ftableschema

Ftableschema is a compact, machine-readable descriptor format for describing the schema of tabular data, particularly flat tables used in data exchange and processing pipelines. The goal is to provide a language-agnostic way to convey what columns exist, how they should be interpreted, and how they relate to other tables, independent of any specific database dialect.

A typical ftableschema document specifies one table and includes: table name, a list of columns with fields such as name, type, and nullable, default values; optional constraints such as primary keys and unique constraints; foreign keys; and metadata like description and version. Each column definition covers name, data type, nullability, and, optionally, length or precision. The schema may also describe indexes and table-level metadata.

Formats and interoperability: ftableschema is designed to be expressed in JSON or YAML, or in a compact inline form used by lightweight tools. The core principle is to describe the structure rather than the data, enabling validation, code generation, and cross-system interoperability. Implementations may provide utilities to translate to SQL DDL, JSON Schema, or language-specific data models.

Usage: It is used to exchange table schemas between systems, validate incoming data against the expected structure, generate data ingestion pipelines, or drive schema-driven code generation.

Example (conceptual): Table users with columns id (integer, not null), email (varchar(255), not null), created_at (timestamp, nullable). Primary key: id.

See also: Database schema, SQL, JSON Schema, data interchange formats.