Molecule and PS bulk import by file

[XingY]

Rationale

Adds a DataClass-specific DataIterator transformer extension point so modules can inject computed columns (e.g., transforming flat import columns into JSON). This handles all import paths uniformly — file upload, API insertRows, folder import, and async import.
Also propagates the useTransactionAuditCache flag through extraScriptContext so nested trigger inserts (component DataClass rows created by MoleculeTriggerFactory) share the outer transaction's audit context.

Related Pull Requests

• Molecule and PS bulk import by file labkey-ui-components#1998
• https://github.com/LabKey/labkey-ui-premium/pull/948
• Molecule and PS bulk import by file #7651
• https://github.com/LabKey/limsModules/pull/2183

Changes

• ExperimentService: registerDataClassDataIteratorTransformer and getDataClassDataIteratorTransformer.
• ExpDataClassDataTableImpl.PreTriggerDataIteratorBuilder**: Applies the registered transformer so computed columns are available to all downstream stages.
• ExpDataIterators: useTransactionAuditCache for INSERT operations is now enabled unless explicitly opted out via Boolean.FALSE config parameter (was unconditional). This ensures insert/update called from a trigger matches behavior of outer insert/update.
• AbstractQueryUpdateService._importRowsUsingDIB: reads useTransactionAuditCache from extraScriptContext and propagates to DataIteratorContext for nested trigger inserts
• AuditController.GetTransactionRowIds: API response now includes dataTypeRowCounts map alongside rowIds