Feature/web api

06-web-api/full-text-search.sln

@@ -0,0 +1,34 @@
+
+Microsoft Visual Studio Solution File, Format Version 12.00
+# Visual Studio Version 17
+VisualStudioVersion = 17.0.31903.59
+MinimumVisualStudioVersion = 10.0.40219.1
+Project("{9A19103F-16F7-4668-BE54-9A1E7A4F7556}") = "DocumentManagement", "src\DocumentManagement\DocumentManagement.csproj", "{D471FD38-26BA-4DEF-96A2-982F235AEA01}"
+EndProject
+Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "test", "test\test.csproj", "{FD776EED-DBA5-4789-A1A7-FF97EF11FBD3}"
+EndProject
+Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "MyWebApplication", "src\MyWebApplication\MyWebApplication.csproj", "{FD4AE171-A581-4FD2-899C-C1BB5EAFE641}"
+EndProject
+Global
+	GlobalSection(SolutionConfigurationPlatforms) = preSolution
+		Debug|Any CPU = Debug|Any CPU
+		Release|Any CPU = Release|Any CPU
+	EndGlobalSection
+	GlobalSection(ProjectConfigurationPlatforms) = postSolution
+		{D471FD38-26BA-4DEF-96A2-982F235AEA01}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
+		{D471FD38-26BA-4DEF-96A2-982F235AEA01}.Debug|Any CPU.Build.0 = Debug|Any CPU
+		{D471FD38-26BA-4DEF-96A2-982F235AEA01}.Release|Any CPU.ActiveCfg = Release|Any CPU
+		{D471FD38-26BA-4DEF-96A2-982F235AEA01}.Release|Any CPU.Build.0 = Release|Any CPU
+		{FD776EED-DBA5-4789-A1A7-FF97EF11FBD3}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
+		{FD776EED-DBA5-4789-A1A7-FF97EF11FBD3}.Debug|Any CPU.Build.0 = Debug|Any CPU
+		{FD776EED-DBA5-4789-A1A7-FF97EF11FBD3}.Release|Any CPU.ActiveCfg = Release|Any CPU
+		{FD776EED-DBA5-4789-A1A7-FF97EF11FBD3}.Release|Any CPU.Build.0 = Release|Any CPU
+		{FD4AE171-A581-4FD2-899C-C1BB5EAFE641}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
+		{FD4AE171-A581-4FD2-899C-C1BB5EAFE641}.Debug|Any CPU.Build.0 = Debug|Any CPU
+		{FD4AE171-A581-4FD2-899C-C1BB5EAFE641}.Release|Any CPU.ActiveCfg = Release|Any CPU
+		{FD4AE171-A581-4FD2-899C-C1BB5EAFE641}.Release|Any CPU.Build.0 = Release|Any CPU
+	EndGlobalSection
+	GlobalSection(SolutionProperties) = preSolution
+		HideSolutionNode = FALSE
+	EndGlobalSection
+EndGlobal

06-web-api/src/DocumentManagement/DocumentManagement.csproj

@@ -0,0 +1,10 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+    <PropertyGroup>
+        <TargetFramework>net8.0</TargetFramework>
+        <ImplicitUsings>enable</ImplicitUsings>
+        <Nullable>enable</Nullable>
+        <RootNamespace>Mohaymen.FullTextSearch.DocumentManagement</RootNamespace>
+    </PropertyGroup>
+
+</Project>

06-web-api/src/DocumentManagement/Interfaces/IFileReader.cs

@@ -0,0 +1,8 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+public interface IFileReader
+{
+    FileCollection ReadAllFiles(string folderPath);
+}

06-web-api/src/DocumentManagement/Interfaces/IInvertedIndex.cs

@@ -0,0 +1,9 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+public interface IInvertedIndex
+{
+    HashSet<string> AllDocuments { get; }
+    HashSet<string> GetDocumentsByKeyword(Keyword keyword);
+}

06-web-api/src/DocumentManagement/Interfaces/IInvertedIndexBuilder.cs

@@ -0,0 +1,9 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+public interface IInvertedIndexBuilder
+{ 
+    IInvertedIndex Build();
+    IInvertedIndexBuilder IndexFilesWords(FileCollection fileCollection);
+}

06-web-api/src/DocumentManagement/Interfaces/ISearchStrategy.cs

@@ -0,0 +1,8 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+public interface ISearchStrategy
+{
+    void FilterDocuments(HashSet<string> documents, List<Keyword> keywords, IInvertedIndex invertedIndex);
+}

06-web-api/src/DocumentManagement/Interfaces/ISearcher.cs

@@ -0,0 +1,7 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+namespace Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+public interface ISearcher<T>
+{
+    ICollection<T> Search(List<SearchQuery> query);
+}

06-web-api/src/DocumentManagement/Interfaces/ITokenizer.cs

@@ -0,0 +1,8 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+public interface ITokenizer
+{
+    List<Keyword> ExtractKeywords(string text);
+}

06-web-api/src/DocumentManagement/Models/AdvancedInvertedIndex.cs

@@ -0,0 +1,90 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+public class AdvancedInvertedIndex : IInvertedIndex, IEquatable<AdvancedInvertedIndex>
+{
+    private readonly Dictionary<Keyword, HashSet<KeywordInfo>> _invertedIndexMap = [];
+    public HashSet<string> AllDocuments { get; } = [];
+    private ITokenizer _tokenizer;
+
+    public AdvancedInvertedIndex(ITokenizer tokenizer)
+    {
+        _tokenizer = tokenizer;
+    }
+
+    public void AddDocumentToKeyword(Keyword keyword, KeywordInfo keywordInfo)
+    {
+        AllDocuments.Add(keywordInfo.Document);
+
+        if (!_invertedIndexMap.ContainsKey(keyword))
+            _invertedIndexMap.Add(keyword, []);
+
+        _invertedIndexMap[keyword].Add(keywordInfo);
+    }
+    public HashSet<string> GetDocumentsByKeyword(Keyword phrase)
+    {
+        var phraseWords = _tokenizer.ExtractKeywords(phrase.Word);
+
+        if (!phraseWords.Any())
+        {
+            return [];
+        }
+
+        foreach (var phraseWord in phraseWords)
+        {
+            if (!_invertedIndexMap.ContainsKey(phraseWord))
+            {
+                return [];
+            }
+        }
+
+        var keywordInfos = new HashSet<KeywordInfo>(_invertedIndexMap[phraseWords[0]]);
+
+        for (int i=1; i<phraseWords.Count; i++)
+        {
+            var currentKeywordInfos = new HashSet<KeywordInfo>(_invertedIndexMap[phraseWords[i]]);
+            currentKeywordInfos.RemoveWhere(keywordInfo =>
+                !keywordInfos.Contains(new KeywordInfo(keywordInfo.Document, keywordInfo.Position - 1))
+            );
+
+            keywordInfos = currentKeywordInfos;
+        }
+
+        return keywordInfos.Select(keywordInfo => keywordInfo.Document).ToHashSet();
+    }
+
+
+    public bool Equals(AdvancedInvertedIndex? other)
+    {
+        if (other is null) return false;
+
+        if (_invertedIndexMap.Count != other._invertedIndexMap.Count)
+            return false;
+
+        foreach (var kvp in _invertedIndexMap)
+        {
+            if (!other._invertedIndexMap.TryGetValue(kvp.Key, out var otherSet))
+                return false;
+
+            if (!kvp.Value.SetEquals(otherSet))
+                return false;
+        }
+
+        var areDocumentsEqual = AllDocuments.SetEquals(other.AllDocuments);
+        return areDocumentsEqual;
+    }
+
+    public override bool Equals(object? obj)
+    {
+        if (ReferenceEquals(null, obj)) return false;
+        if (ReferenceEquals(this, obj)) return true;
+        if (obj.GetType() != this.GetType()) return false;
+        return Equals((AdvancedInvertedIndex)obj);
+    }
+
+    public override int GetHashCode()
+    {
+        return HashCode.Combine(_invertedIndexMap, AllDocuments);
+    }
+}

06-web-api/src/DocumentManagement/Models/FileCollection.cs

@@ -0,0 +1,21 @@
+namespace Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+public class FileCollection
+{
+    private readonly Dictionary<string, string> _filesDataDictionary = [];
+
+    public void AddFile(string filePath, string fileContent) =>
+        _filesDataDictionary.Add(filePath, fileContent);
+
+    public List<string> GetFilesPath() =>
+        _filesDataDictionary.Keys.ToList();
+
+    public string GetFileContent(string filePath) =>
+        _filesDataDictionary[filePath];
+
+    public bool ContainsFile(string filePath) =>
+        _filesDataDictionary.ContainsKey(filePath);
+
+    public int FilesCount() =>
+        _filesDataDictionary.Count;
+}

06-web-api/src/DocumentManagement/Models/InvertedIndex.cs

@@ -0,0 +1,26 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+public class InvertedIndex : IInvertedIndex
+{
+    private readonly Dictionary<Keyword, HashSet<string>> _invertedIndexMap = [];
+    public HashSet<string> AllDocuments { get; } = [];
+
+    public void AddDocumentToKeyword(Keyword keyword, string document)
+    {
+        AllDocuments.Add(document);
+
+        if (!_invertedIndexMap.ContainsKey(keyword))
+            _invertedIndexMap.Add(keyword, []);
+
+        _invertedIndexMap[keyword].Add(document);
+    }
+
+    public HashSet<string> GetDocumentsByKeyword(Keyword keyword)
+    {
+        _invertedIndexMap.TryGetValue(keyword, out HashSet<string>? documents);
+
+        return documents ?? [];
+    }
+}

06-web-api/src/DocumentManagement/Models/Keyword.cs

@@ -0,0 +1,11 @@
+namespace Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+public record Keyword
+{
+    public Keyword(string word)
+    {
+        Word = word.ToUpper();
+    }
+
+    public string Word { get; init; }
+}

06-web-api/src/DocumentManagement/Models/KeywordInfo.cs

@@ -0,0 +1,3 @@
+namespace Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+public record KeywordInfo(string Document, int Position);

06-web-api/src/DocumentManagement/Models/SearchQuery.cs

@@ -0,0 +1,5 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+public record SearchQuery(ISearchStrategy SearchStrategy, List<Keyword> Keywords);

06-web-api/src/DocumentManagement/Services/Files/FileReader.cs

@@ -0,0 +1,25 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+namespace Mohaymen.FullTextSearch.DocumentManagement.Services.FilesService;
+
+public class FileReader : IFileReader
+{
+    public FileCollection ReadAllFiles(string folderPath)
+    {
+        var files = Directory.GetFiles(folderPath);
+        var fileCollection = files.Aggregate(new FileCollection(), AddFileToCollection);
+
+        return fileCollection;
+    }
+
+    private FileCollection AddFileToCollection(FileCollection collection, string filePath)
+    {
+        if (!collection.ContainsFile(filePath))
+            collection.AddFile(
+                filePath,
+                File.ReadAllText(filePath)
+            );
+
+        return collection;
+    }
+}

06-web-api/src/DocumentManagement/Services/InvertedIndex/FilesAdvancedInvertedIndexBuilder.cs

@@ -0,0 +1,41 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Services.InvertedIndexService;
+
+public class FilesAdvancedInvertedIndexBuilder : IInvertedIndexBuilder
+{
+    private AdvancedInvertedIndex _advancedInvertedIndex;
+    private ITokenizer _tokenizer;
+
+    public FilesAdvancedInvertedIndexBuilder(ITokenizer tokenizer)
+    {
+        _advancedInvertedIndex = new(tokenizer);
+        _tokenizer = tokenizer;
+    }
+
+    public IInvertedIndexBuilder IndexFilesWords(FileCollection fileCollection)
+    {
+        foreach (var filePath in fileCollection.GetFilesPath())
+        {
+            var keywords = _tokenizer.ExtractKeywords(fileCollection.GetFileContent(filePath));
+            UpdateInvertedIndexMap(keywords, filePath);
+        }
+
+        return this;
+    }
+
+    private void UpdateInvertedIndexMap(List<Keyword> keywords, string filePath)
+    {
+        for (var i = 0; i < keywords.Count; i++)
+        {
+            _advancedInvertedIndex.AddDocumentToKeyword(keywords[i], new KeywordInfo(filePath, i));    
+        }
+    }
+
+
+    public IInvertedIndex Build()
+    {
+        return _advancedInvertedIndex;
+    }
+}

06-web-api/src/DocumentManagement/Services/InvertedIndex/FilesInvertedIndexBuilder.cs

@@ -0,0 +1,40 @@
+using Mohaymen.FullTextSearch.DocumentManagement.Interfaces;
+using Mohaymen.FullTextSearch.DocumentManagement.Models;
+using Mohaymen.FullTextSearch.DocumentManagement.Utilities;
+
+namespace Mohaymen.FullTextSearch.DocumentManagement.Services.InvertedIndexService;
+
+public class FilesInvertedIndexBuilder : IInvertedIndexBuilder
+{
+    private readonly InvertedIndex _invertedIndex = new();
+    private readonly ITokenizer _tokenizer;
+
+    public FilesInvertedIndexBuilder(ITokenizer tokenizer)
+    {
+        _tokenizer = tokenizer;
+    }
+
+    public IInvertedIndexBuilder IndexFilesWords(FileCollection fileCollection)
+    {
+        foreach (var filePath in fileCollection.GetFilesPath())
+        {
+            var keywords = _tokenizer.ExtractKeywords(fileCollection.GetFileContent(filePath));
+            UpdateInvertedIndexMap(keywords, filePath);
+        }
+
+        return this;
+    }
+
+    private void UpdateInvertedIndexMap(List<Keyword> keywords, string filePath)
+    {
+        foreach (var keyword in keywords)
+        {
+            _invertedIndex.AddDocumentToKeyword(keyword, filePath);
+        }
+    }
+
+    public IInvertedIndex Build()
+    {
+        return _invertedIndex;
+    }
+}