cdsCompile.md

Core Data Service (CDS) Build Tool

What is CDS?

The concept behind Core Data Services is to offer a three-layer approach to user application development. These three layers build upon each other and provide a means for defining a:

1. data model in a database agnostic manner
2. service that exposes your data model to the outside world in the form of an OData service
3. basic end-user application that consumes the generated OData service

As a consequence of these three layers, within an SAP Cloud Platform Business Application Project, the directory naming convention described below is expected.

You are free to use as many of these three layers as are relevant for your requirements - but always following the strict order of Data Model -> Service -> Application,

Directory Name Conventions

Directory Name	Description
db	Contains the CDS definitions of all the entities in your data model. This directory can also contain any .csv files needed to load the generated tables with data once your database has been deployed to HANA
srv	Contains the service definition that exposes the contents of the HANA database as an OData service
app	Contains a user interface definition to consume the OData service as an end-user application

CDS Compiler Conventions

In Web IDE, when you select "Build CDS" from the project context menu, or "Build" from the db folder's context menu, the CDS compiler will be invoked. Although it is possible to configure the behaviour of this compiler, in the absence of any explicit configuration, it will follow certain conventions concerning directory and file names.

In keeping with the strict order described above, the CDS compiler first examines the contents of the db folder, then the srv folder, then finally the app folder.

The index.cds file

The CDS compiler's default behaviour is to look in each of the above directories for an index.cds file. If this file is found, then the compiler behaves in the following manner:

1. Only the index.cds file will be opened explicitly. Since this is a normal CDS file, it can contain any CDS statements you like - which includes references to other .cds files
2. If your data model is described by multiple .cds files, then these files must be referenced within index.cds, otherwise they will not be processed
3. Any .cds file that is not explicitly referenced by index.cds will be ignored.

Q: What happens to the CDS compiler if I don't create my own index.cds file?
A: Nothing bad... :-)

No error will occur if index.cds does not exist. In this case, the CDS compiler will simply open and parse all files with an extension of .cds.

The advantage of using an index.cds file is that should you wish to, you can exclude certain .cds files from your data model at compile time. This is a useful feature if you need to switch on or off certain sections of your data model.