8.5. "heapq" --- 堆队列算法
***************************

**源码：**Lib/heapq.py

======================================================================

这个模块提供了堆队列算法的实现，也称为优先队列算法。

堆是一个二叉树，它的每个父节点的值都只会小于或大于所有孩子节点（的值）
。它使用了数组来实现：从零开始计数，对于所有的 *k* ，都有``heap[k] <=
heap[2*k+1]`` 和 "heap[k] <= heap[2*k+2]" 。为了便于比较，不存在的元素
被认为是无限大。堆最有趣的特性在于最小的元素总是在根结点："heap[0]" 。

这个API与教材的堆算法实现不太一样，在于两方面：（a）我们使用了基于零开
始进行索引。这使得节点和其孩子节点索引之间的关系不太直观，但是更加适合

基于这两方面，把堆看作原生的Python list也没什么奇怪的： "heap[0]" 表示
最小的元素，同时 "heap.sort()" 维护了堆的不变性！

要创建一个堆，可以使用list来初始化为 "[]" ，或者你可以通过一个函数
"heapify()" ，来把一个list转换成堆。

定义了以下函数：

heapq.heappush(heap, item)

   将 *item* 的值加入 *heap* 中，保持堆的不变性。

heapq.heappop(heap)

   弹出并返回 *heap* 的最小的元素，保持堆的不变性。如果堆为空，抛出
   "IndexError" 。使用 "heap[0]" ，可以只访问最小的元素而不弹出它。

heapq.heappushpop(heap, item)

   将 *item* 放入堆中，然后弹出并返回 *heap* 的最小元素。该组合操作比
   先调用  "heappush()" 再调用 "heappop()" 运行起来更有效率。

heapq.heapify(x)

   将list *x* 转换成堆，原地，线性时间内。

heapq.heapreplace(heap, item)

   Pop and return the smallest item from the *heap*, and also push the
   new *item*. The heap size doesn't change. If the heap is empty,
   "IndexError" is raised.

   This one step operation is more efficient than a "heappop()"
   followed by "heappush()" and can be more appropriate when using a
   fixed-size heap. The pop/push combination always returns an element
   from the heap and replaces it with *item*.

   The value returned may be larger than the *item* added.  If that
   isn't desired, consider using "heappushpop()" instead.  Its
   push/pop combination returns the smaller of the two values, leaving
   the larger value on the heap.

The module also offers three general purpose functions based on heaps.

heapq.merge(*iterables, key=None, reverse=False)

   Merge multiple sorted inputs into a single sorted output (for
   example, merge timestamped entries from multiple log files).
   Returns an *iterator* over the sorted values.

   Similar to "sorted(itertools.chain(*iterables))" but returns an
   iterable, does not pull the data into memory all at once, and
   assumes that each of the input streams is already sorted (smallest
   to largest).

   具有两个可选参数，它们都必须指定为关键字参数。

   *key* specifies a *key function* of one argument that is used to
   extract a comparison key from each input element.  The default
   value is "None" (compare the elements directly).

   *reverse* is a boolean value.  If set to "True", then the input
   elements are merged as if each comparison were reversed.

   在 3.5 版更改: Added the optional *key* and *reverse* parameters.

heapq.nlargest(n, iterable, key=None)

   Return a list with the *n* largest elements from the dataset
   defined by *iterable*.  *key*, if provided, specifies a function of
   one argument that is used to extract a comparison key from each
   element in *iterable* (for example, "key=str.lower").  Equivalent
   to:  "sorted(iterable, key=key, reverse=True)[:n]".

heapq.nsmallest(n, iterable, key=None)

   Return a list with the *n* smallest elements from the dataset
   defined by *iterable*.  *key*, if provided, specifies a function of
   one argument that is used to extract a comparison key from each
   element in *iterable* (for example, "key=str.lower").  Equivalent
   to:  "sorted(iterable, key=key)[:n]".

The latter two functions perform best for smaller values of *n*.  For
larger values, it is more efficient to use the "sorted()" function.
Also, when "n==1", it is more efficient to use the built-in "min()"
and "max()" functions.  If repeated usage of these functions is
required, consider turning the iterable into an actual heap.


8.5.1. 基本示例
===============

A heapsort can be implemented by pushing all values onto a heap and
then popping off the smallest values one at a time:

   >>> def heapsort(iterable):
   ...     h = []
   ...     for value in iterable:
   ...         heappush(h, value)
   ...     return [heappop(h) for i in range(len(h))]
   ...
   >>> heapsort([1, 3, 5, 7, 9, 2, 4, 6, 8, 0])
   [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

This is similar to "sorted(iterable)", but unlike "sorted()", this
implementation is not stable.

Heap elements can be tuples.  This is useful for assigning comparison
values (such as task priorities) alongside the main record being
tracked:

   >>> h = []
   >>> heappush(h, (5, 'write code'))
   >>> heappush(h, (7, 'release product'))
   >>> heappush(h, (1, 'write spec'))
   >>> heappush(h, (3, 'create tests'))
   >>> heappop(h)
   (1, 'write spec')


8.5.2. Priority Queue Implementation Notes
==========================================

A priority queue is common use for a heap, and it presents several
implementation challenges:

* Sort stability:  how do you get two tasks with equal priorities to
  be returned in the order they were originally added?

* Tuple comparison breaks for (priority, task) pairs if the
  priorities are equal and the tasks do not have a default comparison
  order.

* If the priority of a task changes, how do you move it to a new
  position in the heap?

* Or if a pending task needs to be deleted, how do you find it and
  remove it from the queue?

A solution to the first two challenges is to store entries as
3-element list including the priority, an entry count, and the task.
The entry count serves as a tie-breaker so that two tasks with the
same priority are returned in the order they were added. And since no
two entry counts are the same, the tuple comparison will never attempt
to directly compare two tasks.

The remaining challenges revolve around finding a pending task and
making changes to its priority or removing it entirely.  Finding a
task can be done with a dictionary pointing to an entry in the queue.

Removing the entry or changing its priority is more difficult because
it would break the heap structure invariants.  So, a possible solution
is to mark the entry as removed and add a new entry with the revised
priority:

   pq = []                         # list of entries arranged in a heap
   entry_finder = {}               # mapping of tasks to entries
   REMOVED = '<removed-task>'      # placeholder for a removed task
   counter = itertools.count()     # unique sequence count

   def add_task(task, priority=0):
       'Add a new task or update the priority of an existing task'
       if task in entry_finder:
           remove_task(task)
       count = next(counter)
       entry = [priority, count, task]
       entry_finder[task] = entry
       heappush(pq, entry)

   def remove_task(task):
       'Mark an existing task as REMOVED.  Raise KeyError if not found.'
       entry = entry_finder.pop(task)
       entry[-1] = REMOVED

   def pop_task():
       'Remove and return the lowest priority task. Raise KeyError if empty.'
       while pq:
           priority, count, task = heappop(pq)
           if task is not REMOVED:
               del entry_finder[task]
               return task
       raise KeyError('pop from an empty priority queue')


8.5.3. Theory
=============

Heaps are arrays for which "a[k] <= a[2*k+1]" and "a[k] <= a[2*k+2]"
for all *k*, counting elements from 0.  For the sake of comparison,
non-existing elements are considered to be infinite.  The interesting
property of a heap is that "a[0]" is always its smallest element.

The strange invariant above is meant to be an efficient memory
representation for a tournament.  The numbers below are *k*, not
"a[k]":

                                  0

                 1                                 2

         3               4                5               6

     7       8       9       10      11      12      13      14

   15 16   17 18   19 20   21 22   23 24   25 26   27 28   29 30

In the tree above, each cell *k* is topping "2*k+1" and "2*k+2". In a
usual binary tournament we see in sports, each cell is the winner over
the two cells it tops, and we can trace the winner down the tree to
see all opponents s/he had.  However, in many computer applications of
such tournaments, we do not need to trace the history of a winner. To
be more memory efficient, when a winner is promoted, we try to replace
it by something else at a lower level, and the rule becomes that a
cell and the two cells it tops contain three different items, but the
top cell "wins" over the two topped cells.

If this heap invariant is protected at all time, index 0 is clearly
the overall winner.  The simplest algorithmic way to remove it and
find the "next" winner is to move some loser (let's say cell 30 in the
diagram above) into the 0 position, and then percolate this new 0 down
the tree, exchanging values, until the invariant is re-established.
This is clearly logarithmic on the total number of items in the tree.
By iterating over all items, you get an O(n log n) sort.

A nice feature of this sort is that you can efficiently insert new
items while the sort is going on, provided that the inserted items are
not "better" than the last 0'th element you extracted.  This is
especially useful in simulation contexts, where the tree holds all
incoming events, and the "win" condition means the smallest scheduled
time.  When an event schedules other events for execution, they are
scheduled into the future, so they can easily go into the heap.  So, a
heap is a good structure for implementing schedulers (this is what I
used for my MIDI sequencer :-).

Various structures for implementing schedulers have been extensively
studied, and heaps are good for this, as they are reasonably speedy,
the speed is almost constant, and the worst case is not much different
than the average case. However, there are other representations which
are more efficient overall, yet the worst cases might be terrible.

Heaps are also very useful in big disk sorts.  You most probably all
know that a big sort implies producing "runs" (which are pre-sorted
sequences, whose size is usually related to the amount of CPU memory),
followed by a merging passes for these runs, which merging is often
very cleverly organised [1]. It is very important that the initial
sort produces the longest runs possible.  Tournaments are a good way
to achieve that.  If, using all the memory available to hold a
tournament, you replace and percolate items that happen to fit the
current run, you'll produce runs which are twice the size of the
memory for random input, and much better for input fuzzily ordered.

Moreover, if you output the 0'th item on disk and get an input which
may not fit in the current tournament (because the value "wins" over
the last output value), it cannot fit in the heap, so the size of the
heap decreases.  The freed memory could be cleverly reused immediately
for progressively building a second heap, which grows at exactly the
same rate the first heap is melting.  When the first heap completely
vanishes, you switch heaps and start a new run.  Clever and quite
effective!

In a word, heaps are useful memory structures to know.  I use them in
a few applications, and I think it is good to keep a 'heap' module
around. :-)

-[ 脚注 ]-

[1] The disk balancing algorithms which are current, nowadays, are
    more annoying than clever, and this is a consequence of the
    seeking capabilities of the disks. On devices which cannot seek,
    like big tape drives, the story was quite different, and one had
    to be very clever to ensure (far in advance) that each tape
    movement will be the most effective possible (that is, will best
    participate at "progressing" the merge).  Some tapes were even
    able to read backwards, and this was also used to avoid the
    rewinding time. Believe me, real good tape sorts were quite
    spectacular to watch! From all times, sorting has always been a
    Great Art! :-)
