import * as pulumi from "@pulumi/pulumi"; import * as inputs from "../types/input"; import * as outputs from "../types/output"; /** * A DataObject is a single item of data (with optional vectors) stored in a * Vector Search Collection. Each DataObject conforms to the parent * Collection's `dataSchema` and `vectorSchema`. * * This resource always issues one `CreateDataObject` request per Terraform * resource block. It does NOT use the `batchCreate` REST endpoint -- * Terraform's resource lifecycle is inherently per-object, so batching * across resources is not modeled. When you use `forEach` or `count`, * Terraform will still issue individual requests, up to `-parallelism` * in parallel. * * For ingesting more than a few hundred items, prefer one of the * following out-of-band paths instead of Terraform: * * * `importDataObjects` (bulk ingest from Cloud Storage) -- highest * throughput, but only available *before* any Index is created on * the Collection. * * `batchCreate` (up to ~1000 items per call) -- available at any * time, but must be driven from your own client code, not Terraform. * * Once an Index exists on the Collection, `importDataObjects` is no * longer available and DataObjects must be created via `CreateDataObject` * (as this resource does) or via `batchCreate`. * * ## Example Usage * * ### Vectorsearch Data Object Basic * * ```typescript * import * as pulumi from "@pulumi/pulumi"; * import * as gcp from "@pulumi/gcp"; * * // NOTE: This resource issues one CreateDataObject request per block. * // It does NOT batch across resources. Terraform will parallelize a * // 'for_each' up to '-parallelism', but each item is still a separate * // HTTP call. * // * // For bulk ingestion of many items, prefer one of these out-of-band * // paths instead of Terraform: * // * 'importDataObjects' (from Cloud Storage) -- highest throughput, * // but only available *before* any Index is created on the Collection. * // * 'batchCreate' (up to ~1000 items per call) -- available at any * // time, but must be driven from client code, not Terraform. * const parent = new gcp.vectorsearch.Collection("parent", { * location: "us-central1", * collectionId: "example-collection", * displayName: "My Awesome Collection", * description: "This collection stores important data.", * dataSchema: `{ * \\"type\\": \\"object\\", * \\"properties\\": { * \\"title\\": { * \\"type\\": \\"string\\" * }, * \\"plot\\": { * \\"type\\": \\"string\\" * } * } * } * `, * vectorSchemas: [{ * fieldName: "text_embedding", * denseVector: { * dimensions: 768, * vertexEmbeddingConfig: { * modelId: "text-embedding-005", * taskType: "RETRIEVAL_DOCUMENT", * textTemplate: "Title: {title} ---- Plot: {plot}", * }, * }, * }], * }); * // Because the parent Collection's 'text_embedding' field is configured * // with a 'vertex_embedding_config', the server will populate the vector * // automatically from 'data.title' and 'data.plot' -- no explicit * // 'vectors' block is required. * const example_data_object = new gcp.vectorsearch.DataObject("example-data-object", { * location: "us-central1", * collectionId: parent.collectionId, * dataObjectId: "example-data-object", * data: JSON.stringify({ * title: "The Matrix", * plot: "A computer hacker learns about the true nature of reality.", * }), * }); * ``` * ### Vectorsearch Data Object With Vectors * * ```typescript * import * as pulumi from "@pulumi/pulumi"; * import * as gcp from "@pulumi/gcp"; * * // NOTE: This resource issues one CreateDataObject request per block. * // It does NOT batch across resources. Terraform will parallelize a * // 'for_each' up to '-parallelism', but each item is still a separate * // HTTP call. * // * // For bulk ingestion of many items, prefer one of these out-of-band * // paths instead of Terraform: * // * 'importDataObjects' (from Cloud Storage) -- highest throughput, * // but only available *before* any Index is created on the Collection. * // * 'batchCreate' (up to ~1000 items per call) -- available at any * // time, but must be driven from client code, not Terraform. * const parent = new gcp.vectorsearch.Collection("parent", { * location: "us-central1", * collectionId: "example-vectors-collection", * displayName: "My BYO-Embedding Collection", * description: "Collection whose vectors are supplied by the client.", * dataSchema: `{ * \\"type\\": \\"object\\", * \\"properties\\": { * \\"title\\": { * \\"type\\": \\"string\\" * }, * \\"category\\": { * \\"type\\": \\"string\\" * } * } * } * `, * vectorSchemas: [ * { * fieldName: "dense_embedding", * denseVector: { * dimensions: 4, * }, * }, * { * fieldName: "sparse_embedding", * sparseVector: {}, * }, * ], * }); * const example_vectors_data_object = new gcp.vectorsearch.DataObject("example-vectors-data-object", { * location: "us-central1", * collectionId: parent.collectionId, * dataObjectId: "example-vectors-data-object", * data: JSON.stringify({ * title: "The Matrix", * category: "movie", * }), * vectors: [ * { * fieldName: "dense_embedding", * dense: { * values: [ * 0.11, * 0.22, * 0.33, * 0.44, * ], * }, * }, * { * fieldName: "sparse_embedding", * sparse: { * values: [ * 0.9, * 0.5, * 0.1, * ], * indices: [ * 3, * 17, * 42, * ], * }, * }, * ], * }); * ``` * * ## Import * * DataObject can be imported using any of these accepted formats: * * * `projects/{{project}}/locations/{{location}}/collections/{{collection_id}}/dataObjects/{{data_object_id}}` * * `{{project}}/{{location}}/{{collection_id}}/{{data_object_id}}` * * `{{location}}/{{collection_id}}/{{data_object_id}}` * * When using the `pulumi import` command, DataObject can be imported using one of the formats above. For example: * * ```sh * $ pulumi import gcp:vectorsearch/dataObject:DataObject default projects/{{project}}/locations/{{location}}/collections/{{collection_id}}/dataObjects/{{data_object_id}} * $ pulumi import gcp:vectorsearch/dataObject:DataObject default {{project}}/{{location}}/{{collection_id}}/{{data_object_id}} * $ pulumi import gcp:vectorsearch/dataObject:DataObject default {{location}}/{{collection_id}}/{{data_object_id}} * ``` */ export declare class DataObject extends pulumi.CustomResource { /** * Get an existing DataObject resource's state with the given name, ID, and optional extra * properties used to qualify the lookup. * * @param name The _unique_ name of the resulting resource. * @param id The _unique_ provider ID of the resource to lookup. * @param state Any extra arguments used during the lookup. * @param opts Optional settings to control the behavior of the CustomResource. */ static get(name: string, id: pulumi.Input, state?: DataObjectState, opts?: pulumi.CustomResourceOptions): DataObject; /** * Returns true if the given object is an instance of DataObject. This is designed to work even * when multiple copies of the Pulumi SDK have been loaded into the same process. */ static isInstance(obj: any): obj is DataObject; /** * The ID of the parent Collection. */ readonly collectionId: pulumi.Output; /** * [Output only] Create time stamp */ readonly createTime: pulumi.Output; /** * The JSON data of the DataObject. Must be a JSON object whose field * names match the fields defined in the parent Collection's * `dataSchema`. */ readonly data: pulumi.Output; /** * ID of the DataObject to create. * The id must be 1-63 characters long, and comply with * [RFC1035](https://www.ietf.org/rfc/rfc1035.txt). * Specifically, it must be 1-63 characters long and match the regular * expression `a-z?`. */ readonly dataObjectId: pulumi.Output; /** * Whether Terraform will be prevented from destroying the resource. Defaults to DELETE. * When a 'terraform destroy' or 'pulumi up' would delete the resource, * the command will fail if this field is set to "PREVENT" in Terraform state. * When set to "ABANDON", the command will remove the resource from Terraform * management without updating or deleting the resource in the API. * When set to "DELETE", deleting the resource is allowed. */ readonly deletionPolicy: pulumi.Output; /** * The etag of the DataObject, used for optimistic concurrency * control on updates and deletes. */ readonly etag: pulumi.Output; /** * Resource ID segment making up resource `name`. It identifies the resource within its parent collection as described in https://google.aip.dev/122. */ readonly location: pulumi.Output; /** * Identifier. name of resource */ readonly name: pulumi.Output; /** * The ID of the project in which the resource belongs. * If it is not provided, the provider project is used. */ readonly project: pulumi.Output; /** * [Output only] Update time stamp */ readonly updateTime: pulumi.Output; /** * The vectors of the DataObject, keyed by the vector field name as * defined in the parent Collection's `vectorSchema`. * If a vector field is configured with a `vertexEmbeddingConfig` on * the Collection, the server will populate the vector automatically * from the corresponding text in `data` and the field should be * omitted here. * Structure is documented below. */ readonly vectors: pulumi.Output; /** * Create a DataObject resource with the given unique name, arguments, and options. * * @param name The _unique_ name of the resource. * @param args The arguments to use to populate this resource's properties. * @param opts A bag of options that control this resource's behavior. */ constructor(name: string, args: DataObjectArgs, opts?: pulumi.CustomResourceOptions); } /** * Input properties used for looking up and filtering DataObject resources. */ export interface DataObjectState { /** * The ID of the parent Collection. */ collectionId?: pulumi.Input; /** * [Output only] Create time stamp */ createTime?: pulumi.Input; /** * The JSON data of the DataObject. Must be a JSON object whose field * names match the fields defined in the parent Collection's * `dataSchema`. */ data?: pulumi.Input; /** * ID of the DataObject to create. * The id must be 1-63 characters long, and comply with * [RFC1035](https://www.ietf.org/rfc/rfc1035.txt). * Specifically, it must be 1-63 characters long and match the regular * expression `a-z?`. */ dataObjectId?: pulumi.Input; /** * Whether Terraform will be prevented from destroying the resource. Defaults to DELETE. * When a 'terraform destroy' or 'pulumi up' would delete the resource, * the command will fail if this field is set to "PREVENT" in Terraform state. * When set to "ABANDON", the command will remove the resource from Terraform * management without updating or deleting the resource in the API. * When set to "DELETE", deleting the resource is allowed. */ deletionPolicy?: pulumi.Input; /** * The etag of the DataObject, used for optimistic concurrency * control on updates and deletes. */ etag?: pulumi.Input; /** * Resource ID segment making up resource `name`. It identifies the resource within its parent collection as described in https://google.aip.dev/122. */ location?: pulumi.Input; /** * Identifier. name of resource */ name?: pulumi.Input; /** * The ID of the project in which the resource belongs. * If it is not provided, the provider project is used. */ project?: pulumi.Input; /** * [Output only] Update time stamp */ updateTime?: pulumi.Input; /** * The vectors of the DataObject, keyed by the vector field name as * defined in the parent Collection's `vectorSchema`. * If a vector field is configured with a `vertexEmbeddingConfig` on * the Collection, the server will populate the vector automatically * from the corresponding text in `data` and the field should be * omitted here. * Structure is documented below. */ vectors?: pulumi.Input[] | undefined>; } /** * The set of arguments for constructing a DataObject resource. */ export interface DataObjectArgs { /** * The ID of the parent Collection. */ collectionId: pulumi.Input; /** * The JSON data of the DataObject. Must be a JSON object whose field * names match the fields defined in the parent Collection's * `dataSchema`. */ data?: pulumi.Input; /** * ID of the DataObject to create. * The id must be 1-63 characters long, and comply with * [RFC1035](https://www.ietf.org/rfc/rfc1035.txt). * Specifically, it must be 1-63 characters long and match the regular * expression `a-z?`. */ dataObjectId: pulumi.Input; /** * Whether Terraform will be prevented from destroying the resource. Defaults to DELETE. * When a 'terraform destroy' or 'pulumi up' would delete the resource, * the command will fail if this field is set to "PREVENT" in Terraform state. * When set to "ABANDON", the command will remove the resource from Terraform * management without updating or deleting the resource in the API. * When set to "DELETE", deleting the resource is allowed. */ deletionPolicy?: pulumi.Input; /** * The etag of the DataObject, used for optimistic concurrency * control on updates and deletes. */ etag?: pulumi.Input; /** * Resource ID segment making up resource `name`. It identifies the resource within its parent collection as described in https://google.aip.dev/122. */ location: pulumi.Input; /** * The ID of the project in which the resource belongs. * If it is not provided, the provider project is used. */ project?: pulumi.Input; /** * The vectors of the DataObject, keyed by the vector field name as * defined in the parent Collection's `vectorSchema`. * If a vector field is configured with a `vertexEmbeddingConfig` on * the Collection, the server will populate the vector automatically * from the corresponding text in `data` and the field should be * omitted here. * Structure is documented below. */ vectors?: pulumi.Input[] | undefined>; } //# sourceMappingURL=dataObject.d.ts.map