Search Results generate_entities_work_units

Overview

HZ_IMP_LOAD_WRAPPER is the master orchestration package for the Oracle EBS Trading Community Architecture (TCA) bulk import and batch data load process. Owned by APPS and classified as OTHER, it serves as the entry point through which customer, contact, address, and classification records staged in the HZ_IMP_* interface tables are validated, deduplicated, and ultimately written into the production TCA registry tables. The package encapsulates the full lifecycle of a batch run — preprocessing, entity and work-unit generation, parallel worker dispatch, and cleanup — so that callers (typically concurrent programs and the batch loader forms) need only invoke a single wrapper procedure rather than coordinate the multiple lower-level APIs individually.

The package header defines several PL/SQL collection types and the DML_RECORD_TYPE record, which carries the runtime parameters of a batch execution: batch identifier, originating system, "what-if" analysis flag, registry deduplication settings, batch mode flag, worker count, error limit, rerun flag, and the standard concurrent request context (user, login, program, application, and request identifiers). Two further fields govern flexible-value validation and data-security enforcement, and one permits disabled lookup values during processing.

Key Procedures and Functions

• DATA_LOAD — The main wrapper for running a data load. It accepts the batch-level parameters described in DML_RECORD_TYPE (batch ID, originating system, what-if analysis, registry dedup options, sysdate, batch mode, number of workers, error limit, rerun flag, and concurrent request context) and drives the complete load for a single batch.
• BATCH_DATA_LOAD — The wrapper for running a batch data load; it delegates to DATA_LOAD. This is the procedure most commonly associated with the "batch_data_load" search term and the standard concurrent-program invocation.
• ONLINE_DATA_LOAD — Handles the online (interactive) variant of the load, used when records are submitted individually rather than as a pre-staged batch.
• RETRIEVE_WORK_UNIT and RETRIEVE_PP_WORK_UNIT — Return individual units of work to parallel worker processes, including the post-processing (PP) variant used to service the worker pool.
• GENERATE_ENTITIES_WORK_UNITS — Divides the staged batch into discrete work units so that multiple workers can process entities concurrently.
• DATA_LOAD_PREPROCESSING — Runs setup and validation steps before entity processing begins.
• CLEANUP_STAGING — Removes or resets staging rows after a load completes or fails.
• STAGING_DATA_EXISTS — A check used to determine whether interface staging data is present for a given batch.

Tables Accessed

The package reads and writes the HZ_IMP_* interface tables, including HZ_IMP_BATCH_SUMMARY and HZ_IMP_BATCH_DETAILS for batch control and status, and the paired _INT/_SG staging tables for addresses, address uses, contact points, classifications, and contact roles (for example HZ_IMP_ADDRESSES_INT, HZ_IMP_ADDRESSES_SG, HZ_IMP_CONTACTPTS_INT, HZ_IMP_CLASSIFICS_INT, and HZ_IMP_CONTACTROLES_INT). It references HZ_CLASS_CATEGORIES and HZ_CONTACT_POINTS for validation and lookup during entity construction. Concurrent-program context and reprocessing messages are drawn from FND_CONCURRENT_REQUESTS and FND_NEW_MESSAGES. All access is through APPS synonyms.

Usage Notes

HZ_IMP_LOAD_WRAPPER is referenced by fifteen other packages and is normally invoked from the TCA batch import concurrent programs rather than directly. Custom code should call BATCH_DATA_LOAD (or DATA_LOAD) only after staging rows have been populated, and should pass a valid batch ID, originating system, and concurrent request context. The package is AUTHID CURRENT_USER, so it executes with the privileges of the calling schema. Because the "Header" revision dates to 2005, the API surface is stable across both 12.1.1 and 12.2.2.