StringSimilarity

Introduction

This package was designed as a place where TUO stores "all" the different string similarity routines around. To see how these routines work internally you are invited to read the code ;-) or look up at wikipedia :-D

Details

By now there are two different methods of string similarity routines in the package. Each unit contains one or the other ;-) simples

Phonetic comparison

These routines compare the string similarity by the "sound" of the word.

• uKoelnerPhonetik.pas
• uSoundEx.pas
• uMetaphone.pas
• DoubleMetaphone.pas

As there is no way of telling that a word sounds 20 or 80 per cent alike the function returns only a boolean value:

function SoundsSimilar(const AStr1, AStr2: String): Boolean;

Similarity ratio

The other way of string comparison is to calculate how similar two strings are in per cent.

uNeedlemanWunch in '....\uNeedlemanWunch.pas',

• uDiceSimilarity.pas
• FuzzyMatching.pas ... by NGram
• uJaroWinkler.pas
• uKoelnerPhonetik.pas
• uLevenshtein.pas
• uSmithWatermanGotoh.pas
• uMongeElkan.pas

The appropriate routine returns a double value between 0 and 1:

function StringSimilarityRatio(const Str1, Str2: String; IgnoreCase: Boolean): Double;