def __init__(self, iterable=None, key=None, load=1000, _set=None):
"""
A `SortedSet` provides the same methods as a `set`. Additionally, a
`SortedSet` maintains its items in sorted order, allowing the
`SortedSet` to be indexed.
An optional *iterable* provides an initial series of items to populate
the `SortedSet`.
An optional *key* argument defines a callable that, like the `key`
argument to Python's `sorted` function, extracts a comparison key from
each set item. If no function is specified, the default compares the
set items directly.
An optional *load* specifies the load-factor of the set. The default
load factor of '1000' works well for sets from tens to tens of millions
of elements. Good practice is to use a value that is the cube root of
the set size. With billions of elements, the best load factor depends
on your usage. It's best to leave the load factor at the default until
you start benchmarking.
"""
self._key = key
self._load = load
self._set = set() if _set is None else _set
_set = self._set
self.isdisjoint = _set.isdisjoint
self.issubset = _set.issubset
self.issuperset = _set.issuperset
if key is None:
self._list = SortedList(self._set, load=load)
else:
self._list = SortedListWithKey(self._set, key=key, load=load)
_list = self._list
self.bisect_left = _list.bisect_left
self.bisect = _list.bisect
self.bisect_right = _list.bisect_right
self.index = _list.index
self.irange = _list.irange
self.islice = _list.islice
if key is not None:
self.bisect_key_left = _list.bisect_key_left
self.bisect_key_right = _list.bisect_key_right
self.bisect_key = _list.bisect_key
self.irange_key = _list.irange_key
if iterable is not None:
self._update(iterable)
def __init__(self, *args, **kwargs):
"""
A SortedDict provides the same methods as a dict. Additionally, a
SortedDict efficiently maintains its keys in sorted order. Consequently,
the keys method will return the keys in sorted order, the popitem method
will remove the item with the highest key, etc.
An optional *key* argument defines a callable that, like the `key`
argument to Python's `sorted` function, extracts a comparison key from
each dict key. If no function is specified, the default compares the
dict keys directly. The `key` argument must be provided as a positional
argument and must come before all other arguments.
An optional *load* argument defines the load factor of the internal list
used to maintain sort order. If present, this argument must come before
an iterable. The default load factor of '1000' works well for lists from
tens to tens of millions of elements. Good practice is to use a value
that is the cube root of the list size. With billions of elements, the
best load factor depends on your usage. It's best to leave the load
factor at the default until you start benchmarking.
An optional *iterable* argument provides an initial series of items to
populate the SortedDict. Each item in the series must itself contain
two items. The first is used as a key in the new dictionary, and the
second as the key's value. If a given key is seen more than once, the
last value associated with it is retained in the new dictionary.
If keyword arguments are given, the keywords themselves with their
associated values are added as items to the dictionary. If a key is
specified both in the positional argument and as a keyword argument, the
value associated with the keyword is retained in the dictionary. For
example, these all return a dictionary equal to ``{"one": 2, "two":
3}``:
* ``SortedDict(one=2, two=3)``
* ``SortedDict({'one': 2, 'two': 3})``
* ``SortedDict(zip(('one', 'two'), (2, 3)))``
* ``SortedDict([['two', 3], ['one', 2]])``
The first example only works for keys that are valid Python
identifiers; the others work with any valid keys.
"""
if len(args) > 0 and (args[0] is None or callable(args[0])):
self._key = args[0]
args = args[1:]
else:
self._key = None
if len(args) > 0 and type(args[0]) == int:
self._load = args[0]
args = args[1:]
else:
self._load = 1000
if self._key is None:
self._list = SortedList(load=self._load)
else:
self._list = SortedListWithKey(key=self._key, load=self._load)
# Cache function pointers to dict methods.
_dict = super(SortedDict, self)
self._dict = _dict
self._clear = _dict.clear
self._delitem = _dict.__delitem__
self._iter = _dict.__iter__
self._pop = _dict.pop
self._setdefault = _dict.setdefault
self._setitem = _dict.__setitem__
self._dict_update = _dict.update
# Cache function pointers to SortedList methods.
_list = self._list
self._list_add = _list.add
self.bisect_left = _list.bisect_left
self.bisect = _list.bisect_right
self.bisect_right = _list.bisect_right
self._list_clear = _list.clear
self.index = _list.index
self._list_pop = _list.pop
self._list_remove = _list.remove
self._list_update = _list.update
self.irange = _list.irange
self.islice = _list.islice
if self._key is not None:
self.bisect_key_left = _list.bisect_key_left
self.bisect_key_right = _list.bisect_key_right
self.bisect_key = _list.bisect_key
self.irange_key = _list.irange_key
self.iloc = _IlocWrapper(self)
self._update(*args, **kwargs)
class SortedSet(MutableSet, Sequence):
"""
A `SortedSet` provides the same methods as a `set`. Additionally, a
`SortedSet` maintains its items in sorted order, allowing the `SortedSet` to
be indexed.
Unlike a `set`, a `SortedSet` requires items be hashable and comparable.
"""
def __init__(self, iterable=None, key=None, load=1000, _set=None):
"""
A `SortedSet` provides the same methods as a `set`. Additionally, a
`SortedSet` maintains its items in sorted order, allowing the
`SortedSet` to be indexed.
An optional *iterable* provides an initial series of items to populate
the `SortedSet`.
An optional *key* argument defines a callable that, like the `key`
argument to Python's `sorted` function, extracts a comparison key from
each set item. If no function is specified, the default compares the
set items directly.
An optional *load* specifies the load-factor of the set. The default
load factor of '1000' works well for sets from tens to tens of millions
of elements. Good practice is to use a value that is the cube root of
the set size. With billions of elements, the best load factor depends
on your usage. It's best to leave the load factor at the default until
you start benchmarking.
"""
self._key = key
self._load = load
self._set = set() if _set is None else _set
_set = self._set
self.isdisjoint = _set.isdisjoint
self.issubset = _set.issubset
self.issuperset = _set.issuperset
if key is None:
self._list = SortedList(self._set, load=load)
else:
self._list = SortedListWithKey(self._set, key=key, load=load)
_list = self._list
self.bisect_left = _list.bisect_left
self.bisect = _list.bisect
self.bisect_right = _list.bisect_right
self.index = _list.index
self.irange = _list.irange
self.islice = _list.islice
if key is not None:
self.bisect_key_left = _list.bisect_key_left
self.bisect_key_right = _list.bisect_key_right
self.bisect_key = _list.bisect_key
self.irange_key = _list.irange_key
if iterable is not None:
self._update(iterable)
def __contains__(self, value):
"""Return True if and only if *value* is an element in the set."""
return (value in self._set)
def __getitem__(self, index):
"""
Return the element at position *index*.
Supports slice notation and negative indexes.
"""
return self._list[index]
def __delitem__(self, index):
"""
Remove the element at position *index*.
Supports slice notation and negative indexes.
"""
_list = self._list
if isinstance(index, slice):
values = _list[index]
self._set.difference_update(values)
else:
value = _list[index]
self._set.remove(value)
del _list[index]
def _make_cmp(set_op, doc):
def comparer(self, that):
if isinstance(that, SortedSet):
return set_op(self._set, that._set)
elif isinstance(that, Set):
return set_op(self._set, that)
else:
return NotImplemented
comparer.__name__ = '__{0}__'.format(set_op.__name__)
doc_str = 'Return True if and only if Set is {0} `that`.'
comparer.__doc__ = doc_str.format(doc)
return comparer
__eq__ = _make_cmp(op.eq, 'equal to')
__ne__ = _make_cmp(op.ne, 'not equal to')
__lt__ = _make_cmp(op.lt, 'a proper subset of')
__gt__ = _make_cmp(op.gt, 'a proper superset of')
__le__ = _make_cmp(op.le, 'a subset of')
__ge__ = _make_cmp(op.ge, 'a superset of')
def __len__(self):
"""Return the number of elements in the set."""
return len(self._set)
def __iter__(self):
"""
Return an iterator over the Set. Elements are iterated in their sorted
order.
Iterating the Set while adding or deleting values may raise a
`RuntimeError` or fail to iterate over all entries.
"""
return iter(self._list)
def __reversed__(self):
"""
Return an iterator over the Set. Elements are iterated in their reverse
sorted order.
Iterating the Set while adding or deleting values may raise a
`RuntimeError` or fail to iterate over all entries.
"""
return reversed(self._list)
def add(self, value):
"""Add the element *value* to the set."""
if value not in self._set:
self._set.add(value)
self._list.add(value)
def clear(self):
"""Remove all elements from the set."""
self._set.clear()
self._list.clear()
def copy(self):
"""Create a shallow copy of the sorted set."""
return self.__class__(key=self._key,
load=self._load,
_set=set(self._set))
__copy__ = copy
def count(self, value):
"""Return the number of occurrences of *value* in the set."""
return 1 if value in self._set else 0
def discard(self, value):
"""
Remove the first occurrence of *value*. If *value* is not a member,
does nothing.
"""
if value in self._set:
self._set.remove(value)
self._list.discard(value)
def pop(self, index=-1):
"""
Remove and return item at *index* (default last). Raises IndexError if
set is empty or index is out of range. Negative indexes are supported,
as for slice indices.
"""
value = self._list.pop(index)
self._set.remove(value)
return value
def remove(self, value):
"""
Remove first occurrence of *value*. Raises ValueError if
*value* is not present.
"""
self._set.remove(value)
self._list.remove(value)
def difference(self, *iterables):
"""
Return a new set with elements in the set that are not in the
*iterables*.
"""
diff = self._set.difference(*iterables)
new_set = self.__class__(key=self._key, load=self._load, _set=diff)
return new_set
__sub__ = difference
__rsub__ = __sub__
def difference_update(self, *iterables):
"""
Update the set, removing elements found in keeping only elements
found in any of the *iterables*.
"""
values = set(chain(*iterables))
if (4 * len(values)) > len(self):
self._set.difference_update(values)
self._list.clear()
self._list.update(self._set)
else:
_discard = self.discard
for value in values:
_discard(value)
return self
__isub__ = difference_update
def intersection(self, *iterables):
"""
Return a new set with elements common to the set and all *iterables*.
"""
comb = self._set.intersection(*iterables)
new_set = self.__class__(key=self._key, load=self._load, _set=comb)
return new_set
__and__ = intersection
__rand__ = __and__
def intersection_update(self, *iterables):
"""
Update the set, keeping only elements found in it and all *iterables*.
"""
self._set.intersection_update(*iterables)
self._list.clear()
self._list.update(self._set)
return self
__iand__ = intersection_update
def symmetric_difference(self, that):
"""
Return a new set with elements in either *self* or *that* but not both.
"""
diff = self._set.symmetric_difference(that)
new_set = self.__class__(key=self._key, load=self._load, _set=diff)
return new_set
__xor__ = symmetric_difference
__rxor__ = __xor__
def symmetric_difference_update(self, that):
"""
Update the set, keeping only elements found in either *self* or *that*,
but not in both.
"""
self._set.symmetric_difference_update(that)
self._list.clear()
self._list.update(self._set)
return self
__ixor__ = symmetric_difference_update
def union(self, *iterables):
"""
Return a new SortedSet with elements from the set and all *iterables*.
"""
return self.__class__(
chain(
iter(self), *iterables),
key=self._key,
load=self._load)
__or__ = union
__ror__ = __or__
def update(self, *iterables):
"""Update the set, adding elements from all *iterables*."""
values = set(chain(*iterables))
if (4 * len(values)) > len(self):
self._set.update(values)
self._list.clear()
self._list.update(self._set)
else:
_add = self.add
for value in values:
_add(value)
return self
__ior__ = update
_update = update
def __reduce__(self):
return (self.__class__, ((), self._key, self._load, self._set))
@recursive_repr
def __repr__(self):
temp = '{0}({1}, key={2}, load={3})'
return temp.format(self.__class__.__name__, repr(list(self)),
repr(self._key), repr(self._load))
def _check(self):
self._list._check()
assert len(self._set) == len(self._list)
_set = self._set
assert all(val in _set for val in self._list)
class SortedDict(dict):
"""
A SortedDict provides the same methods as a dict. Additionally, a
SortedDict efficiently maintains its keys in sorted order. Consequently, the
keys method will return the keys in sorted order, the popitem method will
remove the item with the highest key, etc.
"""
def __init__(self, *args, **kwargs):
"""
A SortedDict provides the same methods as a dict. Additionally, a
SortedDict efficiently maintains its keys in sorted order. Consequently,
the keys method will return the keys in sorted order, the popitem method
will remove the item with the highest key, etc.
An optional *key* argument defines a callable that, like the `key`
argument to Python's `sorted` function, extracts a comparison key from
each dict key. If no function is specified, the default compares the
dict keys directly. The `key` argument must be provided as a positional
argument and must come before all other arguments.
An optional *load* argument defines the load factor of the internal list
used to maintain sort order. If present, this argument must come before
an iterable. The default load factor of '1000' works well for lists from
tens to tens of millions of elements. Good practice is to use a value
that is the cube root of the list size. With billions of elements, the
best load factor depends on your usage. It's best to leave the load
factor at the default until you start benchmarking.
An optional *iterable* argument provides an initial series of items to
populate the SortedDict. Each item in the series must itself contain
two items. The first is used as a key in the new dictionary, and the
second as the key's value. If a given key is seen more than once, the
last value associated with it is retained in the new dictionary.
If keyword arguments are given, the keywords themselves with their
associated values are added as items to the dictionary. If a key is
specified both in the positional argument and as a keyword argument, the
value associated with the keyword is retained in the dictionary. For
example, these all return a dictionary equal to ``{"one": 2, "two":
3}``:
* ``SortedDict(one=2, two=3)``
* ``SortedDict({'one': 2, 'two': 3})``
* ``SortedDict(zip(('one', 'two'), (2, 3)))``
* ``SortedDict([['two', 3], ['one', 2]])``
The first example only works for keys that are valid Python
identifiers; the others work with any valid keys.
"""
if len(args) > 0 and (args[0] is None or callable(args[0])):
self._key = args[0]
args = args[1:]
else:
self._key = None
if len(args) > 0 and type(args[0]) == int:
self._load = args[0]
args = args[1:]
else:
self._load = 1000
if self._key is None:
self._list = SortedList(load=self._load)
else:
self._list = SortedListWithKey(key=self._key, load=self._load)
# Cache function pointers to dict methods.
_dict = super(SortedDict, self)
self._dict = _dict
self._clear = _dict.clear
self._delitem = _dict.__delitem__
self._iter = _dict.__iter__
self._pop = _dict.pop
self._setdefault = _dict.setdefault
self._setitem = _dict.__setitem__
self._dict_update = _dict.update
# Cache function pointers to SortedList methods.
_list = self._list
self._list_add = _list.add
self.bisect_left = _list.bisect_left
self.bisect = _list.bisect_right
self.bisect_right = _list.bisect_right
self._list_clear = _list.clear
self.index = _list.index
self._list_pop = _list.pop
self._list_remove = _list.remove
self._list_update = _list.update
self.irange = _list.irange
self.islice = _list.islice
if self._key is not None:
self.bisect_key_left = _list.bisect_key_left
self.bisect_key_right = _list.bisect_key_right
self.bisect_key = _list.bisect_key
self.irange_key = _list.irange_key
self.iloc = _IlocWrapper(self)
self._update(*args, **kwargs)
def clear(self):
"""Remove all elements from the dictionary."""
self._clear()
self._list_clear()
def __delitem__(self, key):
"""
Remove ``d[key]`` from *d*. Raises a KeyError if *key* is not in the
dictionary.
"""
self._delitem(key)
self._list_remove(key)
def __iter__(self):
"""
Return an iterator over the sorted keys of the dictionary.
Iterating the Mapping while adding or deleting keys may raise a
`RuntimeError` or fail to iterate over all entries.
"""
return iter(self._list)
def __reversed__(self):
"""
Return a reversed iterator over the sorted keys of the dictionary.
Iterating the Mapping while adding or deleting keys may raise a
`RuntimeError` or fail to iterate over all entries.
"""
return reversed(self._list)
def __setitem__(self, key, value):
"""Set `d[key]` to *value*."""
if key not in self:
self._list_add(key)
self._setitem(key, value)
def copy(self):
"""Return a shallow copy of the sorted dictionary."""
return self.__class__(self._key, self._load, self._iteritems())
__copy__ = copy
@classmethod
def fromkeys(cls, seq, value=None):
"""
Create a new dictionary with keys from *seq* and values set to *value*.
"""
return cls((key, value) for key in seq)
if hexversion < 0x03000000:
def items(self):
"""
Return a list of the dictionary's items (``(key, value)`` pairs).
"""
return list(self._iteritems())
else:
def items(self):
"""
Return a new ItemsView of the dictionary's items. In addition to
the methods provided by the built-in `view` the ItemsView is
indexable (e.g. ``d.items()[5]``).
"""
return ItemsView(self)
def iteritems(self):
"""
Return an iterator over the items (``(key, value)`` pairs).
Iterating the Mapping while adding or deleting keys may raise a
`RuntimeError` or fail to iterate over all entries.
"""
return iter((key, self[key]) for key in self._list)
_iteritems = iteritems
if hexversion < 0x03000000:
def keys(self):
"""Return a SortedSet of the dictionary's keys."""
return SortedSet(self._list, key=self._key, load=self._load)
else:
def keys(self):
"""
Return a new KeysView of the dictionary's keys. In addition to the
methods provided by the built-in `view` the KeysView is indexable
(e.g. ``d.keys()[5]``).
"""
return KeysView(self)
def iterkeys(self):
"""
Return an iterator over the sorted keys of the Mapping.
Iterating the Mapping while adding or deleting keys may raise a
`RuntimeError` or fail to iterate over all entries.
"""
return iter(self._list)
if hexversion < 0x03000000:
def values(self):
"""Return a list of the dictionary's values."""
return list(self._itervalues())
else:
def values(self):
"""
Return a new :class:`ValuesView` of the dictionary's values.
In addition to the methods provided by the built-in `view` the
ValuesView is indexable (e.g., ``d.values()[5]``).
"""
return ValuesView(self)
def itervalues(self):
"""
Return an iterator over the values of the Mapping.
Iterating the Mapping while adding or deleting keys may raise a
`RuntimeError` or fail to iterate over all entries.
"""
return iter(self[key] for key in self._list)
_itervalues = itervalues
def pop(self, key, default=_NotGiven):
"""
If *key* is in the dictionary, remove it and return its value,
else return *default*. If *default* is not given and *key* is not in
the dictionary, a KeyError is raised.
"""
if key in self:
self._list_remove(key)
return self._pop(key)
else:
if default is _NotGiven:
raise KeyError(key)
else:
return default
def popitem(self, last=True):
"""
Remove and return a ``(key, value)`` pair from the dictionary. If
last=True (default) then remove the *greatest* `key` from the
diciontary. Else, remove the *least* key from the dictionary.
If the dictionary is empty, calling `popitem` raises a
KeyError`.
"""
if not len(self):
raise KeyError('popitem(): dictionary is empty')
key = self._list_pop(-1 if last else 0)
value = self._pop(key)
return (key, value)
def setdefault(self, key, default=None):
"""
If *key* is in the dictionary, return its value. If not, insert *key*
with a value of *default* and return *default*. *default* defaults to
``None``.
"""
if key in self:
return self[key]
else:
self._setitem(key, default)
self._list_add(key)
return default
def update(self, *args, **kwargs):
"""
Update the dictionary with the key/value pairs from *other*, overwriting
existing keys.
*update* accepts either another dictionary object or an iterable of
key/value pairs (as a tuple or other iterable of length two). If
keyword arguments are specified, the dictionary is then updated with
those key/value pairs: ``d.update(red=1, blue=2)``.
"""
if not len(self):
self._dict_update(*args, **kwargs)
self._list_update(self._iter())
return
if (len(kwargs) == 0 and len(args) == 1 and isinstance(args[0], dict)):
pairs = args[0]
else:
pairs = dict(*args, **kwargs)
if (10 * len(pairs)) > len(self):
self._dict_update(pairs)
self._list_clear()
self._list_update(self._iter())
else:
for key in pairs:
self[key] = pairs[key]
_update = update
@not26
def viewkeys(self):
"""
In Python 2.7 and later, return a new `KeysView` of the dictionary's
keys.
In Python 2.6, raise a NotImplementedError.
"""
return KeysView(self)
@not26
def viewvalues(self):
"""
In Python 2.7 and later, return a new `ValuesView` of the dictionary's
values.
In Python 2.6, raise a NotImplementedError.
"""
return ValuesView(self)
@not26
def viewitems(self):
"""
In Python 2.7 and later, return a new `ItemsView` of the dictionary's
items.
In Python 2.6, raise a NotImplementedError.
"""
return ItemsView(self)
def __reduce__(self):
return (self.__class__,
(self._key, self._load, list(self._iteritems())))
@recursive_repr
def __repr__(self):
temp = '{0}({1}, {2}, {{{3}}})'
items = ', '.join('{0}: {1}'.format(
repr(key), repr(self[key])) for key in self._list)
return temp.format(self.__class__.__name__, repr(self._key),
repr(self._load), items)
def _check(self):
self._list._check()
assert len(self) == len(self._list)
assert all(val in self for val in self._list)
