CrawlableDataset and Catalog GenerationThe CrawlableDataset interface was designed to be an abstract
representation of a hierarchical collection of datasets. The goal was
to allow various data storage systems to be mapped into the
CrawlableDataset model and from there used to generate THREDDS catalogs
and provide access to datasets so they can be served by the TDS.
public interface CrawlableDataset {
public String getPath();
public String getName();
public boolean exists();
public CrawlableDataset getParentDataset();
public boolean isCollection();
public CrawlableDataset getDescendant( String relativePath);
public List listDatasets() throws IOException;
public List listDatasets( CrawlableDatasetFilter filter ) throws IOException;
public long length();
public Date lastModified();
}
It is a very simple interface similar to (and based upon)
java.io.File. The reference implementation, CrawlableDatasetFile, is
built on top of java.io.File. Two other implementations exist:
The CatalogBuilder interface provides methods for generating THREDDS
catalogs. The main implementation generates catalogs based on the
datasetScan configuration
The goal for the CrawlableDataset and CatalogBuilder family of
interfaces (see diagram below) is to allow users to change how catalogs
are built, beyond what is currently available in the datasetScan
configuration files.
CrawlableDataset and CatalogBuilder Architecture
This document is maintained by Unidata and was
last updated July 2007. Send comments to THREDDS
support.