Types and description of GRAVITY operators

Think of the GRAVITY application as a collection of operators that perform operations on data passed between them through the data bus. Each operator in GRAVITY is specialized in performing a specific type of action—ranging from retrieving data, transforming it, to storing or sending it.

With operators, you can:

retrieve data from various sources (databases, files, web services),
process and transform data in different ways (filtering, enriching, joining, cleaning, branching, etc.),
save processing results in many formats and locations (databases, files, emails, queuing systems, etc.),
temporarily store data in memory for reuse in other parts of a project or across projects.

You can use any number of operators in a single project. The data stream carried by the bus can contain any number of records and any number of bits (i.e., columns)—there are no limits on its width or volume.

For clarity and design convenience, operators in GRAVITY are divided into three main groups:

Input operators

They are used to retrieve data from external sources—such as relational databases, files (e.g., CSV, Excel), web services (REST/SOAP), FTP, AS2, or messages from queuing systems.

Processing operators

These operators transform data as it flows through the bus—most often acting on each input record. In this group, you’ll find operators for filtering, grouping, calculating, mapping, splitting paths, as well as integrating projects and applications.

Output operators

They handle the final storage of data or its delivery outside the GRAVITY environment—to files, databases, queueing systems (RabbitMQ, TCP), web services, or emails.

Every path (a line of data flow in a project) in GRAVITY must end with an output operator. This ensures that the system recognizes the project segment as fully designed and that the data from this path has been properly processed.

In the following sections you will find detailed descriptions of individual operators, their configuration, and typical use cases.

You can invoke a project and test (debug) operators using Manual Invocation. When the project is ready to run and you expect automatic executions, use Triggers.