Elamur Groupantsprag-roles-token-projection
The groupant beside prag/src/roles.rs in the Mova repository: what the code computes, what it never does, how to go back from its outputs, its invariants, cost and checks.
Facts
Implementation material. A Mova code model describes this code at one source revision: partial models, finite checks and their stated scope. It is not a reviewed general exposition, and its labels are shown as written, never raised.
Components
| component | label | where |
|---|---|---|
| approximant | partial source model | forward |
| ast | partial | ../syntax.json |
| composition | partial source model | composition |
| cost | partial source model | cost |
| descriptant | partial source model | epoch-memory |
| invariant | partial source model | invariant |
| tests | partial source model | checks |
Obernants
The typed relations around the forward model: the behaviour complement (what the code never does), the reverse reconstruction (every input compatible with an output) and specialised views.
| id | type | label | where |
|---|---|---|---|
prag-roles-token-projection-behavior-complement | behavior_complement | partial source model | behavior-complement |
prag-roles-token-projection-reverse-reconstruction | reverse_reconstruction | partial source model | reverse-reconstruction |
prag-roles-token-projection-coupling | constraint_relation | partial source model | coupling |
Model text
forward
Read supplied tree into event/speech frames. Choose first unused eligible predicate matching the first verb lemma token. For each filled slot choose closest nonpredicate form/lemma match; pronoun slots retain Ent.pron. Sort and deduplicate role tuples.
invariant
Output positions are 1-based; tuples sorted unique. Equal-distance ties choose earliest token. Missing lexical matches produce no tuple.
cost
External frame reading plus O(FN) token scans per fixed frame slot; O(R log R) role sorting.
composition
Native Word -> Tree::of_words -> frames_of_tree -> roles/world_docs executes. CLI annotation/cache paths inspected only; temporary Ctx is not retained learned state.
epoch-memory
2026-10-30 initial source-bound epoch. Irony malformed-head native witness panicked before bounds repair. Missing parent evidence now abstains while linear evidence remains. Lens/roles production unchanged. Costs are prospective asymptotic models; no optimization, percentage guarantee or timing claim.
checks
cargo test -p prag --lib roles::wave12_models -- --test-threads=1. File-root evidence.md binds repeated profiles and mutants. No live parser/teacher/corpus or persisted exam evaluation.
behavior-complement
Domain: finite inputs satisfying this fragment's forward premises and the parent API tag/state constraints; output is its scoped observation. Let J_fragment(x,y) impose the concrete relations: Read supplied tree into event/speech frames. Choose first unused eligible predicate matching the first verb lemma token. For each filled slot choose closest nonpredicate form/lemma match; pronoun slots retain Ent.pron. Sort and deduplicate role tuples. Also require: Output positions are 1-based; tuples sorted unique. Equal-distance ties choose earliest token. Missing lexical matches produce no tuple. The inner absent-behavior model is {(x,y): not J_fragment(x,y)}. This excludes wrong indices, selections or outputs, rather than treating a scope disclaimer as a behavioral predicate. Finite checks support selected cases; unknown external semantics remain outside this claim.
reverse-reconstruction
Same tuples admit different parses/context states and duplicate frame emissions. Repeated lexical mentions are not uniquely recoverable from roles. Relation B(y) retains every modeled compatible antecedent; no unique inverse or complete executed solver.
Formally retain every parent-tagged input and compatible intermediate state satisfying the forward relation with the observed fragment output. Do not merge antecedents merely because their projected result agrees; no unique inverse or complete executed solver is claimed.
coupling
Native supplied cat/eat/fish tree has exactly agent (1,2) and patient (3,2). Parser quality and repeated-reference semantic correctness remain unproved.
Typed entries bind the manifest revision, stated domain and observation, assuming stable compiled dependencies and bounded execution. Status is partial source-derived relation with finite named checks. Partial syntax leaves macro bodies opaque and is no semantic proof.
This page in Markdown → https://md.groupants.elamur.ai/free/en/mova/prag/src/roles.rs.groupant/token-projection.groupant/