This library provides some of the most common component development functionality in a simple, reusable way.
To install, type
npm install @elastic.io/component-commons-library
A number of Platform API Client classes are available to use and extend them to create Clients for Platform API.
PlatformApiRestClient class extends BasicAuthRestClient class. The method inside this class checks for the status code 200, if not, then throws an error. And also checks that the response came with the correct data in the JSON format and the other expected response headers.
- emitter - EIO emitting context.
- cfg - configuration of EIO component object.
const Client = new PlatformApiRestClient(emitter, cfg);
PlatformApiLogicClient class extends PlatformApiRestClient class. Contains useful methods to manipulate flow's state to set it either to active running or to inactive stopped, searching flows, workspaces, credentials and more.
- emitter - EIO emitting context.
- cfg - configuration of EIO component object.
const Client = new PlatformApiLogicClient(emitter, cfg);
- fetchAllFlowsForWorkspace(options) - Fetch all flows for a given workspace
- fetchAllCredentialsForWorkspace(options) - Fetch all credentials for a given workspace
- fetchAllSecretsForWorkspace - Fetch all secrets for a given workspace
- fetchSecretById - Fetch secret by id for a given workspace
- refreshTokenBySecretId - Refresh token by secret id for a given workspace
- fetchComponentsAccessibleFromContract(options) - Fetch All Components Accessible From a Given Workspace
- splitParallelization(maxParallelization, splitFactor) - Helping method to calculate right number of parallel calls
- fetchFlowList(options) - Fetches a list of flows
- fetchWorkspaceList(options) - Fetch a list of all workspaces across all contracts for a user
- fetchWorkspaceId(workspaceUniqueCriteria) - Given a set of unique criteria, find the workspace that matches
- removeNonWritableProperties(flow, includeDataSamples) - Given a flow, remove the properties of the flow that are regularly changed by the system such as last executed time
- fetchFlowId(flowUniqueCriteria) - Fetch flow by it's unique criteria
- fetchFlowById(id) - Fetch flow by it's id
- fetchFlowByNameAndWorkspaceId(flowName, workspaceId) - Fetch flow by flow name and workspace id
- changeFlowState(options) - Given a flow, change the flow to a given state (running, stopped, etc)
- startFlow(flowId, options) - sets the flow to active running state
- stopFlow(flowId, options) - sets the flow to inactive stopped state
- hydrateFlow(options) - Hydrates the flow using removeNonWritableProperties method, but additionally enriches the flow with all data samples, credential names, command and component Id fields.
Contains tools for JSON metadata generation
convertJsonSchemaToEioSchema(keyToReturn, completeSchemaOriginal)- converts general JSON schema to platform-friendly JSON metadatamakeSchemaInline(json, availableSchemas)- replaces $ref recursively with full object description for provided json object using availableSchemas schemas map.
Contains functions to transform platform data that contains JSONata expressions
jsonataTransform(msg, cfg)- returns themsg.bodywith the JSONata expressions evaluated to values
The attachment processor function can be used to store attachments on the platform. It exposes the following functions
uploadAttachment(getAttachment, retryOptions, contentType), which will upload an attachment to the platformMaesterstorage and return the result object. Where:
getAttachment- async function which returns stream
retryOptions- (optional): parameters for retrying of upload attachment, if request failed. Parameters areretriesCountandrequestTimeout(ms)contentType- (optional):Content-Typeof attachment. By defaultContent-Typewill be calculated automatically.getAttachment(url, contentType), which will retrieve an attachment fromstewardormaesterstorage. To specify the storage - query parameterstorage_typemust be provided. To get items frommaesterstorage -?storage_type=maestershould added to theurlargument. By default attachments are retrieved fromstewardstorage, so?storage_type=stewardis not obligated to be added to theurlargument.contentType- one of [stream,arraybuffer]
Example:
const { AttachmentProcessor } = require('@elastic.io/component-commons-library');
const getAttachAsStream = async () => (
await axios.get('http://sample.pdf', { responseType: 'stream' })
).data;
const result = await new AttachmentProcessor().uploadAttachment(getAttachAsStream, 'application/pdf');
const { objectId } = result.data;const { AttachmentProcessor } = require('@elastic.io/component-commons-library');
const result = await new AttachmentProcessor().getAttachment('http://example.com', 'stream'); // steward storage
const result = await new AttachmentProcessor().getAttachment('http://example.com?storage_type=steward', 'arraybuffer'); // steward storage
const result = await new AttachmentProcessor().getAttachment('http://example.com?storage_type=maester', 'stream'); // maester storage- API_RETRIES_COUNT (defaults to 3): maximum amount of retries for 5xx errors. If server still responding 5xx, error will be thrown.
- API_REQUEST_TIMEOUT (defaults to 15000, min: 500, max: 120000): specifies the number of milliseconds before the request times out. If the request takes longer than timeout, the request will be aborted.
axiosReqWithRetryOnServerError(use with.call()to pass context, implement it as a method of class withloggerandcfg(value of configuration object for current action) values in a constructor) - function which makes axios request by specified request-config, making logging and error handling:- If 5xx error occurred, it will be retried maximum
API_RETRIES_COUNTtimes, each retry will be delayed withexponentialSleepfunction. - If 4xx error occurred - error will be thrown.
- If action
cfghasdoNotThrow404set to true: 404 error won't be treated as error.
Look on examples below.
- If 5xx error occurred, it will be retried maximum
getErrMsg- forms error message from axios-response.getRetryOptions- return valid values for envsAPI_RETRIES_COUNTandAPI_REQUEST_TIMEOUT. If values are higher or lower the limit - they'll be overwritten by default values.sleep- return promise which resolves after N time.exponentialDelay- returns number of milliseconds depending to current retry. See exponential backoff to explanation.exponentialSleep- return promise which resolves after N time. Where N is number of milliseconds fromexponentialDelayexecution.
Example for axiosReqWithRetryOnServerError function:
class Client {
private logger: any;
private cfg: any;
constructor(emitter, cfg) {
this.logger = emitter.logger;
this.cfg = cfg;
}
public async apiRequest(options: AxiosRequestConfig): Promise<any> {
try {
const response = await axiosReq.axiosReqWithRetryOnServerError(this, requestOptions);
return response.data;
} catch (error) {
if (error.response?.status === 401) {
// update token
}
throw error;
}
}
public async getUserById(id) {
return this.apiRequest({
url: `/users/${id}`,
method: 'GET',
});
}
}The built in logger uses Bunyan Logger as its base implementation. The available logger methods can be found here.
Example:
const { Logger } = require('@elastic.io/component-commons-library');
const logger = Logger.getLogger();
logger.info('Hello, world');The getLogger() method takes an optional parameter loggerName that lets you declare multiple different loggers.
Apache-2.0 © elastic.io GmbH
