Introduced with XNAT 1.6.2, the refresh catalog service is used to update catalog xmls that are out of sync with the file system. This service can be used to store checksums for entries that are missing the, remove entries that no longer have valid files, or add new entries for files that have been manually added to the archive directory.
When using this feature to add files that have been manually added to the archive directory, you must have placed the files in the appropriate archive directory (in the same directory as the generated catalog xml or a sub-directory). The catalog xml should already exist before triggering this service. If you haven't generated the catalog yet, you can do so by doing a PUT to the resource URL (i.e. /data/archive/experiments/ID/resources/TEST).
You can reference an object that contains resources instead of referencing resources themselves. So, if you want to update all of the resources for an entire MR session, you could reference the MR URL as your input, /services/refresh/catalog?resource=/archive/experiments/EXPT.
POST - /data/services/refresh/catalog
Required querystring parameter.
- /archive/projects/PROJECT/subjects/SUBJECT -- will not cascade to children (experiments)
- /archive/projects/PROJECT-- will not cascade to children (subjects or experiments)
- /archive/subjects/SUBJECT -- will not cascade to children (experiments)
You can reference multiple resources by including multiple resource parameters. However, the refresh operation interacts with the file system and can be an expensive transaction, so you are encouraged to limit the number of resources you are updating at one time.
- checksum: generate checksums for any entries that are missing them
- delete: remove entries that do not reference valid files
- append. add entries for files in the catalog directory (or sub-directory)
- populateStats: updates the statistics for the resource in the XNAT abstract resource table.
See: Catalog Refresh API