Elamur Groupantsworld-quant-expressions
The groupant beside world/src/quant.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-quant-expressions-behavior-complement | behavior_complement | partial source model | behavior-complement |
world-quant-expressions-reverse-reconstruction | reverse_reconstruction | partial source model | reverse-reconstruction |
world-quant-expressions-rational-oracle | constraint_relation | partial source model | rational-oracle |
world-quant-expressions-reference-normalization | information_loss | partial source model | reference-normalization |
Model text
Tokenizer, precedence parser, evaluator and literal collection.
forward
Sentence-prefix validation applies only to numeric literals. References read the current command-order state and carry no sentence provenance. Thus @2 set a.x=19 followed by @1 set b.x=a.x and @1 ask b.x remains green when 19 occurs only in sentence 2. This is a documented current limitation, not complete temporal provenance or a proposed behavior change.
Tokenize operators/whitespace; parse term before expression; atoms are parentheses, rationals or lowercase dotted references. eval recursively applies exact rational arithmetic; nums_of traverses literals left-to-right.
invariant
Accepted parse consumes all tokens. Undefined references and zero divisors fail. Direct Ex::Bin with an unknown operator falls into division; the model restricts to parser-produced operators.
cost
O(L) tokenization and O(E) node construction, recursive rational costs dependent on bit length; depth may reach O(E).
composition
Parsing feeds ordered execution; literal collection precedes evaluation against state, and check merges parse errors, execution diagnostics and gold inequality. Text extraction is an external exact dependency. Numeric-origin errors do not suppress effects, unlike evaluation and anchor errors; a parent must preserve this distinction. Dataset/teacher adapters do not establish script semantics.
epoch-memory
2026-10-30: regressions first demonstrated backward-anchor number leakage and destination credit from missing transfer sources. Prefix endpoints now restrict each command to its own sentence prefix while preserving incremental extraction; missing-source transfer skips effects. Diagnostic-only literal errors remain unchanged. No corpus or teacher calls.
checks
See ../tests.md and native quant::wave8_models tests. Models are source-derived partial relations, with finite evidence rather than general proof. Syntax projection excludes macro expansion, name resolution and inferred types. No executable approximant: exact gates and categorical choices have no justified percentage metric here.
behavior-complement
Universe: source strings and proposed parsed/evaluated results under fixed state. Trailing tokens, missing parentheses and unary minus exclude successful parsing; these form an inner absent set, not a full resource-safety model.
reverse-reconstruction
For state s and rational r retain all accepted expressions evaluating to r. Equivalent parenthesizations and arithmetic identities produce multiple preimages; AST still loses whitespace.
rational-oracle
For a,c in 0..8 and b in 1..8, ((a+c)/b)-ca=(a+c-ca*b)/b exactly. Tests compare 648 evaluations plus six malformed strings and zero division.
reference-normalization
Case variants map to the same lowercase Ref, whereas Num retains original numeric spelling. Reverse formatting therefore has different fibers for reference and numeric atoms.
This page in Markdown → https://md.groupants.elamur.ai/free/en/mova/world/src/quant.rs.groupant/expressions.groupant/