Elamur Groupantsworld-principles-render-selection
The groupant beside world/src/principles.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 |
|---|---|---|---|
world-principles-render-selection-behavior-complement | behavior_complement | partial source model | behavior-complement |
world-principles-render-selection-reverse-reconstruction | reverse_reconstruction | partial source model | reverse-reconstruction |
world-principles-render-selection-coupling | constraint_relation | partial source model | coupling |
Model text
forward
statement maps link and negation to English templates. key projects subject, Debug link, object and polarity with pipe separators. pending loads checked keys, then pending_with filters source prefix 'derived from', excludes matching keys, stable sorts by key, deduplicates equal keys retaining the first encounter, and truncates to n. Encounter order is BTreeMap subject buckets then vector order, not source chronology.
invariant
No assessment/training boundary is inferred from the source prefix. Frequency, count and source are absent from keys and statement; equal keys collapse such distinctions. Unescaped pipe components can collide. Selection is deterministic for supplied Knowledge and checked-key snapshots, not semantic ranking or confidence ordering.
cost
F candidates: O(F log F) key comparisons with fresh string allocations; O(F) cloned facts plus key bytes. Checked-file load/parse is external. This epoch refactors pure selection without optimization.
composition
Source callers: main dispatches main_args; family invokes round once per research round and reports returned counts/errors; round consumes pending and statement then external teacher/cache/world operations. Native public join: caller-supplied know::Fact -> statement. Selection tests call the exact private pure helper used by pending with supplied keys, avoiding filesystem state.
epoch-memory
2026-10-30: initial source-bound review. Prospective cost model predicts no asymptotic change: selection helper extraction preserves sorting and cloning; list repair replaces one scalar division with checked division. Old zero-limit native witness divides by zero. Final implementation returns a nonzero stride and take(0) produces no rows. Mind production behavior is unchanged. No surrogate, speed or full-system correctness claim.
checks
cargo test -p world --lib principles::wave12_models -- --test-threads=1. File evidence.md binds repeated native profiles and compiled production controls. Partial syntax leaves macros opaque; source-bound syntax is not semantic AST verification.
behavior-complement
Domain and observation scope: statement(F) and pending_with(K,C,n) on supplied finite Facts/Knowledge/checked keys, fixed key/link-name projection; no cache I/O.
Using root relations S,Q, exclude template/polarity or subject/object mismatches in S; exclude ineligible/checked Facts, repeated/descending keys, length above n, wrong truncated prefix or a non-first duplicate's full Fact in Q. This does not forbid key collisions or infer source truth: equal projected keys collapse even when count/frequency/provenance differ. Eight polarity fixtures and 40 independently enumerated selections execute; fallback link templates are inspected only.
reverse-reconstruction
S fibers retain all compatible template decompositions and lost frequency/count/provenance. Q fibers retain every K,C,n producing exactly the observed ascending eligible unique-key prefix, with each full Fact first for its key. Unselected records and tail candidates remain free subject to this prefix constraint; no unique knowledge base or checked file is recovered.
coupling
40 subset/limit cases independently enumerate the expected a,b,c selection, excluded nonderived d and stable duplicate provenance. Eight polarity rows use literal grammar templates and demonstrate information loss when source/frequency/count change. Public statement joins supplied Fact values; pending itself additionally performs unchecked filesystem loading.
Typed entries bind the manifest source revision, this section's domain/operator/observation, and the named finite checks. Status is partial source-derived model. No complete reconstruction search, universal proof or factual-truth guarantee is claimed.
This page in Markdown → https://md.groupants.elamur.ai/free/en/mova/world/src/principles.rs.groupant/render-selection.groupant/