The Ticketmaster handler (inc/Steps/EventImport/Handlers/Ticketmaster/Ticketmaster.php, TicketmasterSettings) plugs into EventImportStep through HandlerRegistrationTrait so it can be selected inside a Data Machine pipeline. Each execution paginates through Ticketmaster results up to MAX_PAGE = 19, rejects previously imported events before packet fan-out, and returns the remaining events as individual DataPacket objects.

Configuration & Authentication

• Auth: Ticketmaster uses an auth provider (TicketmasterAuth) that supplies an api_key.
• Handler settings (from TicketmasterSettings): classification_type (required), location (lat,lng string), radius (miles), optional genre, optional venue_id, optional search, optional exclude_keywords, and optional max_items.

Data Mapping

• The handler maps Ticketmaster API responses into a standardized event payload (title, dates/times, venue name, ticket URL, etc.), plus a separate venue_metadata array.
• The handler stores venue context into EventEngineData::storeVenueContext() and merges price into engine data (when available).
• The handler returns a DataPacket whose body contains JSON: { event, venue_metadata, import_source: "ticketmaster" }.

Unique Capabilities

Ticketmaster uses its stable upstream event ID plus a revision hash of the mapped event fields. A shared atomic source claim prevents neighboring city flows from selecting the same ID concurrently. Successful child jobs persist the revision in Data Machine’s tracked-item ledger; failed jobs release the claim without persisting it. Unchanged revisions are skipped before fan-out, while changed dates, times, venues, titles, prices, and ticket URLs continue to EventUpsert. The generic datamachine_should_reprocess_item policy can also make an unchanged revision eligible again.

Pagination

The handler paginates Ticketmaster API results up to MAX_PAGE = 19. Fan-out defaults to 100 child jobs per run; flows can set max_items explicitly when a lower rollout bound is appropriate. Items beyond the cap remain unclaimed so later runs can process them.

Each run logs an Import fan-out summary with fetched, unchanged, contended, overflow, source-claimed, and packets-ready counts. source_claimed and packets_ready are measured after the handler’s bound and atomic claims; Data Machine’s pipeline-batch log remains authoritative for child jobs actually scheduled. When adding cities, start with max_items between 25 and 50, compare pre_fanout_deduped with packets_ready, and confirm parent completion time and worker queue health before increasing toward the 100-item default.

Event Flow

1. EventImportStep instantiates Ticketmaster and reads TicketmasterSettings values.
2. Handler dedupes stable Ticketmaster IDs and existing event identities before returning bounded DataPacket candidates.
3. EventEngineData carries the structured payload into the pipeline.
4. EventUpsert receives the data, merges engine parameters, runs field-by-field change detection, assigns venue/promoter via TaxonomyHandler, syncs the datamachine_event_dates table, and optionally downloads featured images.

Every EventUpsert run uses the same identifier hash so duplicates do not slip through.