feat: add EPUB input support (#310)
Adds EPUB as a first-class input source for skill generation.
- EpubToSkillConverter (epub_scraper.py, ~1200 lines) following PDF scraper pattern
- Dublin Core metadata, spine items, code blocks, tables, images extraction
- DRM detection (Adobe ADEPT, Apple FairPlay, Readium LCP) with fail-fast
- EPUB 3 NCX TOC bug workaround (ignore_ncx=True)
- ebooklib as optional dep: pip install skill-seekers[epub]
- Wired into create command with .epub auto-detection
- 104 tests, all passing
Review fixes: removed 3 empty test stubs, fixed SVG double-counting in
_extract_images(), added logger.debug to bare except pass.
Based on PR #310 by @christianbaumann.
Co-authored-by: Christian Baumann <mail@chriss-baumann.de>
2e30970dfb