SquashQL Codegen

This tool is inspired by Kysely Codegen

squashql-codegen generates type definitions that represent tables and fields from your database in ts file named tables.ts. It fetches from a database the list of available tables and fields and creates a typescript file with this information. The types can be used with the TypeScript library to build SQL-like queries compatible with SquashQL.

NPM

https://www.npmjs.com/package/@squashql/squashql-codegen

Support

Only the following databases are supported for the time being:

• BigQuery
• Snowflake

Example:

import {TableField} from "@squashql/squashql-js"

export interface SquashQLTable {
    _fields: TableField[]
    _name: string
}

class Product implements SquashQLTable {
    readonly _name = "product"
    readonly productId: TableField = new TableField("product.product_id")
    readonly price: TableField = new TableField("product.price")
    readonly brand: TableField = new TableField("product.brand")
    readonly categoryId: TableField = new TableField("product.category_id")
    readonly _fields: TableField[] = [this.productId, this.price, this.brand, this.categoryId]
}

class Category implements SquashQLTable {
    readonly _name = "category"
    readonly categoryId: TableField = new TableField("category.category_id")
    readonly status: TableField = new TableField("category.status")
    readonly _fields: TableField[] = [this.categoryId, this.status]
}

const product = new Product()
const category = new Category()

export {
    product, category
}

And then you can use it for your calculations that need to use the Field interface or the constant . See the documentation.

import {product, category} from "./table"

const price: TableField = product.price
const status: TableField = category.status

const priceName: string = price.name // name of the field is accessible
// ...

Installation & Execution

To install the CLI globally

npm install -g @squashql/squashql-codegen

To execute the CLI without installing anything

npx @squashql/squashql-codegen

Configuration

Create an .env to define a couple of environment variables needed to connect to the database and correctly execute the script. It is also possible to pass the values to the CLI

In each case, a client needs to be created to connect to the DB by setting the env. variable SQUASHQL_CLIENT.

To choose where the file should be written, the following env. variable can be set SQUASHQL_PATH.

# .env file
SQUASHQL_CLIENT="bigquery" # mandatory
SQUASHQL_PATH="/Users/john/tmp" # optional, table.ts will be written in /Users/john/tmp directory

BigQuery

The CLI uses Application Default Credentials to authenticate.

# .env file
SQUASHQL_CLIENT="bigquery" # mandatory
SQUASHQL_BIGQUERY_DATASET_ID="nameofthedataset" # mandatory
GOOGLE_APPLICATION_CREDENTIALS="/path/to/key.json" # mandatory if this authentication method is used https://cloud.google.com/docs/authentication/application-default-credentials#GAC

Snowflake

# .env file
SQUASHQL_CLIENT="snowflake"

# Snowflake
SQUASHQL_SNOWFLAKE_ACCOUNT="abc123.north-europe.azure" # mandatory
SQUASHQL_SNOWFLAKE_USERNAME="john" # mandatory
SQUASHQL_SNOWFLAKE_PASSWORD="mypassword" # mandatory
SQUASHQL_SNOWFLAKE_DATABASE="DATABASE_NAME" # mandatory
SQUASHQL_SNOWFLAKE_SCHEMA="SCHEMA_NAME" # mandatory