Targets and exporters

Targets define where a product goes. Use CSV or another artifact exporter for deliverables, LogExporter for bounded diagnostics, and configured client IDs for databases or message brokers.

Target families

database

Use for an explicit relational write operation through a configured target client.

Operations: clear, delete, insert, update

Complete workflows:

• Assure a finite heterogeneous fixture across one declared sink — Use only for a small finite fixture whose intentionally different rows cannot be expressed by one declarative generated product.
• Execute one explicit database operation — Use operation only for a reviewed source-free clear or MongoDB drop statement.
• Copy a bounded relational selection — Use when rows move from one configured relational source into an explicit database target.

file_export

Use to serialize generated rows with a runtime-declared file exporter.

Complete workflows:

• Transform a project file into an artifact — Use for a bounded project-relative input whose transformed rows must become a file artifact.
• Transform an object into a stored artifact — Use when both input and generated file artifacts live behind one object-storage client.

kafka

Use to publish each generated row through a configured Kafka exporter.

Complete workflows:

• Forward a bounded Kafka selection — Use when a finite topic selection must be transformed and published through another exporter.

log

Use only for bounded diagnostics while developing or reviewing a model.

Complete workflows:

• Consume a bounded persisted ML model — Use after ml-train when a bounded product must generate records from the persisted model.

memstore

Use to make generated rows available to a later memstore source in the same run.

Complete workflows:

• Chain generated and source products through memstore — Use for bounded multi-stage assembly when a later product consumes an earlier product in the same run.

mongodb

Use for an explicit MongoDB write operation through a configured target client.

Operations: clear, delete, drop, insert, update, upsert

Complete workflows:

• Project a bounded MongoDB collection — Use for a bounded collection read whose selected fields are written to another collection.

rabbitmq

Use to publish through a configured RabbitMQ exporter with bounded publisher confirms.

Complete workflows:

• Forward a bounded RabbitMQ queue — Use for count-bounded FIFO/SP queue consumption followed by confirmed publication.

Exporters

ConsoleExporter diagnostic

Print generated rows for embedded debugging.

Why: Use only outside Platform-oriented artifact and log workflows.

Output: diagnostic; none.

<generate name="items" count="1" target="ConsoleExporter"/>

Options: None

targetEntity: not supported

LogExporter diagnostic

Write a bounded diagnostic projection to the task log.

Why: Use for small human-readable diagnostics, not deliverable datasets.

Output: diagnostic; none.

<generate name="items" count="1" target="LogExporter"/>

Options: None

targetEntity: not supported

Preview diagnostic

Capture a bounded preview for runtime inspection.

Why: Use for a small inspection sample while designing a descriptor.

Output: diagnostic; none.

<generate name="items" count="1" target="Preview"/>

Options: None

targetEntity: not supported

TestResultExporter internal

Capture test assertions for the EE test-result integration.

Why: Reserved for test execution; use LogExporter or a file artifact in user descriptors.

Output: runtime; none.

<generate name="items" count="1" target="TestResultExporter"/>

Options: None

targetEntity: not supported

CSV public

Write tabular rows as a CSV artifact.

Why: Use for interoperable tabular artifacts.

Output: artifact; per_worker.

<generate name="items" count="1" target="CSV"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: encoding, delimiter, quotechar, quoting, line_terminator, fieldnames

targetEntity: not supported

JSON public

Write rows as a JSON or NDJSON artifact.

Why: Use when a downstream consumer expects structured JSON records.

Output: artifact; per_worker.

<generate name="items" count="1" target="JSON"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: encoding, use_ndjson

targetEntity: not supported

JSONSingle public

Write one JSON artifact for each routed record.

Why: Use when each record must be delivered as an independently named JSON document.

Output: artifact; per_record.

<generate name="items" count="1" target="JSONSingle" targetEntity="{{this.id}}"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: None

targetEntity: supported

OpenSearchBulk public

Write records in OpenSearch bulk-ingest format.

Why: Use when a downstream OpenSearch bulk loader consumes the artifact.

Output: artifact; per_worker.

<generate name="items" count="1" target="OpenSearchBulk"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: use_ndjson

targetEntity: not supported

PARQUET public

Write typed tabular rows as a Parquet artifact.

Why: Use for efficient analytical-storage artifacts.

Output: artifact; per_worker.

<generate name="items" count="1" target="PARQUET"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: compression

targetEntity: not supported

SQL public

Write rows as SQL statements in an artifact.

Why: Use when a downstream system applies generated SQL statements.

Output: artifact; per_worker.

<generate name="items" count="1" target="SQL"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: dialect, database_type, table_name, columns, column_map, batch_size, mode

targetEntity: not supported

TXT public

Write rendered values as a plain-text artifact.

Why: Use for line-oriented integration files.

Output: artifact; per_worker.

<generate name="items" count="1" target="TXT"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: encoding, delimiter, line_terminator, value_only

targetEntity: not supported

Template public

Render rows through a descriptor-local template into an artifact.

Why: Use when a consumer requires a domain-specific text layout.

Output: artifact; per_worker.

<generate name="items" source="templates/customer.template" target="Template" count="1"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: encoding, error_handling, template_separator

targetEntity: not supported

TemplateSingle public

Render one template artifact for each routed record.

Why: Use only when every record requires its own independently named template artifact.

Output: artifact; per_record.

<generate name="items" count="1" target="TemplateSingle" targetEntity="{{this.id}}"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: extension, error_handling

targetEntity: supported

XLSX public

Write rows as an Excel workbook artifact.

Why: Use when reviewers or business users need a spreadsheet deliverable.

Output: artifact; per_worker.

<generate name="items" count="1" target="XLSX"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: sheet_name

targetEntity: not supported

XML public

Write rows as an XML artifact.

Why: Use for XML-based integration contracts.

Output: artifact; per_worker.

<generate name="items" count="1" target="XML"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: encoding, root_element, item_element

targetEntity: not supported

XMLSingle public

Write one XML artifact for each routed record.

Why: Use when each record must be delivered as an independently named XML document.

Output: artifact; per_record.

<generate name="items" count="1" target="XMLSingle" targetEntity="{{this.id}}"/>

Platform Artifact View

On DATAMIMIC Platform, leave storageId and exportUri unset. Platform-managed artifact storage chooses the task-scoped location read by Artifact View.

Set exportUri only together with a storageId that names your own <object-storage>. It is a relative path prefix, not a filename. Outputs in custom object storage are outside Platform Artifact View.

Options: root_element, item_element

targetEntity: supported

mem public

Retain a named in-run dataset for a later memstore source.

Why: Use for deterministic same-run assembly without a persistent artifact.

Output: runtime; none.

<generate name="items" count="1" target="mem" targetEntity="orders-buffer"/>

Options: None

targetEntity: supported