Reflect the columns of an existing table into Column objects

Inspects an existing database table and returns a named list of [Column] objects, one per column. Used by `Engine$reflect()` to build a TableModel from a table that already exists in the database.

Usage

reflect_columns.mssql(x, tablename, ...)

reflect_columns.postgres(x, tablename, ...)

reflect_columns(x, tablename, ...)

reflect_columns.default(x, tablename, ...)

Arguments

x

An Engine (or other oRm object) used for dialect dispatch.

tablename

Character. Name of the table to reflect, already qualified.

...

Additional arguments for dialect-specific implementations.

Value

A named list of [Column] objects, keyed by column name.

Details

The default implementation is dialect-agnostic: it issues a zero-row query and reads `DBI::dbColumnInfo()` to recover column names and best-effort types. The reported type is backend-dependent (e.g. RSQLite reports R types such as "double"); since reflection targets existing tables, the type is not used to create the table and so is informational only. Dialects may provide richer implementations (capturing primary keys, nullability, and defaults).

Functions

• reflect_columns.mssql(): SQL Server reflection via the `sys` catalog views, capturing declared types (including IDENTITY), primary keys (composite keys included), nullability, defaults, and foreign keys (returned as [ForeignKey] objects, schema-qualified when the target lives in another schema).

• reflect_columns.postgres(): PostgreSQL reflection via `pg_catalog`, capturing canonical types, primary keys, nullability, defaults, and foreign keys (returned as [ForeignKey] objects, schema-qualified when the target lives in another schema).