Implement field merge strategy for multi-source data (#28)
Add a per-field priority system that resolves conflicts when the same book appears in multiple data sources (Audible, Hardcover, manual). - Create scripts/backend/matchers/merger.ts with mergeBook() and mergeAllSources() functions that apply source priority per field (manual > audible > hardcover), skip null overwrites, and flag low-confidence matches for review - Add mergeAndUpsertBook() to db/index.ts for merge-aware ingestion, plus remergeBook() and getMultiSourceBookIds() for batch re-merging - Wire the CORRECT pipeline stage to re-merge all multi-source books using stored raw_data blobs from book_sources
Parent8be15bb
3 files changed+547-3
StackedSplit