feat: Ingestion Pipeline Stabilization and WASM Service Proxies (#42)

This PR stabilizes the Nexus Ingestion Engine by implementing functional service proxies for the Blazor WASM client and refining the backend infrastructure for real-time progress tracking and database compatibility. ### Key Changes - **Infrastructure Stabilization**: - Implemented production-grade `EbookRepository` with PostgreSQL `EF.Functions.ILike` support. - Enforced `IsReadyForReading = false` state for newly added ebooks (resolves #35). - Updated `SignalRSyncBroadcaster` to support targeted user messaging and ingestion-specific progress updates (resolves #37). - **WASM Client Functional Proxies**: - Replaced "Throwing" dummy services with `WasmEbookRepository`, `WasmSyncBroadcaster`, `WasmBookStorageService`, and `WasmEmbeddingGenerator`. - These services proxy requests to the backend via a new set of Minimal API endpoints in `NexusReader.Web`. - **Domain Refinement**: - Added `IsReadyForReading` flag to the `Ebook` entity to manage background AI processing states. ### Related Issues - Fixes #35 - Fixes #36 - Fixes #37 --------- Co-authored-by: Marek Jasiński <jasins.marek@gmail.com> Reviewed-on: #42 Co-authored-by: Antigravity <antigravity@google.com> Co-committed-by: Antigravity <antigravity@google.com>
2026-05-13 18:24:24 +00:00
parent d5c2952bec
commit 5a2223a4c8
39 changed files with 6134 additions and 301 deletions
@@ -0,0 +1,205 @@
+using System.Text.RegularExpressions;
+using FluentResults;
+using Microsoft.EntityFrameworkCore;
+using Microsoft.Extensions.Logging;
+using NexusReader.Application.Abstractions.Services;
+using NexusReader.Application.Queries.Reader;
+using NexusReader.Data.Persistence;
+using VersOne.Epub;
+
+namespace NexusReader.Infrastructure.Services;
+
+/// <summary>
+/// Reads and parses EPUB files from the storage path recorded in the database.
+/// </summary>
+public class EpubReaderService : IEpubReader
+{
+    private readonly IDbContextFactory<AppDbContext> _dbContextFactory;
+    private readonly ILogger<EpubReaderService> _logger;
+    private const int WordThreshold = 1000;
+
+    public EpubReaderService(
+        IDbContextFactory<AppDbContext> dbContextFactory,
+        ILogger<EpubReaderService> logger)
+    {
+        _dbContextFactory = dbContextFactory;
+        _logger = logger;
+    }
+
+    /// <inheritdoc />
+    public async Task<Result<ReaderPageViewModel>> GetEpubContentAsync(
+        Guid ebookId,
+        int chapterIndex,
+        string? userId = null,
+        CancellationToken cancellationToken = default)
+    {
+        try
+        {
+            // 1. Resolve the file path from the database
+            using var context = await _dbContextFactory.CreateDbContextAsync(cancellationToken);
+
+            var ebook = await context.Ebooks
+                .AsNoTracking()
+                .FirstOrDefaultAsync(
+                    e => e.Id == ebookId && (userId == null || e.UserId == userId),
+                    cancellationToken);
+
+            if (ebook == null)
+            {
+                return Result.Fail($"Ebook '{ebookId}' not found for user '{userId}'.");
+            }
+
+            // FilePath is stored as a web-relative path (e.g. "uploads/guid_title.epub").
+            // Resolve against the content root, then against the wwwroot sub-directory.
+            var fullPath = ResolvePath(ebook.FilePath);
+            if (fullPath == null || !File.Exists(fullPath))
+            {
+                _logger.LogError("EPUB file for ebook {EbookId} not found at path '{FilePath}'.", ebookId, ebook.FilePath);
+                return Result.Fail($"The EPUB file for this book could not be found on the server.");
+            }
+
+            // 2. Parse the EPUB
+            using var bookRef = await EpubReader.OpenBookAsync(fullPath);
+            var readingOrder = bookRef.GetReadingOrder();
+
+            if (readingOrder == null || !readingOrder.Any())
+            {
+                return Result.Fail("The EPUB has no readable content files in ReadingOrder.");
+            }
+
+            if (chapterIndex < 0 || chapterIndex >= readingOrder.Count)
+            {
+                chapterIndex = 0;
+            }
+
+            var chapterRef = readingOrder[chapterIndex];
+            var navigation = bookRef.GetNavigation();
+            var chapterTitle = FindTitleInNavigation(navigation, chapterRef.FilePath)
+                               ?? Path.GetFileNameWithoutExtension(chapterRef.FilePath)
+                               ?? $"Chapter {chapterIndex + 1}";
+
+            var chapterContent = await chapterRef.ReadContentAsTextAsync();
+
+            // 3. Build content blocks
+            var blocks = new List<ContentBlock>();
+            int totalWordCount = 0;
+            int blockCounter = 0;
+
+            var paragraphs = ExtractParagraphs(chapterContent);
+            foreach (var p in paragraphs)
+            {
+                var sanitizedContent = SanitizeParagraph(p);
+                if (string.IsNullOrWhiteSpace(sanitizedContent)) continue;
+
+                blocks.Add(new TextSegmentBlock($"seg-{blockCounter++}", sanitizedContent));
+
+                int wordsInP = CountWords(sanitizedContent);
+                totalWordCount += wordsInP;
+
+                if (totalWordCount >= WordThreshold)
+                {
+                    blocks.Add(CreateAiTrigger($"trigger-{blockCounter++}"));
+                    totalWordCount = 0;
+                }
+            }
+
+            if (blocks.Any() && blocks.Last() is not AiActionTriggerBlock)
+            {
+                blocks.Add(CreateAiTrigger($"trigger-{blockCounter++}"));
+            }
+
+            return Result.Ok(new ReaderPageViewModel(blocks, chapterIndex, readingOrder.Count, chapterTitle, ebook.Id));
+        }
+        catch (Exception ex)
+        {
+            _logger.LogError(ex, "Failed to process EPUB for ebook {EbookId}.", ebookId);
+            return Result.Fail(new Error($"Failed to process EPUB: {ex.Message}").CausedBy(ex));
+        }
+    }
+
+    /// <summary>
+    /// Attempts to resolve a web-relative storage path to an absolute filesystem path.
+    /// Searches upward from the app base directory to handle both dev and production layouts.
+    /// </summary>
+    private static string? ResolvePath(string relativePath)
+    {
+        // Normalize forward-slashes to OS separator for file system access
+        var normalized = relativePath.Replace('/', Path.DirectorySeparatorChar);
+
+        var currentDir = new DirectoryInfo(AppDomain.CurrentDomain.BaseDirectory);
+        while (currentDir != null)
+        {
+            var candidate = Path.Combine(currentDir.FullName, "wwwroot", normalized);
+            if (File.Exists(candidate)) return candidate;
+
+            // Also try src/NexusReader.Web/wwwroot (development layout)
+            var devCandidate = Path.Combine(currentDir.FullName, "src", "NexusReader.Web", "wwwroot", normalized);
+            if (File.Exists(devCandidate)) return devCandidate;
+
+            currentDir = currentDir.Parent;
+        }
+
+        return null;
+    }
+
+    private static List<string> ExtractParagraphs(string html)
+    {
+        var bodyMatch = Regex.Match(html, @"<body\b[^>]*>(.*?)</body>", RegexOptions.IgnoreCase | RegexOptions.Singleline);
+        var content = bodyMatch.Success ? bodyMatch.Groups[1].Value : html;
+
+        var paragraphs = new List<string>();
+        var matches = Regex.Matches(content, @"<(p|h[1-6]|ul|ol|blockquote|pre)\b[^>]*>.*?</\1>|<hr\b[^>]*>", RegexOptions.IgnoreCase | RegexOptions.Singleline);
+
+        foreach (Match match in matches)
+        {
+            paragraphs.Add(match.Value);
+        }
+
+        if (paragraphs.Count == 0)
+        {
+            paragraphs = content.Split(new[] { "<br />", "<br>", "\n\n", "\r\n\r\n" }, StringSplitOptions.RemoveEmptyEntries).ToList();
+        }
+
+        return paragraphs;
+    }
+
+    private static string SanitizeParagraph(string html)
+    {
+        var clean = Regex.Replace(html, @"<(style|script)\b[^>]*>.*?</\1>", "", RegexOptions.IgnoreCase | RegexOptions.Singleline);
+        clean = Regex.Replace(clean, @"<(?!/?(b|i|strong|em|h[1-6]|p|ul|ol|li|blockquote|pre|code|br|hr)\b)[^>]+>", "", RegexOptions.IgnoreCase);
+        clean = Regex.Replace(clean, @"<(b|i|strong|em|h[1-6]|p|ul|ol|li|blockquote|pre|code|br|hr)\b[^>]*>", "<$1>", RegexOptions.IgnoreCase);
+        clean = System.Net.WebUtility.HtmlDecode(clean);
+        return clean.Trim();
+    }
+
+    private static int CountWords(string text)
+    {
+        if (string.IsNullOrWhiteSpace(text)) return 0;
+        return text.Split(new[] { ' ', '\r', '\n', '\t' }, StringSplitOptions.RemoveEmptyEntries).Length;
+    }
+
+    private static AiActionTriggerBlock CreateAiTrigger(string id) =>
+        new(id,
+            "Wykryto ciekawy fragment! Czy chcesz, abym wygenerował podsumowanie lub quiz z tego rozdziału?",
+            new List<string> { "Podsumuj", "Generuj Quiz", "Pomiń" });
+
+    private static string? FindTitleInNavigation(IEnumerable<EpubNavigationItemRef> navigation, string? filePath)
+    {
+        if (string.IsNullOrEmpty(filePath)) return null;
+        var fileName = Path.GetFileName(filePath);
+
+        foreach (var item in navigation)
+        {
+            if (item.Link?.ContentFilePath == filePath || item.Link?.ContentFilePath == fileName)
+                return item.Title;
+
+            if (item.NestedItems?.Any() == true)
+            {
+                var childTitle = FindTitleInNavigation(item.NestedItems, filePath);
+                if (childTitle != null) return childTitle;
+            }
+        }
+
+        return null;
+    }
+}