Glossary

For the formal definitions of Omnist's terms (node, edge, Document, Schema, Scalar, cardinality, satisfiability, normalize, extract, infer, and the rest), see the Glossary chapter of spec.omnist.dev (and the Document Model and Schema Model chapters for the full formal treatment). This page only maps those terms to their Python names in this library.

Concept → Python name

Concept (spec.omnist.dev)	Python name here
node / Document	Doc — the wrapper class (.edges(), .get(), .add(), ...); see the API reference
Scalar	Scalar class; ready-made instances under t (t.string, ...) or as top-level names (STRING, ...)
any type	AnyType class; singleton instance t.any
field	Field(label, type, min, max)
record (concept)	Record, built by the record(*fields) builder
Ref	Ref, built by the ref(name) builder
Schema	Schema — root Ref plus an env dict of Records; see the API reference
a value's runtime classification	value_kind(v) / matches_kind(value, name) in omnist/schema.py
compatible_with / equivalent	Schema.compatible_with() / Schema.equivalent()
normalize	Schema.normalize()
extract	Schema.extract(*labels) — raises SchemaError if a kept label set would drop a mandatory field
prune	Schema.prune()
infer	top-level infer(samples) function
lint	Schema.lint(), returning LintFinding(code, severity, location, message) — see the CLI reference
OSD	parsed by parse_schema(), produced by to_osd()
OML	read/written via read_oml/write_oml

Python-only terms

These have no spec counterpart — they're implementation details of this library:

• adjustment / WriteReport — a recorded change a writer made because the target format can't hold a value losslessly (e.g. TOML dropping null). strict=True raises WriteError instead of adjusting silently. See the API reference.
• codec — a Format's registered read/write/check functions (register_format), letting Doc.from_format/to_format/check_format use it like a built-in format. See the API reference.