Elamur GroupantsโŒ•
โ˜ฐ

mova-topic-corpus-library

The groupant beside tools/topic-corpus/src/library.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

componentlabelwhere
approximantpartialForward approximant
astpartialsyntax.json
compositionpartialComposition
costpartialCost
descriptantpresentDescriptant
invariantpartialInvariant
testsfinite evidencetests.md

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.

idtypelabelwhere
behavior_complementbehavior_complementpartial source derived modelbehavior_complement
reverse_reconstructionreverse_reconstructionpartial source derived modelreverse_reconstruction

Model text

Descriptant

Read-only view of the paper library: export rows, citation edges, resolved reference lists, per-paper refs.json bibliographies, OpenAlex topic taxonomy, author positions; id/DOI/arXiv/ACL/title resolution and text discovery. Source: tools/topic-corpus/src/library.rs.

Forward approximant

load: stream export JSONL -> Work rows with keyword ids (title+abstract words, stop words out) and document frequencies -> indices by id, DOI (+ACL DOI), arXiv, ACL, normalized title (>=3 words). resolve(entry): DOI > arXiv > ACL > title with year +-1. text(w): text.md (open unless local-only) > text.local.md > eprint.md.

Invariant

First row of an id wins; stale layout paths are matched by their last component; no network.

Cost

One pass over the export (~186k rows) plus one pass over citations.tsv (2.57M lines) or refs.jsonl per query.

Composition

Parent: topic-corpus.

behavior_complement

Does not write anything; does not fetch; title matches need three words and the year within one. Evidence status: source-derived model with finite checks, not a proof.

reverse_reconstruction

From a Work the export row's graph fields are recoverable; abstracts are reduced to keyword ids and not recoverable. No unique inverse is claimed.

tests.md

library::tests::loads_resolves_and_finds_texts (fixture library).

cargo test --manifest-path tools/topic-corpus/Cargo.toml --offline with WORLD_EN_MODEL=$MOVA_DATA/download/0/0.4/en.bin (18 unit tests; fixture library, fake papers binary, fake terminal, local Sea Spider store; two identical runs on 2026-11-04). Without the model the model-dependent tests report a skip.

Negative controls 2026-11-04 (each red, then restored): public admission without the origin condition; assert_public without the Sea Spider/draft/note refusals; seeding without the learner lock; shadow-library filter off; public citing that ignores refusals; cache eviction ignoring the hot pin.

These are finite fixtures and live samples; not a proof.

This page in Markdown โ†’ https://md.groupants.elamur.ai/free/en/mova/tools/topic-corpus/src/library.rs.groupant/