All Classes and Interfaces
Class
Description
Bean wiring for the batch-indexing processing pipeline.
Invalidates the tag and child-association caches in the Alfresco metadata repository before every batch indexing job run.
Observer
ItemWriter that scans events for cascade signals and stashes enriched sources for the post-cycle sweep.Bean wiring for the batch-indexing cascade support.
Configuration properties for the batch-indexing cascade support.
Snapshot of cascade sources drained from
CascadeSourceStash at the end of a reindex cycle.Thread-safe, cycle-scoped bag of cascade sources observed during a batch-indexing cycle.
Post-cycle orchestrator that drains the
CascadeSourceStash once a reindexByDate cycle finishes successfully and dispatches each observed cascade source to the appropriate cascade service.Cascades a category delete by pruning the deleted UUID from every referring document's
CategoryDeleteCascadeService.CATEGORIES_FIELD (cm%3Acategories), CATEGORY_ANCESTOR and PARENT arrays, and stripping PATH / UNPREFIXED_PATH entries that contain any of the deleted category's paths, via a single _update_by_query.Descriptor for a delete event that corresponds to a
cm:category node.Configuration for Transform Core AIO direct integration.
Structured, non-retryable content-transform failure surfaced by
ContentTransformClient so the writer can hand it to the DLQ bridge without re-inspecting the underlying HTTP exception.Hook invoked by
ContentTransformDirectWriter when a per-node content transform fails in a non-retryable way (currently 4xx from ACS).Outcome of a single
ContentTransformClient.transformContent(Long) call.Configuration for continuous reindexing mode — always active in the batch-indexing module.
Configuration properties for continuous reindexing mode (live indexing via polling).
Continuously polls the Alfresco DB for new/changed nodes and launches
reindexByDate Spring Batch jobs one at a time.Wires the dead-letter queue components.
One record per give-up event the indexer has surfaced for operator visibility.
Stage of the pipeline that produced the skip event.
Configuration properties for the dead-letter queue that records poison nodes the indexer has given up on.
Bridges per-item bulk failures from the metadata writer into
DlqWriter.Bridges non-retryable content-transform failures (currently ACS 4xx) into
DlqWriter.Bridges page-level read failures (a bad row in a JDBC page fails the whole page — see reindexing
AbstractAlfrescoNodeReader) into DlqWriter as a READER record keyed by the failed time window.Bridges per-item bulk failures from the path writer into
DlqWriter.Bridges Spring Batch skip events into
DlqWriter.Writes dead-letter records to a dedicated Elasticsearch index.
Persists the reindexing watermark as a typed
WatermarkDocument in a dedicated Elasticsearch index.Attaches the
CacheInvalidationListener to the reindexByDateJob bean (defined in the reindexing module) without overriding it.Cascades a rename to descendants by swapping the old path segment for the new one in each descendant's
PATH / UNPREFIXED_PATH arrays.Descriptor for a rename cascade: the old and new terminal path segments to swap in descendants'
PATH / UNPREFIXED_PATH arrays.Batched mget helper that snapshots cascade-relevant fields (
PATH, UNPREFIXED_PATH, TYPE) from source docs before the downstream writers overwrite them.Snapshot of a source doc's cascade-relevant fields.
Checked exception indicating the watermark store cannot be reached.
One-shot startup guard and cursor seed for the batch-indexing watermark.
Unchecked exception thrown when a watermark write fails with HTTP 409 (version conflict).
Typed watermark document persisted in Elasticsearch.