Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 1 | __all__ = ['Counter', 'deque', 'defaultdict', 'namedtuple', 'OrderedDict'] |
Raymond Hettinger | 88880b2 | 2007-12-18 00:13:45 +0000 | [diff] [blame] | 2 | # For bootstrapping reasons, the collection ABCs are defined in _abcoll.py. |
| 3 | # They should however be considered an integral part of collections.py. |
| 4 | from _abcoll import * |
| 5 | import _abcoll |
| 6 | __all__ += _abcoll.__all__ |
Raymond Hettinger | eb97988 | 2007-02-28 18:37:52 +0000 | [diff] [blame] | 7 | |
| 8 | from _collections import deque, defaultdict |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 9 | from operator import itemgetter as _itemgetter, eq as _eq |
Raymond Hettinger | abfd8df | 2007-10-16 21:28:32 +0000 | [diff] [blame] | 10 | from keyword import iskeyword as _iskeyword |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 11 | import sys as _sys |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 12 | import heapq as _heapq |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 13 | from itertools import repeat as _repeat, chain as _chain, starmap as _starmap, \ |
Georg Brandl | 7d4b759 | 2010-02-06 22:49:47 +0000 | [diff] [blame] | 14 | ifilter as _ifilter, imap as _imap |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 15 | |
| 16 | ################################################################################ |
| 17 | ### OrderedDict |
| 18 | ################################################################################ |
| 19 | |
| 20 | class OrderedDict(dict, MutableMapping): |
Raymond Hettinger | e980d2d | 2009-03-19 23:12:41 +0000 | [diff] [blame] | 21 | 'Dictionary that remembers insertion order' |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 22 | # An inherited dict maps keys to values. |
Raymond Hettinger | e980d2d | 2009-03-19 23:12:41 +0000 | [diff] [blame] | 23 | # The inherited dict provides __getitem__, __len__, __contains__, and get. |
| 24 | # The remaining methods are order-aware. |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 25 | # Big-O running times for all methods are the same as for regular dictionaries. |
| 26 | |
| 27 | # The internal self.__map dictionary maps keys to links in a doubly linked list. |
| 28 | # The circular doubly linked list starts and ends with a sentinel element. |
| 29 | # The sentinel element never gets deleted (this simplifies the algorithm). |
Raymond Hettinger | aba2293 | 2010-03-09 09:58:53 +0000 | [diff] [blame] | 30 | # Each link is stored as a list of length three: [PREV, NEXT, KEY]. |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 31 | |
| 32 | def __init__(self, *args, **kwds): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 33 | '''Initialize an ordered dictionary. Signature is the same as for |
| 34 | regular dictionaries, but keyword arguments are not recommended |
| 35 | because their insertion order is arbitrary. |
| 36 | |
| 37 | ''' |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 38 | if len(args) > 1: |
| 39 | raise TypeError('expected at most 1 arguments, got %d' % len(args)) |
Raymond Hettinger | 9353ea2 | 2009-03-03 20:53:51 +0000 | [diff] [blame] | 40 | try: |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 41 | self.__root |
Raymond Hettinger | 9353ea2 | 2009-03-03 20:53:51 +0000 | [diff] [blame] | 42 | except AttributeError: |
Raymond Hettinger | aba2293 | 2010-03-09 09:58:53 +0000 | [diff] [blame] | 43 | self.__root = root = [None, None, None] # sentinel node |
Raymond Hettinger | e30bc38 | 2010-03-09 11:29:10 +0000 | [diff] [blame] | 44 | PREV = 0 |
| 45 | NEXT = 1 |
Raymond Hettinger | aba2293 | 2010-03-09 09:58:53 +0000 | [diff] [blame] | 46 | root[PREV] = root[NEXT] = root |
Raymond Hettinger | 2dc90fd | 2009-03-25 22:41:32 +0000 | [diff] [blame] | 47 | self.__map = {} |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 48 | self.update(*args, **kwds) |
| 49 | |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 50 | def __setitem__(self, key, value): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 51 | 'od.__setitem__(i, y) <==> od[i]=y' |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 52 | # Setting a new item creates a new link which goes at the end of the linked |
| 53 | # list, and the inherited dictionary is updated with the new key/value pair. |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 54 | if key not in self: |
Raymond Hettinger | e30bc38 | 2010-03-09 11:29:10 +0000 | [diff] [blame] | 55 | PREV = 0 |
| 56 | NEXT = 1 |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 57 | root = self.__root |
Raymond Hettinger | aba2293 | 2010-03-09 09:58:53 +0000 | [diff] [blame] | 58 | last = root[PREV] |
| 59 | last[NEXT] = root[PREV] = self.__map[key] = [last, root, key] |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 60 | dict.__setitem__(self, key, value) |
| 61 | |
| 62 | def __delitem__(self, key): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 63 | 'od.__delitem__(y) <==> del od[y]' |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 64 | # Deleting an existing item uses self.__map to find the link which is |
| 65 | # then removed by updating the links in the predecessor and successor nodes. |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 66 | dict.__delitem__(self, key) |
Raymond Hettinger | e30bc38 | 2010-03-09 11:29:10 +0000 | [diff] [blame] | 67 | PREV = 0 |
| 68 | NEXT = 1 |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 69 | link = self.__map.pop(key) |
Raymond Hettinger | 3928276 | 2010-04-03 00:39:26 +0000 | [diff] [blame] | 70 | link_prev = link[PREV] |
| 71 | link_next = link[NEXT] |
| 72 | link_prev[NEXT] = link_next |
| 73 | link_next[PREV] = link_prev |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 74 | |
| 75 | def __iter__(self): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 76 | 'od.__iter__() <==> iter(od)' |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 77 | # Traverse the linked list in order. |
Raymond Hettinger | e30bc38 | 2010-03-09 11:29:10 +0000 | [diff] [blame] | 78 | NEXT = 1 |
| 79 | KEY = 2 |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 80 | root = self.__root |
Raymond Hettinger | aba2293 | 2010-03-09 09:58:53 +0000 | [diff] [blame] | 81 | curr = root[NEXT] |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 82 | while curr is not root: |
Raymond Hettinger | aba2293 | 2010-03-09 09:58:53 +0000 | [diff] [blame] | 83 | yield curr[KEY] |
| 84 | curr = curr[NEXT] |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 85 | |
| 86 | def __reversed__(self): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 87 | 'od.__reversed__() <==> reversed(od)' |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 88 | # Traverse the linked list in reverse order. |
Raymond Hettinger | e30bc38 | 2010-03-09 11:29:10 +0000 | [diff] [blame] | 89 | PREV = 0 |
| 90 | KEY = 2 |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 91 | root = self.__root |
Raymond Hettinger | aba2293 | 2010-03-09 09:58:53 +0000 | [diff] [blame] | 92 | curr = root[PREV] |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 93 | while curr is not root: |
Raymond Hettinger | aba2293 | 2010-03-09 09:58:53 +0000 | [diff] [blame] | 94 | yield curr[KEY] |
| 95 | curr = curr[PREV] |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 96 | |
| 97 | def __reduce__(self): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 98 | 'Return state information for pickling' |
Alexandre Vassalotti | cb73bda | 2009-06-12 23:03:35 +0000 | [diff] [blame] | 99 | items = [[k, self[k]] for k in self] |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 100 | tmp = self.__map, self.__root |
| 101 | del self.__map, self.__root |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 102 | inst_dict = vars(self).copy() |
Raymond Hettinger | f1e2df9 | 2009-03-23 00:08:09 +0000 | [diff] [blame] | 103 | self.__map, self.__root = tmp |
Alexandre Vassalotti | cb73bda | 2009-06-12 23:03:35 +0000 | [diff] [blame] | 104 | if inst_dict: |
| 105 | return (self.__class__, (items,), inst_dict) |
| 106 | return self.__class__, (items,) |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 107 | |
Raymond Hettinger | 3928276 | 2010-04-03 00:39:26 +0000 | [diff] [blame] | 108 | def clear(self): |
| 109 | 'od.clear() -> None. Remove all items from od.' |
Raymond Hettinger | 6b96ecb | 2010-04-03 03:14:28 +0000 | [diff] [blame^] | 110 | try: |
| 111 | for node in self.__map.itervalues(): |
| 112 | del node[:] |
| 113 | self.__root[:] = [self.__root, self.__root, None] |
| 114 | self.__map.clear() |
| 115 | except AttributeError: |
| 116 | pass |
| 117 | dict.clear(self) |
Raymond Hettinger | 3928276 | 2010-04-03 00:39:26 +0000 | [diff] [blame] | 118 | |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 119 | setdefault = MutableMapping.setdefault |
| 120 | update = MutableMapping.update |
| 121 | pop = MutableMapping.pop |
Raymond Hettinger | a61ae69 | 2009-03-18 22:13:20 +0000 | [diff] [blame] | 122 | keys = MutableMapping.keys |
| 123 | values = MutableMapping.values |
| 124 | items = MutableMapping.items |
| 125 | iterkeys = MutableMapping.iterkeys |
| 126 | itervalues = MutableMapping.itervalues |
| 127 | iteritems = MutableMapping.iteritems |
| 128 | __ne__ = MutableMapping.__ne__ |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 129 | |
Raymond Hettinger | e980d2d | 2009-03-19 23:12:41 +0000 | [diff] [blame] | 130 | def popitem(self, last=True): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 131 | '''od.popitem() -> (k, v), return and remove a (key, value) pair. |
| 132 | Pairs are returned in LIFO order if last is true or FIFO order if false. |
| 133 | |
| 134 | ''' |
Raymond Hettinger | e980d2d | 2009-03-19 23:12:41 +0000 | [diff] [blame] | 135 | if not self: |
| 136 | raise KeyError('dictionary is empty') |
Raymond Hettinger | 1355a3d | 2009-04-08 08:26:55 +0000 | [diff] [blame] | 137 | key = next(reversed(self) if last else iter(self)) |
Raymond Hettinger | e980d2d | 2009-03-19 23:12:41 +0000 | [diff] [blame] | 138 | value = self.pop(key) |
| 139 | return key, value |
| 140 | |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 141 | def __repr__(self): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 142 | 'od.__repr__() <==> repr(od)' |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 143 | if not self: |
| 144 | return '%s()' % (self.__class__.__name__,) |
Raymond Hettinger | 0b15541 | 2009-03-03 21:13:51 +0000 | [diff] [blame] | 145 | return '%s(%r)' % (self.__class__.__name__, self.items()) |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 146 | |
| 147 | def copy(self): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 148 | 'od.copy() -> a shallow copy of od' |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 149 | return self.__class__(self) |
| 150 | |
| 151 | @classmethod |
| 152 | def fromkeys(cls, iterable, value=None): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 153 | '''OD.fromkeys(S[, v]) -> New ordered dictionary with keys from S |
| 154 | and values equal to v (which defaults to None). |
| 155 | |
| 156 | ''' |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 157 | d = cls() |
| 158 | for key in iterable: |
| 159 | d[key] = value |
| 160 | return d |
| 161 | |
| 162 | def __eq__(self, other): |
Raymond Hettinger | a5cd637 | 2009-04-08 05:39:38 +0000 | [diff] [blame] | 163 | '''od.__eq__(y) <==> od==y. Comparison to another OD is order-sensitive |
| 164 | while comparison to a regular mapping is order-insensitive. |
| 165 | |
| 166 | ''' |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 167 | if isinstance(other, OrderedDict): |
Raymond Hettinger | e5b7856 | 2009-03-23 18:26:59 +0000 | [diff] [blame] | 168 | return len(self)==len(other) and \ |
| 169 | all(_imap(_eq, self.iteritems(), other.iteritems())) |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 170 | return dict.__eq__(self, other) |
| 171 | |
Raymond Hettinger | 3928276 | 2010-04-03 00:39:26 +0000 | [diff] [blame] | 172 | def __del__(self): |
| 173 | self.clear() # eliminate cyclical references |
Raymond Hettinger | bc512d3 | 2009-03-03 04:45:34 +0000 | [diff] [blame] | 174 | |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 175 | |
Raymond Hettinger | ae3f068 | 2009-01-20 03:36:36 +0000 | [diff] [blame] | 176 | ################################################################################ |
| 177 | ### namedtuple |
| 178 | ################################################################################ |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 179 | |
Raymond Hettinger | 322daea | 2009-02-10 01:24:05 +0000 | [diff] [blame] | 180 | def namedtuple(typename, field_names, verbose=False, rename=False): |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 181 | """Returns a new subclass of tuple with named fields. |
| 182 | |
Raymond Hettinger | 01a0957 | 2007-10-23 20:37:41 +0000 | [diff] [blame] | 183 | >>> Point = namedtuple('Point', 'x y') |
Raymond Hettinger | d36a60e | 2007-09-17 00:55:00 +0000 | [diff] [blame] | 184 | >>> Point.__doc__ # docstring for the new class |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 185 | 'Point(x, y)' |
Raymond Hettinger | d36a60e | 2007-09-17 00:55:00 +0000 | [diff] [blame] | 186 | >>> p = Point(11, y=22) # instantiate with positional args or keywords |
Raymond Hettinger | 8777bca | 2007-12-18 22:21:27 +0000 | [diff] [blame] | 187 | >>> p[0] + p[1] # indexable like a plain tuple |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 188 | 33 |
Raymond Hettinger | 88880b2 | 2007-12-18 00:13:45 +0000 | [diff] [blame] | 189 | >>> x, y = p # unpack like a regular tuple |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 190 | >>> x, y |
| 191 | (11, 22) |
Raymond Hettinger | d36a60e | 2007-09-17 00:55:00 +0000 | [diff] [blame] | 192 | >>> p.x + p.y # fields also accessable by name |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 193 | 33 |
Raymond Hettinger | 42da874 | 2007-12-14 02:49:47 +0000 | [diff] [blame] | 194 | >>> d = p._asdict() # convert to a dictionary |
Raymond Hettinger | a7fc4b1 | 2007-10-05 02:47:07 +0000 | [diff] [blame] | 195 | >>> d['x'] |
| 196 | 11 |
| 197 | >>> Point(**d) # convert from a dictionary |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 198 | Point(x=11, y=22) |
Raymond Hettinger | 42da874 | 2007-12-14 02:49:47 +0000 | [diff] [blame] | 199 | >>> p._replace(x=100) # _replace() is like str.replace() but targets named fields |
Raymond Hettinger | d36a60e | 2007-09-17 00:55:00 +0000 | [diff] [blame] | 200 | Point(x=100, y=22) |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 201 | |
| 202 | """ |
| 203 | |
Raymond Hettinger | 5816714 | 2008-01-08 02:02:05 +0000 | [diff] [blame] | 204 | # Parse and validate the field names. Validation serves two purposes, |
| 205 | # generating informative error messages and preventing template injection attacks. |
Raymond Hettinger | 2115bbc | 2007-10-08 09:14:28 +0000 | [diff] [blame] | 206 | if isinstance(field_names, basestring): |
Raymond Hettinger | 0e1d606 | 2007-10-08 10:11:51 +0000 | [diff] [blame] | 207 | field_names = field_names.replace(',', ' ').split() # names separated by whitespace and/or commas |
Raymond Hettinger | 6ee7bc0 | 2008-09-25 23:31:52 +0000 | [diff] [blame] | 208 | field_names = tuple(map(str, field_names)) |
Raymond Hettinger | 322daea | 2009-02-10 01:24:05 +0000 | [diff] [blame] | 209 | if rename: |
| 210 | names = list(field_names) |
| 211 | seen = set() |
| 212 | for i, name in enumerate(names): |
| 213 | if (not all(c.isalnum() or c=='_' for c in name) or _iskeyword(name) |
| 214 | or not name or name[0].isdigit() or name.startswith('_') |
| 215 | or name in seen): |
Raymond Hettinger | 756ab67 | 2009-04-02 22:25:40 +0000 | [diff] [blame] | 216 | names[i] = '_%d' % i |
Raymond Hettinger | 322daea | 2009-02-10 01:24:05 +0000 | [diff] [blame] | 217 | seen.add(name) |
| 218 | field_names = tuple(names) |
Raymond Hettinger | 050afbf | 2007-10-16 19:18:30 +0000 | [diff] [blame] | 219 | for name in (typename,) + field_names: |
Raymond Hettinger | 2e1af25 | 2007-12-05 18:11:08 +0000 | [diff] [blame] | 220 | if not all(c.isalnum() or c=='_' for c in name): |
Raymond Hettinger | 050afbf | 2007-10-16 19:18:30 +0000 | [diff] [blame] | 221 | raise ValueError('Type names and field names can only contain alphanumeric characters and underscores: %r' % name) |
Raymond Hettinger | abfd8df | 2007-10-16 21:28:32 +0000 | [diff] [blame] | 222 | if _iskeyword(name): |
| 223 | raise ValueError('Type names and field names cannot be a keyword: %r' % name) |
Raymond Hettinger | 050afbf | 2007-10-16 19:18:30 +0000 | [diff] [blame] | 224 | if name[0].isdigit(): |
| 225 | raise ValueError('Type names and field names cannot start with a number: %r' % name) |
Raymond Hettinger | 163f622 | 2007-10-09 01:36:23 +0000 | [diff] [blame] | 226 | seen_names = set() |
| 227 | for name in field_names: |
Raymond Hettinger | 322daea | 2009-02-10 01:24:05 +0000 | [diff] [blame] | 228 | if name.startswith('_') and not rename: |
Raymond Hettinger | 42da874 | 2007-12-14 02:49:47 +0000 | [diff] [blame] | 229 | raise ValueError('Field names cannot start with an underscore: %r' % name) |
Raymond Hettinger | 163f622 | 2007-10-09 01:36:23 +0000 | [diff] [blame] | 230 | if name in seen_names: |
Raymond Hettinger | 050afbf | 2007-10-16 19:18:30 +0000 | [diff] [blame] | 231 | raise ValueError('Encountered duplicate field name: %r' % name) |
Raymond Hettinger | 163f622 | 2007-10-09 01:36:23 +0000 | [diff] [blame] | 232 | seen_names.add(name) |
Raymond Hettinger | 2115bbc | 2007-10-08 09:14:28 +0000 | [diff] [blame] | 233 | |
| 234 | # Create and fill-in the class template |
Raymond Hettinger | 02740f7 | 2008-01-05 01:35:43 +0000 | [diff] [blame] | 235 | numfields = len(field_names) |
Raymond Hettinger | 2b03d45 | 2007-09-18 03:33:19 +0000 | [diff] [blame] | 236 | argtxt = repr(field_names).replace("'", "")[1:-1] # tuple repr without parens or quotes |
Raymond Hettinger | 5a41daf | 2007-05-19 01:11:16 +0000 | [diff] [blame] | 237 | reprtxt = ', '.join('%s=%%r' % name for name in field_names) |
| 238 | template = '''class %(typename)s(tuple): |
Raymond Hettinger | 48eca67 | 2007-12-14 18:08:20 +0000 | [diff] [blame] | 239 | '%(typename)s(%(argtxt)s)' \n |
| 240 | __slots__ = () \n |
Raymond Hettinger | e0734e7 | 2008-01-04 03:22:53 +0000 | [diff] [blame] | 241 | _fields = %(field_names)r \n |
Raymond Hettinger | a68cad1 | 2009-05-27 02:24:45 +0000 | [diff] [blame] | 242 | def __new__(_cls, %(argtxt)s): |
Raymond Hettinger | 9bd3508 | 2010-03-09 09:01:46 +0000 | [diff] [blame] | 243 | 'Create new instance of %(typename)s(%(argtxt)s)' |
Raymond Hettinger | a68cad1 | 2009-05-27 02:24:45 +0000 | [diff] [blame] | 244 | return _tuple.__new__(_cls, (%(argtxt)s)) \n |
Raymond Hettinger | 02740f7 | 2008-01-05 01:35:43 +0000 | [diff] [blame] | 245 | @classmethod |
Raymond Hettinger | 844f71b | 2008-01-06 22:11:54 +0000 | [diff] [blame] | 246 | def _make(cls, iterable, new=tuple.__new__, len=len): |
Raymond Hettinger | 02740f7 | 2008-01-05 01:35:43 +0000 | [diff] [blame] | 247 | 'Make a new %(typename)s object from a sequence or iterable' |
Raymond Hettinger | 844f71b | 2008-01-06 22:11:54 +0000 | [diff] [blame] | 248 | result = new(cls, iterable) |
Raymond Hettinger | 02740f7 | 2008-01-05 01:35:43 +0000 | [diff] [blame] | 249 | if len(result) != %(numfields)d: |
| 250 | raise TypeError('Expected %(numfields)d arguments, got %%d' %% len(result)) |
| 251 | return result \n |
Raymond Hettinger | 5a41daf | 2007-05-19 01:11:16 +0000 | [diff] [blame] | 252 | def __repr__(self): |
Raymond Hettinger | 9bd3508 | 2010-03-09 09:01:46 +0000 | [diff] [blame] | 253 | 'Return a nicely formatted representation string' |
Raymond Hettinger | 48eca67 | 2007-12-14 18:08:20 +0000 | [diff] [blame] | 254 | return '%(typename)s(%(reprtxt)s)' %% self \n |
Raymond Hettinger | 88a9164 | 2009-03-03 04:51:24 +0000 | [diff] [blame] | 255 | def _asdict(self): |
| 256 | 'Return a new OrderedDict which maps field names to their values' |
| 257 | return OrderedDict(zip(self._fields, self)) \n |
Raymond Hettinger | a68cad1 | 2009-05-27 02:24:45 +0000 | [diff] [blame] | 258 | def _replace(_self, **kwds): |
Raymond Hettinger | eeeb9c4 | 2007-11-15 02:44:53 +0000 | [diff] [blame] | 259 | 'Return a new %(typename)s object replacing specified fields with new values' |
Raymond Hettinger | a68cad1 | 2009-05-27 02:24:45 +0000 | [diff] [blame] | 260 | result = _self._make(map(kwds.pop, %(field_names)r, _self)) |
Raymond Hettinger | 1b50fd7 | 2008-01-05 02:17:24 +0000 | [diff] [blame] | 261 | if kwds: |
| 262 | raise ValueError('Got unexpected field names: %%r' %% kwds.keys()) |
Raymond Hettinger | e98839a | 2008-06-09 01:28:30 +0000 | [diff] [blame] | 263 | return result \n |
| 264 | def __getnewargs__(self): |
Raymond Hettinger | 9bd3508 | 2010-03-09 09:01:46 +0000 | [diff] [blame] | 265 | 'Return self as a plain tuple. Used by copy and pickle.' |
Raymond Hettinger | e98839a | 2008-06-09 01:28:30 +0000 | [diff] [blame] | 266 | return tuple(self) \n\n''' % locals() |
Raymond Hettinger | 5a41daf | 2007-05-19 01:11:16 +0000 | [diff] [blame] | 267 | for i, name in enumerate(field_names): |
Raymond Hettinger | 9bd3508 | 2010-03-09 09:01:46 +0000 | [diff] [blame] | 268 | template += " %s = _property(_itemgetter(%d), doc='Alias for field number %d')\n" % (name, i, i) |
Raymond Hettinger | 2b03d45 | 2007-09-18 03:33:19 +0000 | [diff] [blame] | 269 | if verbose: |
| 270 | print template |
Raymond Hettinger | 2115bbc | 2007-10-08 09:14:28 +0000 | [diff] [blame] | 271 | |
Raymond Hettinger | 3c2523c | 2008-05-30 07:16:53 +0000 | [diff] [blame] | 272 | # Execute the template string in a temporary namespace and |
| 273 | # support tracing utilities by setting a value for frame.f_globals['__name__'] |
Raymond Hettinger | a68cad1 | 2009-05-27 02:24:45 +0000 | [diff] [blame] | 274 | namespace = dict(_itemgetter=_itemgetter, __name__='namedtuple_%s' % typename, |
| 275 | OrderedDict=OrderedDict, _property=property, _tuple=tuple) |
Raymond Hettinger | 2115bbc | 2007-10-08 09:14:28 +0000 | [diff] [blame] | 276 | try: |
Raymond Hettinger | 0e1d606 | 2007-10-08 10:11:51 +0000 | [diff] [blame] | 277 | exec template in namespace |
Raymond Hettinger | 2115bbc | 2007-10-08 09:14:28 +0000 | [diff] [blame] | 278 | except SyntaxError, e: |
| 279 | raise SyntaxError(e.message + ':\n' + template) |
Raymond Hettinger | 0e1d606 | 2007-10-08 10:11:51 +0000 | [diff] [blame] | 280 | result = namespace[typename] |
Raymond Hettinger | 2115bbc | 2007-10-08 09:14:28 +0000 | [diff] [blame] | 281 | |
| 282 | # For pickling to work, the __module__ variable needs to be set to the frame |
| 283 | # where the named tuple is created. Bypass this step in enviroments where |
Benjamin Peterson | 7c67b03 | 2009-05-05 00:55:24 +0000 | [diff] [blame] | 284 | # sys._getframe is not defined (Jython for example) or sys._getframe is not |
| 285 | # defined for arguments greater than 0 (IronPython). |
| 286 | try: |
Raymond Hettinger | ecf252a | 2009-01-27 10:03:04 +0000 | [diff] [blame] | 287 | result.__module__ = _sys._getframe(1).f_globals.get('__name__', '__main__') |
Benjamin Peterson | 7c67b03 | 2009-05-05 00:55:24 +0000 | [diff] [blame] | 288 | except (AttributeError, ValueError): |
| 289 | pass |
Raymond Hettinger | 2115bbc | 2007-10-08 09:14:28 +0000 | [diff] [blame] | 290 | |
Raymond Hettinger | 5a41daf | 2007-05-19 01:11:16 +0000 | [diff] [blame] | 291 | return result |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 292 | |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 293 | |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 294 | ######################################################################## |
Raymond Hettinger | ae3f068 | 2009-01-20 03:36:36 +0000 | [diff] [blame] | 295 | ### Counter |
| 296 | ######################################################################## |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 297 | |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 298 | class Counter(dict): |
| 299 | '''Dict subclass for counting hashable items. Sometimes called a bag |
| 300 | or multiset. Elements are stored as dictionary keys and their counts |
| 301 | are stored as dictionary values. |
| 302 | |
| 303 | >>> c = Counter('abracadabra') # count elements from a string |
| 304 | |
| 305 | >>> c.most_common(3) # three most common elements |
| 306 | [('a', 5), ('r', 2), ('b', 2)] |
| 307 | >>> sorted(c) # list all unique elements |
| 308 | ['a', 'b', 'c', 'd', 'r'] |
| 309 | >>> ''.join(sorted(c.elements())) # list elements with repetitions |
| 310 | 'aaaaabbcdrr' |
| 311 | >>> sum(c.values()) # total of all counts |
| 312 | 11 |
| 313 | |
| 314 | >>> c['a'] # count of letter 'a' |
| 315 | 5 |
| 316 | >>> for elem in 'shazam': # update counts from an iterable |
| 317 | ... c[elem] += 1 # by adding 1 to each element's count |
| 318 | >>> c['a'] # now there are seven 'a' |
| 319 | 7 |
| 320 | >>> del c['r'] # remove all 'r' |
| 321 | >>> c['r'] # now there are zero 'r' |
| 322 | 0 |
| 323 | |
| 324 | >>> d = Counter('simsalabim') # make another counter |
| 325 | >>> c.update(d) # add in the second counter |
| 326 | >>> c['a'] # now there are nine 'a' |
| 327 | 9 |
| 328 | |
| 329 | >>> c.clear() # empty the counter |
| 330 | >>> c |
| 331 | Counter() |
| 332 | |
| 333 | Note: If a count is set to zero or reduced to zero, it will remain |
| 334 | in the counter until the entry is deleted or the counter is cleared: |
| 335 | |
| 336 | >>> c = Counter('aaabbc') |
| 337 | >>> c['b'] -= 2 # reduce the count of 'b' by two |
| 338 | >>> c.most_common() # 'b' is still in, but its count is zero |
| 339 | [('a', 3), ('c', 1), ('b', 0)] |
| 340 | |
| 341 | ''' |
| 342 | # References: |
| 343 | # http://en.wikipedia.org/wiki/Multiset |
| 344 | # http://www.gnu.org/software/smalltalk/manual-base/html_node/Bag.html |
| 345 | # http://www.demo2s.com/Tutorial/Cpp/0380__set-multiset/Catalog0380__set-multiset.htm |
| 346 | # http://code.activestate.com/recipes/259174/ |
| 347 | # Knuth, TAOCP Vol. II section 4.6.3 |
| 348 | |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 349 | def __init__(self, iterable=None, **kwds): |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 350 | '''Create a new, empty Counter object. And if given, count elements |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 351 | from an input iterable. Or, initialize the count from another mapping |
| 352 | of elements to their counts. |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 353 | |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 354 | >>> c = Counter() # a new, empty counter |
Raymond Hettinger | 783d73f | 2009-01-13 04:13:53 +0000 | [diff] [blame] | 355 | >>> c = Counter('gallahad') # a new counter from an iterable |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 356 | >>> c = Counter({'a': 4, 'b': 2}) # a new counter from a mapping |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 357 | >>> c = Counter(a=4, b=2) # a new counter from keyword args |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 358 | |
| 359 | ''' |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 360 | self.update(iterable, **kwds) |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 361 | |
| 362 | def __missing__(self, key): |
| 363 | 'The count of elements not in the Counter is zero.' |
| 364 | # Needed so that self[missing_item] does not raise KeyError |
| 365 | return 0 |
| 366 | |
| 367 | def most_common(self, n=None): |
| 368 | '''List the n most common elements and their counts from the most |
| 369 | common to the least. If n is None, then list all element counts. |
| 370 | |
| 371 | >>> Counter('abracadabra').most_common(3) |
| 372 | [('a', 5), ('r', 2), ('b', 2)] |
| 373 | |
| 374 | ''' |
| 375 | # Emulate Bag.sortedByCount from Smalltalk |
| 376 | if n is None: |
| 377 | return sorted(self.iteritems(), key=_itemgetter(1), reverse=True) |
| 378 | return _heapq.nlargest(n, self.iteritems(), key=_itemgetter(1)) |
| 379 | |
| 380 | def elements(self): |
| 381 | '''Iterator over elements repeating each as many times as its count. |
| 382 | |
| 383 | >>> c = Counter('ABCABC') |
| 384 | >>> sorted(c.elements()) |
| 385 | ['A', 'A', 'B', 'B', 'C', 'C'] |
| 386 | |
Raymond Hettinger | 783d73f | 2009-01-13 04:13:53 +0000 | [diff] [blame] | 387 | # Knuth's example for prime factors of 1836: 2**2 * 3**3 * 17**1 |
| 388 | >>> prime_factors = Counter({2: 2, 3: 3, 17: 1}) |
| 389 | >>> product = 1 |
| 390 | >>> for factor in prime_factors.elements(): # loop over factors |
| 391 | ... product *= factor # and multiply them |
| 392 | >>> product |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 393 | 1836 |
| 394 | |
Raymond Hettinger | 783d73f | 2009-01-13 04:13:53 +0000 | [diff] [blame] | 395 | Note, if an element's count has been set to zero or is a negative |
| 396 | number, elements() will ignore it. |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 397 | |
| 398 | ''' |
| 399 | # Emulate Bag.do from Smalltalk and Multiset.begin from C++. |
Raymond Hettinger | 35288c6 | 2009-01-13 04:50:35 +0000 | [diff] [blame] | 400 | return _chain.from_iterable(_starmap(_repeat, self.iteritems())) |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 401 | |
| 402 | # Override dict methods where necessary |
| 403 | |
| 404 | @classmethod |
| 405 | def fromkeys(cls, iterable, v=None): |
| 406 | # There is no equivalent method for counters because setting v=1 |
| 407 | # means that no element can have a count greater than one. |
| 408 | raise NotImplementedError( |
| 409 | 'Counter.fromkeys() is undefined. Use Counter(iterable) instead.') |
| 410 | |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 411 | def update(self, iterable=None, **kwds): |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 412 | '''Like dict.update() but add counts instead of replacing them. |
| 413 | |
Raymond Hettinger | 783d73f | 2009-01-13 04:13:53 +0000 | [diff] [blame] | 414 | Source can be an iterable, a dictionary, or another Counter instance. |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 415 | |
| 416 | >>> c = Counter('which') |
Raymond Hettinger | 783d73f | 2009-01-13 04:13:53 +0000 | [diff] [blame] | 417 | >>> c.update('witch') # add elements from another iterable |
| 418 | >>> d = Counter('watch') |
| 419 | >>> c.update(d) # add elements from another counter |
| 420 | >>> c['h'] # four 'h' in which, witch, and watch |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 421 | 4 |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 422 | |
| 423 | ''' |
| 424 | # The regular dict.update() operation makes no sense here because the |
| 425 | # replace behavior results in the some of original untouched counts |
| 426 | # being mixed-in with all of the other counts for a mismash that |
| 427 | # doesn't have a straight-forward interpretation in most counting |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 428 | # contexts. Instead, we implement straight-addition. Both the inputs |
| 429 | # and outputs are allowed to contain zero and negative counts. |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 430 | |
| 431 | if iterable is not None: |
| 432 | if isinstance(iterable, Mapping): |
Raymond Hettinger | 1bc1c8a | 2009-01-22 09:05:43 +0000 | [diff] [blame] | 433 | if self: |
Raymond Hettinger | 5dfc7f9 | 2009-06-29 19:10:29 +0000 | [diff] [blame] | 434 | self_get = self.get |
Raymond Hettinger | 1bc1c8a | 2009-01-22 09:05:43 +0000 | [diff] [blame] | 435 | for elem, count in iterable.iteritems(): |
Raymond Hettinger | 5dfc7f9 | 2009-06-29 19:10:29 +0000 | [diff] [blame] | 436 | self[elem] = self_get(elem, 0) + count |
Raymond Hettinger | 1bc1c8a | 2009-01-22 09:05:43 +0000 | [diff] [blame] | 437 | else: |
| 438 | dict.update(self, iterable) # fast path when counter is empty |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 439 | else: |
Raymond Hettinger | 5dfc7f9 | 2009-06-29 19:10:29 +0000 | [diff] [blame] | 440 | self_get = self.get |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 441 | for elem in iterable: |
Raymond Hettinger | 5dfc7f9 | 2009-06-29 19:10:29 +0000 | [diff] [blame] | 442 | self[elem] = self_get(elem, 0) + 1 |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 443 | if kwds: |
| 444 | self.update(kwds) |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 445 | |
| 446 | def copy(self): |
| 447 | 'Like dict.copy() but returns a Counter instance instead of a dict.' |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 448 | return Counter(self) |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 449 | |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 450 | def __delitem__(self, elem): |
| 451 | 'Like dict.__delitem__() but does not raise KeyError for missing values.' |
| 452 | if elem in self: |
| 453 | dict.__delitem__(self, elem) |
| 454 | |
Raymond Hettinger | f94d7fa | 2009-01-12 22:58:41 +0000 | [diff] [blame] | 455 | def __repr__(self): |
| 456 | if not self: |
| 457 | return '%s()' % self.__class__.__name__ |
Raymond Hettinger | 35288c6 | 2009-01-13 04:50:35 +0000 | [diff] [blame] | 458 | items = ', '.join(map('%r: %r'.__mod__, self.most_common())) |
Raymond Hettinger | aaa6e63 | 2009-01-13 01:05:03 +0000 | [diff] [blame] | 459 | return '%s({%s})' % (self.__class__.__name__, items) |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 460 | |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 461 | # Multiset-style mathematical operations discussed in: |
| 462 | # Knuth TAOCP Volume II section 4.6.3 exercise 19 |
| 463 | # and at http://en.wikipedia.org/wiki/Multiset |
| 464 | # |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 465 | # Outputs guaranteed to only include positive counts. |
| 466 | # |
| 467 | # To strip negative and zero counts, add-in an empty counter: |
| 468 | # c += Counter() |
| 469 | |
| 470 | def __add__(self, other): |
| 471 | '''Add counts from two counters. |
| 472 | |
| 473 | >>> Counter('abbb') + Counter('bcc') |
| 474 | Counter({'b': 4, 'c': 2, 'a': 1}) |
| 475 | |
| 476 | ''' |
| 477 | if not isinstance(other, Counter): |
| 478 | return NotImplemented |
| 479 | result = Counter() |
| 480 | for elem in set(self) | set(other): |
| 481 | newcount = self[elem] + other[elem] |
| 482 | if newcount > 0: |
| 483 | result[elem] = newcount |
| 484 | return result |
| 485 | |
| 486 | def __sub__(self, other): |
| 487 | ''' Subtract count, but keep only results with positive counts. |
| 488 | |
| 489 | >>> Counter('abbbc') - Counter('bccd') |
| 490 | Counter({'b': 2, 'a': 1}) |
| 491 | |
| 492 | ''' |
| 493 | if not isinstance(other, Counter): |
| 494 | return NotImplemented |
| 495 | result = Counter() |
Raymond Hettinger | 4571f34 | 2009-01-21 20:31:50 +0000 | [diff] [blame] | 496 | for elem in set(self) | set(other): |
| 497 | newcount = self[elem] - other[elem] |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 498 | if newcount > 0: |
| 499 | result[elem] = newcount |
| 500 | return result |
| 501 | |
| 502 | def __or__(self, other): |
| 503 | '''Union is the maximum of value in either of the input counters. |
| 504 | |
| 505 | >>> Counter('abbb') | Counter('bcc') |
| 506 | Counter({'b': 3, 'c': 2, 'a': 1}) |
| 507 | |
| 508 | ''' |
| 509 | if not isinstance(other, Counter): |
| 510 | return NotImplemented |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 511 | result = Counter() |
| 512 | for elem in set(self) | set(other): |
Raymond Hettinger | e3bc557 | 2009-04-04 08:46:58 +0000 | [diff] [blame] | 513 | p, q = self[elem], other[elem] |
| 514 | newcount = q if p < q else p |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 515 | if newcount > 0: |
| 516 | result[elem] = newcount |
| 517 | return result |
| 518 | |
| 519 | def __and__(self, other): |
| 520 | ''' Intersection is the minimum of corresponding counts. |
| 521 | |
| 522 | >>> Counter('abbb') & Counter('bcc') |
| 523 | Counter({'b': 1}) |
| 524 | |
| 525 | ''' |
| 526 | if not isinstance(other, Counter): |
| 527 | return NotImplemented |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 528 | result = Counter() |
| 529 | if len(self) < len(other): |
| 530 | self, other = other, self |
| 531 | for elem in _ifilter(self.__contains__, other): |
Raymond Hettinger | e3bc557 | 2009-04-04 08:46:58 +0000 | [diff] [blame] | 532 | p, q = self[elem], other[elem] |
| 533 | newcount = p if p < q else q |
Raymond Hettinger | bad1eb2 | 2009-01-20 01:19:26 +0000 | [diff] [blame] | 534 | if newcount > 0: |
| 535 | result[elem] = newcount |
| 536 | return result |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 537 | |
| 538 | |
| 539 | if __name__ == '__main__': |
Raymond Hettinger | d36a60e | 2007-09-17 00:55:00 +0000 | [diff] [blame] | 540 | # verify that instances can be pickled |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 541 | from cPickle import loads, dumps |
Raymond Hettinger | 01a0957 | 2007-10-23 20:37:41 +0000 | [diff] [blame] | 542 | Point = namedtuple('Point', 'x, y', True) |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 543 | p = Point(x=10, y=20) |
| 544 | assert p == loads(dumps(p)) |
| 545 | |
Raymond Hettinger | eeeb9c4 | 2007-11-15 02:44:53 +0000 | [diff] [blame] | 546 | # test and demonstrate ability to override methods |
Raymond Hettinger | b8e0072 | 2008-01-07 04:24:49 +0000 | [diff] [blame] | 547 | class Point(namedtuple('Point', 'x y')): |
Raymond Hettinger | e165508 | 2008-01-10 19:15:10 +0000 | [diff] [blame] | 548 | __slots__ = () |
Raymond Hettinger | b8e0072 | 2008-01-07 04:24:49 +0000 | [diff] [blame] | 549 | @property |
| 550 | def hypot(self): |
| 551 | return (self.x ** 2 + self.y ** 2) ** 0.5 |
Raymond Hettinger | 9a35921 | 2008-01-07 20:07:38 +0000 | [diff] [blame] | 552 | def __str__(self): |
Raymond Hettinger | 15b5e55 | 2008-01-10 23:00:01 +0000 | [diff] [blame] | 553 | return 'Point: x=%6.3f y=%6.3f hypot=%6.3f' % (self.x, self.y, self.hypot) |
Raymond Hettinger | b8e0072 | 2008-01-07 04:24:49 +0000 | [diff] [blame] | 554 | |
Raymond Hettinger | e165508 | 2008-01-10 19:15:10 +0000 | [diff] [blame] | 555 | for p in Point(3, 4), Point(14, 5/7.): |
Raymond Hettinger | 9a35921 | 2008-01-07 20:07:38 +0000 | [diff] [blame] | 556 | print p |
Raymond Hettinger | eeeb9c4 | 2007-11-15 02:44:53 +0000 | [diff] [blame] | 557 | |
Raymond Hettinger | dc55f35 | 2008-01-07 09:03:49 +0000 | [diff] [blame] | 558 | class Point(namedtuple('Point', 'x y')): |
| 559 | 'Point class with optimized _make() and _replace() without error-checking' |
Raymond Hettinger | e165508 | 2008-01-10 19:15:10 +0000 | [diff] [blame] | 560 | __slots__ = () |
Raymond Hettinger | dc55f35 | 2008-01-07 09:03:49 +0000 | [diff] [blame] | 561 | _make = classmethod(tuple.__new__) |
| 562 | def _replace(self, _map=map, **kwds): |
Raymond Hettinger | f5e8af1 | 2008-01-07 20:56:05 +0000 | [diff] [blame] | 563 | return self._make(_map(kwds.get, ('x', 'y'), self)) |
Raymond Hettinger | dc55f35 | 2008-01-07 09:03:49 +0000 | [diff] [blame] | 564 | |
| 565 | print Point(11, 22)._replace(x=100) |
| 566 | |
Raymond Hettinger | e850c46 | 2008-01-10 20:37:12 +0000 | [diff] [blame] | 567 | Point3D = namedtuple('Point3D', Point._fields + ('z',)) |
| 568 | print Point3D.__doc__ |
| 569 | |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 570 | import doctest |
Raymond Hettinger | 01a0957 | 2007-10-23 20:37:41 +0000 | [diff] [blame] | 571 | TestResults = namedtuple('TestResults', 'failed attempted') |
Raymond Hettinger | c37e5e0 | 2007-03-01 06:16:43 +0000 | [diff] [blame] | 572 | print TestResults(*doctest.testmod()) |