[derek73]

Closes #857.

sa leaves the vocabulary: its military reading (Seaman Apprentice, a title) and its credential reading are both refused on the word (excluded), so it reads as any name word does. The Portuguese surname Sá written without its accent is the one bearer the corpora show.

How it got here: marking sa ambiguous fixed Maria de Sa but left all 255 SÁ / SA pairs in tse-donors (#843) parting (a one-case name ending in SA still read a suffix, S6), and sa's title reading took the slot in front of a name and alone after a family comma (Smith, Sa → title, where Smith, Ma → given). Neither reading has a bearer in the corpora: every name read with Sa as a title is a person named Sa (Sa Liu, Sa ad Hariri), and every SA / S.A. suffix is a company, the South African qualifier FCP(SA), or initials typed without stops. The 167 NPPES providers who write it have not been read by item.

Readings that change

• Maria de Sa, MARIA DE SA, maria de sa → family de Sa / DE SA / de sa (was family de, suffix Sa); Joao Sa, ANA SA, ANA SILVA SA → family Sa / SA.
• Mrs Sa Li → title Mrs, given Sa (the second title is gone); SA Smith → given SA.
• Costs, accepted: John Smith SA → middle Smith, family SA (was suffix SA); Smith, SA → given SA (was suffix, Should parse("Smith, Jr.") put "Jr." in title rather than suffix? #296's audit row); John Smith, SA → given SA, family John Smith (the name merged into the family, the reading any unlisted two-letter caps word gets; 1.4.0 through 2.5.0 read suffix SA), Smith, John SA → middle SA, , SA → given SA (about 25 after-comma names across ctgov-officials and fec-individuals; NPPES's 167 providers who write it have not been read by item); John Smith S.A. and Smith, S.A. go through the unlisted-dotted rule and report suffix-or-name; a parenthesised FCP (SA) is a nickname SA (25 names in ctgov-officials). A caller who needs SA as a credential adds it to their own lexicon; the exclusions are removable if corpus evidence turns up.

Local corpora (snapshot.py --local, against base 8c965d7, 2026-10-25)

• tse-donors: 1,613 of 848,351 names move, about 700 gaining the SA as part of the family name (ADAO DE SA FERREIRA → family DE SA FERREIRA, was family DE, suffix SA FERREIRA). local_corpora.py pairs tse-donors --agreement reads 0 accent-only pairs with two surnames where the base read 255 SÁ / SA.
• ctgov-officials: 54 of 352,104 (25 the (SA) nickname); psc-individuals: 40 of 1,034,888; ukec-individuals: 0.

Docs and records

• decisions/C8 entry dated 2026-10-25, with supersession pointers in suffix-acronym-collisions, comma-suffix-arc, vocabulary-exclusions and AGENTS.md; release-log bullet.
• tests/v2/rules_budget.json re-recorded (rules/C8 back to its old size); snapshot and rules corpus re-recorded; field-sweep count 1403 → 1406; the pairs-view test now uses RÁ / RA.
• Release checklist: the bullet says "1.4.0 through 2.5.0", so migrate.rst needs a "Changed in" entry at release.

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.07%. Comparing base (8c965d7) to head (68f2797).

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #879   +/-   ##
=======================================
  Coverage   99.07%   99.07%           
=======================================
  Files          47       47           
  Lines        4781     4781           
=======================================
  Hits         4737     4737           
  Misses         44       44           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.