Feature #43: Extract and display cover images from EPUB/AZW3

[lllyys]

Description

EPUB and AZW3 books show format placeholder icons instead of actual cover images in the library.

Current State

• Custom user covers (uploaded via PhotosPicker): ✅ Working (Feature Bug #103: TOC jumps to wrong chapter + wrong position + position restore broken #30)
• Extracted EPUB covers: Not implemented — EPUBMetadataExtractor is a stub returning nil
• Extracted AZW3 covers: Not implemented — AZW3MetadataExtractor is a stub returning nil

Infrastructure Already In Place (~40%)

• Book.coverImagePath: String? — model field exists, always nil
• BookMetadata.coverImagePath: String? — metadata field exists, never populated
• CustomCoverStore — JPEG compression, disk storage, cleanup — fully working
• BookCardView / BookRowView — already check for cover image, fall back to placeholder
• Import pipeline passes coverImagePath through, just receives nil

What's Missing

1. EPUB cover extraction — Parse content.opf for cover image reference, extract from ZIP
2. AZW3 cover extraction — Read EXTH records for embedded cover, or use Foliate-js
3. Save extracted cover to disk during import (via CustomCoverStore or similar)
4. Populate Book.coverImagePath in BookImporter.importFile()

Notes

• EPUB cover extraction may be addressed by Feature Bug #11: Large TXT file nearly impossible to scroll #42 (Foliate-js engine) if Foliate exposes covers
• AZW3 cover extraction depends on Foliate-js or native MOBI header parsing

[lllyys]

Implemented and merged in commit a33b5e9 — EPUB OPF parsing + AZW3 MOBI header parsing extract cover images during import. Book.coverImagePath now populated; BookCardView/BookRowView render extracted covers. 46 new tests. Tracked as Feature #43 (DONE) in docs/features.md. Note: bug #107 tracks the white-edge padding illusion as a follow-up.