Usage

Matcher

The pfzy package provides an async entry point fuzzy_match to perform fuzzy matching using a given string against a given list of strings and will perform ranking automatically.

main.py

import asyncio

from pfzy import fuzzy_match

async def main():
  return await fuzzy_match("ab", ["acb", "acbabc"])

if __name__ == "__main__":
  print(asyncio.run(main()))

>>> python main.py
[{"value": "acbabc", "indices": [3, 4]}, {"value": "acb", "indices": [0, 2]}]

Matching against dictionaries

The second argument can also be a list of dictionary but you’ll have to also specify the argument key so that the function knows which key in the dictionary contains the value to match.

import asyncio

from pfzy import fuzzy_match

result = asyncio.run(fuzzy_match("ab", [{"val": "acb"}, {"val": "acbabc"}], key="val"))

>>> print(result)
[{"val": "acbabc", "indices": [3, 4]}, {"val": "acb", "indices": [0, 2]}]

Using different scorer

By default, it uses the fzy_scorer to perform string matching if not specified. You can explicitly set a different scorer using the argument scorer. Reference #Scorer for a list of available scorers.

import asyncio

from pfzy import fuzzy_match, substr_scorer

result = asyncio.run(fuzzy_match("ab", ["acb", "acbabc"], scorer=substr_scorer))

>>> print(result)
[{'value': 'acbabc', 'indices': [3, 4]}]

Scorer

fzy_scorer

Tip

The higher the score, the higher the string similarity.

The fzy_scorer uses fzy matching logic to perform string fuzzy matching.

The returned value is a tuple with the matching score and the matching indices.

from pfzy import fzy_scorer

score, indices = fzy_scorer("ab", "acbabc")

>>> print(score)
0.98
>>> print(indices)
[3, 4]

substr_scorer

Note

The score returned by substr_scorer might be negative value, but it doesn’t mean its not a match. As a rule of thumb, the higher the score, the higher the string similarity.

Use this scorer when exact substring matching is preferred. Different than the fzy_scorer, substr_scorer only performs exact matching and the score calculation works differently.

The returned value is a tuple with the matching score and the matching indices.

from pfzy import substr_scorer

score, indices = substr_scorer("ab", "awsab")

>>> print(score)
-1.3
>>> print(indices)
[3, 4]

from pfzy import substr_scorer

score, indices = substr_scorer("ab", "asdafswabc")

>>> print(score)
-1.6388888888888888
>>> print(indices)
[7, 8]