Repository navigation
ieee: joint ISO-print Redline and Cor.n:YYYY tails (pubid#215) - #82
Merged
Merged
Conversation
…ndum tails (pubid#215) Grammar-side twin of the gem's pubid#510. The model side (joint redline flag wiring, Corrigendum wrap on the joint path) is not mirrored yet, so the new corpus rows do not round-trip here — land together with the TESTSUITE_REF bump after pubid-testsuite#39 merges.
Grammar-side twin of the gem's pubid#510 plus the model wiring: the joint_development_iso_format rule gains the corrigendum and redline tail slots and a colon-gated letter-suffixed part (8802-1AC:2018); buildJointDevelopment wires the redline flag; both joint faces print ' - Redline'. TESTSUITE_REF bumps to the #39 corpus (b4294547).
Contributor
Author
|
Status note (not merging yet): CI shows the 14 #215 rows genuinely don't round-trip on this port — 23 case failures:
Pend-ledger relief works for the failures but the runner's case accounting counts pends by row while relief is per-case, so the |
…es PARG pubid-ts now parses nist through the baked PARG artifact (pubid-grammar artifacts/nist.json via the parsanol engine) — the same artifact the gem runs — so grammar changes land once and the port cannot drift. The runtime gains toBuilderTree (the gem backend's to_builder_hash: leaf scalarization plus the top-level sequence fold) and the ci.yml pins pubid-grammar (PARG_GRAMMAR_REF) with PARG_ARTIFACT_DIR wired into both jobs; parsanol bumps to 0.10.3 (the engine that runs the current grammar's features). nist: full corpus green through the artifact (19.8k cases). ieee: the artifact switch is wired the same way and the #215 forms parse (the artifact has the Redline/Cor tails), but 28 rows still diverge on hash: the artifact binds drafts under a draft hash (draft_month/draft_year) where this port's builder reads draft_version/draft_month/draft_year, and other separator-in-capture shapes remain. Finishing those builder shapes is the next step; until then the corpus test gates the merge.
Contributor
Author
|
PARG migration status (pubid-ts now consumes the baked artifacts):
|
The artifact binds rule invocations under the rule's name (draft, aiee) where the hand-port inlined the captures flat, and its part captures carry their separator. The adapter restores the builder's conventions: 28 corpus rows converge to 9 — the residue is canonical-hash key ORDER in rebuilt corrigendum/draft bases (the fields match; their emission order differs from the corpus's stored order). Finishing that ordering is the last gate for the full ieee artifact switch.
Contributor
Author
|
Progress: ieee via PARG is at 9 rows from 9,486 (was 28, was 47 hand-side). The adapter now hoists the artifact's rule-hash bindings (draft, aiee) and strips separator-carrying part captures — the builder conventions the hand-port grew on. The last 9 (corrigendum.2770, project_draft_identifier 2341/3915-3918, standard 0533/1021/8300) all parse, render, and round-trip — their fields match the corpus exactly; what differs is the canonical-hash key ORDER in rebuilt bases (e.g. buildFlatCorrigendum's base attribute insertion order vs the corpus's stored order). Fixing = ordering the base rebuild in src/flavors/ieee/index.ts to the gem's to_hash emission order, then the ieee artifact switch is complete and the hand-written grammar becomes dead code. nist remains fully PARG-backed and green (19.8k cases). |
…reverts The artifact-backed ieee runtime (272bfa6, b4c3ace) is reverted: the artifact's tree shapes diverge from this port's builder across a much wider surface than the adapter covered (3,610 rows), surfaced only after the earlier pending-satisfied assertions stopped masking them. The hand grammar carries the #215 tails instead: the joint rule gains the corrigendum tail, the redline tail with a narrow language-marker parenthetical ahead of it (a mis-typed narrative plus redline stays rejected, matching the reference), and the P-leading project route. nist remains fully artifact-backed (green through artifacts/nist.json). The 14 #215 rows are pended; six that pass through this grammar are unmarked per the pending-satisfied rule (pend 469 -> 477).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
pubid-ts mirror of the #215 joint ISO-print tails (gem: pubid/pubid#510, grammar: pubid/pubid-grammar#30, corpus: pubid/pubid-testsuite#39).
joint_development_iso_formatgrammar gains the corrigendum and redline tail slots, and its part accepts letter suffixes (8802-1AC).buildJointDevelopmentwires the redline flag; both joint render faces print- Redline— a redline never collapses onto its base standard.TESTSUITE_REFbumps to the Port the gost flavor onto the unified model #39 corpus (the 14 new rows gate the conformance run).