LCS speedup, add previous faster version of partial ratio

FuzzySharp.Benchmarks/BenchmarkAll.cs

@@ -8,6 +8,12 @@ namespace Raffinert.FuzzySharp.Benchmarks;
 [MemoryDiagnoser]
 public class BenchmarkAll
 {
+    [GlobalSetup]
+    public void GlobalSetup()
+    {
+        GlobalConfig.PartialRatioAccuracy = PartialRatioAccuracy.Strict;
+    }
+
     [Benchmark]
     public int Ratio()
     {
@@ -18,7 +24,7 @@ public int Ratio()
     [Benchmark]
     public int PartialRatio()
     {
-        return Fuzz.PartialRatio("similar", "somewhresimlrbetweenthisstring");
+        return Fuzz.PartialRatio("Supplier: ACME Corp. International, Address: 221B Baker St., London NW1 6XE", "Order: PO-100923, Supplier: Acme Corporation International, Address: 221B Baker Street, London NW1 6XE, VAT: GB123456789, Contact: accounting@acme.example");
     }
 
     [Benchmark]
@@ -96,7 +102,7 @@ public int RatioClassic()
     [Benchmark]
     public int PartialRatioClassic()
     {
-        return Classic.Fuzz.PartialRatio("similar", "somewhresimlrbetweenthisstring");
+        return Classic.Fuzz.PartialRatio("Supplier: ACME Corp. International, Address: 221B Baker St., London NW1 6XE", "Order: PO-100923, Supplier: Acme Corporation International, Address: 221B Baker Street, London NW1 6XE, VAT: GB123456789, Contact: accounting@acme.example");
     }
 
     [Benchmark]

FuzzySharp.Benchmarks/BenchmarkFastPartial.cs

@@ -0,0 +1,50 @@
+using BenchmarkDotNet.Attributes;
+using Raffinert.FuzzySharp.PreProcess;
+
+namespace Raffinert.FuzzySharp.Benchmarks;
+
+[MemoryDiagnoser]
+public class BenchmarkFastPartial
+{
+    [GlobalSetup]
+    public void GlobalSetup()
+    {
+        GlobalConfig.PartialRatioAccuracy = PartialRatioAccuracy.Fast;
+    }
+
+    [Benchmark]
+    public int PartialRatio()
+    {
+        return Fuzz.PartialRatio("Supplier: ACME Corp. International, Address: 221B Baker St., London NW1 6XE", "Order: PO-100923, Supplier: Acme Corporation International, Address: 221B Baker Street, London NW1 6XE, VAT: GB123456789, Contact: accounting@acme.example");
+    }
+
+    [Benchmark]
+    public int PartialTokenSortRatio()
+    {
+        return Fuzz.PartialTokenSortRatio("order words out of", "  words out of order");
+    }
+
+    [Benchmark]
+    public int PartialTokenSetRatio()
+    {
+        return Fuzz.PartialTokenSetRatio("fuzzy was a bear", "fuzzy fuzzy fuzzy bear");
+    }
+
+    [Benchmark]
+    public int WeightedRatio()
+    {
+        return Fuzz.WeightedRatio("The quick brown fox jimps ofver the small lazy dog", "the quick brown fox jumps over the small lazy dog");
+    }
+
+    [Benchmark]
+    public int PartialTokenInitialismRatio()
+    {
+        return Fuzz.PartialTokenInitialismRatio("NASA", "National Aeronautics Space Administration, Kennedy Space Center, Cape Canaveral, Florida 32899");
+    }
+
+    [Benchmark]
+    public int PartialTokenAbbreviationRatio()
+    {
+        return Fuzz.PartialTokenAbbreviationRatio("bl 420", "Baseline section 420", PreprocessMode.Full);
+    }
+}

FuzzySharp.Benchmarks/FuzzySharp.Benchmarks.csproj

@@ -2,7 +2,7 @@
 
   <PropertyGroup>
     <OutputType>Exe</OutputType>
-    <TargetFramework>net9.0</TargetFramework>
+    <TargetFramework>NET10.0</TargetFramework>
     <ImplicitUsings>enable</ImplicitUsings>
     <Nullable>enable</Nullable>
     <AssemblyName>$(MSBuildProjectName)</AssemblyName>
@@ -13,6 +13,7 @@
     <PackageReference Include="BenchmarkDotNet" Version="0.15.2" />
     <PackageReference Include="Fastenshtein" Version="1.0.10" />
     <PackageReference Include="FuzzySharp" Version="2.0.2" />
+    <PackageReference Include="Microsoft.VisualStudio.DiagnosticsHub.BenchmarkDotNetDiagnosers" Version="18.3.36812.1" />
     <PackageReference Include="Quickenshtein" Version="1.5.1" />
   </ItemGroup>
 

FuzzySharp.Benchmarks/Program.cs

@@ -1,43 +1,9 @@
 using BenchmarkDotNet.Configs;
 using BenchmarkDotNet.Jobs;
 using BenchmarkDotNet.Running;
-
-//using Raffinert.FuzzySharp;
-//using Raffinert.FuzzySharp.SimilarityRatio;
-//using Raffinert.FuzzySharp.SimilarityRatio.Scorer.Composite;
-//using Classic = FuzzySharp;
+using Raffinert.FuzzySharp.Benchmarks;
 
 var config = ManualConfig.Create(DefaultConfig.Instance)
-    .AddJob(Job.ShortRun);  // ← built-in short run
-
-BenchmarkRunner.Run(typeof(Program).Assembly, config);
-
-//var input1 = "+30.0% Damage to Close Enemies [30.01%";
-//var input2Collection = new[]
-//{
-//    "+#% Damage",
-//    "+#% Damage to Crowd Controlled Enemies",
-//    "+#% Damage to Close Enemies",
-//    "+#% Damage to Chilled Enemies",
-//    "+#% Damage to Poisoned Enemies",
-//    "#% Block Chance#% Blocked Damage Reduction",
-//    "#% Damage Reduction from Bleeding Enemies",
-//    "#% Damage Reduction",
-//    "+#% Cold Damage"
-//};
-
-//var classicScorer = Classic.SimilarityRatio.ScorerCache.Get<Classic.SimilarityRatio.Scorer.Composite.WeightedRatioScorer>();
-
-//Func<string, int> classicScorerFunc = input2 => classicScorer.Score(input1, input2);
-
-//var classicResult = input2Collection.Select(classicScorerFunc).ToList();
-
-//var scorer = ScorerCache.Get<WeightedRatioScorer>();
-
-//Func<string, int> scorerFunc = input2 => scorer.Score(input1, input2);
-
-//var result = input2Collection.Select(scorerFunc).ToList();
-
-//Console.WriteLine();
+    .AddJob(Job.ShortRun);
 
-//Console.WriteLine(Fuzz.WeightedRatio("The quick brown fox jimps ofver the small lazy dog", "the quick brown fox jumps over the small lazy dog"));
+BenchmarkSwitcher.FromAssembly(typeof(BenchmarkFastPartial).Assembly).Run(args, config);

FuzzySharp.Test/FuzzySharp.Test.csproj

@@ -1,7 +1,7 @@
 <Project Sdk="Microsoft.NET.Sdk">
 
   <PropertyGroup>
-    <TargetFrameworks>netframework4.6.2;netframework4.7.2;NET8.0;NET9.0</TargetFrameworks>
+    <TargetFrameworks>netframework4.6.2;netframework4.7.2;NET8.0;NET10.0</TargetFrameworks>
     <IsPackable>false</IsPackable>
 	<LangVersion>12.0</LangVersion>
 	<AssemblyName>Raffinert.$(MSBuildProjectName)</AssemblyName>

FuzzySharp.Test/FuzzyTests/FastPartialRatioTests.cs

@@ -0,0 +1,237 @@
+using System;
+using NUnit.Framework;
+using Raffinert.FuzzySharp.PreProcess;
+
+namespace Raffinert.FuzzySharp.Test.FuzzyTests;
+
+[TestFixture]
+public class FastPartialRatioTests : IDisposable
+{
+    #region Private Fields
+    private string _s1,
+        _s1A,
+        _s2,
+        _s3,
+        _s4,
+        _s5,
+        _s6,
+        _s7,
+        _s8,
+        _s8A,
+        _s9,
+        _s9A,
+        _s10,
+        _s10A;
+
+    private string[] _cirqueStrings, _baseballStrings;
+    #endregion
+
+    [SetUp]
+    public void Setup()
+    {
+        GlobalConfig.PartialRatioAccuracy = PartialRatioAccuracy.Fast;
+        _s1  = "new york mets";
+        _s1A = "new york mets";
+        _s2  = "new YORK mets";
+        _s3  = "the wonderful new york mets";
+        _s4  = "new york mets vs atlanta braves";
+        _s5  = "atlanta braves vs new york mets";
+        _s6  = "new york mets - atlanta braves";
+        _s7  = "new york city mets - atlanta braves";
+        // Edge cases
+        _s8   = "{";
+        _s8A  = "{";
+        _s9   = "{a";
+        _s9A  = "{a";
+        _s10  = "a{";
+        _s10A = "{b";
+    }
+
+    public void Dispose()
+    {
+        GlobalConfig.PartialRatioAccuracy = PartialRatioAccuracy.Strict;
+    }
+
+    [Test]
+    public void Test_Equal()
+    {
+        Assert.AreEqual(Fuzz.Ratio(_s1, _s1A), 100);
+        Assert.AreEqual(Fuzz.Ratio(_s8, _s8A), 100);
+        Assert.AreEqual(Fuzz.Ratio(_s9, _s9A), 100);
+    }
+
+    [Test]
+    public void Test_Case_Insensitive()
+    {
+        Assert.AreNotEqual(Fuzz.Ratio(_s1, _s2), 100);
+        Assert.AreEqual(Fuzz.Ratio(_s1, _s2, PreprocessMode.Full), 100);
+    }
+
+    [Test]
+    public void Test_Partial()
+    {
+        Assert.AreEqual(Fuzz.PartialRatio(_s1, _s3), 100);
+    }
+
+    [Test]
+    public void TestTokenSortRatio()
+    {
+        Assert.AreEqual(Fuzz.TokenSortRatio(_s1, _s1A), 100);
+    }
+
+    [Test]
+    public void TestPartialTokenSortRatio()
+    {
+        Assert.AreEqual(Fuzz.PartialTokenSortRatio(_s1, _s1A, PreprocessMode.Full), 100);
+        Assert.AreEqual(Fuzz.PartialTokenSortRatio(_s4, _s5, PreprocessMode.Full), 100);
+        Assert.AreEqual(Fuzz.PartialTokenSortRatio(_s8, _s8A), 100);
+        Assert.AreEqual(Fuzz.PartialTokenSortRatio(_s9, _s9A, PreprocessMode.Full), 100);
+        Assert.AreEqual(Fuzz.PartialTokenSortRatio(_s9, _s9A), 100);
+
+        //var al =  Fuzz1.PartialRatioAlignment("a certain string".AsSpan(), "cetain".AsSpan());
+
+        Assert.AreEqual(Fuzz.PartialTokenSortRatio(_s10, _s10A), 50); // 67 in strict mode
+        Assert.AreEqual(Fuzz.PartialTokenSortRatio(_s10, _s10A, PreprocessMode.Full), 0);
+    }
+
+    [Test]
+    public void TestTokenSetRatio()
+    {
+        Assert.AreEqual(Fuzz.TokenSetRatio(_s4, _s5, PreprocessMode.Full), 100);
+        Assert.AreEqual(Fuzz.TokenSetRatio(_s8, _s8A), 100);
+        Assert.AreEqual(Fuzz.TokenSetRatio(_s9, _s9A, PreprocessMode.Full), 100);
+        Assert.AreEqual(Fuzz.TokenSetRatio(_s9, _s9A), 100);
+        Assert.AreEqual(Fuzz.TokenSetRatio(_s10, _s10A), 50);
+    }
+
+    [Test]
+    public void TestTokenAbbreviationRatio()
+    {
+        Assert.AreEqual(Fuzz.TokenAbbreviationRatio("bl 420", "Baseline section 420", PreprocessMode.Full), 40);
+        Assert.AreEqual(Fuzz.PartialTokenAbbreviationRatio("bl 420", "Baseline section 420", PreprocessMode.Full), 50); // 67 in strict mode
+    }
+
+    [Test]
+    public void TestPartialTokenSetRatio()
+    {
+        Assert.AreEqual(Fuzz.PartialTokenSetRatio(_s4, _s7), 100);
+    }
+
+    [Test]
+    public void TestWeightedRatioEqual()
+    {
+        Assert.AreEqual(Fuzz.WeightedRatio(_s1, _s1A), 100);
+    }
+
+    [Test]
+    public void TestWeightedRatioCaseInsensitive()
+    {
+        Assert.AreEqual(Fuzz.WeightedRatio(_s1, _s2, PreprocessMode.Full), 100);
+    }
+
+    [Test]
+    public void TestWeightedRatioPartialMatch()
+    {
+        Assert.AreEqual(Fuzz.WeightedRatio(_s1, _s3), 90);
+    }
+
+    [Test]
+    public void TestWeightedRatioMisorderedMatch()
+    {
+        Assert.AreEqual(Fuzz.WeightedRatio(_s4, _s5), 95);
+    }
+
+    [Test]
+    public void TestEmptyStringsScore0()
+    {
+        Assert.That(Fuzz.Ratio("test_string", ""), Is.EqualTo(0));
+        Assert.That(Fuzz.PartialRatio("test_string", ""), Is.EqualTo(0));
+        Assert.That(Fuzz.Ratio("", ""), Is.EqualTo(0));
+        Assert.That(Fuzz.PartialRatio("", ""), Is.EqualTo(0));
+    }
+
+    [Test]
+    public void TestIssueSeven()
+    {
+        _s1 = "HSINCHUANG";
+        _s2 = "SINJHUAN";
+        _s3 = "LSINJHUANG DISTRIC";
+        _s4 = "SINJHUANG DISTRICT";
+
+        Assert.IsTrue(Fuzz.PartialRatio(_s1, _s2) > 75);
+        Assert.IsTrue(Fuzz.PartialRatio(_s1, _s3) > 75);
+        Assert.IsTrue(Fuzz.PartialRatio(_s1, _s4) > 75);
+    }
+
+    [Test]
+    public void TestIssueEight()
+    {
+        // https://github.com/JakeBayer/FuzzySharp/issues/8
+        Assert.AreEqual(85, Fuzz.PartialRatio("Partnernummer", "Partne\nrnum\nmerASDFPartnernummerASDF")); // 100 in strict mode
+        Assert.AreEqual(77, Fuzz.PartialRatio("Partnernummer", "PartnerrrrnummerASDFPartnernummerASDF"));  // 100 in strict mode
+
+        // https://github.com/xdrop/fuzzywuzzy/issues/39
+        Assert.AreEqual(57, Fuzz.PartialRatio("kaution", "kdeffxxxiban:de1110010060046666666datum:16.11.17zeit:01:12uft0000899999tan076601testd.-20-maisonette-z4-jobas-hagkautionauszug")); // 100 in strict mode
+
+        // https://github.com/seatgeek/fuzzywuzzy/issues/79
+        Assert.AreEqual(93, Fuzz.PartialRatio("this is a test", "is this is a not really thing this is a test!"));  // 100 in strict mode
+
+        // https://github.com/Raffinert/FuzzySharp/issues/2
+        Assert.AreEqual(100, Fuzz.PartialRatio("sh", "Growing eshops without a popular platform", PreprocessMode.Full));
+        Assert.AreEqual(100, Fuzz.PartialRatio("shop", "Growing eshops without a popular platform", PreprocessMode.Full));
+    }
+
+    [Test]
+    public void MorePartialRatio()
+    {
+        Assert.AreEqual(100, Fuzz.PartialRatio("geeks for geeks", "geeks for geeks!"));
+        Assert.AreEqual(64, Fuzz.PartialRatio("geeks for geeks", "geeks geeks")); // 71 in strict mode
+        Assert.AreEqual(100, Fuzz.TokenSortRatio("geeks for geeks", "for geeks geeks"));
+    }
+
+    [Test]
+    public void TestPartialRatioUnicodeString()
+    {
+        _s1 = "\u00C1";
+        _s2 = "ABCD";
+        var score = Fuzz.PartialRatio(_s1, _s2);
+        Assert.AreEqual(0, score);
+    }
+
+    [Test]
+    public void TestZeroRatio()
+    {
+        var ratio = Fuzz.PartialTokenSortRatio("abc", "def");
+
+        Assert.True(ratio == 0);
+    }
+
+    [Test]
+    public void Test03()
+    {
+        var ratio = Fuzz.PartialTokenSortRatio("new york mets", "atlanta braves vs new york mets");
+
+        Assert.True(ratio == 77);
+    }
+
+    [Test]
+    public void TestRatioUnicodeString()
+    {
+        _s1 = "\u00C1";
+        _s2 = "ABCD";
+        var score = Fuzz.WeightedRatio(_s1, _s2);
+        Assert.AreEqual(0, score);
+
+        // Cyrillic.
+        _s1   = "\u043f\u0441\u0438\u0445\u043e\u043b\u043e\u0433";
+        _s2   = "\u043f\u0441\u0438\u0445\u043e\u0442\u0435\u0440\u0430\u043f\u0435\u0432\u0442";
+        score = Fuzz.WeightedRatio(_s1, _s2);
+        Assert.AreNotEqual(0, score);
+
+        // Chinese.
+        _s1   = "\u6211\u4e86\u89e3\u6570\u5b66";
+        _s2   = "\u6211\u5b66\u6570\u5b66";
+        score = Fuzz.WeightedRatio(_s1, _s2);
+        Assert.AreNotEqual(0, score);
+    }
+}