You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
John van der Berg V reads family van der Berg V — the particle chain and the maiden walk take a trailing numeral as a name word #424
Two other sites stop at a suffix with the same test #401 corrected for the bound-given reserve, and both swallow a bare trailing roman numeral:
John van der Berg Jr -> family='van der Berg' suffix='Jr'
John van der Berg V -> family='van der Berg V' suffix='' chain took the numeral
John van der Berg X -> family='van der Berg X'
abdul van der Berg V -> family='van der Berg V'
John née Jones Smith Jr -> maiden='Jones Smith' suffix='Jr'
John née Jones Smith V -> maiden='Jones Smith V' suffix='' walk took the numeral
Jane Smith née Jones V -> maiden='Jones V'
Compare John Smith V → family Smith, suffix V.
Why
P2's chain runs "until … a trailing suffix begins", and M2's walk takes the words after the marker "up to any trailing suffix". Both test that boundary with _is_suffix_piece, which vetoes on the initial tag — right for assign's middle-initial question, wrong here for the reason #401 recorded: a bare V is suffix vocabulary and initial-shaped, so neither site sees the suffix assign will then read. Once the chain or the walk has taken the numeral into its piece, assign never sees a lone trailing V at all.
Scope
Pre-existing, independent of #423 (measured at 2.2.0dev). Single-letter numerals only — V, I, X — after a particle chain or a maiden marker; II, Jr, PhD are unaffected. No differential corpus name has either shape; tests are the verification.
Fix
_vocab.is_trailing_numeral_suffix (shared by assign and the bound-given reserve since #423) answers the question both sites ask; each stop needs the same "last piece of the walk, preceded by a non-initial-shaped piece" framing the reserve uses. Worth doing as one change with one set of examples, since it is one predicate at three sites — and the same altitude question as the acronym follow-up: if assign's suffix verdicts were computed once and shared, none of the three sites would carry its own copy.
Activity
refactor(group): say the P5 mechanism once, and cite the rule
refactor(group): one definition of a merged piece's tags, read by the…
docs(design): record the review round of #424
Merge pull request #427 from derek73/fix/424-chain-and-walk-ask-the-peel
fix(group): only a piece that renders into the run may continue an entry
docs(post_rules,decisions,release_log,differential): the toolkit revi…
fix(group): the chain and the maiden walk stop where assign's peel be…
docs(design): record derek73#424, and correct the derek73#425 entry f…
fix(vocab,assign): give each title/suffix overlap word its true class…
refactor(group,assign): move _segment_holds_no_name down beside its p…
docs(rules,decisions,release_log): the review round
test(differential): classify the names the marker-led clause moves
fix(post_rules,rules,differential): a maiden marker makes the lone na…
test(differential,docs): the absorber the revisit note missed
fix(differential,guards): stop the catch-all declaring a role it neve…
docs(decisions): record the fields-only ledger arc
fix(differential,guards): narrow three 1.4.0 rules to the roles they …
test(benchmark,tools): a per-interpreter band, and a harness in the tree
docs(rules)+ledger: the case signal and the shape class, classified
test(derek73#533): record the ledger diffs at every baseline
docs(derek73#533): the rule, the decision and the release note
fix(derek73#535): a trailing title ends the maiden clause
docs(derek73#535): bound the transparency and numeral claims to what …
test(derek73#544): case rows, rules.md examples, corpora and ledgers
Fix #632's second-round review: reach, controls, stale prose
Fix #632's second-round review: reach, controls, stale prose
Check rule citations in every shape, ledgers included; fix the stale …