Author std/lexing.dag + std/grammar.dag: move LexRule/Grammar/Production canonical types out of compiler/tokenize.dag + compiler/parse.dag into new std/ homes so compiler and extdeps both import std, neither imports the other

scripts/strict_deprose_dag.py

@@ -652,23 +652,23 @@ def main() -> None:
             "src/v4/extdeps/languages/typescript.dag",
             "// Scope: TypeScript 5.9 + ECMA-262 ES2025 numeric primitive fact-bundles and ModelCore wave-1.",
             "// Anchor: https://www.typescriptlang.org/docs/handbook/2/everyday-types.html — ECMA-262 https://tc39.es/ecma262/2025/multipage/",
-            "// Consumes: v4.compiler.parse, v4.compiler.tokenize, v4.std.collection, v4.std.node, v4.std.logic, v4.std.algebra, v4.std.model_core, v4.std.target_model, v4.std.text.",
+            "// Consumes: v4.std.grammar, v4.std.lexing, v4.std.collection, v4.std.node, v4.std.logic, v4.std.algebra, v4.std.model_core, v4.std.target_model, v4.std.text.",
             "// Status: T-4 typescript slice — ECMA `number` (IEEE-754 binary64) + `bigint` (exact unbounded ℤ) fact-bundles; `core: ModelCore` primitives from `ts_numeric_facts_catalog` via fold; canonical_symbols = catalog surface spellings + wave-1 lex/grammar/MVP; target_model edge keys from std/target_model.dag; MVP-1 grammar/token substrate for T-11; bool canonical-B decl-ref — 🟡 E-6(b) staging.",
             "// 🟡",
         ),
         (
             "src/v4/extdeps/languages/swift.dag",
             "// Scope: Swift language/standard-library scalar fact-bundles and ModelCore wave-1.",
             "// Anchor: https://docs.swift.org/swift-book/documentation/the-swift-programming-language/thebasics/",
-            "// Consumes: v4.compiler.parse, v4.compiler.tokenize, v4.std.collection, v4.std.model_core, v4.std.target_model, v4.std.node, v4.std.logic, v4.std.algebra, v4.std.text.",
+            "// Consumes: v4.std.grammar, v4.std.lexing, v4.std.collection, v4.std.model_core, v4.std.target_model, v4.std.node, v4.std.logic, v4.std.algebra, v4.std.text.",
             "// Status: T-4 Swift wave-1; fixed-width integer spellings are explicit, Int/UInt stay platform-word-width facts, Float/Double carry IEEE-754 precision; canonical_symbols = catalog surface spellings + wave-1 lex/grammar/MVP; target_model edge keys from std/target_model.dag; MVP-1 grammar/token substrate for T-11; bool canonical-B decl-ref is E-6(b) staging.",
             "// 🟡",
         ),
         (
             "src/v4/extdeps/languages/wasm.dag",
             "// Scope: WebAssembly Core numeric value types — LanguageModel fact-bundles (Shape A).",
             "// Anchor: https://webassembly.github.io/spec/core/types.html#number-types",
-            "// Consumes: v4.compiler.parse, v4.compiler.tokenize, v4.std.collection, v4.std.model_core, v4.std.target_model, v4.std.node, v4.std.logic, v4.std.algebra, v4.std.text.",
+            "// Consumes: v4.std.grammar, v4.std.lexing, v4.std.collection, v4.std.model_core, v4.std.target_model, v4.std.node, v4.std.logic, v4.std.algebra, v4.std.text.",
             "// Status: T-4 wasm slice — Core §2.3.1 number types (i32/i64/f32/f64 wave-1); `core: ModelCore` (#3474); canonical_symbols = catalog surface spellings + wave-1 lex/grammar/MVP; target_model edge keys from std/target_model.dag; MVP-1 WAT grammar/token substrate for T-11; v128/funcref/externref 🟡 wave-2; integer sign-agnostic per spec (width + modular wrap, not signed/unsigned partition).",
             "// 🟡",
         ),

src/v4/TASKS.md

@@ -492,6 +492,22 @@ conceptual `(Source, LanguageModel)` spelling.
 - merged: `src/v4/compiler/01_tokenize.dag` (B2-OMNI, E0 contract)
 - v3 L2.5 design: `docs/r3-path-b-tokenize-parse-brief-set.md` PB-2
 
+#### T-6.1: LexRule token-class query dissolution
+
+**Owner/lane:** T-6 std lexing substrate follow-up.
+
+**Missing primitive:** substrate-derived `LexRule` token-class projection plus
+`TokenRule` discriminant query in `src/v4/std/lexing.dag`.
+
+**Interim surface:** `lex_rule_token_class_member` / `lex_rule_set_token_class_member`
+are allowed only as the shared std query while language slices need canonical symbol
+membership and no derived `LexRule` projection exists.
+
+**Dissolve-on-arrival:** the first PR that lands the projection/query primitive
+must delete the hand-written `LexRule` arm match in `lex_rule_token_class_member`
+and make `lex_rule_set_token_class_member` consume the substrate query in the same
+commit train. No extdeps/compiler-local duplicate predicate may be added meanwhile.
+
 ---
 
 ### T-7: compiler/02_parse.dag
@@ -515,6 +531,23 @@ syntax-side projection of the landed `LanguageModel = Node` authority in
 - merged: `src/v4/compiler/02_parse.dag` (B2-OMNI, G0 contract)
 - v3 L2.5 design: `docs/r3-path-b-tokenize-parse-brief-set.md` PB-3
 
+#### T-7.1: GrammarExpr node projection dissolution
+
+**Owner/lane:** T-7 std grammar substrate follow-up.
+
+**Missing primitive:** substrate-derived `GrammarExpr` fold/projection to canonical
+`Node` edges in `src/v4/std/grammar.dag`.
+
+**Interim surface:** `grammar_expr_to_node` is allowed only as the shared std
+projection used by `grammar_to_node` and symbol-membership checks while no
+substrate fold exists for `GrammarExpr`.
+
+**Dissolve-on-arrival:** the first PR that lands the `GrammarExpr` fold/projection
+primitive must delete the recursive hand-written coproduct walker in
+`grammar_expr_to_node` and route `grammar_to_node` through the substrate projection
+in the same commit train. No compiler/extdeps-local `GrammarExpr` re-encoder may
+be added meanwhile.
+
 ---
 
 ### T-8: compiler/03_normalize.dag + 03_resolve.dag

src/v4/compiler/00_compile.dag

@@ -20,7 +20,8 @@ import v4.compiler.normalize { normalize }
 import v4.compiler.parse { parse }
 import v4.compiler.resolve { resolve }
 import v4.compiler.tokenize { tokenize }
-import v4.compiler.target_carriers { TargetModel, TargetSource }
+import v4.compiler.target_carriers { TargetSource }
+import v4.std.target_model { TargetModel }
 import v4.extdeps.languages.dag { dag_language_model_wave1, qualified_name_from_module_node }
 import v4.lens.fact_density { fact_density_hollow_alias_gate }
 import v4.std.algebra { Cons, Empty, FreeMonoid, list_append, list_snoc_item }

src/v4/compiler/01_tokenize.dag

@@ -16,62 +16,37 @@ import v4.std.algebra { Cons, Empty, ListTailResult, TailAbsent, TailFound, fold
 import v4.std.text { Char, CharAbsent, CharFound, String, string_head, string_is_empty, string_tail }
 import v4.std.nat { Nat, nat_lte, nat_gte }
 import v4.std.logic { Bool }
-
-// 🟡 coproduct dissolution — CP-1b item 13a; lexical data is modeled, per-arm lexer morphism witness remains gated.
-type LexCharClass
-  = WhitespaceChar
-  | DigitChar
-  | HexDigitChar
-  | IdentStartChar
-  | IdentContinueChar
-  | StringTextChar
-  | LineCommentTextChar
-  | BlockCommentTextChar
-
-
-// 🟡 coproduct dissolution — CP-1b item 13b; declarative parse/emit pattern data is modeled, bidirectional grammar carrier convergence remains gated.
-type LexPattern
-  = EmptyPattern
-  | LiteralPattern { text: String }
-  | KeywordPattern { text: String }
-  | CharPattern { char: Char }
-  | CharClassPattern { class: LexCharClass }
-  | SequencePattern { left: LexPattern, right: LexPattern }
-  | ChoicePattern { left: LexPattern, right: LexPattern }
-  | OptionalPattern { element: LexPattern }
-  | RepeatPattern { element: LexPattern }
-  | OneOrMorePattern { element: LexPattern }
-  | DelimitedPattern { open: LexPattern, body: LexPattern, close: LexPattern }
-
-
-// 🟡 coproduct dissolution — CP-1b item 13c; token/trivia rule data is modeled, per-arm lexer morphism witness remains gated.
-type LexRule
-  = TokenRule { token_class: Symbol, pattern: LexPattern }
-  | TriviaRule { pattern: LexPattern }
-
-
-// Maximal munch: lex_try_rules selects the accepting rule with the longest lexeme;
-// equal-length ties prefer the later rule (specific literals/keywords after ident).
-type LexRuleSet {
-  rules: List<LexRule>
-}
-
-
-// 🟡 coproduct dissolution — CP-1b item 13d; modeled rule data is present, generic lexical walk realized.
-type LexRules
-  = VoidLexRules
-  | ModeledLexRules { root: LexRuleSet }
-
-
-type Token {
-  class: Symbol
-  lexeme: String
-  file: Symbol
-  start: Int
-  end: Int
-}
-type TokenStream {
-  tokens: List<Token>
+import v4.std.lexing {
+  BlockCommentTextChar,
+  CharClassPattern,
+  CharPattern,
+  ChoicePattern,
+  DelimitedPattern,
+  DigitChar,
+  EmptyPattern,
+  HexDigitChar,
+  IdentContinueChar,
+  IdentStartChar,
+  KeywordPattern,
+  LexCharClass,
+  LexPattern,
+  LexRule,
+  LexRuleSet,
+  LexRules,
+  LineCommentTextChar,
+  LiteralPattern,
+  ModeledLexRules,
+  OneOrMorePattern,
+  OptionalPattern,
+  RepeatPattern,
+  SequencePattern,
+  StringTextChar,
+  Token,
+  TokenRule,
+  TokenStream,
+  TriviaRule,
+  VoidLexRules,
+  WhitespaceChar
 }
 
 // 🟡 coproduct dissolution — CP-1b item 13e; per-pattern match result; dissolves when LexPattern converges with GrammarExpr on single bidirectional grammar carrier.
@@ -89,25 +64,6 @@ data tokenize_lex_e0_source_not_exhausted: Symbol = tokenize_lex_e0_source_not_e
 data tokenize_lex_e1_unrecognized_char: Symbol = tokenize_lex_e1_unrecognized_char
 
 
-fn lex_rule_token_class_member(rule: LexRule, token_class: Symbol) -> Bool {
-  match rule {
-    TokenRule { token_class: c, pattern: _ } => c == token_class
-    TriviaRule { pattern: _ } => false
-  }
-}
-
-
-fn lex_rule_set_token_class_member(rules: LexRuleSet, token_class: Symbol) -> Bool {
-  fold_list(
-    xs: rules.rules,
-    empty: false,
-    cons: fn(acc, rule) {
-      acc || lex_rule_token_class_member(rule: rule, token_class: token_class)
-    }
-  )
-}
-
-
 fn tokenize_lex_e0_source_not_exhausted_diagnostic(at: Locus) -> Diagnostic {
   Diagnostic {
     reason: tokenize_lex_e0_source_not_exhausted,