Search Results po_attribute_values_draft

Overview

PO_ATTRIBUTE_VALUES_DRAFT is a table in the PO (Purchasing) schema of Oracle E-Business Suite, documented as the draft counterpart to PO_ATTRIBUTE_VALUES. It is a wide, denormalized staging table that holds pending or uncommitted attribute values for purchasable inventory items as they are entered, edited, or imported through the Purchasing attribute-value workflow. The table is registered as VALID in both Oracle EBS 12.1.1 and 12.2.2, and its structure is stable across those releases; the documented physical schema contains 331 columns in ETRM 12.2.2.

The "_DRAFT" suffix indicates that rows represent in-progress data rather than committed transactional state. Users editing item attributes in the Supplier Portal, iProcurement, or Purchasing item setup flows write rows here first, then a concurrent process validates and promotes the data. This design isolates unvalidated content from the live attribute tables and supports draft approvals, bulk imports, and interface loads.

Under a heuristic Data Vault classification, this object is assessed as standalone. It is a modeling suggestion only: because the mined foreign-key structure shows no enforced relationships to other tables, the object is best modeled as an independent entity rather than a hub, link, or satellite keyed to external hubs. Its natural integration points are the business identifiers it carries (item, line, template, org) rather than an explicit FK constraint.

Key Information Stored

Two columns form the composite primary key, enforced by SYS_C00204874:

• ATTRIBUTE_VALUES_ID — surrogate identifier for the attribute-value record. This is the column referenced by the user search "attribute_values_id" and the primary handle for the row.
• DRAFT_ID — identifier of the draft session or batch to which the row belongs.

Two unique indexes define business-key candidates. PO_ATTR_VALUES_DRAFT_U1 covers (ATTRIBUTE_VALUES_ID, DRAFT_ID), which mirrors the primary key. PO_ATTR_VALUES_DRAFT_U2 covers (INVENTORY_ITEM_ID, PO_LINE_ID, REQ_TEMPLATE_NAME, REQ_TEMPLATE_LINE_NUM, DRAFT_ID, ORG_ID), and is the more meaningful business key for locating a draft row for a specific item, requisition template line, and operating unit.

The most operationally significant non-key columns are INVENTORY_ITEM_ID (the item whose attributes are being drafted), PO_LINE_ID (the purchasing line context), ORG_ID (operating unit), REQ_TEMPLATE_NAME and REQ_TEMPLATE_LINE_NUM (requisition template context), IP_CATEGORY_ID (iProcurement category), and the workflow-control flags CHANGE_ACCEPTED_FLAG, DELETE_FLAG, and CHANGE_STATUS. Descriptive commerce data is also stored, including MANUFACTURER_PART_NUM, SUPPLIER_URL, MANUFACTURER_URL, ATTACHMENT_URL, PICTURE, THUMBNAIL_IMAGE, UNSPSC, AVAILABILITY, and LEAD_TIME. The bulk of the 331 columns are numbered attribute slots: TEXT_BASE_ATTRIBUTE1–100, NUM_BASE_ATTRIBUTE1–100, TEXT_CAT_ATTRIBUTE1–50, and NUM_CAT_ATTRIBUTE1–50, plus WHO/audit columns (CREATED_BY, CREATION_DATE, LAST_UPDATED_BY, LAST_UPDATE_DATE, LAST_UPDATE_LOGIN) and concurrent-program columns (REQUEST_ID, PROGRAM_ID, PROGRAM_APPLICATION_ID, PROGRAM_UPDATE_DATE, LAST_UPDATED_PROGRAM, REBUILD_SEARCH_INDEX_FLAG).

Common Use Cases and Queries

Typical scenarios include reviewing pending attribute changes before promotion, diagnosing rows rejected during validation, and reporting on incomplete draft item data. A query to retrieve a specific draft row uses the primary key:

• SELECT attribute_values_id, draft_id, inventory_item_id, po_line_id, change_status FROM po.po_attribute_values_draft WHERE attribute_values_id = :id;

To locate draft rows for a specific item and operating unit, the second unique index is the efficient access path:

• SELECT * FROM po.po_attribute_values_draft WHERE inventory_item_id = :item AND org_id = :org AND draft_id = :draft;

Validation and cleanup reporting commonly filters on the control flags, for example selecting rows where CHANGE_STATUS indicates pending action or DELETE_FLAG is set, or joining to PO_ATTRIBUTE_VALUES on ATTRIBUTE_VALUES_ID (and DRAFT_ID where applicable) to compare draft against committed values. Because the table carries 331 columns, reports should project only the needed attribute slots rather than using SELECT *.

Related Objects

The documented FK structure is standalone, so no enforced foreign keys exist. The following objects are the principal logical relations:

• PO_ATTRIBUTE_VALUES — the committed counterpart; joined on ATTRIBUTE_VALUES_ID (and DRAFT_ID in the draft context) to reconcile pending and approved values.
• PO_LINES_ALL — joined on PO_LINE_ID, supplying purchasing line context for drafted attributes.
• MTL_SYSTEM_ITEMS_B — joined on INVENTORY_ITEM_ID and ORG_ID for item description and organization context.
• PO_REQ_TEMPLATES (and template lines) — joined on REQ_TEMPLATE_NAME and REQ_TEMPLATE_LINE_NUM.
• MTL_CATEGORIES_B — joined on IP_CATEGORY_ID for category descriptions.
• FND_CONCURRENT_REQUESTS — joined on REQUEST_ID to identify the concurrent program that created or last updated the draft row.
• FND_USER — joined on CREATED_BY and LAST_UPDATED_BY for audit attribution.

These relationships are logical joins based on shared business identifiers, not enforced constraints, reflecting the table's role as an isolated staging area within the Purchasing attribute workflow.