MySQL Data Generator

This is a tool to easily fill a SQL database. It is able to analyse a schema and generate a settings/schema.jsonc which will be used to generate accurate data. It does its best to handle foreign keys. You can provide a settings/schema_custom.jsonc to customize settings/schema.jsonc during the generation phase. This will allow you to override datatype for a column, force the use of a foreign key or specify a list of values.

functionalities

• analyse a table and generate a schema
• allow for customization on data types, foreign keys, values, uniqueness etc.
• handle foreign keys
• define a number of rows to generate per table
• specify a seed to always generate the same dataset
• disable/enable triggers during process

1. Analysis

The first step is to analyse your database to generate a settings/schema.jsonc by providing database credentials:

The schema parameter allows you to specify a name for the output files and differentiate between multiple schemas and setups.

npx @corteks/mysql-data-generator --db mysql://user:[email protected]:3306/database --analyse --schema schema

The --schema parameter allows you to generate mutliple configuration with different names.

If you want to customize the schema, modify the default settings/schema_custom.jsonc that has also be generated.

2. Data generation

Next step is to fill the database with randomly generated values:

mysqldatagen --db mysql://user:[email protected]:3306/database

If any .sql scripts are provided within the settings/scripts folder, they will be played before generation. Those scripts can contains DELIMITER caommands as they will be run directly by MySQL client executable. As they will be run every time the generation is launched you have to take care of the cleanup.

For every tables listed in settings/schema.jsonc, the tool will:

• get the values of foreign keys if needed
• generate batches of 1000 rows
• insert rows until it reaches the defined table limit
• columns in table are ordered accordingly to your custom schema so you can rely on other column value in the same row.

Available options in schema_custom.json:

• settings: Global settings
  • disableTriggers: boolean // disable triggers per table during process and recreate them afterward
  • engine: "MariaDB" // only MariaDB is supported for the time being but it should also be compatible with MySQL.
  • ignoredTables: string[] // list of table name that should not be analysed nor filled
  • options: Array<[key: string]: any[]> // an array of column options to configure specific generators for the whole file generator is an array of string to allow multiple settings at once
  • maxLengthValue: number? // Hard limit of the maximum number of characters in string column type. This will override your custom column max value if it's bigger than maxLengthValue.
  • seed: number // The seed used by the random generator. This is optional. filling process.
  • tablesToFill: string[] // list of table name that should be analysed and filled. You can set this parameter or ignoredTables depending on the number of table to work with
  • values: [key: string]: any[] // an object of user defined array of values
• tables: Table[] // list of tables handled by the tool
  • Table.name: string // table name
  • Table.lines: number // Deprecated in favor of maxLines
  • Table.maxLines: number // Maximum number of rows this table should contains
  • Table.addLines: number // Number of rows to be inserted on a single run. The number of lines resulting in the table will not exceed Table.maxLines
  • Table.columns: Column[] // list of columns handled by the tool
    • Column.name: string // column name
    • Column.generator: bit | boolean | date | foreignKey | integer | real | time | string | values | function | faker // data type generator used for this column
    • Column.[key: string]: any[] // list of options for this column
    • Column.foreignKey: { table: string, column: string, where: string } // link to the table.column referenced by this foreign key. A custom clause can ba added to filter value from the foreign column
    • Column.values: string | any[] | { [key: string]: number } // Name of the list of values to use for this column. // You can also directly specify an array of strings for values. // Or you can use an object to specify a ratio per value. Ratio will be a number between 0 and 1.
    • Column.customFunction: (rowIndex: number, row: { [key: string]: string | number } // a string representing a javascript custom function. It will receive the row index and the full row as arguments.
    • Column.template: string // a template string for faker generator. See fakerjs for more information.
    • Column.locale: string // locale used by the faker generator.