Epub Text Bleeding Finally Fixed! (#4086)

Co-authored-by: Amelia <77553571+Fesaa@users.noreply.github.com> Co-authored-by: Gazy Mahomar <gmahomarf@users.noreply.github.com> Co-authored-by: Stefans.A <104719225+privatestefans@users.noreply.github.com>
2026-05-27 18:12:34 -04:00 · 2025-10-11 09:18:54 -05:00
parent 75e844404c
commit f7dca3806f
48 changed files with 4616 additions and 186 deletions
@@ -35,7 +35,8 @@ public class ScannerHelper
    private readonly string _testDirectory = Path.Join(Directory.GetCurrentDirectory(), "../../../Services/Test Data/ScannerService/ScanTests");
    private readonly string _testcasesDirectory = Path.Join(Directory.GetCurrentDirectory(), "../../../Services/Test Data/ScannerService/TestCases");
    private readonly string _imagePath = Path.Join(Directory.GetCurrentDirectory(), "../../../Services/Test Data/ScannerService/1x1.png");
-    private static readonly string[] ComicInfoExtensions = new[] { ".cbz", ".cbr", ".zip", ".rar" };
+    private static readonly string[] ComicInfoExtensions = [".cbz", ".cbr", ".zip", ".rar"];
+    private static readonly string[] EpubExtensions = [".epub"];

    public ScannerHelper(IUnitOfWork unitOfWork, ITestOutputHelper testOutputHelper)
    {
@@ -43,7 +44,7 @@ public class ScannerHelper
        _testOutputHelper = testOutputHelper;
    }

-    public async Task<Library> GenerateScannerData(string testcase, Dictionary<string, ComicInfo> comicInfos = null)
+    public async Task<Library> GenerateScannerData(string testcase, Dictionary<string, ComicInfo>? comicInfos = null)
    {
        var testDirectoryPath = await GenerateTestDirectory(Path.Join(_testcasesDirectory, testcase), comicInfos);

@@ -113,7 +114,7 @@ public class ScannerHelper



-    private async Task<string> GenerateTestDirectory(string mapPath, Dictionary<string, ComicInfo> comicInfos = null)
+    private async Task<string> GenerateTestDirectory(string mapPath, Dictionary<string, ComicInfo>? comicInfos = null)
    {
        // Read the map file
        var mapContent = await File.ReadAllTextAsync(mapPath);
@@ -130,7 +131,7 @@ public class ScannerHelper
        Directory.CreateDirectory(testDirectory);

        // Generate the files and folders
-        await Scaffold(testDirectory, filePaths, comicInfos);
+        await Scaffold(testDirectory, filePaths ?? [], comicInfos);

        _testOutputHelper.WriteLine($"Test Directory Path: {testDirectory}");

@@ -138,7 +139,7 @@ public class ScannerHelper
    }


-    public async Task Scaffold(string testDirectory, List<string> filePaths, Dictionary<string, ComicInfo> comicInfos = null)
+    public async Task Scaffold(string testDirectory, List<string> filePaths, Dictionary<string, ComicInfo>? comicInfos = null)
    {
        foreach (var relativePath in filePaths)
        {
@@ -157,6 +158,10 @@ public class ScannerHelper
            {
                CreateMinimalCbz(fullPath, info);
            }
+            else if (EpubExtensions.Contains(ext) && comicInfos != null && comicInfos.TryGetValue(Path.GetFileName(relativePath), out var epubInfo))
+            {
+                CreateMinimalEpub(fullPath, epubInfo);
+            }
            else
            {
                // Create an empty file
@@ -205,4 +210,165 @@ public class ScannerHelper
        return stringWriter.ToString().Replace("""<?xml version="1.0" encoding="utf-16"?>""",
            @"<?xml version='1.0' encoding='utf-8'?>");
    }
+
+    private void CreateMinimalEpub(string filePath, ComicInfo? comicInfo = null)
+    {
+        using (var archive = ZipFile.Open(filePath, ZipArchiveMode.Create))
+        {
+            // EPUB requires a mimetype file as the first entry (uncompressed)
+            var mimetypeEntry = archive.CreateEntry("mimetype", CompressionLevel.NoCompression);
+            using (var mimetypeStream = mimetypeEntry.Open())
+            using (var writer = new StreamWriter(mimetypeStream, Encoding.ASCII))
+            {
+                writer.Write("application/epub+zip");
+            }
+
+            // Create META-INF/container.xml
+            var containerEntry = archive.CreateEntry("META-INF/container.xml");
+            using (var containerStream = containerEntry.Open())
+            using (var writer = new StreamWriter(containerStream, Encoding.UTF8))
+            {
+                writer.Write("""
+                    <?xml version="1.0" encoding="UTF-8"?>
+                    <container version="1.0" xmlns="urn:oasis:names:tc:opendocument:xmlns:container">
+                        <rootfiles>
+                            <rootfile full-path="OEBPS/content.opf" media-type="application/oebps-package+xml"/>
+                        </rootfiles>
+                    </container>
+                    """);
+            }
+
+            // Create content.opf with metadata
+            var contentOpf = GenerateContentOpf(comicInfo);
+            var contentEntry = archive.CreateEntry("OEBPS/content.opf");
+            using (var contentStream = contentEntry.Open())
+            using (var writer = new StreamWriter(contentStream, Encoding.UTF8))
+            {
+                writer.Write(contentOpf);
+            }
+
+            // Add a minimal chapter XHTML file
+            var chapterEntry = archive.CreateEntry("OEBPS/chapter1.xhtml");
+            using (var chapterStream = chapterEntry.Open())
+            using (var writer = new StreamWriter(chapterStream, Encoding.UTF8))
+            {
+                writer.Write("""
+                    <?xml version="1.0" encoding="UTF-8"?>
+                    <!DOCTYPE html>
+                    <html xmlns="http://www.w3.org/1999/xhtml">
+                    <head>
+                        <title>Chapter 1</title>
+                    </head>
+                    <body>
+                        <p>Test content.</p>
+                    </body>
+                    </html>
+                    """);
+            }
+
+            // Add the cover image
+            archive.CreateEntryFromFile(_imagePath, "OEBPS/cover.png");
+        }
+        Console.WriteLine($"Created minimal EPUB archive: {filePath} with{(comicInfo != null ? "" : "out")} metadata.");
+    }
+
+    private static string GenerateContentOpf(ComicInfo? comicInfo)
+    {
+        var sb = new StringBuilder();
+        sb.AppendLine("""<?xml version="1.0" encoding="UTF-8"?>""");
+        sb.AppendLine("""<package xmlns="http://www.idpf.org/2007/opf" version="3.0" unique-identifier="book-id">""");
+
+        // Metadata section
+        sb.AppendLine("    <metadata xmlns:dc=\"http://purl.org/dc/elements/1.1/\" xmlns:opf=\"http://www.idpf.org/2007/opf\" xmlns:calibre=\"http://calibre.kovidgoyal.net/2009/metadata\">");
+
+        if (comicInfo != null)
+        {
+            if (!string.IsNullOrEmpty(comicInfo.Title))
+                sb.AppendLine($"        <dc:title>{EscapeXml(comicInfo.Title)}</dc:title>");
+            else
+                sb.AppendLine("        <dc:title>Untitled</dc:title>");
+
+            if (!string.IsNullOrEmpty(comicInfo.Series))
+            {
+                sb.AppendLine($"        <meta property=\"belongs-to-collection\" id=\"collection\">{EscapeXml(comicInfo.Series)}</meta>");
+                sb.AppendLine("        <meta refines=\"#collection\" property=\"collection-type\">series</meta>");
+            }
+
+            if (!string.IsNullOrEmpty(comicInfo.Writer))
+                sb.AppendLine($"        <dc:creator opf:role=\"aut\">{EscapeXml(comicInfo.Writer)}</dc:creator>");
+
+            if (!string.IsNullOrEmpty(comicInfo.Publisher))
+                sb.AppendLine($"        <dc:publisher>{EscapeXml(comicInfo.Publisher)}</dc:publisher>");
+
+            if (!string.IsNullOrEmpty(comicInfo.Summary))
+                sb.AppendLine($"        <dc:description>{EscapeXml(comicInfo.Summary)}</dc:description>");
+
+            if (!string.IsNullOrEmpty(comicInfo.LanguageISO))
+                sb.AppendLine($"        <dc:language>{EscapeXml(comicInfo.LanguageISO)}</dc:language>");
+            else
+                sb.AppendLine("        <dc:language>en</dc:language>");
+
+            if (!string.IsNullOrEmpty(comicInfo.Isbn))
+                sb.AppendLine($"        <dc:identifier id=\"book-id\" opf:scheme=\"ISBN\">{EscapeXml(comicInfo.Isbn)}</dc:identifier>");
+            else
+                sb.AppendLine($"        <dc:identifier id=\"book-id\">urn:uuid:{Guid.NewGuid()}</dc:identifier>");
+
+            if (comicInfo.Year > 0)
+            {
+                var date = $"{comicInfo.Year:D4}";
+                if (comicInfo.Month > 0)
+                {
+                    date += $"-{comicInfo.Month:D2}";
+                    if (comicInfo.Day > 0)
+                        date += $"-{comicInfo.Day:D2}";
+                }
+                sb.AppendLine($"        <dc:date>{date}</dc:date>");
+            }
+
+            if (!string.IsNullOrEmpty(comicInfo.TitleSort))
+                sb.AppendLine($"        <meta name=\"calibre:title_sort\" content=\"{EscapeXml(comicInfo.TitleSort)}\"/>");
+
+            if (!string.IsNullOrEmpty(comicInfo.SeriesSort))
+                sb.AppendLine($"        <meta name=\"calibre:series_sort\" content=\"{EscapeXml(comicInfo.SeriesSort)}\"/>");
+
+            if (!string.IsNullOrEmpty(comicInfo.Number))
+                sb.AppendLine($"        <meta name=\"calibre:series_index\" content=\"{EscapeXml(comicInfo.Number)}\"/>");
+        }
+        else
+        {
+            sb.AppendLine("        <dc:title>Untitled</dc:title>");
+            sb.AppendLine("        <dc:language>en</dc:language>");
+            sb.AppendLine($"        <dc:identifier id=\"book-id\">urn:uuid:{Guid.NewGuid()}</dc:identifier>");
+        }
+
+        sb.AppendLine("    </metadata>");
+
+        // Manifest section
+        sb.AppendLine("    <manifest>");
+        sb.AppendLine("        <item id=\"chapter1\" href=\"chapter1.xhtml\" media-type=\"application/xhtml+xml\"/>");
+        sb.AppendLine("        <item id=\"cover\" href=\"cover.png\" media-type=\"image/png\" properties=\"cover-image\"/>");
+        sb.AppendLine("    </manifest>");
+
+        // Spine section
+        sb.AppendLine("    <spine>");
+        sb.AppendLine("        <itemref idref=\"chapter1\"/>");
+        sb.AppendLine("    </spine>");
+
+        sb.AppendLine("</package>");
+
+        return sb.ToString();
+    }
+
+    private static string EscapeXml(string text)
+    {
+        if (string.IsNullOrEmpty(text)) return text;
+
+        return text
+            .Replace("&", "&amp;")
+            .Replace("<", "&lt;")
+            .Replace(">", "&gt;")
+            .Replace("\"", "&quot;")
+            .Replace("'", "&apos;");
+    }
+
 }