String Distance

What it does

Analyzes the similarity and relationship between two strings. It provides multiple ways to measure how close two strings are to being identical, including exact matches, case-insensitive matches, and structural matches (whitespace-normalized).

It implements two widely-used algorithms for string similarity: - Damerau-Levenshtein Distance: Measures the number of single-character edits (insertions, deletions, substitutions, and transpositions of adjacent characters) required to change one string into another. - Jaro-Winkler Score: A measure of similarity between two strings, where 0.0 is completely different and 1.0 is identical. It’s particularly effective for short strings like names.

Python usage

from csvsmith.utils.distance import analyze_pair

# Compare two strings
res = analyze_pair("Ames, IA", "Ames IA", ignore_case=True)

print(f"Relation: {res.get_relation_string()}")
print(f"Similarity: {res.similarity_percentage:.2f}%")
print(f"Jaro-Winkler: {res.jaro_winkler_score:.4f}")

CLI usage

csvsmith string-distance "Apple Inc." "apple inc" --ignore-case

Behavior notes

  • Classifications:

    • Identical: Exact character-for-character match.

    • Case-Insensitive Match: Matches if case is ignored.

    • Similar (Trimmed): Matches after removing leading/trailing whitespace.

    • Synonymous (No Spaces): Matches after removing ALL internal whitespace.

    • Different: No structural match found.

  • Similarity Percentage: A normalized score derived from the Damerau-Levenshtein distance relative to the length of the longer string.