plugin-v1.0.0

What's new

Performance

• Build history map once per sync instead of scanning history.lua per book — O(1) lookups for annotations and ISBN extraction, significantly faster for large libraries

Improvements

• Raise auto-sync page threshold from 10 to 25 turns to reduce session fragmentation (fewer duplicate sessions from incremental syncs during long reading sittings)
• Support CREngine string-format identifiers for ISBN extraction (e.g. isbn:978... in newline-separated strings)

Refactoring

• Extract _buildHistoryMap() for shared md5→file_path resolution across annotations and ISBN lookup
• Extract _extractIsbnFromProps() for cleaner, testable ISBN parsing from doc_props