---
title: "Pipeline step processors"
slug: "pipeline-step-processors"
abstract: "Introduction to the base classes you can use when creating pipeline step processors in Sitecore Data Exchange Framework."
---
# Pipeline step processors

Writing custom pipeline step processors is a significant part of building custom providers for Data Exchange Framework. This topic describes elements you can use to simplify the task of creating custom pipeline step processors.

<details>
<summary><h2>BasePipelineStepProcessor class</h2></summary>

The step processors implement the `BasePipelineStepProcessor` class. The `BasePipelineStepProcessor`, and the processors that implement it, has the following protected properties:

- PipelineStep
- PipelineContext
- Logger

Most of the methods allow their corresponding parameters to have null values.

You can use the properties as method parameters or protected properties. Best practice recommendations are:

- If you create new pipeline step processors directly based on the `BasePipelineStepProcessor`, we recommend you use `PipelineStep`, `PipelineContext`, and `Logger` as properties.
- If you override already existing pipeline step processors, where `PipelineStep`, `PipelineContext`, or `Logger` are passed as arguments in methods, we recommend you pass the properties as arguments.
- In the processing stage, each step has its own `PipelineStepContext`, while `PipelineStep` objects can be shared between threads. Updating `PipelineStep`, including its plugins, is therefore not considered thread-safe. You can, instead, safely store changing data in its context.

</details>

<details>
<summary><h2>BaseReadDataStep processor</h2></summary>

This class provides methods to simplify writing pipeline step processors that read data from an endpoint.

| Type | `Sitecore.DataExchange.Processors.PipelineSteps.BaseReadDataStepProcessor,Sitecore.DataExchange` |
| --- | --- |

### ReadData(Endpoint, PipelineStep, PipelineContext, ILogger)

This method is responsible for reading data from the specified endpoint. What the method does with the data is up to the developer. Usually the data is added to a plugin that is added to the `PipelineContext`.

</details>