Add HeapDict to collections module #136898
Description
Feature or enhancement
Proposal: adding a new data structure called HeapDict to the Python collections module
Description
I would like to propose adding a new data structure called HeapDict to the Python collections module. This data structure would combine the functionality of a dictionary with a heap, providing efficient access to key-value pairs while maintaining a heap property for priority-based operations.
Motivation
Currently, Python offers heapq for heap operations and dictionaries for key-value storage, but there's no built-in data structure that efficiently combines both functionalities. A HeapDict would be useful in many scenarios:
- Priority queues where items need to be accessed by key
- Caching mechanisms with priority-based eviction
- Algorithms requiring both random access and priority-based operations (e.g., Dijkstra's algorithm)
- Task schedulers with dynamic priority updates
Proposed API
class HeapDict(collections.abc.MutableMapping):
"""Dictionary that maintains heap property based on values."""
def __init__(self, *args, **kwargs):
"""Initialize a new HeapDict with optional initial values."""
def popmin(self):
"""Remove and return the (key, value) pair with minimum value."""
def peekmin(self):
"""Return the (key, value) pair with minimum value without removing it."""
def update_priority(self, key, new_value):
"""Update the value/priority of an existing key."""Implementation Considerations
- The implementation would likely use a combination of a dictionary for O(1) key lookups and a heap for maintaining the priority order
- Special attention would be needed for handling duplicate priorities
- The implementation should ensure that all operations maintain both the dictionary and heap properties efficiently
Related Work
- Python's
heapqmodule provides heap operations but doesn't support key-based access - Third-party packages like
heapdictandprioritydictexist but aren't part of the standard library - Other languages have similar structures (e.g., Java's PriorityQueue with custom comparators)
Has this already been discussed elsewhere?
No response given
Links to previous discussion of this feature:
No response