Control deletion and file processing

HDL provides instructions for deleting application data and controlling how individual files are processed. Before applying these controls, confirm that the business-object component supports the intended action and determine whether the load requires any processing overrides or special data-handling rules.

Delete an individual record

First use View Business Objects to confirm that the component supports the Delete action.

To delete a record, use the DELETE instruction and supply only the attributes required to identify it.

For example:

The DELETE line removes the specified Answer.

Including the parent with a MERGE line is recommended because it groups the records into the same logical object. Some object hierarchies require the top-level record when child records are changed or removed.

Remember:

Delete support is a component capability; DELETE is the HDL instruction used when that capability exists.

Delete a complete logical object

To remove a complete object, identify the top-level record with DELETE.

For example:

For many objects, deleting the top-level component also removes its children, so you don't need separate DELETE lines for every child.

This section covers ordinary HDL deletion using the DELETE instruction. Some business objects have additional deletion rules, prerequisites, or dedicated purge processes. Follow the object-specific playbook where one exists.

Find and use SET instructions

A SET instruction changes default HDL processing for the .dat file in which it appears.

SET instructions:

  • are optional
  • don't have a file discriminator
  • can be repeated when more than one processing override is required
  • must appear before the first METADATA line in the file

Find the available SET instructions

Review both enterprise-level configuration and business-object-specific parameters before using a SET instruction. The available sources provide different information about the processing options that can be overridden for a load.

There are two places to review the SET instructions relevant to a load.

Where What you'll find
Configure HCM Data Loader Common HDL processing parameters and their configured enterprise defaults
View Business Objects > Actions > View Parameters Parameters that apply to the selected business object, including the default value and the SET instruction used to override it

Use View Parameters whenever it's available for the business object. This prevents you from relying on an instruction that doesn't apply to the object or overlooking object-specific processing options.

Understand defaults and file overrides

Many HDL processing behaviors have an enterprise default. Administrators can change supported defaults using Configure HCM Data Loader.

A SET instruction lets you override the applicable default for a particular load rather than changing the enterprise configuration. Oracle's current SET guidance explicitly distinguishes the configured enterprise default from the per-file SET override.

For example, if a processing option is normally disabled for the enterprise but one load requires it, use the corresponding SET instruction in that file rather than changing the default for every HDL load.

Don't copy SET instructions into every file routinely. Review what the instruction controls, its default, and whether the scenario genuinely requires an override.

Some SET instructions control specialist processing such as date-effective maintenance, postload processing, auditing, retention, or Oracle Search updates. Before using those instructions, follow the specialist guidance listed later in this playbook.

Follow the standard data formatting rules

HDL expects data values to use consistent formats unless the business-object attribute specifies otherwise. Apply these standard formats when preparing values for dates, numbers, lookups, and null values.

Data Guideline
Date YYYY/MM/DD
Date and time YYYY/MM/DD HH24:MI:SS
Numeric values Decimal notation only; don't include currency symbols or thousands separators
Lookup values Prefer lookup codes over translated meanings
Nulling an existing value #NULL

Lookup meanings can be language-dependent. Codes are normally safer for repeatable integrations.

For special characters, delimiters, embedded line breaks, and complex text, see Handle special characters and complex text.

For translations, see Load and maintain translated data.

Supply source references for reconciliation

Source references are not keys.

They are optional values held with HDL staging information to help you identify, filter, report on, and reconcile imported records.

You can define up to ten reference columns using SourceRef001 through SourceRef010.

Use source references for values that help answer questions such as:

  • Which source record produced this HDL object?
  • Which assignment or transaction does this failure relate to?
  • What source amount should be reconciled?
  • Which batch or source file supplied this record?

The values aren't written to the target application record and aren't validated as business-object attributes.

Don't query HDL staging tables directly for application reconciliation.

Check whether your load needs specialist guidance

Before you submit the file, check whether the scenario has requirements that need more specific guidance than this general file-authoring playbook.

Requirement See:
Date-effective corrections, updates, end dating, Replace/Retain behavior, or PURGE_FUTURE_CHANGES Make date-effective changes with HCM Data Loader
Flexfield segments, images, or attachments Load flexfield segments, images, and attachments with HCM Data Loader
Special characters, delimiters, embedded line breaks, or complex text Handle special characters and complex text
Translated values Load and maintain translated data
Automated postload processing or suppressing postprocesses for a file Understand, configure, and control postload processing
Auditing HDL or HSDL processing Configure auditing for HDL and HSDL
Data-set retention or extended staging retention Maintain stage tables and data-set retention
Oracle Search index updates after bulk loading Manage Oracle Search index updates after bulk loading