Add collections.NamedTuple

Doc/lib/libcollections.tex

@@ -9,14 +9,16 @@
 
 
 This module implements high-performance container datatypes.  Currently,
-there are two datatypes, deque and defaultdict.
+there are two datatypes, deque and defaultdict, and one datatype factory
+function, \function{NamedTuple}.
 Future additions may include balanced trees and ordered dictionaries.
 \versionchanged[Added defaultdict]{2.5}
+\versionchanged[Added NamedTuple]{2.6}
 
 \subsection{\class{deque} objects \label{deque-objects}}
 
 \begin{funcdesc}{deque}{\optional{iterable}}
-  Returns a new deque objected initialized left-to-right (using
+  Returns a new deque object initialized left-to-right (using
   \method{append()}) with data from \var{iterable}.  If \var{iterable}
   is not specified, the new deque is empty.
 
@@ -339,3 +341,50 @@ Setting the \member{default_factory} to \class{set} makes the
 >>> d.items()
 [('blue', set([2, 4])), ('red', set([1, 3]))]
 \end{verbatim}
+
+
+
+\subsection{\function{NamedTuple} datatype factory function \label{named-tuple-factory}}
+
+\begin{funcdesc}{NamedTuple}{typename, fieldnames}
+  Returns a new tuple subclass named \var{typename}.  The new subclass is used
+  to create tuple-like objects that have fields accessable by attribute
+  lookup as well as being indexable and iterable.  Instances of the subclass
+  also have a helpful docstring (with typename and fieldnames) and a helpful
+  \method{__repr__()} method which lists the tuple contents in a \code{name=value}
+  format.
+  \versionadded{2.6}
+
+  The \var{fieldnames} are specified in a single string and are separated by spaces.
+  Any valid Python identifier may be used for a field name.
+
+  Example:
+  \begin{verbatim}
+    >>> Point = NamedTuple('Point', 'x y')
+    >>> Point.__doc__           # docstring for the new datatype
+    'Point(x, y)'
+    >>> p = Point(11, y=22)     # instantiate with positional or keyword arguments
+    >>> p[0] + p[1]             # works just like the tuple (11, 22)
+    33
+    >>> x, y = p                # unpacks just like a tuple
+    >>> x, y
+    (11, 22)
+    >>> p.x + p.y               # fields also accessable by name
+    33
+    >>> p                       # readable __repr__ with name=value style
+    Point(x=11, y=22)  
+  \end{verbatim}
+
+  The use cases are the same as those for tuples.  The named factories
+  assign meaning to each tuple position and allow for more readable,
+  self-documenting code.  Can also be used to assign field names to tuples
+  returned by the \module{csv} or \module{sqlite3} modules.  For example:
+
+  \begin{verbatim}
+    import csv
+    EmployeeRecord = NamedTuple('EmployeeRecord', 'name age title deparment paygrade')
+    for tup in csv.reader(open("employees.csv", "rb")):
+        print EmployeeRecord(*tup)
+  \end{verbatim}
+
+\end{funcdesc}

Lib/collections.py

-__all__ = ['deque', 'defaultdict']
+__all__ = ['deque', 'defaultdict', 'NamedTuple']
 
 from _collections import deque, defaultdict
+from operator import itemgetter as _itemgetter
+import sys as _sys
+
+def NamedTuple(typename, s):
+    """Returns a new subclass of tuple with named fields.
+
+    >>> Point = NamedTuple('Point', 'x y')
+    >>> Point.__doc__           # docstring for the new class
+    'Point(x, y)'
+    >>> p = Point(11, y=22)     # instantiate with positional args or keywords
+    >>> p[0] + p[1]             # works just like the tuple (11, 22)
+    33
+    >>> x, y = p                # unpacks just like a tuple
+    >>> x, y
+    (11, 22)
+    >>> p.x + p.y               # fields also accessable by name
+    33
+    >>> p                       # readable __repr__ with name=value style
+    Point(x=11, y=22)
+
+    """
+
+    field_names = s.split()
+    nargs = len(field_names)
+
+    def __new__(cls, *args, **kwds):
+        if kwds:
+            try:
+                args += tuple(kwds[name] for name in field_names[len(args):])
+            except KeyError, name:
+                raise TypeError('%s missing required argument: %s' % (typename, name))
+        if len(args) != nargs:
+            raise TypeError('%s takes exactly %d arguments (%d given)' % (typename, nargs, len(args)))
+        return tuple.__new__(cls, args)
+
+    repr_template = '%s(%s)' % (typename, ', '.join('%s=%%r' % name for name in field_names))
+
+    m = dict(vars(tuple))       # pre-lookup superclass methods (for faster lookup)
+    m.update(__doc__= '%s(%s)' % (typename, ', '.join(field_names)),
+             __slots__ = (),    # no per-instance dict (so instances are same size as tuples)
+             __new__ = __new__,
+             __repr__ = lambda self, _format=repr_template.__mod__: _format(self),
+             __module__ = _sys._getframe(1).f_globals['__name__'],
+             )
+    m.update((name, property(_itemgetter(index))) for index, name in enumerate(field_names))
+
+    return type(typename, (tuple,), m)
+
+
+if __name__ == '__main__':
+    # verify that instances are pickable
+    from cPickle import loads, dumps
+    Point = NamedTuple('Point', 'x y')
+    p = Point(x=10, y=20)
+    assert p == loads(dumps(p))
+
+    import doctest
+    TestResults = NamedTuple('TestResults', 'failed attempted')
+    print TestResults(*doctest.testmod())

Lib/test/test_collections.py

+import unittest
+from test import test_support
+from collections import NamedTuple
+
+class TestNamedTuple(unittest.TestCase):
+
+    def test_factory(self):
+        Point = NamedTuple('Point', 'x y')
+        self.assertEqual(Point.__name__, 'Point')
+        self.assertEqual(Point.__doc__, 'Point(x, y)')
+        self.assertEqual(Point.__slots__, ())
+        self.assertEqual(Point.__module__, __name__)
+        self.assertEqual(Point.__getitem__, tuple.__getitem__)
+        self.assert_('__getitem__' in Point.__dict__)                       # superclass methods localized
+
+    def test_instance(self):
+        Point = NamedTuple('Point', 'x y')
+        p = Point(11, 22)
+        self.assertEqual(p, Point(x=11, y=22))
+        self.assertEqual(p, Point(11, y=22))
+        self.assertEqual(p, Point(y=22, x=11))
+        self.assertEqual(p, Point(*(11, 22)))
+        self.assertEqual(p, Point(**dict(x=11, y=22)))
+        self.assertRaises(TypeError, Point, 1)                              # too few args
+        self.assertRaises(TypeError, Point, 1, 2, 3)                        # too many args
+        self.assertRaises(TypeError, eval, 'Point(XXX=1, y=2)', locals())   # wrong keyword argument
+        self.assertRaises(TypeError, eval, 'Point(x=1)', locals())          # missing keyword argument
+        self.assertEqual(repr(p), 'Point(x=11, y=22)')
+        self.assert_('__dict__' not in dir(p))                              # verify instance has no dict
+        self.assert_('__weakref__' not in dir(p))
+
+    def test_tupleness(self):
+        Point = NamedTuple('Point', 'x y')
+        p = Point(11, 22)
+
+        self.assert_(isinstance(p, tuple))
+        self.assertEqual(p, (11, 22))                                       # matches a real tuple
+        self.assertEqual(tuple(p), (11, 22))                                # coercable to a real tuple
+        self.assertEqual(list(p), [11, 22])                                 # coercable to a list
+        self.assertEqual(max(p), 22)                                        # iterable
+        self.assertEqual(max(*p), 22)                                       # star-able
+        x, y = p
+        self.assertEqual(p, (x, y))                                         # unpacks like a tuple
+        self.assertEqual((p[0], p[1]), (11, 22))                            # indexable like a tuple
+        self.assertRaises(IndexError, p.__getitem__, 3)
+
+        self.assertEqual(p.x, x)
+        self.assertEqual(p.y, y)
+        self.assertRaises(AttributeError, eval, 'p.z', locals())
+
+
+def test_main(verbose=None):
+    test_classes = [TestNamedTuple]
+    test_support.run_unittest(*test_classes)
+
+if __name__ == "__main__":
+    test_main(verbose=True)

Misc/NEWS

@@ -138,6 +138,8 @@ Library
 
 - Added heapq.merge() for merging sorted input streams.
 
+- Added collections.NamedTuple() for assigning field names to tuples.
+
 - Added itertools.izip_longest().
 
 - Have the encoding package's search function dynamically import using absolute