Merge pull request Ehco1996#2 from Ehco1996/dev

update

Author: Ehco1996

ehco/__init__.py

@@ -0,0 +1,39 @@
+from array import array
+
+arr = array('u', 'td')
+
+print(arr[0], arr[1])
+print(arr)
+
+
+class Array:
+    '''定长的array'''
+
+    def __init__(self, size=32):
+        self._size = size
+        self._items = [None] * size
+
+    def __getitem__(self, index):
+        return self._items[index]
+
+    def __setitem__(self, index, value):
+        self._items[index] = value
+
+    def __len__(self):
+        return len(self._items)
+
+    def clear(self, value=None):
+        for i in range(self._items):
+            self._items[i] = value
+
+    def __iter__(self):
+        for item in self._items:
+            yield item
+
+
+def test_array():
+    size = 10
+    a = Array(size)
+    a[0] = 1
+    assert a[0] == 1
+    assert len(a) == 10

ehco/array_list/array_and_list.py

@@ -0,0 +1,46 @@
+class Bag:
+
+    def __init__(self, max_size=10):
+        self.max_size = max_size
+        self._items = list()
+
+    def add(self, item):
+        if len(self._items) >= self.max_size:
+            raise Exception('Bag is Full')
+        self._items.append(item)
+
+    def remove(self, item):
+        self._items.remove(item)
+
+    def __len__(self):
+        return len(self._items)
+
+    def __iter__(self):
+        for item in self._items:
+            yield item
+
+
+def test_bag():
+    bag = Bag()
+
+    for i in range(10):
+        bag.add(i)
+
+    bag.remove(3)
+    assert len(bag) == 9
+
+    bag.add(1)
+    assert len(bag) == 10
+
+    try:
+        bag.add(1)
+        print('len', len(bag))
+    except Exception as e:
+        print(e)
+
+    for i in bag:
+        print(i)
+
+
+if __name__ == '__main__':
+    test_bag()

ehco/base/bag_adt.py

@@ -0,0 +1,52 @@
+from ehco.hashtable.hashtable_adt import HashTable
+
+
+class DictADT(HashTable):
+
+    def _iter_slot(self):
+        for slot in self._table:
+            if slot not in (HashTable.EMPTY, HashTable.UNUSED):
+                yield slot
+
+    def __setitem__(self, key, value):
+        self.add(key, value)
+
+    def __getitem__(self, key):
+        if key not in self:
+            raise KeyError()
+        else:
+            return self.get(key)
+
+    def items(self):
+        for slot in self._iter_slot():
+            yield (slot.key, slot.value)
+
+    def keys(self):
+        for slot in self._iter_slot():
+            yield slot.key
+
+    def values(self):
+        for slot in self._iter_slot():
+            yield slot.value
+
+
+def test_dict_adt():
+    import random
+
+    d = DictADT()
+
+    d['a'] = 1
+    assert d['a'] == 1
+    d.remove('a')
+
+    assert len(d) == 0
+
+    l = list(range(30))
+    random.shuffle(l)
+    for i in l:
+        d.add(i, i)
+
+    for i in range(30):
+        assert d.get(i) == i
+
+    assert sorted(list(d.keys())) == sorted(l)

ehco/dict/__init__.py

@@ -0,0 +1,155 @@
+class Array(object):
+
+    def __init__(self, size=32, init=None):
+        self._size = size
+        self._items = [init] * size
+
+    def __getitem__(self, index):
+        return self._items[index]
+
+    def __setitem__(self, index, value):
+        self._items[index] = value
+
+    def __len__(self):
+        return self._size
+
+    def clear(self, value=None):
+        for i in range(self._items):
+            self._items[i] = value
+
+    def __iter__(self):
+        for item in self._items:
+            yield item
+
+
+class Slot(object):
+    """定义一个 hash 表 数组的槽
+    注意，一个槽有三种状态，看你能否想明白。相比链接法解决冲突，二次探查法删除一个 key 的操作稍微复杂。
+
+    1.从未使用 HashMap.UNUSED。此槽没有被使用和冲突过，查找时只要找到 UNUSED 就不用再继续探查了
+    2.使用过但是 remove 了，此时是 HashMap.EMPTY，该探查点后边的元素扔可能是有key
+    3.槽正在使用 Slot 节点
+    """
+
+    def __init__(self, key, value):
+        self.key, self.value = key, value
+
+
+class HashTable(object):
+
+    # 没有被使用过的
+    UNUSED = None
+    # 使用过却被删除了
+    EMPTY = Slot(None, None)
+
+    def __init__(self):
+        self._table = Array(8, init=HashTable.UNUSED)
+        self.length = 0
+
+    @property
+    def _load_factor(self):
+        # load factor 超过0.8的时候重新分配
+        return self.length / float(len(self._table))
+
+    def __len__(self):
+        return self.length
+
+    def _hash(self, key):
+        return abs(hash(key)) % len(self._table)
+
+    def _find_key(self, key):
+        index = self._hash(key)
+        _len = len(self._table)
+        while self._table[index] is not HashTable.UNUSED:
+            if self._table[index] is HashTable.EMPTY:
+                index = (index * 5 + 1) % _len
+                continue
+            elif self._table[index].key == key:
+                return index
+            else:
+                index = (index * 5 + 1) % _len
+        return None
+
+    def _find_slot_for_insert(self, key):
+        index = self._hash(key)
+        _len = len(self._table)
+        while not self._slot_can_insert(index):
+            index = (index * 5 + 1) % _len
+        return index
+
+    def _slot_can_insert(self, index):
+        return self._table[index] in (HashTable.EMPTY, HashTable.UNUSED)
+
+    def __contains__(self, key):
+        index = self._find_key(key)
+        return index is not None
+
+    def add(self, key, value):
+        if key in self:
+            index = self._find_key(key)
+            self._table[index].value = value
+            return False
+        else:
+            index = self._find_slot_for_insert(key)
+            self._table[index] = Slot(key, value)
+            self.length += 1
+            if self._load_factor >= 0.8:
+                self._rehash()
+            return True
+
+    def _rehash(self):
+        old_table = self._table
+        newsize = len(self._table) * 2
+        self._table = Array(newsize, HashTable.UNUSED)
+        self.length = 0
+
+        for slot in old_table:
+            if slot not in (HashTable.UNUSED, HashTable.EMPTY):
+                index = self._find_slot_for_insert(slot.key)
+                self._table[index] = slot
+                self.length += 1
+
+    def get(self, key, default=None):
+        index = self._find_key(key)
+        if index is None:
+            return default
+        else:
+            return self._table[index].value
+
+    def remove(self, key):
+        index = self._find_key(key)
+        if index is None:
+            raise KeyError()
+        value = self._table[index].value
+        self.length -= 1
+        self._table[index] = HashTable.EMPTY
+        return value
+
+    def __iter__(self):
+        for slot in self._table:
+            if slot not in (HashTable.EMPTY, HashTable.UNUSED):
+                yield slot.key
+
+
+def test_hash_table():
+    h = HashTable()
+
+    h.add('a', 0)
+    h.add('b', 1)
+    h.add('c', 2)
+
+    assert len(h) == 3
+    assert h.get('a') == 0
+    assert h.get('b') == 1
+    assert h.get('hehe') is None
+
+    h.remove('a')
+    assert h.get('a') is None
+    assert sorted(list(h)) == ['b', 'c']
+
+    n = 50
+    for i in range(n):
+        h.add(i, i)
+
+    for i in range(n):
+        assert h.get(i) == i