Elamur Groupantskg-cols-integer-codecs
The groupant beside kg/src/cols.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 | approximant |
| ast | partial | ../syntax.json |
| composition | partial | composition |
| cost | partial | cost |
| descriptant | present | descriptant |
| invariant | partial | invariant |
| tests | executed bounded checks | tests |
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 |
|---|---|---|---|
behavior_complement | behavior_complement | partial source derived model | behavior_complement |
reverse_reconstruction | reverse_reconstruction | partial source derived model | reverse_reconstruction |
modulo | modulo | partial source derived model | modulo |
signed_boundary | signed_boundary | partial source derived model | signed_boundary |
empty_count | empty_count | partial source derived model | empty_count |
error_cursor | error_cursor | partial source derived model | error_cursor |
Model text
descriptant
kg-cols-integer-codecs
Private LEB/delta/zigzag/byteplane transformations.
Modeling epoch — 2026-10-31
Source read before tests and models; retained prospective notes accompany the parent. Current evidence uses bounded invented inputs only. This is partial modeling, not proof or corpus accuracy.
approximant
Forward relation
LEB emits 7-bit groups; delta wraps x-prev; inverse wraps prev+d. Zigzag encodes signed wrapping difference. Planes transpose four little-endian bytes across all values.
Executable approximant gap
Exact categorical structure, identifiers, parser decisions and effects have no useful percentage metric established here. No executable surrogate, automatic substitution or 1/2/5/10% guarantee is claimed.
invariant
Invariant and domain
Decode of encoder-produced values and exact count recovers u64/u32 bitpatterns including MAX and repeated zero.
Cost and resources
O(n), <=10 LEB bytes/u64,4n planes; allocation count arithmetic not checked.
cost
Cost and resources
O(n), <=10 LEB bytes/u64,4n planes; allocation count arithmetic not checked.
composition
Composition
Private LEB/delta/zigzag/byteplane transformations.
LEB emits 7-bit groups; delta wraps x-prev; inverse wraps prev+d. Zigzag encodes signed wrapping difference. Planes transpose four little-endian bytes across all values.
The parent invokes scoped transformations through the actual functions named in their source ranges. Children partition useful observations, not a complete semantic proof; boundaries can omit helpers, tests and effects.
models
behavior_complement
Truncated continuation/too-long LEB and wrong plane length error. Trailing LEB bytes and overlarge terminal payload are not fully rejected.
reverse_reconstruction
Plane transpose at fixed n is bijective; decoded LEB vector has multiple byte preimages from trailing/noncanonical encodings.
modulo
Wrapping arithmetic handles descending src too, despite usual sorted production use.
signed_boundary
u64 endpoints reinterpret as i64 differences; zigzag and inverse preserve two's complement bitpattern.
empty_count
n0 decodes no LEB values even with extra bytes; plane n0 requires empty bytes.
error_cursor
unleb increments position after each consumed byte; no rollback on error.
tests
Evidence
Parent source checks (a fragment inherits only the behavior they actually exercise):
- cols::tests::roundtrip_and_filter
- cols::wave29::codec_boundaries_and_errors
Selected kg tests: 15 named outcomes, twice debug/twice release, equal sorted outcomes. Standalone actual global/build.rs: 3 named outcomes per run, twice debug/twice optimized; build_child is normally no-op, parent fixture invokes it against isolated synthetic seeds. Eight compiled source mutations (one per source) each compiled successfully and exited101 on the focused assertion; exact source restoration followed by positive rerun.
This fragment's untested branches remain source-derived. No live graph, DB, corpus, teacher, GPU or held-out evaluation was used. Command specs and numeric exits are retained in the external wave29 graphlang receipt. Strict kg Clippy reports existing source lints; see receipt.
This page in Markdown → https://md.groupants.elamur.ai/free/en/mova/kg/src/cols.rs.groupant/integer-codecs.groupant/