/AWS1/CL_QSTKBTEMPLATECONF

The template configuration for a knowledge base. This object contains connector-specific configuration that defines how data is crawled and indexed.

CONSTRUCTOR

IMPORTING

Optional arguments:

io_template TYPE REF TO /AWS1/CL_RT_DOCUMENT /AWS1/CL_RT_DOCUMENT

The connector configuration for the knowledge base data source. The structure depends on the connector type of the data source referenced by DataSourceArn.

The template must be a JSON object. All connector types share the following top-level keys. The value of type and the contents of connectionConfiguration vary by connector type.

• type – (Required) The connector type of the data source. This value identifies the connector. Valid values: S3V2, WEBCRAWLERV3, GOOGLEDRIVEV3, ONEDRIVEV3, SHAREPOINTV3. For the fields required by each connector, see the connector-specific list that follows.

• connectionConfiguration – (Required) The connection details for the data source. The keys in this object vary by connector type; see the connector-specific list that follows.

• filterConfiguration – (Optional) Rules that determine which content is crawled, such as inclusion and exclusion prefixes, patterns, or file-size limits.

• accessControlConfiguration – (Optional) Document-level access control (ACL) settings. Supported by all connector types except Web Crawler (WEBCRAWLERV3). The available fields depend on the connector type.

• deletionProtectionConfiguration – (Optional) Deletion-protection settings, supported by all connector types. Contains enableDeletionProtection (Boolean) and deletionProtectionThreshold (String; a value from 1 to 100).

The following list describes the valid type value, the connectionConfiguration contents, and any connector-specific fields for each connector type:

• Amazon S3 (type: S3V2) – The type value must be S3V2. connectionConfiguration is required and contains:

  • bucketName – (Required) The name of the Amazon S3 bucket to crawl. Type: String. Length: 3–63 characters. Pattern: ^[a-z0-9][.-a-z0-9]{1,61}[a-z0-9]$.

  • bucketOwnerAccountId – (Required) The ID of the AWS account that owns the bucket. Type: String. Pattern: ^\d{12}$.

  Amazon S3 supports the following optional filterConfiguration fields:

  • inclusionPrefixes or exclusionPrefixes – Amazon S3 key prefixes to include or exclude. Type: Array of String. Up to 350 items, each 1–1,024 characters.

  • inclusionPatterns or exclusionPatterns – Patterns to include or exclude objects. Type: Array of String. Up to 350 items, each 1–1,024 characters.

  • maxFileSizeInMegaBytes – The maximum size, in MB, of a file to ingest. Type: String. Pattern: ^\d+$.

  For Amazon S3, accessControlConfiguration supports the following fields:

  • crawlAcl – Specifies whether the connector crawls and enforces document access control lists (ACLs). Type: Boolean. When set to true, provide ACLs either in a global ACL configuration file (aclConfigurationFilePath) or in per-document metadata files.

  • aclConfigurationFilePath – The Amazon S3 URI of the global ACL configuration file. Type: String. Length: 1–1,024 characters. Optional. If you don't provide a global ACL configuration file, define ACLs in per-document metadata files.

  • defaultAccessType – The access behavior applied to Amazon S3 prefixes that are not listed in the ACL configuration. Type: String. The only supported value is ALLOW.

  metadataFilesPrefix – (Optional) The Amazon S3 prefix under which per-document metadata files are stored. Each metadata file describes a single source document and its indexable attributes. This is not the global ACL configuration file. For a single global ACL file, use accessControlConfiguration.aclConfigurationFilePath. Type: String. Length: 1–1,024 characters.

• Google Drive (type: GOOGLEDRIVEV3) – Requires connectionConfiguration with authType set to SERVICE_ACCOUNT. Supports dataEntityConfiguration with crawlMyDrive, crawlSharedWithMe, and crawlSharedDrives.

• OneDrive (type: ONEDRIVEV3) – Requires authType at the template root level set to TWO_LEGGED_OAUTH. Requires connectionConfiguration with tenantId in UUID format. Supports dataEntityConfiguration with crawlPersonalDrives and crawlSharedWithMe.

• SharePoint (type: SHAREPOINTV3) – Requires connectionConfiguration with tenantId in UUID format. Supports dataEntityConfiguration with siteUrls, crawlFiles, and crawlPages.

• Web Crawler (type: WEBCRAWLERV3) – Requires connectionConfiguration with seedUrls or siteMapUrls (mutually exclusive) and authType. Supports crawlConfiguration for crawl depth, rate limits, and scope. Supports filterConfiguration for file size limits and URL patterns. Valid values for authType: NO_AUTH, BASIC_AUTH, FORM, SAML.

Enabling document-level access control for Amazon S3

For an Amazon S3 (S3V2) knowledge base, document-level access control is governed by two settings that must both be enabled:

1. In this template, set accessControlConfiguration.crawlAcl to true. Define ACLs either in a global ACL configuration file, referenced by accessControlConfiguration.aclConfigurationFilePath, or in per-document metadata files. To control access for prefixes that are not listed in the ACL file, you can also set accessControlConfiguration.defaultAccessType.

2. In the CreateKnowledgeBase or UpdateKnowledgeBase request, set the top-level AccessControlConfiguration.isACLEnabled to true.

---

Queryable Attributes

template

The connector configuration for the knowledge base data source. The structure depends on the connector type of the data source referenced by DataSourceArn.

The template must be a JSON object. All connector types share the following top-level keys. The value of type and the contents of connectionConfiguration vary by connector type.

• type – (Required) The connector type of the data source. This value identifies the connector. Valid values: S3V2, WEBCRAWLERV3, GOOGLEDRIVEV3, ONEDRIVEV3, SHAREPOINTV3. For the fields required by each connector, see the connector-specific list that follows.

• connectionConfiguration – (Required) The connection details for the data source. The keys in this object vary by connector type; see the connector-specific list that follows.

• filterConfiguration – (Optional) Rules that determine which content is crawled, such as inclusion and exclusion prefixes, patterns, or file-size limits.

• accessControlConfiguration – (Optional) Document-level access control (ACL) settings. Supported by all connector types except Web Crawler (WEBCRAWLERV3). The available fields depend on the connector type.

• deletionProtectionConfiguration – (Optional) Deletion-protection settings, supported by all connector types. Contains enableDeletionProtection (Boolean) and deletionProtectionThreshold (String; a value from 1 to 100).

The following list describes the valid type value, the connectionConfiguration contents, and any connector-specific fields for each connector type:

• Amazon S3 (type: S3V2) – The type value must be S3V2. connectionConfiguration is required and contains:

  • bucketName – (Required) The name of the Amazon S3 bucket to crawl. Type: String. Length: 3–63 characters. Pattern: ^[a-z0-9][.-a-z0-9]{1,61}[a-z0-9]$.

  • bucketOwnerAccountId – (Required) The ID of the AWS account that owns the bucket. Type: String. Pattern: ^\d{12}$.

  Amazon S3 supports the following optional filterConfiguration fields:

  • inclusionPrefixes or exclusionPrefixes – Amazon S3 key prefixes to include or exclude. Type: Array of String. Up to 350 items, each 1–1,024 characters.

  • inclusionPatterns or exclusionPatterns – Patterns to include or exclude objects. Type: Array of String. Up to 350 items, each 1–1,024 characters.

  • maxFileSizeInMegaBytes – The maximum size, in MB, of a file to ingest. Type: String. Pattern: ^\d+$.

  For Amazon S3, accessControlConfiguration supports the following fields:

  • crawlAcl – Specifies whether the connector crawls and enforces document access control lists (ACLs). Type: Boolean. When set to true, provide ACLs either in a global ACL configuration file (aclConfigurationFilePath) or in per-document metadata files.

  • aclConfigurationFilePath – The Amazon S3 URI of the global ACL configuration file. Type: String. Length: 1–1,024 characters. Optional. If you don't provide a global ACL configuration file, define ACLs in per-document metadata files.

  • defaultAccessType – The access behavior applied to Amazon S3 prefixes that are not listed in the ACL configuration. Type: String. The only supported value is ALLOW.

  metadataFilesPrefix – (Optional) The Amazon S3 prefix under which per-document metadata files are stored. Each metadata file describes a single source document and its indexable attributes. This is not the global ACL configuration file. For a single global ACL file, use accessControlConfiguration.aclConfigurationFilePath. Type: String. Length: 1–1,024 characters.

• Google Drive (type: GOOGLEDRIVEV3) – Requires connectionConfiguration with authType set to SERVICE_ACCOUNT. Supports dataEntityConfiguration with crawlMyDrive, crawlSharedWithMe, and crawlSharedDrives.

• OneDrive (type: ONEDRIVEV3) – Requires authType at the template root level set to TWO_LEGGED_OAUTH. Requires connectionConfiguration with tenantId in UUID format. Supports dataEntityConfiguration with crawlPersonalDrives and crawlSharedWithMe.

• SharePoint (type: SHAREPOINTV3) – Requires connectionConfiguration with tenantId in UUID format. Supports dataEntityConfiguration with siteUrls, crawlFiles, and crawlPages.

• Web Crawler (type: WEBCRAWLERV3) – Requires connectionConfiguration with seedUrls or siteMapUrls (mutually exclusive) and authType. Supports crawlConfiguration for crawl depth, rate limits, and scope. Supports filterConfiguration for file size limits and URL patterns. Valid values for authType: NO_AUTH, BASIC_AUTH, FORM, SAML.

Enabling document-level access control for Amazon S3

For an Amazon S3 (S3V2) knowledge base, document-level access control is governed by two settings that must both be enabled:

1. In this template, set accessControlConfiguration.crawlAcl to true. Define ACLs either in a global ACL configuration file, referenced by accessControlConfiguration.aclConfigurationFilePath, or in per-document metadata files. To control access for prefixes that are not listed in the ACL file, you can also set accessControlConfiguration.defaultAccessType.

2. In the CreateKnowledgeBase or UpdateKnowledgeBase request, set the top-level AccessControlConfiguration.isACLEnabled to true.

Accessible with the following methods

Method	Description
GET_TEMPLATE()	Getter for TEMPLATE, with configurable default
ASK_TEMPLATE()	Getter for TEMPLATE w/ exceptions if field has no value
HAS_TEMPLATE()	Determine if TEMPLATE has a value