Engine
Data Sources
Script Document

Script Document Data Source (SCRIPT_DOCUMENT_JS)

A Data Source that handles JSON and YAML documents and delegates the final processing steps to JavaScript.

As explained here the adapter exchanges JSON documents with your JavaScript, via bound objects.
Therefore, your scripts do not have to deal with the document build, since that happens at adapter level, so your scripts have only the logic to interact with other systems, which in most cases means calling its API.

When the engine receives, for example, and UPDATE query, the documents that match the criteria are returned by calling your resultSetScript, then, based on the type of table being updated, the engine updates the document/s and calls the script defined as updateScript.

💡

It is important to note that the resultSetScript must do its best effort to return documents that match the criteria, or WHERE clause, however when the system does not have specific filter mechanisms or it is complex to implement, your script can return documents not matching the criteria, since the result set is further inspected by the engine before returning the rows.

Sample configuration in VDB file:

- name: github
  dataSourceType: SCRIPT_DOCUMENT_JS
  configObject:
    module:
      zipFilePath: kbl://github/0.0.1
  translatorConfig: module:translator-config.yaml
  schema:
    type: PHYSICAL
    ddlFilePaths:
      - module:github.ddl

In the sample above, zipFilePath uses the kbl URL scheme that points to the Module repository configured in the Main application configuration, under moduleRepoBaseURI. By default, it points to https://repo.kubling.com.

translatorConfig and schema.ddlFilePaths point to files inside the module, that is why they use the special module: prefix. For more information please see Virtual Database information file.

Specific TABLE directives

Following directives work only in Document Data Sources (like this one)

DirectiveTypeOptionsDescription
synthetic_parentStringAnyParent TABLE. Must include schema name.
synthetic_pathStringAnyName of the parent TABLE field (array) used to ungroup the content
synthetic_allow_bulk_insertBooleanTrue/FalseIndicates whether an INSERT operation tolerates multiple entries in the synthetic_path or must be strictly one.
supports_idempotencyBooleanTrue/FalseIndicates whether the upstream API supports idempotence.
When false the QP verifies whether the operation changed the original record/document before applying.
If omitted, the default value is true

Specific field directives

DirectiveTypeOptionsDescription
synthetic_typeStringparent, parent_array_keyType of the synthetic field.
parent indicates that exists in and is connected to the parent entity.
parent_array_key indicates that the field is used to locate specific elements within synthetic_path
synthetic_parent_fieldStringAnyIndicates that the related parent field name is not the same as defined in the synthetic entity field. Used mostly when field names overlap.
parser_formatStringparser_format 'option'
Options:
asJsonPretty, asYaml, asKeyValueList		Applies a new format to the value, without altering it.
KubernetesEmbedded