cb73bdac95
Revert r73401 per Raymond Hettinger's request.
...
The rational is the change might cause imcompatiblity problems with
PyYAML. In addition, Raymond wants to kept the different versions of
collections synchronized across Python versions.
2009-06-12 23:03:35 +00:00
450ae573bc
Make pickling of OrderedDict instances more efficient.
2009-06-12 21:52:14 +00:00
a68cad13ae
Fix field name conflicts for named tuples.
2009-05-27 02:24:45 +00:00
7c67b03051
using sys._getframe(x), where x > 0 doesnt' work on IronPython
2009-05-05 00:55:24 +00:00
1355a3dbf6
Minor factoring.
2009-04-08 08:26:55 +00:00
a5cd6375c0
Add docstrings.
2009-04-08 05:39:38 +00:00
e3bc5577e8
Replace the localized min/max calls with normal if/else
2009-04-04 08:46:58 +00:00
756ab67a98
Have namedtuple's field renamer assign names that
...
are consistent with the corresponding tuple index.
2009-04-02 22:25:40 +00:00
2dc90fdfaf
Separate initialization from clearing.
2009-03-25 22:41:32 +00:00
e5b78563b6
Make imported name private and wrap long-line.
2009-03-23 18:26:59 +00:00
906f95e80b
Move initialization of root link to __init__.
2009-03-23 04:42:18 +00:00
f1e2df9780
Add more comments. Improve variable names.
...
Make links clearer by using a Link object
instead of a list. Use proxy links to avoid
circular references.
2009-03-23 00:08:09 +00:00
c8e81ef508
Fix typo
2009-03-19 23:22:25 +00:00
e980d2d44b
* Add implementation notes.
...
* Re-order methods so that those touching the underlying data
structure come first and the derived methods come last.
2009-03-19 23:12:41 +00:00
2412299be9
* Add clearer comment to initialization code.
...
* Add optional argument to popitem() -- modeled
after Anthon van der Neut's C version.
* Fix method markup in docs.
2009-03-19 19:59:58 +00:00
2124599eaa
Improve implementation with better underlying data structure
...
for O(1) deletions. Big-Oh performance now the same as regular
dictionaries. Uses a doubly-linked list instead of a list/seq
to track insertion order.
2009-03-19 15:21:10 +00:00
a61ae6922f
Use mixin methods where possible. (2.7 only -- these don't all exist in 3.0)
2009-03-18 22:13:20 +00:00
131af6505a
Backport 70140, 70141, 70143, and 70144.
...
Adds tests, switches from list to deque, fixes __reduce__
which was unnecessarily copying __keys.
2009-03-03 22:59:25 +00:00
0b15541ae2
Minor simplification.
2009-03-03 21:13:51 +00:00
9353ea2b85
Make the underlying data structure more private.
2009-03-03 20:53:51 +00:00
88a9164cdb
Backport 70106: Add OrderedDict support to collections.namedtuple().
2009-03-03 04:51:24 +00:00
bc512d3abd
Backport PEP 372: OrderedDict()
2009-03-03 04:45:34 +00:00
322daea7c3
Issue 1818: collections.namedtuple() to support automatic renaming of invalid fieldnames.
2009-02-10 01:24:05 +00:00
ecf252abac
Issue 5021: doctest.testfile should set __name__
2009-01-27 10:03:04 +00:00
1bc1c8ac76
Update comments and add an optimized path for Counter.update().
2009-01-22 09:05:43 +00:00
4571f347f9
Simplify explanation of multiset operations by removing restrictions on negative inputs.
2009-01-21 20:31:50 +00:00
ae3f068fb5
Make merging easier by formattng comment blocks the same in Py3.1
2009-01-20 03:36:36 +00:00
562a4ada56
Add Counter() to __all__.
2009-01-20 02:24:38 +00:00
bad1eb2ff3
Build-outs for Counter() class:
...
* Constructor and update() support keyword args (like their dict counterparts).
* The 'del' statement no longer raises KeyError for missing values.
* Add multiset operations: __add__, __sub__, __and__, __or__.
2009-01-20 01:19:26 +00:00
35288c6b6c
Speed-up __repr__. Eliminate duplicate tests. Use a from-irmport.
2009-01-13 04:50:35 +00:00
783d73fc2f
Fixup and simplify docstrings and doctests.
2009-01-13 04:13:53 +00:00
aaa6e630f8
Simplify Counter() API. Replace items keyword argument
...
with a mapping. Makes Counter() idempotent, makes update()
API the same as Counter.__init__(), makes a more readable
repr, makes the API more dict-like, and allows Steven
Bethard's update() example to work.
2009-01-13 01:05:03 +00:00
f94d7fa5fb
Issue 1696199: Add collections.Counter().
2009-01-12 22:58:41 +00:00
6ee7bc04f7
Fix namedtuple bug reported by Glenn Linderman. Template did not form correctly if the field names were input in Unicode.
2008-09-25 23:31:52 +00:00
e98839a1f4
Issue3065: Fixed pickling of named tuples. Added tests.
2008-06-09 01:28:30 +00:00
3c2523c2eb
Issue 2903: Add __name__ in globals for namedtuple namespace.
2008-05-30 07:16:53 +00:00
15b5e55b48
Neaten-up the named tuple docs
2008-01-10 23:00:01 +00:00
e850c466c7
Clarify how to add a field to a named tuple.
2008-01-10 20:37:12 +00:00
e1655088ca
Examples for named tuple subclassing should include __slots__
2008-01-10 19:15:10 +00:00
581671419f
Expand comment.
2008-01-08 02:02:05 +00:00
f5e8af1bb7
Use get() instead of pop() for the optimized version of _replace().
2008-01-07 20:56:05 +00:00
9a359210aa
Cleanup named tuple subclassing example.
2008-01-07 20:07:38 +00:00
dc55f35f38
Add another named tuple subclassing example.
2008-01-07 09:03:49 +00:00
1db6f80cd5
Cleanup subclassing example to more clearly show fixed-width print format.
2008-01-07 05:50:35 +00:00
b8e0072fec
Add subclassing example to docs for named tuples.
2008-01-07 04:24:49 +00:00
844f71b7e4
Speed-up named tuple's _make() constructor.
2008-01-06 22:11:54 +00:00
1166872006
Small code simplification. Forgot that classmethods can be called from intances.
2008-01-06 09:02:24 +00:00
1b50fd7cb3
Add error-checking to namedtuple's _replace() method.
2008-01-05 02:17:24 +00:00
02740f73ff
Improve namedtuple's _cast() method with a docstring, new name, and error-checking.
2008-01-05 01:35:43 +00:00
e0734e7dc0
Minor fix-ups to named tuples:
...
* Make the _replace() method respect subclassing.
* Using property() to make _fields read-only wasn't a good idea.
It caused len(Point._fields) to fail.
* Add note to _cast() about length checking and alternative with the star-operator.
2008-01-04 03:22:53 +00:00
Alexandre Vassalotti
Raymond Hettinger
Benjamin Peterson