kert 001abbe186 fix(pfs): CPT parser drops resequenced placeholder rows (refs #687) ...

A resequenced code prints twice in the AMA book: a bare pointer at its
old numeric position ("Code is out of numerical sequence. See
NNNNN-NNNNN") and its real entry, with descriptor/elements/tail, at its
new print-order position. finalize_entry now recognizes the pointer
row by its own text and emits nothing for it (C8), so only the real
entry survives — a downstream loader can now assert one pfs.cpt_code
row per (edition_year, code) without the pointer creating a spurious
duplicate.

2026-09-09 18:03:04 -04:00

4.2 KiB

Raw Blame History

View Raw