feat(import): move PDF parsing to backend; fix heartbeat/reap for import_tasks

- parsePDF function restored in import.go (body was orphaned outside function)
- ParseImportFile() called at upload time with 3-min timeout; chapters stored as JSON in MinIO
- runner.go: prefer ChaptersKey path (read pre-parsed JSON) over BookImport.Import()
- ImportChapterStore interface added; store wired in runner/main.go
- HeartbeatTask and ReapStaleTasks now include import_tasks collection
- parseImportTask now returns ChaptersKey in domain.ImportTask
- asynq_runner.go handleImportTask passes ChaptersKey
- pb-init-v3.sh: chapters_key field added to import_tasks schema

backend/internal/runner/asynq_runner.go

@@ -199,10 +199,11 @@ func (r *Runner) handleImportTask(ctx context.Context, t *asynq.Task) error {
 		return fmt.Errorf("unmarshal import payload: %w", err)
 	}
 	task := domain.ImportTask{
-		ID:        p.PBTaskID,
-		Slug:      p.Slug,
-		Title:     p.Title,
-		FileType:  p.FileType,
+		ID:          p.PBTaskID,
+		Slug:        p.Slug,
+		Title:       p.Title,
+		FileType:    p.FileType,
+		ChaptersKey: p.ChaptersKey,
 	}
 	r.tasksRunning.Add(1)
 	defer r.tasksRunning.Add(-1)

backend/internal/runner/runner.go

@@ -15,6 +15,7 @@ package runner
 
 import (
 	"context"
+	"encoding/json"
 	"fmt"
 	"log/slog"
 	"os"
@@ -49,6 +50,11 @@ type ChapterIngester interface {
 	IngestChapters(ctx context.Context, slug string, chapters []bookstore.Chapter) error
 }
 
+// ImportChapterStore retrieves pre-parsed chapter JSON blobs from object storage.
+type ImportChapterStore interface {
+	GetImportChapters(ctx context.Context, key string) ([]byte, error)
+}
+
 // Config tunes the runner behaviour.
 type Config struct {
 	// WorkerID uniquely identifies this runner instance in PocketBase records.
@@ -114,7 +120,12 @@ type Dependencies struct {
 	// CoverStore stores book cover images in MinIO.
 	CoverStore bookstore.CoverStore
 	// BookImport handles PDF/EPUB file parsing and chapter extraction.
+	// Kept for backward compatibility when ChaptersKey is not set.
 	BookImport bookstore.BookImporter
+	// ImportChapterStore retrieves pre-parsed chapter JSON blobs from MinIO.
+	// When set and the task has a ChaptersKey, the runner reads from here
+	// instead of calling BookImport.Import() (the new preferred path).
+	ImportChapterStore ImportChapterStore
 	// ChapterIngester persists extracted chapters into MinIO/PocketBase.
 	ChapterIngester ChapterIngester
 	// Notifier creates notifications for users.
@@ -675,6 +686,10 @@ func (r *Runner) runAudioTask(ctx context.Context, task domain.AudioTask) {
 }
 
 // runImportTask executes one PDF/EPUB import task.
+// Preferred path: when task.ChaptersKey is set, it reads pre-parsed chapters
+// JSON from MinIO (written by the backend at upload time) and ingests them.
+// Fallback path: when ChaptersKey is empty, calls BookImport.Import() to
+// parse the raw file on the runner (legacy behaviour, not used for new tasks).
 func (r *Runner) runImportTask(ctx context.Context, task domain.ImportTask, objectKey string) {
 	ctx, span := otel.Tracer("runner").Start(ctx, "runner.import_task")
 	defer span.End()
@@ -682,10 +697,11 @@ func (r *Runner) runImportTask(ctx context.Context, task domain.ImportTask, obje
 		attribute.String("task.id", task.ID),
 		attribute.String("book.slug", task.Slug),
 		attribute.String("file.type", task.FileType),
+		attribute.String("chapters_key", task.ChaptersKey),
 	)
 
 	log := r.deps.Log.With("task_id", task.ID, "slug", task.Slug, "file_type", task.FileType)
-	log.Info("runner: import task starting")
+	log.Info("runner: import task starting", "chapters_key", task.ChaptersKey)
 
 	hbCtx, hbCancel := context.WithCancel(ctx)
 	defer hbCancel()
@@ -714,15 +730,33 @@ func (r *Runner) runImportTask(ctx context.Context, task domain.ImportTask, obje
 		}
 	}
 
-	if r.deps.BookImport == nil {
-		fail("book import not configured (BookImport dependency missing)")
-		return
-	}
+	var chapters []bookstore.Chapter
 
-	chapters, err := r.deps.BookImport.Import(ctx, objectKey, task.FileType)
-	if err != nil {
-		fail(fmt.Sprintf("import file: %v", err))
-		return
+	if task.ChaptersKey != "" && r.deps.ImportChapterStore != nil {
+		// New path: read pre-parsed chapters JSON uploaded by the backend.
+		raw, err := r.deps.ImportChapterStore.GetImportChapters(ctx, task.ChaptersKey)
+		if err != nil {
+			fail(fmt.Sprintf("get chapters JSON: %v", err))
+			return
+		}
+		if err := json.Unmarshal(raw, &chapters); err != nil {
+			fail(fmt.Sprintf("unmarshal chapters JSON: %v", err))
+			return
+		}
+		log.Info("runner: loaded pre-parsed chapters", "count", len(chapters))
+	} else {
+		// Legacy path: parse the raw file on the runner.
+		if r.deps.BookImport == nil {
+			fail("book import not configured (BookImport dependency missing)")
+			return
+		}
+		var err error
+		chapters, err = r.deps.BookImport.Import(ctx, objectKey, task.FileType)
+		if err != nil {
+			fail(fmt.Sprintf("import file: %v", err))
+			return
+		}
+		log.Info("runner: parsed chapters from file (legacy path)", "count", len(chapters))
 	}
 
 	if len(chapters) == 0 {
@@ -730,23 +764,12 @@ func (r *Runner) runImportTask(ctx context.Context, task domain.ImportTask, obje
 		return
 	}
 
-	// Store chapters via BookWriter
-	// Note: BookWriter.WriteChapters expects domain.Chapter, need conversion
-	var domainChapters []bookstore.Chapter
-	for _, ch := range chapters {
-		domainChapters = append(domainChapters, bookstore.Chapter{
-			Number:   ch.Number,
-			Title:    ch.Title,
-			Content:  ch.Content,
-		})
-	}
-
-	// Store chapters via ChapterIngester
+	// Persist chapters via ChapterIngester.
 	if r.deps.ChapterIngester == nil {
 		fail("chapter ingester not configured")
 		return
 	}
-	if err := r.deps.ChapterIngester.IngestChapters(ctx, task.Slug, domainChapters); err != nil {
+	if err := r.deps.ChapterIngester.IngestChapters(ctx, task.Slug, chapters); err != nil {
 		fail(fmt.Sprintf("store chapters: %v", err))
 		return
 	}
@@ -786,7 +809,7 @@ func (r *Runner) runImportTask(ctx context.Context, task domain.ImportTask, obje
 		log.Error("runner: FinishImportTask failed", "err", err)
 	}
 
-	// Create notification for the user who initiated the import
+	// Notify the user who initiated the import.
 	if r.deps.Notifier != nil {
 		msg := fmt.Sprintf("Import completed: %d chapters from %s", len(chapters), task.Title)
 		targetUser := task.InitiatorUserID