Skip to content

Read Sta. and Sto. as saint particles, so Juan Sta. Maria Reyes keeps its surname whole (#725) - #1066

Merged
derek73 merged 5 commits into
masterfrom
claude/particle-vocab-725-694-6c64f6
Nov 5, 2026
Merged

derek73 merged 5 commits into
masterfrom
claude/particle-vocab-725-694-6c64f6

Conversation

@derek73

@derek73 derek73 commented Nov 5, 2026

Copy link
Copy Markdown
Owner

Closes #725

sta and sto join fil.toml as saint particles, with may_be_given = false and a title-case mask. Before this, Juan Sta. Maria read middle 'Sta.', family 'Maria' (1.4.0 through 2.8.1).

Input 1.4.0 – 2.8.1 Now
Juan Sta. Maria middle Sta., family Maria family Sta. Maria
Juan Sta. Maria Reyes middle Sta. Maria, family Reyes family Sta. Maria Reyes
Juan Sto. Tomas Jr. middle Sto., family Tomas family Sto. Tomas
Sta Juan given Sta family Sta Juan
JUAN STA. MARIA (capitalized()) — Juan Sta. Maria
  • Why saint. As a plain particle, rules/P10 (a given name a particle completes) would read Sta. Maria as a second given name, because maria is written after_particle. P10 refuses a saint run, so the phrase stays in the family.
  • Why no given reading. ORCID's 2025 extract has sta in 122 family boxes, 38 of them Philippine. It is in 11 given boxes, but those are five surnames filed with the given names and six junk rows. The Census counts Sta 120 times as a surname, and SSA has fewer than 5 births. ste has had no given reading since Mc Donald and Ste Marie read the particle as the given name #360 for the same reason.
  • Measured (config_diff.py, the recipe is in decisions/H6): of 63 corpus names holding either word, 33 move, every one from middle to family.
    • 30 read right.
    • 2 are garbage input.
    • 1 is wrong and accepted: ERNANI ESPIRITO STO ASSIS, where Sto ends the Portuguese given name Espírito Santo. It is the only Espirito Sto in any local corpus.
  • Accepted costs (decisions/H6):
    • A bare Sta opening a part now starts the family through P1's fold: Reyes, Sta Maria reads family 'Reyes Sta Maria' with no given name.
    • P6 takes a trailing Sta. after a family comma into the family.
    • The two family-first readings of Juan dela Cruz Sta. Ana change. That is P1/P4's area and is left to that work.
  • Snapshot: one view moves. The initials of Sta. Maria, Juan lose the S.
  • Not reached: a glued Sta.Ana reads as before. rules/S3 makes it a suffix where there's a word to spare (Juan Reyes Sta.Ana) and the family otherwise. I'll file it separately.

Docs: decisions/H6 closes its #725 Open with a dated entry. decisions/P1 and P10, rules/P10's saint clause, AGENTS.md and the vocab README list the new words. There is a release-log bullet.

Checks: the full suite passes on 3.11 and 3.15, the serial benchmarks pass, and ruff and mypy are clean. A code review, a design-docs review and a second-round review of the fix commit each ran; their findings are fixed in d0b3570 and 57b1080.

🤖 Generated with Claude Code

derek73 and others added 4 commits November 4, 2026 16:11
… its surname whole (#725)

fil.toml lists sta and sto with the saint kind, may_be_given = false and
a title-case mask. As a saint run, rules/P10 never reads Sta. Maria as a
second given name; with no given reading, a bare Sta opening a name
begins the family, as Ste does. decisions/H6 records the measurements
and closes its #725 Open.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…old, the glued spelling

decisions/H6's entry now counts the moved names as 30 right, two garbage
and one wrong (the given name Espírito Santo written Sto), records the
bare Sta's P1 fold after a comma and with an empty head, P6's trailing
Sta, both family-first readings and the dropped initial, and says where
a glued Sta.Ana reads as a suffix and where as the family. decisions/P1
names #725 among the never-given particle additions, and the old H6 entry
points at its decision. The release bullet drops the glued claim.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ORCID's 11 given boxes holding sta are five surnames filed with the
given names and six junk rows, not given names; the initial is dropped
only where the word stands in the middle or family ahead of a name word;
and the default reading of Juan dela Cruz Sta. Ana moved too. The
release bullet drops the initials claim.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@derek73 derek73 added this to the 3.0 milestone Nov 5, 2026
@derek73 derek73 added enhancement vocabulary Per-word vocabulary model and language packs labels Nov 5, 2026
@derek73 derek73 self-assigned this Nov 5, 2026
@derek73
derek73 merged commit 50cf8ff into master Nov 5, 2026
13 checks passed
@derek73
derek73 deleted the claude/particle-vocab-725-694-6c64f6 branch November 5, 2026 04:35
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

enhancement vocabulary Per-word vocabulary model and language packs

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Should Juan Sta. Maria read family Sta. Maria? Today it reads middle Sta., family Maria

1 participant