Skip to content

List dipl, the diploma degree, as a title (#1115) - #1123

Merged
derek73 merged 3 commits into
masterfrom
feat/issue-1115-diplom
Nov 7, 2026
Merged

derek73 merged 3 commits into
masterfrom
feat/issue-1115-diplom

Conversation

@derek73

@derek73 derek73 commented Nov 7, 2026 •

Copy link
Copy Markdown
Owner

Part of #1115, the Diplom degrees.

What changes

dipl is listed alone (mul.toml, academic). A token whose chunks a period joins already reads as a title when any chunk is a title (v1's rule for Lt.Gov.), so every field written after Dipl with a period now reads as one, whatever the field:

Anselm Crombach, Dipl.-Psych.   given 'Dipl.-Psych.', family 'Anselm Crombach'  →  title 'Dipl.-Psych.', given 'Anselm', family 'Crombach'   (1.4.0 through 2.8.2)
Dipl.-Kfm. Dalimil Ospalý       given 'Dipl.-Kfm.', middle 'Dalimil'             →  title 'Dipl.-Kfm.', given 'Dalimil'
Josef Mareček, dipl.um.         suffix 'dipl.um.'                                →  title 'dipl.um.'

The per-compound listing and the separator-insensitive match proposed on #1115 are declined as unneeded. rules/S3 now states the title-chunk reading, which v1 had but no rule stated. An Except bullet says a period-joined word the vocabulary does not list whole is a title before the name and after a comma when any chunk is a title, read before its suffix chunks (Dr.PH stays its listed credential). An Accepted bullet records the readings that reach beyond the degrees (MD.MS, St.John).

Measured

160 corpus names hold dipl as a word; 58 move (2026-11-06, 8c1cc709):

  • 34 a degree leaving a name field (19 ClinicalTrials.gov officials written Name, Dipl.-Psych. stop reversing)
  • 24 from suffix to title

Costs, in decisions/H6's 2026-11-06 Diplom entry:

  • a spaced degree inside a run of post-nominals splits off its field (3 names, CLDR's Bosnian dipl. stom. among them; its agreement record is re-recorded)
  • a degree ending a credential run leaves ; or & behind (2)
  • a degree glued to a name word takes the word, as Dr. does (1)
  • a glued degree ending a name now reads as a name word where it read as a suffix: with no comma it is the family name (Maxie Bunz Dipl.Psych. → family 'Dipl.Psych.', middle 'Bunz'), and after a family comma a middle name (Doe, John Dipl.Ing.). No corpus name with Dipl takes either shape, and it is what John Smith Lt.Gov. already gives. It is accepted here and recorded as rules/S3's Accepted bullet, and Should Hans Meier Priv.-Doz. read the listed title at the end, as Hans Meier Prof. does? #1116 is widened to cover both shapes, since its fix to H5's trailing run reaches them.

Left as they read: a field joined by a hyphen alone (Dipl-Psych, 3), since the chunk rule splits at periods only and hyphenated surnames hold title words; a spaced field without its period (Dipl. Biol, Dipl Ing, 6); and the spaced form before a family comma (1).

Review

The design-docs review confirmed the 160/58/34/24 measurement, read every moved name, swept every comma shape and order, checked the case views and checked the release claim against eleven wheels from 1.4.0 to 2.8.2. It found the trailing regression above, S3's missing title-chunk precedence, two cut credential runs, an undercount of the spaced-field leftovers, the 20 writers behind the 19 moves, and an unreproducible spelling count. All are fixed in the second commit. A second-round review of that commit's S3 bullets found the Except's lead false for a word listed whole, unscoped by slot, and wider than its argument, plus the family-comma move above; all are fixed in the third commit.

Test plan

  • uv run pytest (39,190 passed) and uv run pytest -n0 -m serial
  • uv run mypy, uv run ruff check
  • uv run sphinx-build -b doctest docs and the html build

🤖 Generated with Claude Code

derek73 and others added 2 commits November 6, 2026 15:15
dipl (mul.toml, academic) is listed alone; the period-joined chunk rule
that reads Lt.Gov. then reads every field written after it with a
period as a title: Dipl.-Psych., Dipl.-Kfm., Dipl.Ing., dipl.um.
"Anselm Crombach, Dipl.-Psych." no longer reverses into given
'Dipl.-Psych.', family 'Anselm Crombach'. 58 of 160 corpus names move.

The separator match proposed on #1115 is declined as unneeded. Left: a
field joined by a hyphen alone (Dipl-Psych), the spaced periodless
form, and a spaced degree in a post-nominal run, whose field splits off
(dipl. stom., CLDR's Bosnian rendering, re-recorded).

Co-Authored-By: Claude Opus 5.5 <[email protected]>
#1115)

The design-docs review found that listing dipl moves a degree glued to
its field and ending a name with no comma from the suffix to the family
(Maxie Bunz Dipl.Psych. -> family 'Dipl.Psych.', middle 'Bunz'): the
chunk verdict is now the title, and H5's trailing run reads no word with
an interior period. No corpus name is written so; accepted here, as
Lt.Gov. already reads, and added to #1116.

rules/S3 now states that a title chunk is read before the suffix chunks
(an Except bullet) and records the trailing case (an Accepted bullet).
The entry gains the two cut credential runs, the six spaced fields
without a period (four read so before), the 20 officials beside the 19
moved, and drops an unreproducible spelling count; dipl's note records
the English Diplomate.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
@derek73 derek73 added design vocabulary Per-word vocabulary model and language packs labels Nov 7, 2026
@derek73 derek73 self-assigned this Nov 7, 2026
The second-round review of the S3 bullets found the Except's lead false
for a word listed whole (Dr.PH, MS.Ed. read as their listed credential,
the chunk test never asked), stated for every slot where the title
reading holds only before the name and after a comma, and argued for
the degrees while reaching every title chunk. The Except now names the
whole-word listing first (a Dr.PH boundary row) and its slots; the
Accepted bullet covers the given part after a family comma, which this
PR moved from suffix to middle (Doe, John Dipl.Ing.), and the name's
inside; a second Accepted bullet records the duals and Saint the
chunk reaches (MD.MS, St.John).

Co-Authored-By: Claude Opus 5.5 <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

design vocabulary Per-word vocabulary model and language packs

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant