SQL from data

Turn JSON, YAML or XML into CREATE TABLE and INSERT scripts for PostgreSQL, SQL Server, MySQL or SQLite.

Load test data into a database without writing a script by hand. Quap reads the records, proposes tables and column types, and writes the SQL.

Steps

Open the SQL target

Click Conv in the preview toolbar and choose → SQL.

Choose mode and dialect

Mode: Insert — one table per record array, with CREATE TABLE and INSERT statements.
Mode: Normalize — splits nested objects and arrays into their own tables linked by foreign keys.
Dialect — PostgreSQL, T-SQL (SQL Server), MySQL or SQLite.

For PostgreSQL, Identifier mode either keeps the original casing (quoted identifiers) or normalises everything to lowercase.

Check the mapping

Field mapping lists every field with its Source, Table, Column and Type:

untick Include to leave a field out,
tick Unique to generate distinct rows by that column,
tick Lookup to move repeating values into a separate lookup table linked by a foreign key.

Use List path to choose which array becomes the rows, set the Root table name, filter long mappings with Search tables or fields…, and Re-infer to start over.

Convert

Click Convert, then Copy the script or Apply to document.

Translate SQL between dialects

To translate an existing SQL script — for example from MySQL to PostgreSQL, or to C# LINQ — open it and use the SQL preview.

With the AI assistant

The suggestion bubbles CREATE TABLE · PostgreSQL and CREATE + INSERT · PostgreSQL (and the other dialects) write the script in one step, and you can say things like "generate a MySQL schema for the orders only".