editdistpy

Warnings

• breaking Support for Python 3.8 was officially dropped in version 0.1.6. Users on Python 3.8 or older must upgrade their Python environment to Python 3.9 or newer to use versions >= 0.1.6.
  Fix: Upgrade Python to version 3.9 or higher.
• gotcha When `max_distance` is provided to the `distance` method, it will return `-1` if the calculated edit distance exceeds this maximum. This is an intentional performance feature, not an error. To always get the full distance, pass `sys.maxsize` as `max_distance`.
  Fix: Handle the `-1` return value appropriately, or explicitly pass `sys.maxsize` for an unbounded calculation.
• gotcha The `damerau_osa` module implements the Damerau Optimal String Alignment (OSA) distance, which is a restricted form of the Damerau-Levenshtein distance. OSA imposes a constraint that no substring may be edited more than once, meaning it does not always satisfy the triangle inequality and might return a different distance in certain cases compared to a full Damerau-Levenshtein implementation.
  Fix: Be aware of the distinction between Damerau OSA and the unrestricted Damerau-Levenshtein distance, and ensure it aligns with your specific use case. For typical typo correction, OSA is usually sufficient.
• deprecated In versions prior to 0.1.3, the library might not correctly handle distance calculations when one of the input strings is `None` or when `max_distance` is set to `0`. These issues were fixed in version 0.1.3.
  Fix: Upgrade to version 0.1.3 or later to ensure robust handling of `None` inputs and `max_distance=0`.

Install

• pip install editdistpy Install latest version

Imports

• levenshtein

  from editdistpy import levenshtein

• damerau_osa

  from editdistpy import damerau_osa

Quickstart

This quickstart demonstrates how to calculate both Levenshtein and Damerau Optimal String Alignment (OSA) distances using `editdistpy`. It also highlights the use of `max_distance` for performance optimization, which returns -1 if the actual distance exceeds the specified maximum.

import sys
from editdistpy import levenshtein, damerau_osa

string_1 = "flintstone"
string_2 = "hanson"

# Levenshtein distance
max_dist_lev = 2
lev_dist_bounded = levenshtein.distance(string_1, string_2, max_dist_lev)
print(f"Levenshtein (max_distance={max_dist_lev}): {lev_dist_bounded}") # Expected: -1

lev_dist_full = levenshtein.distance(string_1, string_2, sys.maxsize)
print(f"Levenshtein (full): {lev_dist_full}") # Expected: 6

# Damerau Optimal String Alignment (OSA) distance
max_dist_osa = 2
osa_dist_bounded = damerau_osa.distance(string_1, string_2, max_dist_osa)
print(f"Damerau OSA (max_distance={max_dist_osa}): {osa_dist_bounded}") # Expected: -1

osa_dist_full = damerau_osa.distance(string_1, string_2, sys.maxsize)
print(f"Damerau OSA (full): {osa_dist_full}") # Expected: 6