feat: Chnaged Node children from list to array

kpol · Mar 29, 2024 · 58a2cfb · 58a2cfb
1 parent ce611a3
commit 58a2cfb
Show file tree

Hide file tree

Showing 8 changed files with 370,198 additions and 72 deletions.
diff --git a/README.md b/README.md
@@ -1,7 +1,7 @@
 Trie
 ------
 **Trie** (a.k.a. prefix tree)  is an ordered tree data structure that is used to store an associative array where the keys are usually strings. All the descendants of a node have a common prefix of the string associated with that node, and the root is associated with the empty string.  
-*Reference*: [Wikipedia &ndash; trie](http://en.wikipedia.org/wiki/Trie)
+*Reference*: [Wikipedia](http://en.wikipedia.org/wiki/Trie)
 
 [![CI Build](https://github.com/kpol/trie/workflows/CI%20Build/badge.svg)](https://github.com/kpol/trie/actions?query=workflow%3A%22CI+Build%22)
 [![Nuget](https://img.shields.io/nuget/v/KTrie.svg?logo=nuget)](https://www.nuget.org/packages/KTrie)
@@ -67,20 +67,21 @@ There are two overloads of the `StartsWith` method:
 
 Benchmark tests
 ------
-For performance tests I used 58110 English words of length from 2 to 22 chars. The table below shows prefix lookup time comparing to the Linq `Where` and `string.StartsWith`. Number of prefixes: 10
-
-| Method                         | Mean          | Error       | StdDev      | Allocated |
-|------------------------------- |--------------:|------------:|------------:|----------:|
-| Trie_StartsWith                |  1,663.334 us |  25.0298 us |  22.1883 us |  782258 B |
-| LinqSimple_StartsWith          | 17,899.727 us | 178.2255 us | 157.9923 us |  675940 B |
-| Linq_StartsWith                |  1,880.081 us |  22.4351 us |  20.9858 us |  676893 B |
-| Linq_DictionaryWithAllPrefixes |    775.352 us |   7.5212 us |   6.6673 us |  673053 B |
-| Trie_Matches                   |      5.389 us |   0.0623 us |   0.0583 us |    9096 B |
-| Trie_PatternStartsWith         |     10.924 us |   0.2181 us |   0.4455 us |   14896 B |
-| String_PatternMatching         |    116.097 us |   2.0039 us |   2.6057 us |     416 B |
-| String_PrefixPatternMatching   |    108.479 us |   1.8731 us |   1.7521 us |    3432 B |
-| Regex_PatternMatching          |  4,410.587 us |  87.8454 us |  90.2107 us |     419 B |
-| Regex_PrefixPatternMatching    |  4,309.215 us |  54.2987 us |  50.7910 us |    3435 B |
+For performance tests I used 370105 English words (from: https://github.com/dwyl/english-words).
+
+| Method                               | Mean          | Error        | StdDev       | Allocated   |
+|------------------------------------- |--------------:|-------------:|-------------:|------------:|
+| Load_Trie                            | 217,385.15 us | 4,059.770 us | 4,343.909 us | 72741.36 KB |
+| Trie_StartsWith                      |  11,394.07 us |   219.067 us |   466.849 us |  3604.64 KB |
+| Linq_StartsWith                      | 113,231.21 us |   780.126 us |   729.730 us |  2843.55 KB |
+| Linq_GroupedByFirstLetter_StartsWith |  10,244.17 us |    91.502 us |    85.591 us |  2844.41 KB |
+| Linq_DictionaryWithAllPrefixes       |   4,194.10 us |    41.829 us |    39.127 us |  2840.66 KB |
+| Trie_Matches                         |      15.03 us |     0.287 us |     0.268 us |    18.05 KB |
+| Trie_PatternStartsWith               |      62.98 us |     0.482 us |     0.451 us |    65.65 KB |
+| String_PatternMatching               |     875.47 us |     7.239 us |     6.045 us |     1.56 KB |
+| String_PrefixPatternMatching         |     895.72 us |     3.407 us |     2.660 us |    33.72 KB |
+| Regex_PatternMatching                |  26,587.50 us |   206.420 us |   182.986 us |     1.57 KB |
+| Regex_PrefixPatternMatching          |  27,545.88 us |   188.291 us |   176.127 us |    33.73 KB |
 
 ------
 &copy; Kirill Polishchuk
diff --git a/src/KTrie.TestBenchmark/KTrie.TestBenchmark.csproj b/src/KTrie.TestBenchmark/KTrie.TestBenchmark.csproj
@@ -3,7 +3,7 @@
 	<PropertyGroup>
 		<OutputType>Exe</OutputType>
 		<TargetFramework>net8.0</TargetFramework>
-		<IsPackable >False</IsPackable>
+		<IsPackable>False</IsPackable>
 	</PropertyGroup>
 
 	<ItemGroup>
@@ -22,6 +22,9 @@
 		<None Update="TestData\vocabulary.txt">
 			<CopyToOutputDirectory>Always</CopyToOutputDirectory>
 		</None>
+		<None Update="TestData\words_alpha.txt">
+		  <CopyToOutputDirectory>Always</CopyToOutputDirectory>
+		</None>
 	</ItemGroup>
 
 </Project>
diff --git a/src/KTrie.TestBenchmark/Program.cs b/src/KTrie.TestBenchmark/Program.cs
@@ -2,5 +2,4 @@
 using KTrie.TestBenchmark;
 
 
-var _ = BenchmarkRunner.Run<StringTrieTest>();
-
+BenchmarkRunner.Run<StringTrieTest>();
diff --git a/src/KTrie.TestBenchmark/StringTrieTest.cs b/src/KTrie.TestBenchmark/StringTrieTest.cs
@@ -9,10 +9,10 @@ namespace KTrie.TestBenchmark;
 [MemoryDiagnoser(false)]
 public class StringTrieTest
 {
-    private readonly string[] _words;
-    private readonly ILookup<char, string> _wordGroups;
-    private readonly Dictionary<string, List<string>> _dictWithAllPrefixes;
-    private readonly Trie _trie;
+    private string[] _words;
+    private ILookup<char, string> _wordGroups;
+    private Dictionary<string, List<string>> _dictWithAllPrefixes;
+    private Trie _trie;
 
     private readonly string[] _prefixes =
     [
@@ -38,7 +38,8 @@ public class StringTrieTest
         "ve"
     ];
 
-    public StringTrieTest()
+    [GlobalSetup]
+    public void GlobalSetup()
     {
         _words = GetWords();
         _wordGroups = PreprocessWords();
@@ -47,6 +48,12 @@ public StringTrieTest()
         _trie = [.. _words];
     }
 
+    [Benchmark]
+    public void Load_Trie()
+    {
+        Trie _ = [.. _words];
+    }
+
     [Benchmark]
     public ICollection<string> Trie_StartsWith()
     {
@@ -64,7 +71,7 @@ public ICollection<string> Trie_StartsWith()
     }
 
     [Benchmark]
-    public ICollection<string> LinqSimple_StartsWith()
+    public ICollection<string> Linq_StartsWith()
     {
         HashSet<string> result = [];
 
@@ -80,13 +87,14 @@ public ICollection<string> LinqSimple_StartsWith()
     }
 
     [Benchmark]
-    public ICollection<string> Linq_StartsWith()
+    public ICollection<string> Linq_GroupedByFirstLetter_StartsWith()
     {
         HashSet<string> result = [];
 
         foreach (var prefix in _prefixes)
         {
             var firstLetter = prefix[0];
+
             foreach (var word in _wordGroups[firstLetter].Where(w => w.StartsWith(prefix)))
             {
                 result.Add(word);
@@ -139,7 +147,7 @@ public ICollection<string> Regex_PatternMatching() =>
     public ICollection<string> Regex_PrefixPatternMatching() =>
         _words.Where(word => Regex.IsMatch(word, "^.c.{2}t")).ToHashSet();
 
-    private static string[] GetWords() => File.ReadAllLines("TestData/vocabulary.txt");
+    private static string[] GetWords() => File.ReadAllLines("TestData/words_alpha.txt");
 
     private ILookup<char, string> PreprocessWords()
     {
Original file line number	Diff line number	Diff line change
Expand Up		@@ -2,5 +2,4 @@
		using KTrie.TestBenchmark;


		var _ = BenchmarkRunner.Run<StringTrieTest>();

		BenchmarkRunner.Run<StringTrieTest>();