Search Results pa_forecast_items_s

Overview

PA_FORECASTITEM_PVT is a private Application Programming Interface (API) package body owned by APPS that supports the forecast item generation and maintenance process within Oracle E-Business Suite Project Management (PA). Forecast items represent forward-looking resource requirements calculated against project assignments, schedules, and resource definitions; this package encapsulates the server-side logic required to create, build, reverse, fetch, and delete both forecast item headers and details.

The package follows the private (PVT) API convention, meaning its procedures are not intended for direct external invocation and are instead called by sibling public packages and other forecast-processing modules. The package exposes 43 documented procedures and functions. It is referenced by eight other packages, confirming its role as a shared internal service layer.

A defining characteristic is the self-service design constraint noted in the source header: backend code executed in self-service cannot rely on package globals. Accordingly, the package avoids persistent global state except for a debug flag initialized from the profile option PA_DEBUG_MODE.

Key Procedures and Functions

The most visible procedure for developers, and the one associated with the search term "print_message", is PRINT_MESSAGE. It is a centralized diagnostic utility that writes messages to the PA debug infrastructure. It first checks the PA_DEBUG_MODE profile option; when this evaluates to "Y", it routes the message to pa_debug.write using the module identifier pa.plsql.pa_forecastitem_pvt and log level 1, and also invokes pa_debug.write_file. The procedure deliberately guards against null input and re-raises any unexpected exception, preserving error propagation semantics.

The core generation logic resides in CREATE_FORECAST_ITEM, which supports multiple process modes: GENERATE for new item creation and regeneration when schedule data changes, RECALCULATE for updates triggered by expenditure organizational changes (operating unit, organization, expenditure type, type class, or borrowed flag), and ERROR for regeneration of items that failed during a prior run.

Supporting procedures include GENERATE_REQUIREMENT_FI, GENERATE_ASSIGNMENT_FI, INITIALIZE_DAY_FI, and BUILD_DAY_FI, which drive day-level forecast building. Header and detail assembly is handled by BUILD_FI_HDR_REQ, BUILD_FI_DTL_REQ, BUILD_FI_HDR_ASG, and BUILD_FI_DTL_ASG. Retrieval and maintenance operations are provided by FETCH_FI_HDR, FETCH_FI_DTL, REVERSE_FI_HDR, REVERSE_FI_DTL, DELETE_FORECAST_ITEM, DELETE_FI, DELETE_FI_HDR, and DELETE_FI_DTL. GET_ASSIGNMENT_DTLS and GET_PROJECT_DTLS supply the descriptive context required during generation.

Tables Accessed

The package reads and writes the forecast item base and interface tables PA_FORECAST_ITEMS and PA_FORECAST_ITEMS_S, and the child table PA_FORECAST_ITEM_DETAILS, through APPS synonyms. Project and assignment context is drawn from PA_PROJECTS_ALL and PA_PROJECT_ASSIGNMENTS. Project typing is resolved via PA_PROJECT_TYPES_ALL, while resource definitions are supplied by PA_RESOURCES_DENORM. Scheduling data is read from PA_SCHEDULES to drive requirement calculations. DUAL is used for single-row computations, and PLITBLM serves as the standard PL/SQL index-by table for bulk operations.

Usage Notes

PA_FORECASTITEM_PVT is not called directly from Oracle Forms or from user-facing concurrent program parameters in normal operation. Rather, it is invoked by the forecast item generation concurrent program and by higher-level public APIs within Project Management, including the eight packages documented as referencing it. Custom code should treat PRINT_MESSAGE as the sanctioned diagnostic hook: setting the PA_DEBUG_MODE profile option to "Y" at the user or site level activates logging without code changes, and messages are emitted with the module tag pa.plsql.pa_forecastitem_pvt. Because the package is classified PVT, extensions should avoid direct calls to CREATE_FORECAST_ITEM and its sibling routines; the supported entry points are the public forecast APIs that wrap them.