Improve and accept PEP 424 (__length_hint__).

pep-0424.txt

@@ -3,7 +3,7 @@ Title: A method for exposing a length hint
 Version: $Revision$
 Last-Modified: $Date
 Author: Alex Gaynor <alex.gaynor@gmail.com>
-Status: Draft
+Status: Accepted
 Type: Standards Track
 Content-Type: text/x-rst
 Created: 14-July-2012
@@ -13,18 +13,19 @@ Post-History: http://mail.python.org/pipermail/python-dev/2012-July/120920.html
 Abstract
 ========
 
-CPython currently defines an ``__length_hint__`` method on several types, such
+CPython currently defines a ``__length_hint__`` method on several
-as various iterators. This method is then used by various other functions (such
+types, such as various iterators. This method is then used by various
-as ``list``) to presize lists based on the estimated returned by
+other functions (such as ``list``) to presize lists based on the
-``__length_hint__``. Types can then define ``__length_hint__`` which are not
+estimate returned by ``__length_hint__``. Types which are not sized,
-sized, and thus should not define ``__len__``, but can estimate or compute a
+and thus should not define ``__len__``, can then define
-size (such as many iterators).
+``__length_hint__``, to allow estimating or computing a size (such as
+many iterators).
 
-Proposal
+Specification
-========
+=============
 
-This PEP proposes formally documenting ``__length_hint__`` for other
+This PEP formally documents ``__length_hint__`` for other
-interpreter and non-standard library Python to implement.
+interpreters and non-standard-library Python modules to implement.
 
 ``__length_hint__`` must return an integer (else a TypeError is raised) or
 ``NotImplemented, and is not required to be accurate. It may return a value
@@ -32,21 +33,23 @@ that is either larger or smaller than the actual size of the container. A
 return value of ``NotImplemented`` indicates that there is no finite length
 estimate. It may not return a negative value (else a ValueError is raised).
 
-In addition, a new function ``operator.length`` hint is added, having the
+In addition, a new function ``operator.length_hint`` hint is added,
-follow semantics (which define how ``__length_hint__`` should be used::
+with the follow semantics (which define how ``__length_hint__`` should
+be used)::
 
-    def length_hint(obj, default):
+    def length_hint(obj, default=0):
-        """
+        """Return an estimate of the number of items in obj.
-        Return an estimate of the number of items in obj. This is
-        useful for presizing containers when building from an iterable.
 
-        If the object supports len(), the result will be exact. Otherwise, it
+        This is useful for presizing containers when building from an
-        may over or underestimate by an arbitrary amount. The result will be an
+        iterable.
-        integer >= 0.
+
+        If the object supports len(), the result will be
+        exact. Otherwise, it may over- or under-estimate by an
+        arbitrary amount. The result will be an integer >= 0.
         """
-        try:
+        if <obj has a __len__ method>:
             return len(obj)
-        except TypeError:
+        else:
             try:
                 get_hint = obj.__length_hint__
             except AttributeError:
@@ -55,13 +58,18 @@ follow semantics (which define how ``__length_hint__`` should be used::
             if hint is NotImplemented:
                 return default
             if not isinstance(hint, int):
-                raise TypeError("Length hint must be an integer, not %r" % type(hint))
+                raise TypeError("Length hint must be an integer, not %r" %
-            if hint < 0:
+                                type(hint))
-                raise ValueError("Length hint (%r) must be >= 0" % hint)
+            return max(hint, 0)
-            return hint
 
-Callers are required to provide a default value, because there is no sane
+Note: there is no good way to spell "obj has a __len__ method" in pure
-return value for objects which do not provide a length or length hint.
+Python. In CPython, this comes down to checking for a ``sq_length``
+slot. Other implementations presumably have their own way of
+checking. Calling ``len(obj)`` and catching TypeError is not quite
+correct (as it would assume no __len__ method exists when in fact one
+exists but calling it raises TypeError); checking ``hasattr(obj,
+'__len__')`` likewise is incorrect if obj is a class defining a
+``__len__`` method for its instances.
 
 
 Rationale