Convert Hana artifacts to CDS

Command line compiling CDS/CSN models from Hana artifacts

Introduction

SAP CAP framework (https://cap.cloud.sap/) - is a framework by SAP which allows to design and build database artifacts using CDS language and also builds native database files, like for example .hdbtable/.hdbview for SAP Hana database. This is a target scenario for a new application. However very often we face another problem - when we need to build a model on top of existing database schema. In this case the table already exists and to be able to expose it as a service we need to have CDS representation

The idea behind this CLI is very simple - we connect to Hana database using native CAP service binding (https://cap.cloud.sap/docs/advanced/hybrid-testing) and then we generate CSN model as a result (https://cap.cloud.sap/docs/cds/csn)

This model may be later converted to CDS files by use of cds compile --to cdl command

Installation

npm install -D hana2cds

Usage

Usage: hana2cds [options]

Generates CDS model from Hana table/view defintion

Options:
  --service <service>               HDI service name
  -s, --schema <schema>             Database schema
  -n, --namespace <namespace>       CDS namespace
  -f, --filter <filter>             Comma-separated list of tables/views
  -o, --output <output>             Name of output file (STDOUT by default)
  -c, --case <properties case>      Convert properties to a specific case
  -p, --prefix <projection prefix>  Prefix for projection names
  -h, --help                        display help for command

Some examples: Generate a model with all tables and views from FLIGHTS schema and writing file in Linux way

hana2cds -s FLIGHTS > flights.csn

Generate a model with specific tables and writing to a specific file (OS independent)

hana2cds -s FLIGHTS --f AIRLINES,FLIGHTS -o flights.csn.json

Default behavior

By default CLI will try to take schema from cds.env.requires.db

So command like hana2cds will be equivalent to hana2cds -s <schema-from-cds.env.requires.db>

This object always contains a first hana database from VCAP_SERVICES. If you have several hana services in your VCAP_SERVICES you can pick up a specific one by using --service parameter:

hana2cds --service=flights-hdi > flights.csn

Storing secrets

There are multiple ways how to store secrets for CDS. I will use the way I test apps using Hana Cloud:

1. login to Hana Cloud instance using cf login or via site
2. Create service key 'hana-hdi-key' for a service instance 'hana-hdi'. Names can be different of course
3. cds bind --to hana-hdi:hana-hdi-key. If successful this step will create you a file .cdsrc-private.json with your service bidning but without credentials. Beware that by default cds bind generates hybrid profile and you may need to set up CDS_ENV=hybrid variable in your environment
4. you may follow documented way (https://cap.cloud.sap/docs/advanced/hybrid-testing#node) and consume credentials with this command: cds env get requires.db.credentials --resolve-bindings.
5. I didn't like the idea to use environment variables or manual copying of that command content and decided to create advanced scenario using file-based secret storage

cds env get requires.db.credentials --resolve-bindings | json2fs --to .cds/requires/db/credentials

This additional utility also developed in this project will parse the file and will store secrets in key-based files with the provided path. In can be local path ( make sure you ignore it from git ) or it can be your home directory for example. In this case CDS will be able to read those secrets automatically if you define this PATH in CDS_CONFIG=/workspace/.cds. Keep in mind that currenlty CDS_CONFIG only supports absolute paths, and even tilda is not parsed properly. Also notice that the folder should have a structure requires->service_name->credentials and CDS config is ponting to the root folder for requires.