between int and long types [1]
• True division becomes default behavior [10]
• Make all strings be Unicode, and have a separate bytes() type [1]
• exec as a statement is not worth it -- make it a function
• Add optional declarations for static typing [11]
• Support only new-style classes; classic classes will be gone [1]
• Return iterators instead of lists where appropriate for atomic type methods (e.g. dict.keys(), dict.values(), dict.items(), etc.) (Do we keep iter*() methods or remove them? I vote remove. -- nn)
• Replace print by a function [16]
• Do something so you can catch multiple exceptions using except E1, E2, E3:. Maybe use except E1, E2, E3 as err: if you want the error variable? [3]
• None, True and False become keywords [4]
• as becomes a keyword [5]
• Have list comprehensions be syntactic sugar for passing an equivalent generator expression to list(); as a consequence the loop variable will no longer be exposed [12]
• Comparisons other than == and != between disparate types will raise an exception unless explicitly supported by the type [6]
• Exceptions will grow an attribute to store the traceback [13]
• floats will not be acceptable as arguments in place of ints for operations where floats are inadvertantly accepted (PyArg_ParseTuple() i & l formats)
• Imports will be absolute by default. Relative imports must be explicitly specified [19]
• __init__.py will be optional in sub-packages. __init__.py will still be required for top-level packages.

To be removed:

• The lambda statement: use nested or named functions [1], [9]

• String exceptions: use instances of an Exception class [2]

• raise Exception, "message": use raise Exception("message") [14]

• `x`: use repr(x) [2]

• The <> operator: use != instead [3]

• Unbound methods [7]

• METH_OLDARGS

• __getslice__, __setslice__, __delslice__ [17]

• C APIs (see code): PyFloat_AsString, PyFloat_AsReprString, PyFloat_AsStringEx, PySequence_In, PyEval_EvalFrame, PyEval_CallObject, _PyObject_Del, _PyObject_GC_Del, _PyObject_GC_Track, _PyObject_GC_UnTrack PyString_AsEncodedString, PyString_AsDecodedString

  typedefs: intargfunc, intintargfunc

Built-in Namespace

• Make built-ins return an iterator where appropriate (e.g. range(), zip(), etc.)
• Relevant functions should consume iterators (e.g. min(), max())
• Introduce trunc(), which would call the __trunc__() method on its argument; suggested use is for objects like float where calling __int__() has data loss, but an integral representation is still desired [8]
• Exception hierarchy changes [20]

To be removed:

• apply(): use f(*args, **kw) instead [2]
• buffer(): must die (use a bytes() type instead) [2]
• callable(): just call the object and catch the exception [2]
• compile(): put in sys (or perhaps in a module of its own) [2]
• coerce(): no longer needed [2]
• execfile(), reload(): use exec() [2]
• input(): use eval(sys.stdin.readline()) [2]
• intern(), id(): put in sys [2]
• map(), filter(): use list comprehensions instead [1], [9]
• reduce(): write a loop instead [2], [9]
• raw_input(): use sys.stdin.readline() [2]
• xrange(): use range() instead [1]

Atomic Types

To be removed:

• basestring.find() and basestring.rfind(); use basestring.index() or basestring.rindex() in a try/except block [15]
• file.xreadlines() method [17]

Standard library

• Reorganize the standard library to not be as shallow
• Move test code to where it belongs, there will be no more test() functions in the standard library

To be removed:

• Deprecated modules, methods, parameters, attributes, etc. [1] [17] [18] There may be other modules, the most common are listed below.

  stdlib modules to be removed (see docstrings and comments in the source):

  • macfs, new, reconvert, stringold, xmllib
  • pcre, pypcre, strop

  stdlib modules to be removed (see PEP 4): [18]

  • posixfile, pre, regsub, rfc822,
  • statcache, string, TERMIOS
  • mimetools, MimeWriter, mimify,
  • mpz, rgbimage
  • Everything in lib-old: [18] * Para.py, addpack.py, cmp.py, cmpcache.py, codehack.py, * dircmp.py, dump.py, find.py, fmt.py, grep.py, lockfile.py, * newdir.py, ni.py, packmail.py, poly.py, rand.py, statcache.py, * tb.py, tzparse.py, util.py, whatsound.py, whrandom.py, zmod.py

• sys.exitfunc: use atexit module instead [17]

• sys.exc_type, sys.exc_values, sys.exc_traceback: not thread-safe; use sys.exc_info() or an attribute of the exception [2] [13] [17]

• array.read, array.write [17]

• operator.isCallable, operator.sequenceIncludes [17]

References

Copyright

This document has been placed in the public domain.