Search Results purge_interface_lines




Overview
Explore in AI Chat

Overview

EGO_ITEM_PEOPLE_IMPORT_PKG is a PL/SQL package owned by APPS in the Oracle E-Business Suite EGO (Enterprise Governance/Item Management) schema family. It supports the Item People Import process, a bulk data-loading utility that brings item-to-person relationship data — such as item owners, planners, buyers, and other assigned parties — into Oracle EBS through a SQL*Loader driven interface. The package manages the lifecycle of an import run: identifying the current dataset, loading staged rows into the interface table, and purging processed interface lines once execution completes.

The specification header (EGOCIPIS.pls, version 120.2) documents the package as implementing the Item People Import HLD. The package declares a broad set of constants governing concurrent program behavior: standard return codes (RETCODE_SUCCESS, RETCODE_WARNING, RETCODE_ERROR), debug modes (FATAL through DEBUG), logging destinations (table only, file only, or both), batch sizing thresholds (small 1000, recommended 5000, large 10000), ORDER BY category codes, and process status flags including G_PS_TO_BE_PROCESSED. It also declares G_CURR_DATASET_ID, used to track the active SQL*Loader dataset, and G_PACKAGE_NAME for consistent log and error identification. This constant structure is characteristic of Oracle's bulk import packages, which are designed to support restartable, auditable loads with controlled commit granularity.

Key Procedures and Functions

  • GET_CURR_DATASET_ID — Returns the dataset identifier associated with the current import run. The dataset ID is used by the SQL*Loader control files and by the interface tables to partition rows belonging to a single load, allowing independent processing and cleanup per run.
  • LOAD_INTERFACE_LINES — Loads staged import records into the interface table. This is the core data-movement routine that populates the interface with the item people data to be validated and subsequently applied to the production entities.
  • PURGE_INTERFACE_LINES — Removes processed interface rows, typically after successful processing or to clean up prior failed runs. Purging prevents accumulation of stale data that could interfere with subsequent imports and helps maintain interface table performance.

Tables Accessed

The package reads and writes EGO_ITEM_PEOPLE_INTF, the primary interface table for item people import data. It obtains sequence values from EGO_IPI_DATASET_ID_S and EGO_IPI_TRANSACTION_ID_S for dataset and transaction identification. Errors captured during loading are recorded in MTL_INTERFACE_ERRORS, the standard interface error repository. Party and relationship data is sourced from HZ_PARTIES and HZ_RELATIONSHIPS. The package references EGO_LOGIN_ITEMS_TEMP as a temporary work table. File-based logging uses UTL_FILE. Several FND tables — FND_FORM_FUNCTIONS, FND_GRANTS, FND_MENUS, FND_MENUS_TL, FND_MENU_ENTRIES, FND_NEW_MESSAGES, and FND_OBJECTS — are referenced, supporting menu/function metadata and message retrieval.

Usage Notes

The package is normally invoked through the Item People Import concurrent program, which orchestrates the load in phases: resolving the dataset ID via GET_CURR_DATASET_ID, loading rows with LOAD_INTERFACE_LINES, processing them, and finally calling PURGE_INTERFACE_LINES for cleanup. Batch size constants (recommended 5000, maximum 10000) guide commit frequency during loading. Debug and logging modes allow DBAs to trace failures either to EGO_LOGIN_ITEMS_TEMP or to an operating-system file via UTL_FILE. Because the package is classified as OTHER rather than a formal public API, customizations should call it only for its documented procedures and should not assume stable signatures across patches. At least one other package references it, so changes to its specification can have downstream impact.