Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(45)

Unified Diff: generated/googleapis_beta/lib/dataproc/v1beta1.dart

Issue 2987103002: Api-Roll 52: 2017-07-31 (Closed)
Patch Set: Created 3 years, 5 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« no previous file with comments | « generated/googleapis_beta/lib/dataflow/v1b3.dart ('k') | generated/googleapis_beta/lib/dlp/v2beta1.dart » ('j') | no next file with comments »
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: generated/googleapis_beta/lib/dataproc/v1beta1.dart
diff --git a/generated/googleapis_beta/lib/dataproc/v1beta1.dart b/generated/googleapis_beta/lib/dataproc/v1beta1.dart
deleted file mode 100644
index d8b829547e65f78edcb6df43c9604eacf96341bc..0000000000000000000000000000000000000000
--- a/generated/googleapis_beta/lib/dataproc/v1beta1.dart
+++ /dev/null
@@ -1,3314 +0,0 @@
-// This is a generated file (see the discoveryapis_generator project).
-
-library googleapis_beta.dataproc.v1beta1;
-
-import 'dart:core' as core;
-import 'dart:async' as async;
-import 'dart:convert' as convert;
-
-import 'package:_discoveryapis_commons/_discoveryapis_commons.dart' as commons;
-import 'package:http/http.dart' as http;
-
-export 'package:_discoveryapis_commons/_discoveryapis_commons.dart' show
- ApiRequestError, DetailedApiRequestError;
-
-const core.String USER_AGENT = 'dart-api-client dataproc/v1beta1';
-
-/** Manages Hadoop-based clusters and jobs on Google Cloud Platform. */
-class DataprocApi {
- /** View and manage your data across Google Cloud Platform services */
- static const CloudPlatformScope = "https://www.googleapis.com/auth/cloud-platform";
-
-
- final commons.ApiRequester _requester;
-
- OperationsResourceApi get operations => new OperationsResourceApi(_requester);
- ProjectsResourceApi get projects => new ProjectsResourceApi(_requester);
-
- DataprocApi(http.Client client, {core.String rootUrl: "https://dataproc.googleapis.com/", core.String servicePath: ""}) :
- _requester = new commons.ApiRequester(client, rootUrl, servicePath, USER_AGENT);
-}
-
-
-class OperationsResourceApi {
- final commons.ApiRequester _requester;
-
- OperationsResourceApi(commons.ApiRequester client) :
- _requester = client;
-
- /**
- * Starts asynchronous cancellation on a long-running operation. The server
- * makes a best effort to cancel the operation, but success is not guaranteed.
- * If the server doesn't support this method, it returns
- * google.rpc.Code.UNIMPLEMENTED. Clients can use operations.get or other
- * methods to check whether the cancellation succeeded or whether the
- * operation completed despite cancellation.
- *
- * [request] - The metadata request object.
- *
- * Request parameters:
- *
- * [name] - The name of the operation resource to be cancelled.
- * Value must have pattern "^operations/.+$".
- *
- * Completes with a [Empty].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Empty> cancel(CancelOperationRequest request, core.String name) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (request != null) {
- _body = convert.JSON.encode((request).toJson());
- }
- if (name == null) {
- throw new core.ArgumentError("Parameter name is required.");
- }
-
- _url = 'v1beta1/' + commons.Escaper.ecapeVariableReserved('$name') + ':cancel';
-
- var _response = _requester.request(_url,
- "POST",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Empty.fromJson(data));
- }
-
- /**
- * Deletes a long-running operation. This method indicates that the client is
- * no longer interested in the operation result. It does not cancel the
- * operation. If the server doesn't support this method, it returns
- * google.rpc.Code.UNIMPLEMENTED.
- *
- * Request parameters:
- *
- * [name] - The name of the operation resource to be deleted.
- * Value must have pattern "^operations/.+$".
- *
- * Completes with a [Empty].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Empty> delete(core.String name) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (name == null) {
- throw new core.ArgumentError("Parameter name is required.");
- }
-
- _url = 'v1beta1/' + commons.Escaper.ecapeVariableReserved('$name');
-
- var _response = _requester.request(_url,
- "DELETE",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Empty.fromJson(data));
- }
-
- /**
- * Gets the latest state of a long-running operation. Clients can use this
- * method to poll the operation result at intervals as recommended by the API
- * service.
- *
- * Request parameters:
- *
- * [name] - The name of the operation resource.
- * Value must have pattern "^operations/.+$".
- *
- * Completes with a [Operation].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Operation> get(core.String name) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (name == null) {
- throw new core.ArgumentError("Parameter name is required.");
- }
-
- _url = 'v1beta1/' + commons.Escaper.ecapeVariableReserved('$name');
-
- var _response = _requester.request(_url,
- "GET",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Operation.fromJson(data));
- }
-
- /**
- * Lists operations that match the specified filter in the request. If the
- * server doesn't support this method, it returns UNIMPLEMENTED.NOTE: the name
- * binding below allows API services to override the binding to use different
- * resource name schemes, such as users / * /operations.
- *
- * Request parameters:
- *
- * [name] - The name of the operation's parent resource.
- * Value must have pattern "^operations$".
- *
- * [filter] - The standard list filter.
- *
- * [pageToken] - The standard list page token.
- *
- * [pageSize] - The standard list page size.
- *
- * Completes with a [ListOperationsResponse].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<ListOperationsResponse> list(core.String name, {core.String filter, core.String pageToken, core.int pageSize}) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (name == null) {
- throw new core.ArgumentError("Parameter name is required.");
- }
- if (filter != null) {
- _queryParams["filter"] = [filter];
- }
- if (pageToken != null) {
- _queryParams["pageToken"] = [pageToken];
- }
- if (pageSize != null) {
- _queryParams["pageSize"] = ["${pageSize}"];
- }
-
- _url = 'v1beta1/' + commons.Escaper.ecapeVariableReserved('$name');
-
- var _response = _requester.request(_url,
- "GET",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new ListOperationsResponse.fromJson(data));
- }
-
-}
-
-
-class ProjectsResourceApi {
- final commons.ApiRequester _requester;
-
- ProjectsClustersResourceApi get clusters => new ProjectsClustersResourceApi(_requester);
- ProjectsJobsResourceApi get jobs => new ProjectsJobsResourceApi(_requester);
-
- ProjectsResourceApi(commons.ApiRequester client) :
- _requester = client;
-}
-
-
-class ProjectsClustersResourceApi {
- final commons.ApiRequester _requester;
-
- ProjectsClustersResourceApi(commons.ApiRequester client) :
- _requester = client;
-
- /**
- * Creates a cluster in a project.
- *
- * [request] - The metadata request object.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * cluster belongs to.
- *
- * Completes with a [Operation].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Operation> create(Cluster request, core.String projectId) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (request != null) {
- _body = convert.JSON.encode((request).toJson());
- }
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/clusters';
-
- var _response = _requester.request(_url,
- "POST",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Operation.fromJson(data));
- }
-
- /**
- * Deletes a cluster in a project.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * cluster belongs to.
- *
- * [clusterName] - Required The cluster name.
- *
- * Completes with a [Operation].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Operation> delete(core.String projectId, core.String clusterName) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (clusterName == null) {
- throw new core.ArgumentError("Parameter clusterName is required.");
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/clusters/' + commons.Escaper.ecapeVariable('$clusterName');
-
- var _response = _requester.request(_url,
- "DELETE",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Operation.fromJson(data));
- }
-
- /**
- * Gets cluster diagnostic information. After the operation completes, the
- * Operation.response field contains DiagnoseClusterOutputLocation.
- *
- * [request] - The metadata request object.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * cluster belongs to.
- *
- * [clusterName] - Required The cluster name.
- *
- * Completes with a [Operation].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Operation> diagnose(DiagnoseClusterRequest request, core.String projectId, core.String clusterName) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (request != null) {
- _body = convert.JSON.encode((request).toJson());
- }
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (clusterName == null) {
- throw new core.ArgumentError("Parameter clusterName is required.");
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/clusters/' + commons.Escaper.ecapeVariable('$clusterName') + ':diagnose';
-
- var _response = _requester.request(_url,
- "POST",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Operation.fromJson(data));
- }
-
- /**
- * Gets the resource representation for a cluster in a project.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * cluster belongs to.
- *
- * [clusterName] - Required The cluster name.
- *
- * Completes with a [Cluster].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Cluster> get(core.String projectId, core.String clusterName) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (clusterName == null) {
- throw new core.ArgumentError("Parameter clusterName is required.");
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/clusters/' + commons.Escaper.ecapeVariable('$clusterName');
-
- var _response = _requester.request(_url,
- "GET",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Cluster.fromJson(data));
- }
-
- /**
- * Lists all clusters in a project.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * cluster belongs to.
- *
- * [pageToken] - The standard List page token.
- *
- * [pageSize] - The standard List page size.
- *
- * [filter] - Optional A filter constraining which clusters to list. Valid
- * filters contain label terms such as: labels.key1 = val1 AND (-labels.k2 =
- * val2 OR labels.k3 = val3)
- *
- * Completes with a [ListClustersResponse].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<ListClustersResponse> list(core.String projectId, {core.String pageToken, core.int pageSize, core.String filter}) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (pageToken != null) {
- _queryParams["pageToken"] = [pageToken];
- }
- if (pageSize != null) {
- _queryParams["pageSize"] = ["${pageSize}"];
- }
- if (filter != null) {
- _queryParams["filter"] = [filter];
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/clusters';
-
- var _response = _requester.request(_url,
- "GET",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new ListClustersResponse.fromJson(data));
- }
-
- /**
- * Updates a cluster in a project.
- *
- * [request] - The metadata request object.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project the
- * cluster belongs to.
- *
- * [clusterName] - Required The cluster name.
- *
- * [updateMask] - Required Specifies the path, relative to
- * <code>Cluster</code>, of the field to update. For example, to change the
- * number of workers in a cluster to 5, the <code>update_mask</code> parameter
- * would be specified as
- * <code>configuration.worker_configuration.num_instances</code>, and the
- * PATCH request body would specify the new value, as follows:
- * {
- * "configuration":{
- * "workerConfiguration":{
- * "numInstances":"5"
- * }
- * }
- * }
- * Similarly, to change the number of preemptible workers in a cluster to 5,
- * the <code>update_mask</code> parameter would be
- * <code>config.secondary_worker_config.num_instances</code>, and the PATCH
- * request body would be set as follows:
- * {
- * "config":{
- * "secondaryWorkerConfig":{
- * "numInstances":"5"
- * }
- * }
- * }
- * <strong>Note:</strong> Currently,
- * <code>config.worker_config.num_instances</code> and
- * <code>config.secondary_worker_config.num_instances</code> are the only
- * fields that can be updated.
- *
- * Completes with a [Operation].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Operation> patch(Cluster request, core.String projectId, core.String clusterName, {core.String updateMask}) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (request != null) {
- _body = convert.JSON.encode((request).toJson());
- }
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (clusterName == null) {
- throw new core.ArgumentError("Parameter clusterName is required.");
- }
- if (updateMask != null) {
- _queryParams["updateMask"] = [updateMask];
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/clusters/' + commons.Escaper.ecapeVariable('$clusterName');
-
- var _response = _requester.request(_url,
- "PATCH",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Operation.fromJson(data));
- }
-
-}
-
-
-class ProjectsJobsResourceApi {
- final commons.ApiRequester _requester;
-
- ProjectsJobsResourceApi(commons.ApiRequester client) :
- _requester = client;
-
- /**
- * Starts a job cancellation request. To access the job resource after
- * cancellation, call jobs.list or jobs.get.
- *
- * [request] - The metadata request object.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * job belongs to.
- *
- * [jobId] - Required The job ID.
- *
- * Completes with a [Job].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Job> cancel(CancelJobRequest request, core.String projectId, core.String jobId) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (request != null) {
- _body = convert.JSON.encode((request).toJson());
- }
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (jobId == null) {
- throw new core.ArgumentError("Parameter jobId is required.");
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/jobs/' + commons.Escaper.ecapeVariable('$jobId') + ':cancel';
-
- var _response = _requester.request(_url,
- "POST",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Job.fromJson(data));
- }
-
- /**
- * Deletes the job from the project. If the job is active, the delete fails,
- * and the response returns FAILED_PRECONDITION.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * job belongs to.
- *
- * [jobId] - Required The job ID.
- *
- * Completes with a [Empty].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Empty> delete(core.String projectId, core.String jobId) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (jobId == null) {
- throw new core.ArgumentError("Parameter jobId is required.");
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/jobs/' + commons.Escaper.ecapeVariable('$jobId');
-
- var _response = _requester.request(_url,
- "DELETE",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Empty.fromJson(data));
- }
-
- /**
- * Gets the resource representation for a job in a project.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * job belongs to.
- *
- * [jobId] - Required The job ID.
- *
- * Completes with a [Job].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Job> get(core.String projectId, core.String jobId) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (jobId == null) {
- throw new core.ArgumentError("Parameter jobId is required.");
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/jobs/' + commons.Escaper.ecapeVariable('$jobId');
-
- var _response = _requester.request(_url,
- "GET",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Job.fromJson(data));
- }
-
- /**
- * Lists jobs in a project.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * job belongs to.
- *
- * [clusterName] - Optional If set, the returned jobs list includes only jobs
- * that were submitted to the named cluster.
- *
- * [filter] - Optional A filter constraining which jobs to list. Valid filters
- * contain job state and label terms such as: labels.key1 = val1 AND
- * (labels.k2 = val2 OR labels.k3 = val3)
- *
- * [jobStateMatcher] - Optional Specifies enumerated categories of jobs to
- * list.
- * Possible string values are:
- * - "ALL" : A ALL.
- * - "ACTIVE" : A ACTIVE.
- * - "NON_ACTIVE" : A NON_ACTIVE.
- *
- * [pageToken] - Optional The page token, returned by a previous call, to
- * request the next page of results.
- *
- * [pageSize] - Optional The number of results to return in each response.
- *
- * Completes with a [ListJobsResponse].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<ListJobsResponse> list(core.String projectId, {core.String clusterName, core.String filter, core.String jobStateMatcher, core.String pageToken, core.int pageSize}) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (clusterName != null) {
- _queryParams["clusterName"] = [clusterName];
- }
- if (filter != null) {
- _queryParams["filter"] = [filter];
- }
- if (jobStateMatcher != null) {
- _queryParams["jobStateMatcher"] = [jobStateMatcher];
- }
- if (pageToken != null) {
- _queryParams["pageToken"] = [pageToken];
- }
- if (pageSize != null) {
- _queryParams["pageSize"] = ["${pageSize}"];
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/jobs';
-
- var _response = _requester.request(_url,
- "GET",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new ListJobsResponse.fromJson(data));
- }
-
- /**
- * Updates a job in a project.
- *
- * [request] - The metadata request object.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * job belongs to.
- *
- * [jobId] - Required The job ID.
- *
- * [updateMask] - Required Specifies the path, relative to <code>Job</code>,
- * of the field to update. For example, to update the labels of a Job the
- * <code>update_mask</code> parameter would be specified as
- * <code>labels</code>, and the PATCH request body would specify the new
- * value. <strong>Note:</strong> Currently, <code>labels</code> is the only
- * field that can be updated.
- *
- * Completes with a [Job].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Job> patch(Job request, core.String projectId, core.String jobId, {core.String updateMask}) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (request != null) {
- _body = convert.JSON.encode((request).toJson());
- }
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
- if (jobId == null) {
- throw new core.ArgumentError("Parameter jobId is required.");
- }
- if (updateMask != null) {
- _queryParams["updateMask"] = [updateMask];
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/jobs/' + commons.Escaper.ecapeVariable('$jobId');
-
- var _response = _requester.request(_url,
- "PATCH",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Job.fromJson(data));
- }
-
- /**
- * Submits a job to a cluster.
- *
- * [request] - The metadata request object.
- *
- * Request parameters:
- *
- * [projectId] - Required The ID of the Google Cloud Platform project that the
- * job belongs to.
- *
- * Completes with a [Job].
- *
- * Completes with a [commons.ApiRequestError] if the API endpoint returned an
- * error.
- *
- * If the used [http.Client] completes with an error when making a REST call,
- * this method will complete with the same error.
- */
- async.Future<Job> submit(SubmitJobRequest request, core.String projectId) {
- var _url = null;
- var _queryParams = new core.Map();
- var _uploadMedia = null;
- var _uploadOptions = null;
- var _downloadOptions = commons.DownloadOptions.Metadata;
- var _body = null;
-
- if (request != null) {
- _body = convert.JSON.encode((request).toJson());
- }
- if (projectId == null) {
- throw new core.ArgumentError("Parameter projectId is required.");
- }
-
- _url = 'v1beta1/projects/' + commons.Escaper.ecapeVariable('$projectId') + '/jobs:submit';
-
- var _response = _requester.request(_url,
- "POST",
- body: _body,
- queryParams: _queryParams,
- uploadOptions: _uploadOptions,
- uploadMedia: _uploadMedia,
- downloadOptions: _downloadOptions);
- return _response.then((data) => new Job.fromJson(data));
- }
-
-}
-
-
-
-/**
- * Specifies the type and number of accelerator cards attached to the instances
- * of an instance group (see GPUs on Compute Engine).
- */
-class AcceleratorConfiguration {
- /**
- * The number of the accelerator cards of this type exposed to this instance.
- */
- core.int acceleratorCount;
- /**
- * Full or partial URI of the accelerator type resource to expose to this
- * instance. See Google Compute Engine AcceleratorTypes(
- * /compute/docs/reference/beta/acceleratorTypes)
- */
- core.String acceleratorTypeUri;
-
- AcceleratorConfiguration();
-
- AcceleratorConfiguration.fromJson(core.Map _json) {
- if (_json.containsKey("acceleratorCount")) {
- acceleratorCount = _json["acceleratorCount"];
- }
- if (_json.containsKey("acceleratorTypeUri")) {
- acceleratorTypeUri = _json["acceleratorTypeUri"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (acceleratorCount != null) {
- _json["acceleratorCount"] = acceleratorCount;
- }
- if (acceleratorTypeUri != null) {
- _json["acceleratorTypeUri"] = acceleratorTypeUri;
- }
- return _json;
- }
-}
-
-/** A request to cancel a job. */
-class CancelJobRequest {
-
- CancelJobRequest();
-
- CancelJobRequest.fromJson(core.Map _json) {
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- return _json;
- }
-}
-
-/** The request message for Operations.CancelOperation. */
-class CancelOperationRequest {
-
- CancelOperationRequest();
-
- CancelOperationRequest.fromJson(core.Map _json) {
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- return _json;
- }
-}
-
-/**
- * Describes the identifying information, configuration, and status of a cluster
- * of Google Compute Engine instances.
- */
-class Cluster {
- /**
- * Required The cluster name. Cluster names within a project must be unique.
- * Names from deleted clusters can be reused.
- */
- core.String clusterName;
- /**
- * Output-only A cluster UUID (Unique Universal Identifier). Cloud Dataproc
- * generates this value when it creates the cluster.
- */
- core.String clusterUuid;
- /**
- * Required The cluster configuration. Note that Cloud Dataproc may set
- * default values, and values may change when clusters are updated.
- */
- ClusterConfiguration configuration;
- /**
- * Optional The labels to associate with this cluster.Label keys must be
- * between 1 and 63 characters long, and must conform to the following PCRE
- * regular expression: \p{Ll}\p{Lo}{0,62}Label values must be between 1 and 63
- * characters long, and must conform to the following PCRE regular expression:
- * \p{Ll}\p{Lo}\p{N}_-{0,63}No more than 64 labels can be associated with a
- * given cluster.
- */
- core.Map<core.String, core.String> labels;
- /** Contains cluster daemon metrics such as HDFS and YARN stats. */
- ClusterMetrics metrics;
- /**
- * Required The Google Cloud Platform project ID that the cluster belongs to.
- */
- core.String projectId;
- /** Output-only Cluster status. */
- ClusterStatus status;
- /** Output-only Previous cluster statuses. */
- core.List<ClusterStatus> statusHistory;
-
- Cluster();
-
- Cluster.fromJson(core.Map _json) {
- if (_json.containsKey("clusterName")) {
- clusterName = _json["clusterName"];
- }
- if (_json.containsKey("clusterUuid")) {
- clusterUuid = _json["clusterUuid"];
- }
- if (_json.containsKey("configuration")) {
- configuration = new ClusterConfiguration.fromJson(_json["configuration"]);
- }
- if (_json.containsKey("labels")) {
- labels = _json["labels"];
- }
- if (_json.containsKey("metrics")) {
- metrics = new ClusterMetrics.fromJson(_json["metrics"]);
- }
- if (_json.containsKey("projectId")) {
- projectId = _json["projectId"];
- }
- if (_json.containsKey("status")) {
- status = new ClusterStatus.fromJson(_json["status"]);
- }
- if (_json.containsKey("statusHistory")) {
- statusHistory = _json["statusHistory"].map((value) => new ClusterStatus.fromJson(value)).toList();
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (clusterName != null) {
- _json["clusterName"] = clusterName;
- }
- if (clusterUuid != null) {
- _json["clusterUuid"] = clusterUuid;
- }
- if (configuration != null) {
- _json["configuration"] = (configuration).toJson();
- }
- if (labels != null) {
- _json["labels"] = labels;
- }
- if (metrics != null) {
- _json["metrics"] = (metrics).toJson();
- }
- if (projectId != null) {
- _json["projectId"] = projectId;
- }
- if (status != null) {
- _json["status"] = (status).toJson();
- }
- if (statusHistory != null) {
- _json["statusHistory"] = statusHistory.map((value) => (value).toJson()).toList();
- }
- return _json;
- }
-}
-
-/** The cluster configuration. */
-class ClusterConfiguration {
- /**
- * Optional A Google Cloud Storage staging bucket used for sharing generated
- * SSH keys and configuration. If you do not specify a staging bucket, Cloud
- * Dataproc will determine an appropriate Cloud Storage location (US, ASIA, or
- * EU) for your cluster's staging bucket according to the Google Compute
- * Engine zone where your cluster is deployed, and then it will create and
- * manage this project-level, per-location bucket for you.
- */
- core.String configurationBucket;
- /**
- * Required The shared Google Compute Engine configuration settings for all
- * instances in a cluster.
- */
- GceClusterConfiguration gceClusterConfiguration;
- /**
- * Optional Commands to execute on each node after configuration is completed.
- * By default, executables are run on master and all worker nodes. You can
- * test a node's <code>role</code> metadata to run an executable on a master
- * or worker node, as shown below:
- * ROLE=$(/usr/share/google/get_metadata_value attributes/role)
- * if [[ "${ROLE}" == 'Master' ]]; then
- * ... master specific actions ...
- * else
- * ... worker specific actions ...
- * fi
- */
- core.List<NodeInitializationAction> initializationActions;
- /**
- * Optional The Google Compute Engine configuration settings for the master
- * instance in a cluster.
- */
- InstanceGroupConfiguration masterConfiguration;
- /**
- * Optional The Google Compute Engine configuration settings for additional
- * worker instances in a cluster.
- */
- InstanceGroupConfiguration secondaryWorkerConfiguration;
- /** Optional The configuration settings for software inside the cluster. */
- SoftwareConfiguration softwareConfiguration;
- /**
- * Optional The Google Compute Engine configuration settings for worker
- * instances in a cluster.
- */
- InstanceGroupConfiguration workerConfiguration;
-
- ClusterConfiguration();
-
- ClusterConfiguration.fromJson(core.Map _json) {
- if (_json.containsKey("configurationBucket")) {
- configurationBucket = _json["configurationBucket"];
- }
- if (_json.containsKey("gceClusterConfiguration")) {
- gceClusterConfiguration = new GceClusterConfiguration.fromJson(_json["gceClusterConfiguration"]);
- }
- if (_json.containsKey("initializationActions")) {
- initializationActions = _json["initializationActions"].map((value) => new NodeInitializationAction.fromJson(value)).toList();
- }
- if (_json.containsKey("masterConfiguration")) {
- masterConfiguration = new InstanceGroupConfiguration.fromJson(_json["masterConfiguration"]);
- }
- if (_json.containsKey("secondaryWorkerConfiguration")) {
- secondaryWorkerConfiguration = new InstanceGroupConfiguration.fromJson(_json["secondaryWorkerConfiguration"]);
- }
- if (_json.containsKey("softwareConfiguration")) {
- softwareConfiguration = new SoftwareConfiguration.fromJson(_json["softwareConfiguration"]);
- }
- if (_json.containsKey("workerConfiguration")) {
- workerConfiguration = new InstanceGroupConfiguration.fromJson(_json["workerConfiguration"]);
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (configurationBucket != null) {
- _json["configurationBucket"] = configurationBucket;
- }
- if (gceClusterConfiguration != null) {
- _json["gceClusterConfiguration"] = (gceClusterConfiguration).toJson();
- }
- if (initializationActions != null) {
- _json["initializationActions"] = initializationActions.map((value) => (value).toJson()).toList();
- }
- if (masterConfiguration != null) {
- _json["masterConfiguration"] = (masterConfiguration).toJson();
- }
- if (secondaryWorkerConfiguration != null) {
- _json["secondaryWorkerConfiguration"] = (secondaryWorkerConfiguration).toJson();
- }
- if (softwareConfiguration != null) {
- _json["softwareConfiguration"] = (softwareConfiguration).toJson();
- }
- if (workerConfiguration != null) {
- _json["workerConfiguration"] = (workerConfiguration).toJson();
- }
- return _json;
- }
-}
-
-/** Contains cluster daemon metrics, such as HDFS and YARN stats. */
-class ClusterMetrics {
- /** The HDFS metrics. */
- core.Map<core.String, core.String> hdfsMetrics;
- /** The YARN metrics. */
- core.Map<core.String, core.String> yarnMetrics;
-
- ClusterMetrics();
-
- ClusterMetrics.fromJson(core.Map _json) {
- if (_json.containsKey("hdfsMetrics")) {
- hdfsMetrics = _json["hdfsMetrics"];
- }
- if (_json.containsKey("yarnMetrics")) {
- yarnMetrics = _json["yarnMetrics"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (hdfsMetrics != null) {
- _json["hdfsMetrics"] = hdfsMetrics;
- }
- if (yarnMetrics != null) {
- _json["yarnMetrics"] = yarnMetrics;
- }
- return _json;
- }
-}
-
-/** Metadata describing the operation. */
-class ClusterOperationMetadata {
- /** Output-only. Name of the cluster for the operation. */
- core.String clusterName;
- /** Output-only. Cluster UUID for the operation. */
- core.String clusterUuid;
- /** Output-only. Short description of operation. */
- core.String description;
- /** Output-only. Labels associated with the operation */
- core.Map<core.String, core.String> labels;
- /** Output-only. The operation type. */
- core.String operationType;
- /** Output-only. Current operation status. */
- ClusterOperationStatus status;
- /** Output-only. The previous operation status. */
- core.List<ClusterOperationStatus> statusHistory;
- /** Output-only. Errors encountered during operation execution. */
- core.List<core.String> warnings;
-
- ClusterOperationMetadata();
-
- ClusterOperationMetadata.fromJson(core.Map _json) {
- if (_json.containsKey("clusterName")) {
- clusterName = _json["clusterName"];
- }
- if (_json.containsKey("clusterUuid")) {
- clusterUuid = _json["clusterUuid"];
- }
- if (_json.containsKey("description")) {
- description = _json["description"];
- }
- if (_json.containsKey("labels")) {
- labels = _json["labels"];
- }
- if (_json.containsKey("operationType")) {
- operationType = _json["operationType"];
- }
- if (_json.containsKey("status")) {
- status = new ClusterOperationStatus.fromJson(_json["status"]);
- }
- if (_json.containsKey("statusHistory")) {
- statusHistory = _json["statusHistory"].map((value) => new ClusterOperationStatus.fromJson(value)).toList();
- }
- if (_json.containsKey("warnings")) {
- warnings = _json["warnings"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (clusterName != null) {
- _json["clusterName"] = clusterName;
- }
- if (clusterUuid != null) {
- _json["clusterUuid"] = clusterUuid;
- }
- if (description != null) {
- _json["description"] = description;
- }
- if (labels != null) {
- _json["labels"] = labels;
- }
- if (operationType != null) {
- _json["operationType"] = operationType;
- }
- if (status != null) {
- _json["status"] = (status).toJson();
- }
- if (statusHistory != null) {
- _json["statusHistory"] = statusHistory.map((value) => (value).toJson()).toList();
- }
- if (warnings != null) {
- _json["warnings"] = warnings;
- }
- return _json;
- }
-}
-
-/** The status of the operation. */
-class ClusterOperationStatus {
- /** Output-only.A message containing any operation metadata details. */
- core.String details;
- /** Output-only. A message containing the detailed operation state. */
- core.String innerState;
- /**
- * Output-only. A message containing the operation state.
- * Possible string values are:
- * - "UNKNOWN" : Unused.
- * - "PENDING" : The operation has been created.
- * - "RUNNING" : The operation is running.
- * - "DONE" : The operation is done; either cancelled or completed.
- */
- core.String state;
- /** Output-only. The time this state was entered. */
- core.String stateStartTime;
-
- ClusterOperationStatus();
-
- ClusterOperationStatus.fromJson(core.Map _json) {
- if (_json.containsKey("details")) {
- details = _json["details"];
- }
- if (_json.containsKey("innerState")) {
- innerState = _json["innerState"];
- }
- if (_json.containsKey("state")) {
- state = _json["state"];
- }
- if (_json.containsKey("stateStartTime")) {
- stateStartTime = _json["stateStartTime"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (details != null) {
- _json["details"] = details;
- }
- if (innerState != null) {
- _json["innerState"] = innerState;
- }
- if (state != null) {
- _json["state"] = state;
- }
- if (stateStartTime != null) {
- _json["stateStartTime"] = stateStartTime;
- }
- return _json;
- }
-}
-
-/** The status of a cluster and its instances. */
-class ClusterStatus {
- /** Optional details of cluster's state. */
- core.String detail;
- /**
- * The cluster's state.
- * Possible string values are:
- * - "UNKNOWN" : The cluster state is unknown.
- * - "CREATING" : The cluster is being created and set up. It is not ready for
- * use.
- * - "RUNNING" : The cluster is currently running and healthy. It is ready for
- * use.
- * - "ERROR" : The cluster encountered an error. It is not ready for use.
- * - "DELETING" : The cluster is being deleted. It cannot be used.
- * - "UPDATING" : The cluster is being updated. It continues to accept and
- * process jobs.
- */
- core.String state;
- /** Time when this state was entered. */
- core.String stateStartTime;
- /**
- * Output-only Additional state information that includes status reported by
- * the agent.
- * Possible string values are:
- * - "UNSPECIFIED"
- * - "UNHEALTHY" : The cluster is known to be in an unhealthy state (for
- * example, critical daemons are not running or HDFS capacity is
- * exhausted).Applies to RUNNING state.
- * - "STALE_STATUS" : The agent-reported status is out of date (may occur if
- * Cloud Dataproc loses communication with Agent).Applies to RUNNING state.
- */
- core.String substate;
-
- ClusterStatus();
-
- ClusterStatus.fromJson(core.Map _json) {
- if (_json.containsKey("detail")) {
- detail = _json["detail"];
- }
- if (_json.containsKey("state")) {
- state = _json["state"];
- }
- if (_json.containsKey("stateStartTime")) {
- stateStartTime = _json["stateStartTime"];
- }
- if (_json.containsKey("substate")) {
- substate = _json["substate"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (detail != null) {
- _json["detail"] = detail;
- }
- if (state != null) {
- _json["state"] = state;
- }
- if (stateStartTime != null) {
- _json["stateStartTime"] = stateStartTime;
- }
- if (substate != null) {
- _json["substate"] = substate;
- }
- return _json;
- }
-}
-
-/** The location of diagnostic output. */
-class DiagnoseClusterOutputLocation {
- /**
- * Output-only The Google Cloud Storage URI of the diagnostic output. This is
- * a plain text file with a summary of collected diagnostics.
- */
- core.String outputUri;
-
- DiagnoseClusterOutputLocation();
-
- DiagnoseClusterOutputLocation.fromJson(core.Map _json) {
- if (_json.containsKey("outputUri")) {
- outputUri = _json["outputUri"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (outputUri != null) {
- _json["outputUri"] = outputUri;
- }
- return _json;
- }
-}
-
-/** A request to collect cluster diagnostic information. */
-class DiagnoseClusterRequest {
-
- DiagnoseClusterRequest();
-
- DiagnoseClusterRequest.fromJson(core.Map _json) {
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- return _json;
- }
-}
-
-/** The location of diagnostic output. */
-class DiagnoseClusterResults {
- /**
- * Output-only. The Google Cloud Storage URI of the diagnostic output. The
- * output report is a plain text file with a summary of collected diagnostics.
- */
- core.String outputUri;
-
- DiagnoseClusterResults();
-
- DiagnoseClusterResults.fromJson(core.Map _json) {
- if (_json.containsKey("outputUri")) {
- outputUri = _json["outputUri"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (outputUri != null) {
- _json["outputUri"] = outputUri;
- }
- return _json;
- }
-}
-
-/** Specifies the configuration of disk options for a group of VM instances. */
-class DiskConfiguration {
- /** Optional Size in GB of the boot disk (default is 500GB). */
- core.int bootDiskSizeGb;
- /**
- * Optional Number of attached SSDs, from 0 to 4 (default is 0). If SSDs are
- * not attached, the boot disk is used to store runtime logs and HDFS data. If
- * one or more SSDs are attached, this runtime bulk data is spread across
- * them, and the boot disk contains only basic configuration and installed
- * binaries.
- */
- core.int numLocalSsds;
-
- DiskConfiguration();
-
- DiskConfiguration.fromJson(core.Map _json) {
- if (_json.containsKey("bootDiskSizeGb")) {
- bootDiskSizeGb = _json["bootDiskSizeGb"];
- }
- if (_json.containsKey("numLocalSsds")) {
- numLocalSsds = _json["numLocalSsds"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (bootDiskSizeGb != null) {
- _json["bootDiskSizeGb"] = bootDiskSizeGb;
- }
- if (numLocalSsds != null) {
- _json["numLocalSsds"] = numLocalSsds;
- }
- return _json;
- }
-}
-
-/**
- * A generic empty message that you can re-use to avoid defining duplicated
- * empty messages in your APIs. A typical example is to use it as the request or
- * the response type of an API method. For instance:
- * service Foo {
- * rpc Bar(google.protobuf.Empty) returns (google.protobuf.Empty);
- * }
- * The JSON representation for Empty is empty JSON object {}.
- */
-class Empty {
-
- Empty();
-
- Empty.fromJson(core.Map _json) {
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- return _json;
- }
-}
-
-/**
- * Common configuration settings for resources of Google Compute Engine cluster
- * instances, applicable to all instances in the cluster.
- */
-class GceClusterConfiguration {
- /**
- * If true, all instances in the cluser will only have internal IP addresses.
- * By default, clusters are not restricted to internal IP addresses, and will
- * have ephemeral external IP addresses assigned to each instance. This
- * restriction can only be enabled for subnetwork enabled networks, and all
- * off-cluster dependencies must be configured to be accessible without
- * external IP addresses.
- */
- core.bool internalIpOnly;
- /** The Google Compute Engine metadata entries to add to all instances. */
- core.Map<core.String, core.String> metadata;
- /**
- * The Google Compute Engine network to be used for machine communications.
- * Cannot be specified with subnetwork_uri. If neither network_uri nor
- * subnetwork_uri is specified, the "default" network of the project is used,
- * if it exists. Cannot be a "Custom Subnet Network" (see
- * https://cloud.google.com/compute/docs/subnetworks for more information).
- * Example:
- * https://www.googleapis.com/compute/v1/projects/[project_id]/regions/global/default.
- */
- core.String networkUri;
- /**
- * Optional The service account of the instances. Defaults to the default
- * Google Compute Engine service account. Custom service accounts need
- * permissions equivalent to the folloing IAM roles:
- * roles/logging.logWriter
- * roles/storage.objectAdmin(see
- * https://cloud.google.com/compute/docs/access/service-accounts#custom_service_accounts
- * for more information). Example:
- * [account_id]@[project_id].iam.gserviceaccount.com
- */
- core.String serviceAccount;
- /**
- * The URIs of service account scopes to be included in Google Compute Engine
- * instances. The following base set of scopes is always included: -
- * https://www.googleapis.com/auth/cloud.useraccounts.readonly -
- * https://www.googleapis.com/auth/devstorage.read_write -
- * https://www.googleapis.com/auth/logging.write If no scopes are specfied,
- * the following defaults are also provided: -
- * https://www.googleapis.com/auth/bigquery -
- * https://www.googleapis.com/auth/bigtable.admin.table -
- * https://www.googleapis.com/auth/bigtable.data -
- * https://www.googleapis.com/auth/devstorage.full_control
- */
- core.List<core.String> serviceAccountScopes;
- /**
- * The Google Compute Engine subnetwork to be used for machine communications.
- * Cannot be specified with network_uri. Example:
- * https://www.googleapis.com/compute/v1/projects/[project_id]/regions/us-east1/sub0.
- */
- core.String subnetworkUri;
- /** The Google Compute Engine tags to add to all instances. */
- core.List<core.String> tags;
- /**
- * Required The zone where the Google Compute Engine cluster will be located.
- * Example:
- * https://www.googleapis.com/compute/v1/projects/[project_id]/zones/[zone].
- */
- core.String zoneUri;
-
- GceClusterConfiguration();
-
- GceClusterConfiguration.fromJson(core.Map _json) {
- if (_json.containsKey("internalIpOnly")) {
- internalIpOnly = _json["internalIpOnly"];
- }
- if (_json.containsKey("metadata")) {
- metadata = _json["metadata"];
- }
- if (_json.containsKey("networkUri")) {
- networkUri = _json["networkUri"];
- }
- if (_json.containsKey("serviceAccount")) {
- serviceAccount = _json["serviceAccount"];
- }
- if (_json.containsKey("serviceAccountScopes")) {
- serviceAccountScopes = _json["serviceAccountScopes"];
- }
- if (_json.containsKey("subnetworkUri")) {
- subnetworkUri = _json["subnetworkUri"];
- }
- if (_json.containsKey("tags")) {
- tags = _json["tags"];
- }
- if (_json.containsKey("zoneUri")) {
- zoneUri = _json["zoneUri"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (internalIpOnly != null) {
- _json["internalIpOnly"] = internalIpOnly;
- }
- if (metadata != null) {
- _json["metadata"] = metadata;
- }
- if (networkUri != null) {
- _json["networkUri"] = networkUri;
- }
- if (serviceAccount != null) {
- _json["serviceAccount"] = serviceAccount;
- }
- if (serviceAccountScopes != null) {
- _json["serviceAccountScopes"] = serviceAccountScopes;
- }
- if (subnetworkUri != null) {
- _json["subnetworkUri"] = subnetworkUri;
- }
- if (tags != null) {
- _json["tags"] = tags;
- }
- if (zoneUri != null) {
- _json["zoneUri"] = zoneUri;
- }
- return _json;
- }
-}
-
-/** A Cloud Dataproc job for running Hadoop MapReduce jobs on YARN. */
-class HadoopJob {
- /**
- * Optional HCFS URIs of archives to be extracted in the working directory of
- * Hadoop drivers and tasks. Supported file types: .jar, .tar, .tar.gz, .tgz,
- * or .zip.
- */
- core.List<core.String> archiveUris;
- /**
- * Optional The arguments to pass to the driver. Do not include arguments,
- * such as -libjars or -Dfoo=bar, that can be set as job properties, since a
- * collision may occur that causes an incorrect job submission.
- */
- core.List<core.String> args;
- /**
- * Optional HCFS URIs of files to be copied to the working directory of Hadoop
- * drivers and distributed tasks. Useful for naively parallel tasks.
- */
- core.List<core.String> fileUris;
- /**
- * Optional Jar file URIs to add to the CLASSPATHs of the Hadoop driver and
- * tasks.
- */
- core.List<core.String> jarFileUris;
- /** Optional The runtime log configuration for job execution. */
- LoggingConfiguration loggingConfiguration;
- /**
- * The name of the driver's main class. The jar file containing the class must
- * be in the default CLASSPATH or specified in jar_file_uris.
- */
- core.String mainClass;
- /**
- * The Hadoop Compatible Filesystem (HCFS) URI of the jar file containing the
- * main class. Examples:
- * 'gs://foo-bucket/analytics-binaries/extract-useful-metrics-mr.jar'
- * 'hdfs:/tmp/test-samples/custom-wordcount.jar'
- * 'file:///home/usr/lib/hadoop-mapreduce/hadoop-mapreduce-examples.jar'
- */
- core.String mainJarFileUri;
- /**
- * Optional A mapping of property names to values, used to configure Hadoop.
- * Properties that conflict with values set by the Cloud Dataproc API may be
- * overwritten. Can include properties set in /etc/hadoop/conf / * -site and
- * classes in user code.
- */
- core.Map<core.String, core.String> properties;
-
- HadoopJob();
-
- HadoopJob.fromJson(core.Map _json) {
- if (_json.containsKey("archiveUris")) {
- archiveUris = _json["archiveUris"];
- }
- if (_json.containsKey("args")) {
- args = _json["args"];
- }
- if (_json.containsKey("fileUris")) {
- fileUris = _json["fileUris"];
- }
- if (_json.containsKey("jarFileUris")) {
- jarFileUris = _json["jarFileUris"];
- }
- if (_json.containsKey("loggingConfiguration")) {
- loggingConfiguration = new LoggingConfiguration.fromJson(_json["loggingConfiguration"]);
- }
- if (_json.containsKey("mainClass")) {
- mainClass = _json["mainClass"];
- }
- if (_json.containsKey("mainJarFileUri")) {
- mainJarFileUri = _json["mainJarFileUri"];
- }
- if (_json.containsKey("properties")) {
- properties = _json["properties"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (archiveUris != null) {
- _json["archiveUris"] = archiveUris;
- }
- if (args != null) {
- _json["args"] = args;
- }
- if (fileUris != null) {
- _json["fileUris"] = fileUris;
- }
- if (jarFileUris != null) {
- _json["jarFileUris"] = jarFileUris;
- }
- if (loggingConfiguration != null) {
- _json["loggingConfiguration"] = (loggingConfiguration).toJson();
- }
- if (mainClass != null) {
- _json["mainClass"] = mainClass;
- }
- if (mainJarFileUri != null) {
- _json["mainJarFileUri"] = mainJarFileUri;
- }
- if (properties != null) {
- _json["properties"] = properties;
- }
- return _json;
- }
-}
-
-/** A Cloud Dataproc job for running Hive queries on YARN. */
-class HiveJob {
- /**
- * Optional Whether to continue executing queries if a query fails. The
- * default value is false. Setting to true can be useful when executing
- * independent parallel queries.
- */
- core.bool continueOnFailure;
- /**
- * Optional HCFS URIs of jar files to add to the CLASSPATH of the Hive server
- * and Hadoop MapReduce (MR) tasks. Can contain Hive SerDes and UDFs.
- */
- core.List<core.String> jarFileUris;
- /**
- * Optional A mapping of property names and values, used to configure Hive.
- * Properties that conflict with values set by the Cloud Dataproc API may be
- * overwritten. Can include properties set in /etc/hadoop/conf / * -site.xml,
- * /etc/hive/conf/hive-site.xml, and classes in user code.
- */
- core.Map<core.String, core.String> properties;
- /** The HCFS URI of the script that contains Hive queries. */
- core.String queryFileUri;
- /** A list of queries. */
- QueryList queryList;
- /**
- * Optional Mapping of query variable names to values (equivalent to the Hive
- * command: SET name="value";).
- */
- core.Map<core.String, core.String> scriptVariables;
-
- HiveJob();
-
- HiveJob.fromJson(core.Map _json) {
- if (_json.containsKey("continueOnFailure")) {
- continueOnFailure = _json["continueOnFailure"];
- }
- if (_json.containsKey("jarFileUris")) {
- jarFileUris = _json["jarFileUris"];
- }
- if (_json.containsKey("properties")) {
- properties = _json["properties"];
- }
- if (_json.containsKey("queryFileUri")) {
- queryFileUri = _json["queryFileUri"];
- }
- if (_json.containsKey("queryList")) {
- queryList = new QueryList.fromJson(_json["queryList"]);
- }
- if (_json.containsKey("scriptVariables")) {
- scriptVariables = _json["scriptVariables"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (continueOnFailure != null) {
- _json["continueOnFailure"] = continueOnFailure;
- }
- if (jarFileUris != null) {
- _json["jarFileUris"] = jarFileUris;
- }
- if (properties != null) {
- _json["properties"] = properties;
- }
- if (queryFileUri != null) {
- _json["queryFileUri"] = queryFileUri;
- }
- if (queryList != null) {
- _json["queryList"] = (queryList).toJson();
- }
- if (scriptVariables != null) {
- _json["scriptVariables"] = scriptVariables;
- }
- return _json;
- }
-}
-
-/**
- * The configuration settings for Google Compute Engine resources in an instance
- * group, such as a master or worker group.
- */
-class InstanceGroupConfiguration {
- /**
- * Optional The Google Compute Engine accelerator configuration for these
- * instances.
- */
- core.List<AcceleratorConfiguration> accelerators;
- /** Disk option configuration settings. */
- DiskConfiguration diskConfiguration;
- /**
- * Output-only The Google Compute Engine image resource used for cluster
- * instances. Inferred from SoftwareConfiguration.image_version.
- */
- core.String imageUri;
- /**
- * The list of instance names. Dataproc derives the names from cluster_name,
- * num_instances, and the instance group if not set by user (recommended
- * practice is to let Dataproc derive the name).
- */
- core.List<core.String> instanceNames;
- /** Specifies that this instance group contains Preemptible Instances. */
- core.bool isPreemptible;
- /**
- * The Google Compute Engine machine type used for cluster instances. Example:
- * https://www.googleapis.com/compute/v1/projects/[project_id]/zones/us-east1-a/machineTypes/n1-standard-2.
- */
- core.String machineTypeUri;
- /**
- * Output-only The configuration for Google Compute Engine Instance Group
- * Manager that manages this group. This is only used for preemptible instance
- * groups.
- */
- ManagedGroupConfiguration managedGroupConfiguration;
- /**
- * The number of VM instances in the instance group. For master instance
- * groups, must be set to 1.
- */
- core.int numInstances;
-
- InstanceGroupConfiguration();
-
- InstanceGroupConfiguration.fromJson(core.Map _json) {
- if (_json.containsKey("accelerators")) {
- accelerators = _json["accelerators"].map((value) => new AcceleratorConfiguration.fromJson(value)).toList();
- }
- if (_json.containsKey("diskConfiguration")) {
- diskConfiguration = new DiskConfiguration.fromJson(_json["diskConfiguration"]);
- }
- if (_json.containsKey("imageUri")) {
- imageUri = _json["imageUri"];
- }
- if (_json.containsKey("instanceNames")) {
- instanceNames = _json["instanceNames"];
- }
- if (_json.containsKey("isPreemptible")) {
- isPreemptible = _json["isPreemptible"];
- }
- if (_json.containsKey("machineTypeUri")) {
- machineTypeUri = _json["machineTypeUri"];
- }
- if (_json.containsKey("managedGroupConfiguration")) {
- managedGroupConfiguration = new ManagedGroupConfiguration.fromJson(_json["managedGroupConfiguration"]);
- }
- if (_json.containsKey("numInstances")) {
- numInstances = _json["numInstances"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (accelerators != null) {
- _json["accelerators"] = accelerators.map((value) => (value).toJson()).toList();
- }
- if (diskConfiguration != null) {
- _json["diskConfiguration"] = (diskConfiguration).toJson();
- }
- if (imageUri != null) {
- _json["imageUri"] = imageUri;
- }
- if (instanceNames != null) {
- _json["instanceNames"] = instanceNames;
- }
- if (isPreemptible != null) {
- _json["isPreemptible"] = isPreemptible;
- }
- if (machineTypeUri != null) {
- _json["machineTypeUri"] = machineTypeUri;
- }
- if (managedGroupConfiguration != null) {
- _json["managedGroupConfiguration"] = (managedGroupConfiguration).toJson();
- }
- if (numInstances != null) {
- _json["numInstances"] = numInstances;
- }
- return _json;
- }
-}
-
-/** A Cloud Dataproc job resource. */
-class Job {
- /**
- * Output-only If present, the location of miscellaneous control files which
- * may be used as part of job setup and handling. If not present, control
- * files may be placed in the same location as driver_output_uri.
- */
- core.String driverControlFilesUri;
- /**
- * Output-only A URI pointing to the location of the stdin of the job's driver
- * program, only set if the job is interactive.
- */
- core.String driverInputResourceUri;
- /**
- * Output-only A URI pointing to the location of the stdout of the job's
- * driver program.
- */
- core.String driverOutputResourceUri;
- /** Job is a Hadoop job. */
- HadoopJob hadoopJob;
- /** Job is a Hive job. */
- HiveJob hiveJob;
- /**
- * Optional If set to true, the driver's stdin will be kept open and
- * driver_input_uri will be set to provide a path at which additional input
- * can be sent to the driver.
- */
- core.bool interactive;
- /**
- * Optional The labels to associate with this job.Label keys must be between 1
- * and 63 characters long, and must conform to the following regular
- * expression: \p{Ll}\p{Lo}{0,62}Label values must be between 1 and 63
- * characters long, and must conform to the following regular expression:
- * \p{Ll}\p{Lo}\p{N}_-{0,63}No more than 64 labels can be associated with a
- * given job.
- */
- core.Map<core.String, core.String> labels;
- /** Job is a Pig job. */
- PigJob pigJob;
- /**
- * Required Job information, including how, when, and where to run the job.
- */
- JobPlacement placement;
- /** Job is a Pyspark job. */
- PySparkJob pysparkJob;
- /**
- * Optional The fully qualified reference to the job, which can be used to
- * obtain the equivalent REST path of the job resource. If this property is
- * not specified when a job is created, the server generates a
- * <code>job_id</code>.
- */
- JobReference reference;
- /** Optional Job scheduling configuration. */
- JobScheduling scheduling;
- /** Job is a Spark job. */
- SparkJob sparkJob;
- /** Job is a SparkSql job. */
- SparkSqlJob sparkSqlJob;
- /**
- * Output-only The job status. Additional application-specific status
- * information may be contained in the <code>type_job</code> and
- * <code>yarn_applications</code> fields.
- */
- JobStatus status;
- /** Output-only The previous job status. */
- core.List<JobStatus> statusHistory;
- /**
- * Output-only The email address of the user submitting the job. For jobs
- * submitted on the cluster, the address is <code>username@hostname</code>.
- */
- core.String submittedBy;
- /** Output-only The collection of YARN applications spun up by this job. */
- core.List<YarnApplication> yarnApplications;
-
- Job();
-
- Job.fromJson(core.Map _json) {
- if (_json.containsKey("driverControlFilesUri")) {
- driverControlFilesUri = _json["driverControlFilesUri"];
- }
- if (_json.containsKey("driverInputResourceUri")) {
- driverInputResourceUri = _json["driverInputResourceUri"];
- }
- if (_json.containsKey("driverOutputResourceUri")) {
- driverOutputResourceUri = _json["driverOutputResourceUri"];
- }
- if (_json.containsKey("hadoopJob")) {
- hadoopJob = new HadoopJob.fromJson(_json["hadoopJob"]);
- }
- if (_json.containsKey("hiveJob")) {
- hiveJob = new HiveJob.fromJson(_json["hiveJob"]);
- }
- if (_json.containsKey("interactive")) {
- interactive = _json["interactive"];
- }
- if (_json.containsKey("labels")) {
- labels = _json["labels"];
- }
- if (_json.containsKey("pigJob")) {
- pigJob = new PigJob.fromJson(_json["pigJob"]);
- }
- if (_json.containsKey("placement")) {
- placement = new JobPlacement.fromJson(_json["placement"]);
- }
- if (_json.containsKey("pysparkJob")) {
- pysparkJob = new PySparkJob.fromJson(_json["pysparkJob"]);
- }
- if (_json.containsKey("reference")) {
- reference = new JobReference.fromJson(_json["reference"]);
- }
- if (_json.containsKey("scheduling")) {
- scheduling = new JobScheduling.fromJson(_json["scheduling"]);
- }
- if (_json.containsKey("sparkJob")) {
- sparkJob = new SparkJob.fromJson(_json["sparkJob"]);
- }
- if (_json.containsKey("sparkSqlJob")) {
- sparkSqlJob = new SparkSqlJob.fromJson(_json["sparkSqlJob"]);
- }
- if (_json.containsKey("status")) {
- status = new JobStatus.fromJson(_json["status"]);
- }
- if (_json.containsKey("statusHistory")) {
- statusHistory = _json["statusHistory"].map((value) => new JobStatus.fromJson(value)).toList();
- }
- if (_json.containsKey("submittedBy")) {
- submittedBy = _json["submittedBy"];
- }
- if (_json.containsKey("yarnApplications")) {
- yarnApplications = _json["yarnApplications"].map((value) => new YarnApplication.fromJson(value)).toList();
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (driverControlFilesUri != null) {
- _json["driverControlFilesUri"] = driverControlFilesUri;
- }
- if (driverInputResourceUri != null) {
- _json["driverInputResourceUri"] = driverInputResourceUri;
- }
- if (driverOutputResourceUri != null) {
- _json["driverOutputResourceUri"] = driverOutputResourceUri;
- }
- if (hadoopJob != null) {
- _json["hadoopJob"] = (hadoopJob).toJson();
- }
- if (hiveJob != null) {
- _json["hiveJob"] = (hiveJob).toJson();
- }
- if (interactive != null) {
- _json["interactive"] = interactive;
- }
- if (labels != null) {
- _json["labels"] = labels;
- }
- if (pigJob != null) {
- _json["pigJob"] = (pigJob).toJson();
- }
- if (placement != null) {
- _json["placement"] = (placement).toJson();
- }
- if (pysparkJob != null) {
- _json["pysparkJob"] = (pysparkJob).toJson();
- }
- if (reference != null) {
- _json["reference"] = (reference).toJson();
- }
- if (scheduling != null) {
- _json["scheduling"] = (scheduling).toJson();
- }
- if (sparkJob != null) {
- _json["sparkJob"] = (sparkJob).toJson();
- }
- if (sparkSqlJob != null) {
- _json["sparkSqlJob"] = (sparkSqlJob).toJson();
- }
- if (status != null) {
- _json["status"] = (status).toJson();
- }
- if (statusHistory != null) {
- _json["statusHistory"] = statusHistory.map((value) => (value).toJson()).toList();
- }
- if (submittedBy != null) {
- _json["submittedBy"] = submittedBy;
- }
- if (yarnApplications != null) {
- _json["yarnApplications"] = yarnApplications.map((value) => (value).toJson()).toList();
- }
- return _json;
- }
-}
-
-/** Cloud Dataproc job configuration. */
-class JobPlacement {
- /** Required The name of the cluster where the job will be submitted. */
- core.String clusterName;
- /**
- * Output-only A cluster UUID generated by the Dataproc service when the job
- * is submitted.
- */
- core.String clusterUuid;
-
- JobPlacement();
-
- JobPlacement.fromJson(core.Map _json) {
- if (_json.containsKey("clusterName")) {
- clusterName = _json["clusterName"];
- }
- if (_json.containsKey("clusterUuid")) {
- clusterUuid = _json["clusterUuid"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (clusterName != null) {
- _json["clusterName"] = clusterName;
- }
- if (clusterUuid != null) {
- _json["clusterUuid"] = clusterUuid;
- }
- return _json;
- }
-}
-
-/** Encapsulates the full scoping used to reference a job. */
-class JobReference {
- /**
- * Required The job ID, which must be unique within the project. The job ID is
- * generated by the server upon job submission or provided by the user as a
- * means to perform retries without creating duplicate jobs. The ID must
- * contain only letters (a-z, A-Z), numbers (0-9), underscores (_), or hyphens
- * (-). The maximum length is 100 characters.
- */
- core.String jobId;
- /**
- * Required The ID of the Google Cloud Platform project that the job belongs
- * to.
- */
- core.String projectId;
-
- JobReference();
-
- JobReference.fromJson(core.Map _json) {
- if (_json.containsKey("jobId")) {
- jobId = _json["jobId"];
- }
- if (_json.containsKey("projectId")) {
- projectId = _json["projectId"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (jobId != null) {
- _json["jobId"] = jobId;
- }
- if (projectId != null) {
- _json["projectId"] = projectId;
- }
- return _json;
- }
-}
-
-/**
- * Job scheduling options.Beta Feature: These options are available for testing
- * purposes only. They may be changed before final release.
- */
-class JobScheduling {
- /**
- * Optional Maximum number of times per hour a driver may be restarted as a
- * result of driver terminating with non-zero code before job is reported
- * failed.A job may be reported as thrashing if driver exits with non-zero
- * code 4 times within 10 minute window.Maximum value is 10.
- */
- core.int maxFailuresPerHour;
-
- JobScheduling();
-
- JobScheduling.fromJson(core.Map _json) {
- if (_json.containsKey("maxFailuresPerHour")) {
- maxFailuresPerHour = _json["maxFailuresPerHour"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (maxFailuresPerHour != null) {
- _json["maxFailuresPerHour"] = maxFailuresPerHour;
- }
- return _json;
- }
-}
-
-/** Cloud Dataproc job status. */
-class JobStatus {
- /**
- * Optional Job state details, such as an error description if the state is
- * <code>ERROR</code>.
- */
- core.String details;
- /**
- * Required A state message specifying the overall job state.
- * Possible string values are:
- * - "STATE_UNSPECIFIED" : The job state is unknown.
- * - "PENDING" : The job is pending; it has been submitted, but is not yet
- * running.
- * - "SETUP_DONE" : Job has been received by the service and completed initial
- * setup; it will shortly be submitted to the cluster.
- * - "RUNNING" : The job is running on the cluster.
- * - "CANCEL_PENDING" : A CancelJob request has been received, but is pending.
- * - "CANCEL_STARTED" : Transient in-flight resources have been canceled, and
- * the request to cancel the running job has been issued to the cluster.
- * - "CANCELLED" : The job cancelation was successful.
- * - "DONE" : The job has completed successfully.
- * - "ERROR" : The job has completed, but encountered an error.
- * - "ATTEMPT_FAILURE" : Job attempt has failed. The detail field contains
- * failure details for this attempt.Applies to restartable jobs only.
- */
- core.String state;
- /** Output-only The time when this state was entered. */
- core.String stateStartTime;
- /**
- * Output-only Additional state information, which includes status reported by
- * the agent.
- * Possible string values are:
- * - "UNSPECIFIED"
- * - "SUBMITTED" : The Job is submitted to the agent.Applies to RUNNING state.
- * - "QUEUED" : The Job has been received and is awaiting execution (it may be
- * waiting for a condition to be met). See the "details" field for the reason
- * for the delay.Applies to RUNNING state.
- * - "STALE_STATUS" : The agent-reported status is out of date, which may be
- * caused by a loss of communication between the agent and Cloud Dataproc. If
- * the agent does not send a timely update, the job will fail.Applies to
- * RUNNING state.
- */
- core.String substate;
-
- JobStatus();
-
- JobStatus.fromJson(core.Map _json) {
- if (_json.containsKey("details")) {
- details = _json["details"];
- }
- if (_json.containsKey("state")) {
- state = _json["state"];
- }
- if (_json.containsKey("stateStartTime")) {
- stateStartTime = _json["stateStartTime"];
- }
- if (_json.containsKey("substate")) {
- substate = _json["substate"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (details != null) {
- _json["details"] = details;
- }
- if (state != null) {
- _json["state"] = state;
- }
- if (stateStartTime != null) {
- _json["stateStartTime"] = stateStartTime;
- }
- if (substate != null) {
- _json["substate"] = substate;
- }
- return _json;
- }
-}
-
-/** The list of all clusters in a project. */
-class ListClustersResponse {
- /** Output-only The clusters in the project. */
- core.List<Cluster> clusters;
- /** The standard List next-page token. */
- core.String nextPageToken;
-
- ListClustersResponse();
-
- ListClustersResponse.fromJson(core.Map _json) {
- if (_json.containsKey("clusters")) {
- clusters = _json["clusters"].map((value) => new Cluster.fromJson(value)).toList();
- }
- if (_json.containsKey("nextPageToken")) {
- nextPageToken = _json["nextPageToken"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (clusters != null) {
- _json["clusters"] = clusters.map((value) => (value).toJson()).toList();
- }
- if (nextPageToken != null) {
- _json["nextPageToken"] = nextPageToken;
- }
- return _json;
- }
-}
-
-/** A list of jobs in a project. */
-class ListJobsResponse {
- /** Output-only Jobs list. */
- core.List<Job> jobs;
- /**
- * Optional This token is included in the response if there are more results
- * to fetch. To fetch additional results, provide this value as the page_token
- * in a subsequent <code>ListJobsRequest</code>.
- */
- core.String nextPageToken;
-
- ListJobsResponse();
-
- ListJobsResponse.fromJson(core.Map _json) {
- if (_json.containsKey("jobs")) {
- jobs = _json["jobs"].map((value) => new Job.fromJson(value)).toList();
- }
- if (_json.containsKey("nextPageToken")) {
- nextPageToken = _json["nextPageToken"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (jobs != null) {
- _json["jobs"] = jobs.map((value) => (value).toJson()).toList();
- }
- if (nextPageToken != null) {
- _json["nextPageToken"] = nextPageToken;
- }
- return _json;
- }
-}
-
-/** The response message for Operations.ListOperations. */
-class ListOperationsResponse {
- /** The standard List next-page token. */
- core.String nextPageToken;
- /** A list of operations that matches the specified filter in the request. */
- core.List<Operation> operations;
-
- ListOperationsResponse();
-
- ListOperationsResponse.fromJson(core.Map _json) {
- if (_json.containsKey("nextPageToken")) {
- nextPageToken = _json["nextPageToken"];
- }
- if (_json.containsKey("operations")) {
- operations = _json["operations"].map((value) => new Operation.fromJson(value)).toList();
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (nextPageToken != null) {
- _json["nextPageToken"] = nextPageToken;
- }
- if (operations != null) {
- _json["operations"] = operations.map((value) => (value).toJson()).toList();
- }
- return _json;
- }
-}
-
-/** The runtime logging configuration of the job. */
-class LoggingConfiguration {
- /**
- * The per-package log levels for the driver. This may include "root" package
- * name to configure rootLogger. Examples: 'com.google = FATAL', 'root =
- * INFO', 'org.apache = DEBUG'
- */
- core.Map<core.String, core.String> driverLogLevels;
-
- LoggingConfiguration();
-
- LoggingConfiguration.fromJson(core.Map _json) {
- if (_json.containsKey("driverLogLevels")) {
- driverLogLevels = _json["driverLogLevels"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (driverLogLevels != null) {
- _json["driverLogLevels"] = driverLogLevels;
- }
- return _json;
- }
-}
-
-/** Specifies the resources used to actively manage an instance group. */
-class ManagedGroupConfiguration {
- /** Output-only The name of the Instance Group Manager for this group. */
- core.String instanceGroupManagerName;
- /**
- * Output-only The name of the Instance Template used for the Managed Instance
- * Group.
- */
- core.String instanceTemplateName;
-
- ManagedGroupConfiguration();
-
- ManagedGroupConfiguration.fromJson(core.Map _json) {
- if (_json.containsKey("instanceGroupManagerName")) {
- instanceGroupManagerName = _json["instanceGroupManagerName"];
- }
- if (_json.containsKey("instanceTemplateName")) {
- instanceTemplateName = _json["instanceTemplateName"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (instanceGroupManagerName != null) {
- _json["instanceGroupManagerName"] = instanceGroupManagerName;
- }
- if (instanceTemplateName != null) {
- _json["instanceTemplateName"] = instanceTemplateName;
- }
- return _json;
- }
-}
-
-/**
- * Specifies an executable to run on a fully configured node and a timeout
- * period for executable completion.
- */
-class NodeInitializationAction {
- /** Required Google Cloud Storage URI of executable file. */
- core.String executableFile;
- /**
- * Optional Amount of time executable has to complete. Default is 10 minutes.
- * Cluster creation fails with an explanatory error message (the name of the
- * executable that caused the error and the exceeded timeout period) if the
- * executable is not completed at end of the timeout period.
- */
- core.String executionTimeout;
-
- NodeInitializationAction();
-
- NodeInitializationAction.fromJson(core.Map _json) {
- if (_json.containsKey("executableFile")) {
- executableFile = _json["executableFile"];
- }
- if (_json.containsKey("executionTimeout")) {
- executionTimeout = _json["executionTimeout"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (executableFile != null) {
- _json["executableFile"] = executableFile;
- }
- if (executionTimeout != null) {
- _json["executionTimeout"] = executionTimeout;
- }
- return _json;
- }
-}
-
-/**
- * This resource represents a long-running operation that is the result of a
- * network API call.
- */
-class Operation {
- /**
- * If the value is false, it means the operation is still in progress. If
- * true, the operation is completed, and either error or response is
- * available.
- */
- core.bool done;
- /** The error result of the operation in case of failure or cancellation. */
- Status error;
- /**
- * Service-specific metadata associated with the operation. It typically
- * contains progress information and common metadata such as create time. Some
- * services might not provide such metadata. Any method that returns a
- * long-running operation should document the metadata type, if any.
- *
- * The values for Object must be JSON objects. It can consist of `num`,
- * `String`, `bool` and `null` as well as `Map` and `List` values.
- */
- core.Map<core.String, core.Object> metadata;
- /**
- * The server-assigned name, which is only unique within the same service that
- * originally returns it. If you use the default HTTP mapping, the name should
- * have the format of operations/some/unique/name.
- */
- core.String name;
- /**
- * The normal response of the operation in case of success. If the original
- * method returns no data on success, such as Delete, the response is
- * google.protobuf.Empty. If the original method is standard
- * Get/Create/Update, the response should be the resource. For other methods,
- * the response should have the type XxxResponse, where Xxx is the original
- * method name. For example, if the original method name is TakeSnapshot(),
- * the inferred response type is TakeSnapshotResponse.
- *
- * The values for Object must be JSON objects. It can consist of `num`,
- * `String`, `bool` and `null` as well as `Map` and `List` values.
- */
- core.Map<core.String, core.Object> response;
-
- Operation();
-
- Operation.fromJson(core.Map _json) {
- if (_json.containsKey("done")) {
- done = _json["done"];
- }
- if (_json.containsKey("error")) {
- error = new Status.fromJson(_json["error"]);
- }
- if (_json.containsKey("metadata")) {
- metadata = _json["metadata"];
- }
- if (_json.containsKey("name")) {
- name = _json["name"];
- }
- if (_json.containsKey("response")) {
- response = _json["response"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (done != null) {
- _json["done"] = done;
- }
- if (error != null) {
- _json["error"] = (error).toJson();
- }
- if (metadata != null) {
- _json["metadata"] = metadata;
- }
- if (name != null) {
- _json["name"] = name;
- }
- if (response != null) {
- _json["response"] = response;
- }
- return _json;
- }
-}
-
-/** Metadata describing the operation. */
-class OperationMetadata {
- /** Name of the cluster for the operation. */
- core.String clusterName;
- /** Cluster UUId for the operation. */
- core.String clusterUuid;
- /** Output-only Short description of operation. */
- core.String description;
- /** Output-only The operation type. */
- core.String operationType;
- /** Output-only Current operation status. */
- OperationStatus status;
- /** Output-only Previous operation status. */
- core.List<OperationStatus> statusHistory;
- /** Output-only Errors encountered during operation execution. */
- core.List<core.String> warnings;
-
- OperationMetadata();
-
- OperationMetadata.fromJson(core.Map _json) {
- if (_json.containsKey("clusterName")) {
- clusterName = _json["clusterName"];
- }
- if (_json.containsKey("clusterUuid")) {
- clusterUuid = _json["clusterUuid"];
- }
- if (_json.containsKey("description")) {
- description = _json["description"];
- }
- if (_json.containsKey("operationType")) {
- operationType = _json["operationType"];
- }
- if (_json.containsKey("status")) {
- status = new OperationStatus.fromJson(_json["status"]);
- }
- if (_json.containsKey("statusHistory")) {
- statusHistory = _json["statusHistory"].map((value) => new OperationStatus.fromJson(value)).toList();
- }
- if (_json.containsKey("warnings")) {
- warnings = _json["warnings"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (clusterName != null) {
- _json["clusterName"] = clusterName;
- }
- if (clusterUuid != null) {
- _json["clusterUuid"] = clusterUuid;
- }
- if (description != null) {
- _json["description"] = description;
- }
- if (operationType != null) {
- _json["operationType"] = operationType;
- }
- if (status != null) {
- _json["status"] = (status).toJson();
- }
- if (statusHistory != null) {
- _json["statusHistory"] = statusHistory.map((value) => (value).toJson()).toList();
- }
- if (warnings != null) {
- _json["warnings"] = warnings;
- }
- return _json;
- }
-}
-
-/** The status of the operation. */
-class OperationStatus {
- /** A message containing any operation metadata details. */
- core.String details;
- /** A message containing the detailed operation state. */
- core.String innerState;
- /**
- * A message containing the operation state.
- * Possible string values are:
- * - "UNKNOWN" : Unused.
- * - "PENDING" : The operation has been created.
- * - "RUNNING" : The operation is running.
- * - "DONE" : The operation is done; either cancelled or completed.
- */
- core.String state;
- /** The time this state was entered. */
- core.String stateStartTime;
-
- OperationStatus();
-
- OperationStatus.fromJson(core.Map _json) {
- if (_json.containsKey("details")) {
- details = _json["details"];
- }
- if (_json.containsKey("innerState")) {
- innerState = _json["innerState"];
- }
- if (_json.containsKey("state")) {
- state = _json["state"];
- }
- if (_json.containsKey("stateStartTime")) {
- stateStartTime = _json["stateStartTime"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (details != null) {
- _json["details"] = details;
- }
- if (innerState != null) {
- _json["innerState"] = innerState;
- }
- if (state != null) {
- _json["state"] = state;
- }
- if (stateStartTime != null) {
- _json["stateStartTime"] = stateStartTime;
- }
- return _json;
- }
-}
-
-/** A Cloud Dataproc job for running Pig queries on YARN. */
-class PigJob {
- /**
- * Optional Whether to continue executing queries if a query fails. The
- * default value is false. Setting to true can be useful when executing
- * independent parallel queries.
- */
- core.bool continueOnFailure;
- /**
- * Optional HCFS URIs of jar files to add to the CLASSPATH of the Pig Client
- * and Hadoop MapReduce (MR) tasks. Can contain Pig UDFs.
- */
- core.List<core.String> jarFileUris;
- /** Optional The runtime log configuration for job execution. */
- LoggingConfiguration loggingConfiguration;
- /**
- * Optional A mapping of property names to values, used to configure Pig.
- * Properties that conflict with values set by the Cloud Dataproc API may be
- * overwritten. Can include properties set in /etc/hadoop/conf / * -site.xml,
- * /etc/pig/conf/pig.properties, and classes in user code.
- */
- core.Map<core.String, core.String> properties;
- /** The HCFS URI of the script that contains the Pig queries. */
- core.String queryFileUri;
- /** A list of queries. */
- QueryList queryList;
- /**
- * Optional Mapping of query variable names to values (equivalent to the Pig
- * command: name=[value]).
- */
- core.Map<core.String, core.String> scriptVariables;
-
- PigJob();
-
- PigJob.fromJson(core.Map _json) {
- if (_json.containsKey("continueOnFailure")) {
- continueOnFailure = _json["continueOnFailure"];
- }
- if (_json.containsKey("jarFileUris")) {
- jarFileUris = _json["jarFileUris"];
- }
- if (_json.containsKey("loggingConfiguration")) {
- loggingConfiguration = new LoggingConfiguration.fromJson(_json["loggingConfiguration"]);
- }
- if (_json.containsKey("properties")) {
- properties = _json["properties"];
- }
- if (_json.containsKey("queryFileUri")) {
- queryFileUri = _json["queryFileUri"];
- }
- if (_json.containsKey("queryList")) {
- queryList = new QueryList.fromJson(_json["queryList"]);
- }
- if (_json.containsKey("scriptVariables")) {
- scriptVariables = _json["scriptVariables"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (continueOnFailure != null) {
- _json["continueOnFailure"] = continueOnFailure;
- }
- if (jarFileUris != null) {
- _json["jarFileUris"] = jarFileUris;
- }
- if (loggingConfiguration != null) {
- _json["loggingConfiguration"] = (loggingConfiguration).toJson();
- }
- if (properties != null) {
- _json["properties"] = properties;
- }
- if (queryFileUri != null) {
- _json["queryFileUri"] = queryFileUri;
- }
- if (queryList != null) {
- _json["queryList"] = (queryList).toJson();
- }
- if (scriptVariables != null) {
- _json["scriptVariables"] = scriptVariables;
- }
- return _json;
- }
-}
-
-/** A Cloud Dataproc job for running PySpark applications on YARN. */
-class PySparkJob {
- /**
- * Optional HCFS URIs of archives to be extracted in the working directory of
- * .jar, .tar, .tar.gz, .tgz, and .zip.
- */
- core.List<core.String> archiveUris;
- /**
- * Optional The arguments to pass to the driver. Do not include arguments,
- * such as --conf, that can be set as job properties, since a collision may
- * occur that causes an incorrect job submission.
- */
- core.List<core.String> args;
- /**
- * Optional HCFS URIs of files to be copied to the working directory of Python
- * drivers and distributed tasks. Useful for naively parallel tasks.
- */
- core.List<core.String> fileUris;
- /**
- * Optional HCFS URIs of jar files to add to the CLASSPATHs of the Python
- * driver and tasks.
- */
- core.List<core.String> jarFileUris;
- /** Optional The runtime log configuration for job execution. */
- LoggingConfiguration loggingConfiguration;
- /**
- * Required The Hadoop Compatible Filesystem (HCFS) URI of the main Python
- * file to use as the driver. Must be a .py file.
- */
- core.String mainPythonFileUri;
- /**
- * Optional A mapping of property names to values, used to configure PySpark.
- * Properties that conflict with values set by the Cloud Dataproc API may be
- * overwritten. Can include properties set in
- * /etc/spark/conf/spark-defaults.conf and classes in user code.
- */
- core.Map<core.String, core.String> properties;
- /**
- * Optional HCFS file URIs of Python files to pass to the PySpark framework.
- * Supported file types: .py, .egg, and .zip.
- */
- core.List<core.String> pythonFileUris;
-
- PySparkJob();
-
- PySparkJob.fromJson(core.Map _json) {
- if (_json.containsKey("archiveUris")) {
- archiveUris = _json["archiveUris"];
- }
- if (_json.containsKey("args")) {
- args = _json["args"];
- }
- if (_json.containsKey("fileUris")) {
- fileUris = _json["fileUris"];
- }
- if (_json.containsKey("jarFileUris")) {
- jarFileUris = _json["jarFileUris"];
- }
- if (_json.containsKey("loggingConfiguration")) {
- loggingConfiguration = new LoggingConfiguration.fromJson(_json["loggingConfiguration"]);
- }
- if (_json.containsKey("mainPythonFileUri")) {
- mainPythonFileUri = _json["mainPythonFileUri"];
- }
- if (_json.containsKey("properties")) {
- properties = _json["properties"];
- }
- if (_json.containsKey("pythonFileUris")) {
- pythonFileUris = _json["pythonFileUris"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (archiveUris != null) {
- _json["archiveUris"] = archiveUris;
- }
- if (args != null) {
- _json["args"] = args;
- }
- if (fileUris != null) {
- _json["fileUris"] = fileUris;
- }
- if (jarFileUris != null) {
- _json["jarFileUris"] = jarFileUris;
- }
- if (loggingConfiguration != null) {
- _json["loggingConfiguration"] = (loggingConfiguration).toJson();
- }
- if (mainPythonFileUri != null) {
- _json["mainPythonFileUri"] = mainPythonFileUri;
- }
- if (properties != null) {
- _json["properties"] = properties;
- }
- if (pythonFileUris != null) {
- _json["pythonFileUris"] = pythonFileUris;
- }
- return _json;
- }
-}
-
-/** A list of queries to run on a cluster. */
-class QueryList {
- /**
- * Required The queries to execute. You do not need to terminate a query with
- * a semicolon. Multiple queries can be specified in one string by separating
- * each with a semicolon. Here is an example of an Cloud Dataproc API snippet
- * that uses a QueryList to specify a HiveJob:
- * "hiveJob": {
- * "queryList": {
- * "queries": [
- * "query1",
- * "query2",
- * "query3;query4",
- * ]
- * }
- * }
- */
- core.List<core.String> queries;
-
- QueryList();
-
- QueryList.fromJson(core.Map _json) {
- if (_json.containsKey("queries")) {
- queries = _json["queries"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (queries != null) {
- _json["queries"] = queries;
- }
- return _json;
- }
-}
-
-/**
- * Specifies the selection and configuration of software inside the cluster.
- */
-class SoftwareConfiguration {
- /**
- * Optional The version of software inside the cluster. It must match the
- * regular expression [0-9]+\.[0-9]+. If unspecified, it defaults to the
- * latest version (see Cloud Dataproc Versioning).
- */
- core.String imageVersion;
- /**
- * Optional The properties to set on daemon configuration files.Property keys
- * are specified in "prefix:property" format, such as "core:fs.defaultFS". The
- * following are supported prefixes and their mappings: core - core-site.xml
- * hdfs - hdfs-site.xml mapred - mapred-site.xml yarn - yarn-site.xml hive
- * - hive-site.xml pig - pig.properties spark - spark-defaults.conf
- */
- core.Map<core.String, core.String> properties;
-
- SoftwareConfiguration();
-
- SoftwareConfiguration.fromJson(core.Map _json) {
- if (_json.containsKey("imageVersion")) {
- imageVersion = _json["imageVersion"];
- }
- if (_json.containsKey("properties")) {
- properties = _json["properties"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (imageVersion != null) {
- _json["imageVersion"] = imageVersion;
- }
- if (properties != null) {
- _json["properties"] = properties;
- }
- return _json;
- }
-}
-
-/** A Cloud Dataproc job for running Spark applications on YARN. */
-class SparkJob {
- /**
- * Optional HCFS URIs of archives to be extracted in the working directory of
- * Spark drivers and tasks. Supported file types: .jar, .tar, .tar.gz, .tgz,
- * and .zip.
- */
- core.List<core.String> archiveUris;
- /**
- * Optional The arguments to pass to the driver. Do not include arguments,
- * such as --conf, that can be set as job properties, since a collision may
- * occur that causes an incorrect job submission.
- */
- core.List<core.String> args;
- /**
- * Optional HCFS URIs of files to be copied to the working directory of Spark
- * drivers and distributed tasks. Useful for naively parallel tasks.
- */
- core.List<core.String> fileUris;
- /**
- * Optional HCFS URIs of jar files to add to the CLASSPATHs of the Spark
- * driver and tasks.
- */
- core.List<core.String> jarFileUris;
- /** Optional The runtime log configuration for job execution. */
- LoggingConfiguration loggingConfiguration;
- /**
- * The name of the driver's main class. The jar file that contains the class
- * must be in the default CLASSPATH or specified in jar_file_uris.
- */
- core.String mainClass;
- /**
- * The Hadoop Compatible Filesystem (HCFS) URI of the jar file that contains
- * the main class.
- */
- core.String mainJarFileUri;
- /**
- * Optional A mapping of property names to values, used to configure Spark.
- * Properties that conflict with values set by the Cloud Dataproc API may be
- * overwritten. Can include properties set in
- * /etc/spark/conf/spark-defaults.conf and classes in user code.
- */
- core.Map<core.String, core.String> properties;
-
- SparkJob();
-
- SparkJob.fromJson(core.Map _json) {
- if (_json.containsKey("archiveUris")) {
- archiveUris = _json["archiveUris"];
- }
- if (_json.containsKey("args")) {
- args = _json["args"];
- }
- if (_json.containsKey("fileUris")) {
- fileUris = _json["fileUris"];
- }
- if (_json.containsKey("jarFileUris")) {
- jarFileUris = _json["jarFileUris"];
- }
- if (_json.containsKey("loggingConfiguration")) {
- loggingConfiguration = new LoggingConfiguration.fromJson(_json["loggingConfiguration"]);
- }
- if (_json.containsKey("mainClass")) {
- mainClass = _json["mainClass"];
- }
- if (_json.containsKey("mainJarFileUri")) {
- mainJarFileUri = _json["mainJarFileUri"];
- }
- if (_json.containsKey("properties")) {
- properties = _json["properties"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (archiveUris != null) {
- _json["archiveUris"] = archiveUris;
- }
- if (args != null) {
- _json["args"] = args;
- }
- if (fileUris != null) {
- _json["fileUris"] = fileUris;
- }
- if (jarFileUris != null) {
- _json["jarFileUris"] = jarFileUris;
- }
- if (loggingConfiguration != null) {
- _json["loggingConfiguration"] = (loggingConfiguration).toJson();
- }
- if (mainClass != null) {
- _json["mainClass"] = mainClass;
- }
- if (mainJarFileUri != null) {
- _json["mainJarFileUri"] = mainJarFileUri;
- }
- if (properties != null) {
- _json["properties"] = properties;
- }
- return _json;
- }
-}
-
-/** A Cloud Dataproc job for running Spark SQL queries. */
-class SparkSqlJob {
- /** Optional HCFS URIs of jar files to be added to the Spark CLASSPATH. */
- core.List<core.String> jarFileUris;
- /** Optional The runtime log configuration for job execution. */
- LoggingConfiguration loggingConfiguration;
- /**
- * Optional A mapping of property names to values, used to configure Spark
- * SQL's SparkConf. Properties that conflict with values set by the Cloud
- * Dataproc API may be overwritten.
- */
- core.Map<core.String, core.String> properties;
- /** The HCFS URI of the script that contains SQL queries. */
- core.String queryFileUri;
- /** A list of queries. */
- QueryList queryList;
- /**
- * Optional Mapping of query variable names to values (equivalent to the Spark
- * SQL command: SET name="value";).
- */
- core.Map<core.String, core.String> scriptVariables;
-
- SparkSqlJob();
-
- SparkSqlJob.fromJson(core.Map _json) {
- if (_json.containsKey("jarFileUris")) {
- jarFileUris = _json["jarFileUris"];
- }
- if (_json.containsKey("loggingConfiguration")) {
- loggingConfiguration = new LoggingConfiguration.fromJson(_json["loggingConfiguration"]);
- }
- if (_json.containsKey("properties")) {
- properties = _json["properties"];
- }
- if (_json.containsKey("queryFileUri")) {
- queryFileUri = _json["queryFileUri"];
- }
- if (_json.containsKey("queryList")) {
- queryList = new QueryList.fromJson(_json["queryList"]);
- }
- if (_json.containsKey("scriptVariables")) {
- scriptVariables = _json["scriptVariables"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (jarFileUris != null) {
- _json["jarFileUris"] = jarFileUris;
- }
- if (loggingConfiguration != null) {
- _json["loggingConfiguration"] = (loggingConfiguration).toJson();
- }
- if (properties != null) {
- _json["properties"] = properties;
- }
- if (queryFileUri != null) {
- _json["queryFileUri"] = queryFileUri;
- }
- if (queryList != null) {
- _json["queryList"] = (queryList).toJson();
- }
- if (scriptVariables != null) {
- _json["scriptVariables"] = scriptVariables;
- }
- return _json;
- }
-}
-
-/**
- * The Status type defines a logical error model that is suitable for different
- * programming environments, including REST APIs and RPC APIs. It is used by
- * gRPC (https://github.com/grpc). The error model is designed to be:
- * Simple to use and understand for most users
- * Flexible enough to meet unexpected needsOverviewThe Status message contains
- * three pieces of data: error code, error message, and error details. The error
- * code should be an enum value of google.rpc.Code, but it may accept additional
- * error codes if needed. The error message should be a developer-facing English
- * message that helps developers understand and resolve the error. If a
- * localized user-facing error message is needed, put the localized message in
- * the error details or localize it in the client. The optional error details
- * may contain arbitrary information about the error. There is a predefined set
- * of error detail types in the package google.rpc that can be used for common
- * error conditions.Language mappingThe Status message is the logical
- * representation of the error model, but it is not necessarily the actual wire
- * format. When the Status message is exposed in different client libraries and
- * different wire protocols, it can be mapped differently. For example, it will
- * likely be mapped to some exceptions in Java, but more likely mapped to some
- * error codes in C.Other usesThe error model and the Status message can be used
- * in a variety of environments, either with or without APIs, to provide a
- * consistent developer experience across different environments.Example uses of
- * this error model include:
- * Partial errors. If a service needs to return partial errors to the client, it
- * may embed the Status in the normal response to indicate the partial errors.
- * Workflow errors. A typical workflow has multiple steps. Each step may have a
- * Status message for error reporting.
- * Batch operations. If a client uses batch request and batch response, the
- * Status message should be used directly inside batch response, one for each
- * error sub-response.
- * Asynchronous operations. If an API call embeds asynchronous operation results
- * in its response, the status of those operations should be represented
- * directly using the Status message.
- * Logging. If some API errors are stored in logs, the message Status could be
- * used directly after any stripping needed for security/privacy reasons.
- */
-class Status {
- /** The status code, which should be an enum value of google.rpc.Code. */
- core.int code;
- /**
- * A list of messages that carry the error details. There will be a common set
- * of message types for APIs to use.
- *
- * The values for Object must be JSON objects. It can consist of `num`,
- * `String`, `bool` and `null` as well as `Map` and `List` values.
- */
- core.List<core.Map<core.String, core.Object>> details;
- /**
- * A developer-facing error message, which should be in English. Any
- * user-facing error message should be localized and sent in the
- * google.rpc.Status.details field, or localized by the client.
- */
- core.String message;
-
- Status();
-
- Status.fromJson(core.Map _json) {
- if (_json.containsKey("code")) {
- code = _json["code"];
- }
- if (_json.containsKey("details")) {
- details = _json["details"];
- }
- if (_json.containsKey("message")) {
- message = _json["message"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (code != null) {
- _json["code"] = code;
- }
- if (details != null) {
- _json["details"] = details;
- }
- if (message != null) {
- _json["message"] = message;
- }
- return _json;
- }
-}
-
-/** A request to submit a job. */
-class SubmitJobRequest {
- /** Required The job resource. */
- Job job;
-
- SubmitJobRequest();
-
- SubmitJobRequest.fromJson(core.Map _json) {
- if (_json.containsKey("job")) {
- job = new Job.fromJson(_json["job"]);
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (job != null) {
- _json["job"] = (job).toJson();
- }
- return _json;
- }
-}
-
-/**
- * A YARN application created by a job. Application information is a subset of
- * <code>org.apache.hadoop.yarn.proto.YarnProtos.ApplicationReportProto</code>.
- */
-class YarnApplication {
- /** Required The application name. */
- core.String name;
- /** Required The numerical progress of the application, from 1 to 100. */
- core.double progress;
- /**
- * Required The application state.
- * Possible string values are:
- * - "STATE_UNSPECIFIED" : Status is unspecified.
- * - "NEW" : Status is NEW.
- * - "NEW_SAVING" : Status is NEW_SAVING.
- * - "SUBMITTED" : Status is SUBMITTED.
- * - "ACCEPTED" : Status is ACCEPTED.
- * - "RUNNING" : Status is RUNNING.
- * - "FINISHED" : Status is FINISHED.
- * - "FAILED" : Status is FAILED.
- * - "KILLED" : Status is KILLED.
- */
- core.String state;
- /**
- * Optional The HTTP URL of the ApplicationMaster, HistoryServer, or
- * TimelineServer that provides application-specific information. The URL uses
- * the internal hostname, and requires a proxy server for resolution and,
- * possibly, access.
- */
- core.String trackingUrl;
-
- YarnApplication();
-
- YarnApplication.fromJson(core.Map _json) {
- if (_json.containsKey("name")) {
- name = _json["name"];
- }
- if (_json.containsKey("progress")) {
- progress = _json["progress"];
- }
- if (_json.containsKey("state")) {
- state = _json["state"];
- }
- if (_json.containsKey("trackingUrl")) {
- trackingUrl = _json["trackingUrl"];
- }
- }
-
- core.Map<core.String, core.Object> toJson() {
- final core.Map<core.String, core.Object> _json = new core.Map<core.String, core.Object>();
- if (name != null) {
- _json["name"] = name;
- }
- if (progress != null) {
- _json["progress"] = progress;
- }
- if (state != null) {
- _json["state"] = state;
- }
- if (trackingUrl != null) {
- _json["trackingUrl"] = trackingUrl;
- }
- return _json;
- }
-}
« no previous file with comments | « generated/googleapis_beta/lib/dataflow/v1b3.dart ('k') | generated/googleapis_beta/lib/dlp/v2beta1.dart » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698