splitBraceに渡すテキストが不適切に分割される問題の修正

Epub/KoeBook.Epub/Contracts/Services/ISplitBraceService.cs

@@ -0,0 +1,6 @@
+namespace KoeBook.Epub.Contracts.Services;
+
+public interface ISplitBraceService
+{
+    IEnumerable<string> SplitBrace(string text);
+}

Epub/KoeBook.Epub/Services/ScrapingAozoraService.cs

@@ -6,15 +6,16 @@
 using KoeBook.Epub.Contracts.Services;
 using KoeBook.Epub.Models;
 using Microsoft.Extensions.DependencyInjection;
-using static KoeBook.Epub.Utility.ScrapingHelper;
 
 
 namespace KoeBook.Epub.Services
 {
-    public partial class ScrapingAozoraService([FromKeyedServices(nameof(ScrapingAozoraService))] IScrapingClientService scrapingClientService) : IScrapingService
+    public partial class ScrapingAozoraService(ISplitBraceService splitBraceService, [FromKeyedServices(nameof(ScrapingAozoraService))] IScrapingClientService scrapingClientService) : IScrapingService
     {
+        private readonly ISplitBraceService _splitBraceService = splitBraceService;
         private readonly IScrapingClientService _scrapingClientService = scrapingClientService;
 
+
         public bool IsMatchSite(Uri uri)
         {
             return uri.Host == "www.aozora.gr.jp";
@@ -175,7 +176,7 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
                                 paragraph.Text += TextProcess(midashi);
                                 document.Chapters[chapterNum].Sections[sectionNum].Elements.Add(new Paragraph());
 
-                                foreach (var splitText in SplitBrace(TextProcess(midashi)))
+                                foreach (var splitText in _splitBraceService.SplitBrace(TextProcess(midashi)))
                                 {
                                     if (document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph1)
                                     {
@@ -194,7 +195,7 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
                             document.EnsureParagraph(chapterNum, sectionNum);
                             if ((document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph))
                             {
-                                var split = SplitBrace(TextProcess(element));
+                                var split = _splitBraceService.SplitBrace(TextProcess(element)).ToList();
                                 for (int i = 0; i < split.Count - 1; i++)
                                 {
                                     if (document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph1)
@@ -232,7 +233,7 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
                             document.EnsureParagraph(chapterNum, sectionNum);
                             if ((document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph))
                             {
-                                foreach (var splitText in SplitBrace(TextProcess(element)))
+                                foreach (var splitText in _splitBraceService.SplitBrace(TextProcess(element)))
                                 {
                                     if (document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph1)
                                     {
@@ -343,7 +344,7 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
                                 document.EnsureParagraph(chapterNum, sectionNum);
                                 if ((document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph))
                                 {
-                                    foreach (var splitText in SplitBrace(TextProcess(element)))
+                                    foreach (var splitText in _splitBraceService.SplitBrace(TextProcess(element)))
                                     {
                                         if (document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph1)
                                         {
@@ -378,7 +379,7 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
                         document.EnsureParagraph(chapterNum, sectionNum);
                         if ((document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph))
                         {
-                            var split = SplitBrace(TextProcess(element));
+                            var split = _splitBraceService.SplitBrace(TextProcess(element)).ToList();
                             for (int i = 0; i < split.Count - 1; i++)
                             {
                                 if (document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph1)
@@ -420,7 +421,7 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
                     {
                         paragraph.Text += TextProcess(element);
 
-                        var split = SplitBrace(TextProcess(element));
+                        var split = _splitBraceService.SplitBrace(TextProcess(element)).ToList();
                         for (int i = 0; i < split.Count - 1; i++)
                         {
                             if (document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph1)
@@ -466,7 +467,7 @@ public async ValueTask<EpubDocument> ScrapingAsync(string url, string coverFileP
                             document.EnsureParagraph(chapterNum, sectionNum);
                             if ((document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph))
                             {
-                                var split = SplitBrace(TextReplace(nextNode.Text()));
+                                var split = _splitBraceService.SplitBrace(TextReplace(nextNode.Text())).ToList();
                                 for (int i = 0; i < split.Count - 1; i++)
                                 {
                                     if (document.Chapters[chapterNum].Sections[sectionNum].Elements[^1] is Paragraph paragraph1)

Epub/KoeBook.Epub/Services/ScrapingNaroService.cs

@@ -4,16 +4,17 @@
 using AngleSharp.Html.Dom;
 using AngleSharp.Io;
 using KoeBook.Core;
+using KoeBook.Core.Utility;
 using KoeBook.Epub.Contracts.Services;
 using KoeBook.Epub.Models;
 using Microsoft.Extensions.DependencyInjection;
-using static KoeBook.Epub.Utility.ScrapingHelper;
 
 namespace KoeBook.Epub.Services
 {
-    public partial class ScrapingNaroService(IHttpClientFactory httpClientFactory, [FromKeyedServices(nameof(ScrapingNaroService))] IScrapingClientService scrapingClientService) : IScrapingService
+    public partial class ScrapingNaroService(IHttpClientFactory httpClientFactory, ISplitBraceService splitBraceService, [FromKeyedServices(nameof(ScrapingNaroService))] IScrapingClientService scrapingClientService) : IScrapingService
     {
         private readonly IHttpClientFactory _httpCliantFactory = httpClientFactory;
+        private readonly ISplitBraceService _splitBraceService = splitBraceService;
         private readonly IScrapingClientService _scrapingClientService = scrapingClientService;
 
         public bool IsMatchSite(Uri uri)
@@ -135,8 +136,10 @@ public record BookInfo(int? allcount, int? noveltype, int? general_all_no);
 
         private record SectionWithChapterTitle(string? title, Section section);
 
-        private static async Task<SectionWithChapterTitle> ReadPageAsync(string url, bool isRensai, string imageDirectory, CancellationToken ct)
+        private async ValueTask<SectionWithChapterTitle> ReadPageAsync(string url, bool isRensai, string imageDirectory, CancellationToken ct)
         {
+            var store = new StringStorer();
+
             var config = Configuration.Default.WithDefaultLoader();
             using var context = BrowsingContext.New(config);
             var doc = await context.OpenAsync(url, ct).ConfigureAwait(false);
@@ -171,7 +174,6 @@ private static async Task<SectionWithChapterTitle> ReadPageAsync(string url, boo
 
             var section = new Section(sectionTitleElement.InnerHtml);
 
-
             var main_text = doc.QuerySelector("#novel_honbun")
                 ?? throw new EbookException(ExceptionType.WebScrapingFailed, "本文がありません");
 
@@ -184,10 +186,7 @@ private static async Task<SectionWithChapterTitle> ReadPageAsync(string url, boo
                 {
                     if (!string.IsNullOrWhiteSpace(item.InnerHtml))
                     {
-                        foreach (var split in SplitBrace(item.InnerHtml))
-                        {
-                            section.Elements.Add(new Paragraph() { Text = split });
-                        }
+                        store.Store(item.InnerHtml);
                     }
                 }
                 else if (item.ChildElementCount == 1)
@@ -221,13 +220,17 @@ private static async Task<SectionWithChapterTitle> ReadPageAsync(string url, boo
                     {
                         if (!string.IsNullOrWhiteSpace(item.InnerHtml))
                         {
-                            foreach (var split in SplitBrace(item.InnerHtml))
-                            {
-                                section.Elements.Add(new Paragraph() { Text = split });
-                            }
+                            store.Store(item.InnerHtml);
                         }
                     }
-                    else if (item.Children[0] is not IHtmlBreakRowElement)
+                    else if (item.Children[0] is IHtmlBreakRowElement)
+                    {
+                        foreach (var split in _splitBraceService.SplitBrace(store.Release()))
+                        {
+                            section.Elements.Add(new Paragraph() { Text = split });
+                        }
+                    }
+                    else
                         throw new EbookException(ExceptionType.UnexpectedStructure);
                 }
                 else
@@ -247,16 +250,18 @@ private static async Task<SectionWithChapterTitle> ReadPageAsync(string url, boo
 
                     if (!string.IsNullOrWhiteSpace(item.InnerHtml))
                     {
-                        foreach (var split in SplitBrace(item.InnerHtml))
-                        {
-                            section.Elements.Add(new Paragraph() { Text = split });
-                        }
+                        store.Store(item.InnerHtml);
                     }
                 }
+                foreach (var split in _splitBraceService.SplitBrace(store.Release()))
+                {
+                    section.Elements.Add(new Paragraph() { Text = split });
+                }
             }
             return new SectionWithChapterTitle(chapterTitle, section);
         }
 
+
         [System.Text.RegularExpressions.GeneratedRegex(@"https://.{5,7}.syosetu.com/(.{7}).?")]
         private static partial System.Text.RegularExpressions.Regex UrlToNcode();
 

Epub/KoeBook.Epub/Utility/ScrapingHelper.cs → Epub/KoeBook.Epub/Services/SplitBraceService.cs

@@ -1,11 +1,21 @@
-namespace KoeBook.Epub.Utility;
+using System.Text;
+using KoeBook.Epub.Contracts.Services;
 
-public static class ScrapingHelper
+namespace KoeBook.Epub.Services;
+
+public class SplitBraceService : ISplitBraceService
 {
-    public static List<string> SplitBrace(string text)
+    public IEnumerable<string> SplitBrace(string text)
     {
-        if (text.Length == 1 && (text == "「" || text == "『" || text == "」" || text == "』"))
-            return [text];
+        // textが空白だった時 paragraph を挿入する処理をスキップ
+        if (string.IsNullOrWhiteSpace(text))
+            yield break;
+
+        if (text.Length == 1)
+        {
+            yield return text;
+            yield break;
+        }
 
         var bracket = 0;
         var brackets = new int[text.Length];
@@ -17,28 +27,25 @@ public static List<string> SplitBrace(string text)
             brackets[i] = bracket;
         }
 
-        var result = new List<string>();
         var mn = Math.Min(0, brackets.Min());
         var startIdx = 0;
         for (var i = 0; i < brackets.Length; i++)
         {
             brackets[i] -= mn;
             if ((text[i] == '「' || text[i] == '『') && brackets[i] == 1 && i != 0 && startIdx != i)
             {
-                result.Add(text[startIdx..i]);
+                yield return text[startIdx..i];
                 startIdx = i;
             }
             if ((text[i] == '」' || text[i] == '』') && brackets[i] == 0)
             {
-                result.Add(text[startIdx..(i + 1)]);
+                yield return text[startIdx..(i + 1)];
                 startIdx = i + 1;
             }
         }
         if (startIdx != text.Length)
         {
-            result.Add(text[startIdx..]);
+            yield return text[startIdx..];
         }
-
-        return result;
     }
 }

KoeBook.Core/Utility/StringStorer.cs

@@ -0,0 +1,25 @@
+using System;
+using System.Collections.Generic;
+using System.Linq;
+using System.Text;
+using System.Threading.Tasks;
+
+namespace KoeBook.Core.Utility
+{
+    public class StringStorer
+    {
+        private StringBuilder _stringBuilder = new();
+
+        public void Store(string text)
+        {
+            _stringBuilder.Append(text);
+        }
+
+        public string Release()
+        {
+            var result = _stringBuilder.ToString();
+            _stringBuilder.Clear();
+            return result;
+        }
+    }
+}

KoeBook.Test/Epub/ScrapingHelperTest.cs

@@ -1,4 +1,4 @@
-using KoeBook.Epub.Utility;
+using KoeBook.Epub.Services;
 
 namespace KoeBook.Test.Epub;
 
@@ -83,7 +83,7 @@ public static object[][] TestCases()
     [MemberData(nameof(TestCases))]
     public void SplitBraceTest(string text, List<string> expected)
     {
-        Assert.Equal(expected, ScrapingHelper.SplitBrace(text));
-
+        var helper = new SplitBraceService();
+        Assert.Equal(expected, helper.SplitBrace(text));
     }
 }

KoeBook/App.xaml.cs

@@ -110,6 +110,7 @@ public App()
                     .AddSingleton<IScrapingService, ScrapingAozoraService>()
                     .AddSingleton<IScrapingService, ScrapingNaroService>();
                 services.AddSingleton<IEpubCreateService, EpubCreateService>();
+                services.AddSingleton<ISplitBraceService, SplitBraceService>();
                 services.AddSingleton<IFileExtensionService, FileExtensionService>();
 
                 // Views and ViewModels