cds-pg - PostgreSQL adapter for SAP CDS (CAP)

First a big thank you to our contributors:

This node module provides an adapter to the PostgreSQL database.

For a short introduction on the background of this project you can check out a short video that has been captured as part of the SAP devtoberfest.

Current status

cds-pg is ready to be used! Still, there's some gaps left to fill - note the list below and please see CONTRIBUTING.md for how to contribute additional capabilities!

Also checkout the following blog posts on how to get started using cds-pg in your local development environment and on SAP Business Technology Platform (BTP), Cloud Foundry:

• Setup an example project using cds-pg
• Prepare and deploy the project to SAP BTP CF

TODO

• implement basic SELECT|READ(~ OData GET)
• implement basic INSERT|CREATE(~ OData POST)
• implement basic UPDATE(~ OData PUT|PATCH)
• implement basic DELETE(~ OData DELETE)
• map OData to PostgreSQL vocabulary
• implement basic cds deployment
• use default query builders for UPDATE/DELETE
• add support for multitenancy (see issue #25)
  • support multiple schemas (at config time)
  • support queries with tenant info (at runtime)
• add support for full OData vocabulary
  • system query options ($filter, $expand, $select)
  • string functions
  • collection functions
  • date + time functions
  • geo functions
  • arithmetic operators + functions
• add draft support (see issue #30)
• add advanced deployment model that supports delta handling/migrations (see issue #27)
• support virtual fields (see issue 204)
• support localized entities (see issue 196)

Usage in your CAP project

Add this package to your SAP Cloud Application Programming Model project by running:

npm install cds-pg

Then add this configuration to the cds section of your `package.json:

  "cds": {
    "requires": {
      "db": {
        "kind": "postgres"
      },
      "postgres": {
        "dialect": "plain", // <- for cds >= 5.1
        "impl": "cds-pg",
        "model": [
          "srv"
        ]
      }
    }
  }

For local development you can provide the credentials in the file default-env.json in the root folder of your project:

{
  "VCAP_SERVICES": {
    "postgres": [
      {
        "name": "postgres",
        "label": "postgres",
        "tags": ["plain", "database"],
        "credentials": {
          "host": "localhost",
          "port": "5432",
          "database": "dbname",
          "user": "postgres",
          "password": "postgres",
          "schema":"public"
        }
      }
    ]
  }
}

CDS deployment

cds-pg contains the database adapter to translate the incoming requests to PostgreSQL during runtime, but also includes a quick and dirty command to deploy the current data model to the PostgreSQL database specified in default-env.json. Initial data will also be filled from the provided .csv files following the approach described in Providing Initial Data. Be aware that the existing tables and views are deleted and then re-created according the CDS model, so this should not be used in production environments:

npx cds-pg deploy srv --to db

For a more sophisticated approach, please check out cds-dbm at https://github.com/mikezaschka/cds-dbm, which offers an advanced deployment model including delta handling of data and models! Please also read the following blogposts for a detailed description, on how cds-dbm can be used in combintation with cds-pg:

• Setup an example project using cds-pg
• Prepare and deploy the project to SAP BTP CF

Projects using cds-pg

• pg-beershop – An example project with multiple deployment scenarios (Cloud Foundry, Kyma, Azure, Google Cloud Platform)
• cap-devtoberfest – A simple project showcasing how to run cds-pg and cds-dbm locally and on SAP BTB Cloud Foundry