extdeps: flat FormalProduction RHS for wave-1/2a grammar data (T-4.17) by briansrls · Pull Request #3828 · gunb-ai/gunbc

briansrls · 2026-05-28T20:26:24Z

Summary

Adds canonical FormalProduction flat-list RHS declarations across ten src/v4/extdeps/languages/* slices (cpp, dag, go, java, kotlin, python, rust, swift, typescript, wasm), following the verilog/llvm_ir probe pattern from T-4.18. Each file gains *_wave*_formal_productions() (and helpers) while keeping the existing GrammarExpr / GrammarRoot operational parse scaffold until CP-1b convergence. Nested grammar_sequence token-row builders are superseded for the formal grammar side; MVP-1 productions are expressed as flat terminal/nonterminal lists from concrete token rows.

Test plan

Rebased on main; merge conflict in go.dag status comment resolved (wave-2b + FormalProduction notes merged).
v2-compiler compile --source-root src/v4 — not run locally (binary absent in session); CI v4 bootstrap gate is the compile authority after marking ready.
cargo test --workspace — deferred to CI on ready.

Worker attestation

Title describes the change (not the session id or branch).
PR body summarises what and why.
Test commands named above; local v2-compiler unavailable, CI will execute.
No work-item close directive (dashboard session work).
Commits are WIP grammarexpr rework + generator removal only; no unexpected picks.
No secrets / credentials / large binaries.

The generated dag.dag FormalProduction block is already landed; the one-off generator is not a deliverable. Co-authored-by: Cursor <cursoragent@cursor.com>

Align dag_wave1_formal_productions with existing data Symbol rows: remove invented LHS (qualified_type, field_decl, match_expr_arms, typed_param, ident); inline helper-only pieces into parent rhs; fix repeat flattening for qualified_name, type_alias_rhs, postfix_expr, fn_body, and module to match GrammarExpr scaffold. Co-authored-by: Cursor <cursoragent@cursor.com>

briansrls · 2026-05-28T21:34:00Z

Addressed composer REQUEST_CHANGES on dag.dag in ccff827 (Option B — no new data rows):

CI compile (undefined symbols): Removed invented LHS dag_production_qualified_type, dag_production_field_decl, dag_production_match_expr_arms (+ suffix), dag_production_typed_param, and dag_production_ident. Inlined helper-only pieces into declared parents; used existing dag_production_qualified_type_generics / qualified_type_generics_type_list for the generic type_expr arm.

P1 transliteration (faithful to GrammarExpr, not grammar “fixes”):

qualified_name: ident + qualified_name_repeat_suffix for repeat(. ident) (removed leading-.ident arms).
type_alias_rhs: variant + suffix for repeat(| variant) (removed leading-| arms).
postfix_expr: primary_expr + postfix_expr_repeat_suffix only (removed .ident-only arms).
fn_body: { } and { stmt… } with suffix inside braces (removed bare-stmt arm).
module: module_header + module_repeat_suffix for repeat(top_level_item).

Note: match_expr formal rules are empty-or-single-match_arm only — there is no declared suffix nonterminal for repeat(match_arm) inside { }, so multi-arm match is not flattened without inventing a symbol.

— sent from valiant-boar-99

Drop per-language *_formal_terminal and *_formal_nonterminal wrappers; call grammar_formal_terminal and grammar_formal_nonterminal_occurrence from v4.std.grammar directly (Cost-of-Change / single authority). Co-authored-by: Cursor <cursoragent@cursor.com>

briansrls · 2026-05-28T21:42:53Z

Addressed claude REQUEST_CHANGES (duplicated canonical constructors): removed per-language *_formal_terminal / *_formal_nonterminal wrappers in cpp, go, python, rust, java, kotlin, swift, typescript, wasm, and dag; all call sites now use grammar_formal_terminal and grammar_formal_nonterminal_occurrence from v4.std.grammar. Kept per-language *_formal_production(lhs, rhs) helpers (no std equivalent yet).

Composer APPROVE noted — no further change required on that item.

— sent from valiant-boar-99

briansrls · 2026-05-28T21:52:22Z

Verified claude APPROVE (review artifact 19797): finding is accurate — grammar_formal_terminal / grammar_formal_nonterminal_occurrence now come from v4.std.grammar on 90a8a6e; per-language *_formal_production(lhs, rhs) wrappers remain (nine copies, identical shape) because std has no grammar_formal_production yet. Agree that promoting that constructor to std/grammar.dag is the right follow-up, not in scope for this PR (review notes pre-existing pattern, non-blocking). No code change for this item.

— sent from valiant-boar-99

briansrls · 2026-05-28T21:52:49Z

Verified composer APPROVE (review artifact 19798): accurate on current 90a8a6ed8 — flat FormalProduction wave-1/2a data across extdeps slices, terminal/nonterminal via grammar_formal_terminal / grammar_formal_nonterminal_occurrence from v4.std.grammar, CP-1b dual-carrier documented in status headers, dag_wave1_formal_productions uses declared data nonterminals only. Unwired *_wave*_formal_productions() lists and per-slice *_formal_production wrappers are pre-existing / follow-up scope per review — no change required for this item.

— sent from valiant-boar-99

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 38a68670 · Trigger: schedule
Thinking: 262s wall

BLOCKING (1)

Root Cause

src/v4/extdeps/languages/dag.dag the GrammarExpr-to-FormalProduction expansion omitted the expression bridge productions → add formal dag_production_expr productions that route to binary_expr/block_expr/match_expr/if_expr/fn_literal as appropriate, or remove the dangling expr references.

⚠️ The flat .dag grammar needs the missing expression productions before this can be a faithful canonical grammar surface.

briansrls · 2026-05-28T22:52:50Z

Sign in to view

+        dag_formal_terminal(token_class: dag_token_ident),
+        dag_formal_terminal(token_class: dag_token_eq),
+        dag_formal_nonterminal(production: dag_production_expr)
+      ]


Valid on pre-6887656db HEAD: dag_production_expr was referenced throughout dag_wave1_formal_productions but had no FormalProduction lhs.

Fixed in 6887656db by adding five choice arms mirroring dag_grammar_expr_expr() — each routes to an existing declared production (match_expr, if_expr, fn_literal, block_expr, binary_expr). Closure audit: every grammar_formal_nonterminal_occurrence(identity: dag_production_*) in the formal table now has a matching lhs: rule.

— sent from valiant-boar-99

Replay PR #3844 onto main after #3828: hoist formal-production helpers to std/grammar, migrate Wave-1/2a extdeps relation rows to FormalProduction + formal_production_to_node (rust/kotlin/dag), CI v4 receipt on ubuntu-latest, strict_deprose status sync, dag ingest claim path move. Co-authored-by: Cursor <cursoragent@cursor.com>

…guage wrappers (follow-on to #3828) (#3847) * WIP: std/grammar.dag: add grammar_formal_production helper + dedup per-langua * WIP: std/grammar.dag: add grammar_formal_production helper + dedup per-langua

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 6887656d · Trigger: schedule
Thinking: 295s wall

Non-blocking — Strengths

src/v4/extdeps/languages/dag.dag The new flat FormalProduction surface is closed over its nonterminal occurrences, and the remaining GrammarExpr scaffold is bounded to CP-1b convergence.

✅ No blocking concerns in the changed .dag grammar-model surface.

briansrls force-pushed the session/valiant-boar-99 branch from 21b5d21 to 0b2a865 Compare May 28, 2026 20:39

briansrls and others added 9 commits May 28, 2026 20:52

WIP: grammarexpr rework

a7751ab

WIP: grammarexpr rework

efe95aa

WIP: grammarexpr rework

f558bc3

WIP: grammarexpr rework

35a69eb

WIP: grammarexpr rework

7d61e32

WIP: grammarexpr rework

7eb82e2

WIP: grammarexpr rework

fa807ce

WIP: grammarexpr rework

e51abaf

chore: remove local dag formal generator script from PR

38a6867

The generated dag.dag FormalProduction block is already landed; the one-off generator is not a deliverable. Co-authored-by: Cursor <cursoragent@cursor.com>

briansrls force-pushed the session/valiant-boar-99 branch from 7e60594 to 38a6867 Compare May 28, 2026 20:52

briansrls changed the title ~~grammarexpr rework~~ extdeps: flat FormalProduction RHS for wave-1/2a grammar data (T-4.17) May 28, 2026

briansrls marked this pull request as ready for review May 28, 2026 20:54

briansrls and others added 6 commits May 28, 2026 21:04

WIP: grammarexpr rework

250c5b8

WIP: grammarexpr rework

29e6de4

WIP: grammarexpr rework

376ecc9

WIP: grammarexpr rework

9e1c7ef

WIP: grammarexpr rework

5750058

briansrls and others added 2 commits May 28, 2026 21:42

WIP: grammarexpr rework

9fc7e80

briansrls commented May 28, 2026

View reviewed changes

WIP: grammarexpr rework

6887656

briansrls merged commit 77e6626 into main May 28, 2026
7 checks passed

briansrls mentioned this pull request May 28, 2026

T-4.17 Wave 2a wasm migration replacement: rebase archived #3802 work after #3807 std lexing/grammar migration; fix recorded wasm index/localidx blockers; keep draft until validated #3833

Merged

1 of 2 tasks

briansrls commented May 28, 2026

View reviewed changes

briansrls mentioned this pull request May 28, 2026

T-4.17 Wave 2a: python lex/grammar data fill (deps T-6/T-7/T-4; cpp/java excluded) #3798

Merged

5 of 6 tasks

briansrls deleted the session/valiant-boar-99 branch June 1, 2026 18:43

Navigation Menu

Repository navigation

extdeps: flat FormalProduction RHS for wave-1/2a grammar data (T-4.17) - #3828

extdeps: flat FormalProduction RHS for wave-1/2a grammar data (T-4.17)#3828
briansrls merged 18 commits into
mainfrom
session/valiant-boar-99

briansrls commented May 28, 2026 •

edited

Loading

Uh oh!

briansrls commented May 28, 2026

Uh oh!

briansrls commented May 28, 2026

Uh oh!

briansrls commented May 28, 2026

Uh oh!

briansrls commented May 28, 2026

Uh oh!

briansrls left a comment

Uh oh!

This comment was marked as resolved.

Uh oh!

briansrls May 28, 2026

Uh oh!

Uh oh!

briansrls left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

briansrls commented May 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Worker attestation

Uh oh!

briansrls commented May 28, 2026

Uh oh!

briansrls commented May 28, 2026

Uh oh!

briansrls commented May 28, 2026

Uh oh!

briansrls commented May 28, 2026

Uh oh!

briansrls left a comment

Choose a reason for hiding this comment

Uh oh!

This comment was marked as resolved.

Uh oh!

briansrls May 28, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

briansrls left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

briansrls commented May 28, 2026 •

edited

Loading