v1.2.0 (2025-03-31)
Features
-
benchmarking: Add benchmark script for query performance with and without indexes (
137cee1) -
Introduce
scripts/benchmark.pywhich benchmarks DictDB's query performance using cProfile. -
Implements three benchmarking scenarios: without an index, with a hash index and with a sorted index.
-
Populates a large dataset and measures the average execution time of a simple equality select query.
-
Module is fully documented in reStructuredText format and adheres to PEP8 with strong typing.
-
indexing: Add support for user-defined indexes and automatic index updates (
d7c9baa) -
Implement create_index() in Table to allow creating indexes on specific fields.
-
Automatically update indexes on INSERT, UPDATE, and DELETE operations.
-
Accelerate SELECT queries by using indexes for simple equality conditions.
-
Update documentation and inline comments to reflect new indexing feature and changes.
This commit introduces indexing functionality to optimize SELECT queries.
-
indexing: Use efficient index data structures for maintaining indices (
d0604a3) -
Introduced a new module "index.py" defining an abstract IndexBase with two implementations: HashIndex (using Python's dict, a hash map) and SortedIndex (a simple sorted index using bisect to simulate B-tree behavior).
-
Updated Table.create_index() to accept an "index_type" parameter ("hash" or "sorted") and instantiate the corresponding index.
-
Refactored _update_indexes_on_insert, _update_indexes_on_update, and _update_indexes_on_delete in Table to delegate to the index object's methods.
-
Modified select() to use index.search() for simple equality conditions.
-
Added new tests in tests/test_indexing.py to verify the behavior of SortedIndex.
This commit evaluates and uses efficient data structures (hash maps and a sorted index) for maintaining indices and optimizes SELECT queries accordingly.
Detailed Changes: v1.1.0...v1.2.0