Skip to content

sct info

Inspect any file produced by sct and print a summary - without needing to open a database or write a query.

Accepts concept .ndjson, payload .refsets.ndjson, .db, and .arrow files.

Usage

sct info <FILE> [--format text|json|yaml]

--format (-f) defaults to text. Use json or yaml for scripts and agents - the field names below (concept_count, hierarchies, tct_row_count, etc.) are stable across all three file types where they apply.

Examples

sct info snomed-uk-20260311.ndjson
sct info snomed-uk-20260311.refsets.ndjson
sct info snomed.db
sct info snomed-embeddings.arrow
sct info snomed.db --format json | jq .tct_row_count

Output by file type

.ndjson

File:           snomed-uk-20260701.ndjson
Size:           1.2 GB
Format:         NDJSON
Schema version: 5
Edition:        uk_sct2mo_42.3.0_20260701000001Z
Release date:   2026-07-01
Release id:     uk_sct2mo_42.3.0_20260701000001Z
Built by:       sct 0.18.2
Concepts:       837,930

Hierarchy breakdown (20 top-level):
  Pharmaceutical / biologic product             231,194
  Physical object                               224,797
  Clinical finding                              137,830
  ...

Reports: - Format (NDJSON) - Total concept count - active only by default; a separate (inactive) line appears when the file was built with --include-inactive - schema_version - Provenance, when present in the file's header: edition label, release date, release id, and the sct version that built it. Falls back to a filename-inferred release date for older pre-provenance NDJSONs. - Hierarchy breakdown

.refsets.ndjson

File:             snomed-uk-20260701.refsets.ndjson
Size:             84.2 MB
Format:           Payload refset NDJSON
Schema version:   1
Edition:          uk_sct2mo_42.3.0_20260701000001Z
Release date:     2026-07-01
Source fingerprint: sha256:...
Records:          1,234,567
  Complex Map:    12,345
  Extended Map:   1,210,000
  AttributeValue: 12,222

Reports the companion schema version and record-family counts, then identifies the exact concept artefact provenance to which the stream is bound. Invalid headers, malformed records, or content-fingerprint mismatches fail rather than reporting a partial summary.

.db

File:              snomed.db
Size:              2.6 GB
Format:            SQLite (sct sqlite)
Schema version:    5
Edition:           uk_sct2mo_42.3.0_20260701000001Z
Release date:      2026-07-01
Release id:        uk_sct2mo_42.3.0_20260701000001Z
Built by:          sct 0.18.2
Concepts:          837,930
FTS5 rows:         837,930
IS-A edges:        1,605,202
TCT rows:          11,607,152

Hierarchy breakdown (20 top-level):
  Pharmaceutical / biologic product             231,194
  Physical object                               224,797
  Clinical finding                              137,830
  ...

Reports: - Format (SQLite (sct sqlite)) - Provenance, when present: edition label, release date, release id, and the sct version that built it - Concept count - schema_version - FTS5 row count - IS-A edge count (concept_isa table) - Complex Map, Extended Map, and Attribute Value row counts when payload-refset tables are populated (complex_map_refset_count, extended_map_refset_count, and attribute_value_refset_count in structured output) - TCT row count and usability (tct_row_count / tct_usable in structured output), or a note explaining how to build or repair it with sct tct - Hierarchy breakdown

.arrow

File:             snomed-embeddings.arrow
Size:             2.6 GB
Format:           Arrow IPC (sct embed)
Edition:          uk_sct2mo_42.3.0_20260701000001Z
Release date:     2026-07-01
Release id:       uk_sct2mo_42.3.0_20260701000001Z
Built by:         sct 0.18.2
Embeddings:       837,930
Dimension:        768

Schema:
  id                   Utf8
  preferred_term       Utf8
  hierarchy            Utf8
  embedding            FixedSizeList(768 x non-null Float32)

Reports: - Format (Arrow IPC (sct embed)) - Provenance, when present: edition label, release date, release id, and the sct version that built it - Embedding count - Embedding dimension - Arrow schema (field names and types) - File size

The Arrow metadata records the Ollama embedding model name (for example nomic-embed-text) so query commands can reject cross-model searches. sct info does not currently display that field.

See also