Skip to main content
Version: v2026.0.0

Connectors

A connector is what bridges ARender with an external document repository. It is responsible for retrieving document content from an external system and making it available to the viewer.

Connector model​

Connectors are Java JARs bundled directly in the viewer's classpath.

A connector:

  1. Implements DocumentAccessor to provide document content, MIME type, and metadata to the viewer.
  2. Exposes a DocumentServiceURLParser that parses the incoming request (via the canParse / parse contract) to determine whether this connector handles the given URL.
  3. Connects to the external system using its native API (CMIS, FileNet P8 CE, HTTP, etc.) and returns a DocumentAccessor.
  4. Optionally provides an AnnotationAccessor so that annotations are stored back into the same external system.

Connectors are loaded as Spring Boot auto-configured beans. Adding a connector to a deployment means placing its JAR on the viewer's classpath (typically /home/arender/lib/).

Connectors are packaged as fat JARs using the -jar-with-dependencies classifier, which bundles the connector code together with its third-party dependencies.

Annotation connectors​

Annotation storage follows the same connector model. Each AnnotationAccessor implementation connects to a storage backend — a SQL database (JDBC), an HTTP endpoint (REST), the local filesystem (XFDF), or a repository-native store (FileNet, CMIS). The viewer picks the annotation connector that matches its configuration.

See Annotations for the annotation model.

Connectors vs. UI plugins​

UI plugins (Alfresco Share plugin, IBM Content Navigator plugin) are not connectors. They are modules installed in the ECM's own interface that open documents in ARender by generating the correct viewer URL. The actual document retrieval still goes through a repository connector (e.g., CMIS, FileNet). See the integration catalog for the full picture.

Last updated on