CSV import model - *.csvim + *.csv

The *.csvim file is the import model: it lists CSV files to load, the target table for each, column mappings, and transformations. The matching *.csv files hold the data. Synchronizer: CsvimSynchronizer (components/data/data-csvim).

File format

{
    "files": [
        {
            "name": "COUNTRY",
            "table": "COUNTRY",
            "schema": "PUBLIC",
            "file": "data/COUNTRY.csv",
            "header": true,
            "delimField": ",",
            "delimEnclosing": "\"",
            "distinguishEmptyFromNull": false,
            "keys": [
                { "column": "ID", "values": [] }
            ],
            "version": "1.0.0"
        }
    ]
}

Fields

Field	Notes
name	Logical entry name.
table	Target table.
schema	Target schema (optional, dialect-dependent).
file	Path to the CSV, relative to the project root.
header	true if the first CSV row is the column list.
delimField	Field delimiter.
delimEnclosing	Field-quote character.
distinguishEmptyFromNull	true -> empty string stays empty; false -> empty -> NULL.
keys	Composite primary key spec for idempotent re-imports.
version	Caller-managed version label.

Reconciliation

On CREATE / UPDATE the synchronizer loads each referenced CSV into the target table. Re-imports use keys to decide insert vs update. DELETE removes the imported rows when the artefact is deleted.

Editor

CSVIM has a dedicated editor in the IDE - see /help/ide/editors/csv-and-csvim.

Tenant isolation

CSV imports land in the tenant-isolated data source. Multi-tenant deployments get one materialised copy of the data per tenant.

See also

• Table
• Schema
• Data transfer