Skip to main content

Customize Exports with Transformers

Endatix exports support a powerful value transformation pipeline. You can register custom transformers to modify data before it is written to the export file (JSON, CSV, or Excel).

Common use cases include:

  • Rewriting URLs (e.g., converting internal storage paths to public CDN links).
  • Flattening Data (e.g., extracting a single field from a complex JSON object).
  • Redacting Information (e.g., removing sensitive keys from a JSON response).

How it works​

During an export, Endatix "lifts" every value (whether it comes from a database string, a JSON column, or a C# property) into a mutable JsonNode.

This JsonNode is passed through a chain of registered transformers. Each transformer can:

  1. Inspect the node.
  2. Modify it in place (e.g., add/remove properties).
  3. Replace it entirely (e.g., replace an Object with a String).
  4. Return it for the next transformer.

This architecture ensures that you don't have to manually parse JSON strings or worry about "double-escaping" bugs. You work with a structured Document Object Model (DOM) at all times.

The IValueTransformer interface​

Transformers implement the IValueTransformer interface. They receive a nullable JsonNode and must return a JsonNode (or null if the value should be removed).

public interface IValueTransformer
{
    JsonNode? Transform<T>(JsonNode? node, TransformationContext<T> context);
}

The TransformationContext<T> provides access to:

  • Row: The entire data row (e.g., SubmissionExportRow).
  • JsonDoc: The raw JSON document of the submission answers.
  • Logger: An ILogger for diagnostics.

Registering a transformer​

You register transformers via the Dependency Injection (DI) container during startup.

builder.Host.ConfigureEndatixWithDefaults(endatix =>
{
    // Transformers run in the order they are registered.
    endatix.Services.AddExportTransformer<FileContentExtractorTransformer>();
});

Note on Ordering: Transformers form a pipeline. For example, if you have a transformer that rewrites URLs and another that extracts them, order matters. Usually, you want to rewrite the URL inside the object first, and then extract the result.

Example: Extract only file URLs​

File upload answers in Endatix are stored as rich JSON objects containing metadata (name, type, size):

{
  "name": "receipt.jpg",
  "type": "image/jpeg",
  "content": "[https://storage.example.com/receipt.jpg](https://storage.example.com/receipt.jpg)"
}

For a CSV export, you might prefer a simple URL string instead of the full JSON object.

The Transformation​

  • Input (Single File): A JSON Object => Result: A JSON String ("https://...").
  • Input (Multiple Files): A JSON Array of Objects => Result: A JSON Array of Strings (["https://...", "https://..."]).

Implementation​

This transformer inspects the node. If it finds a content property, it replaces the complex object with just the value of that property. A complete working sample is available in the Endatix repository:

It includes:

  • The FileContentExtractorTransformer implementation.
  • Startup logic showing how to register it properly.