This is a javascript client library for interacting with the Vbrick Rev API. It should work in node.js 16+, browsers, and deno.
This library is intended for use with VBrick Rev.
The compiled library is at dist/rev-client.js
.
npm install @vbrick/rev-client
fetch
, AbortController
, etc.require
) and ES Module (import
) versions. By default it uses the (node-fetch
) polyfill unless you import using the @vbrick/rev-client/native-fetch
named export (which uses native fetch
).--allow-net
permissions at minimum.By default CORS (Cross Origin Resource Sharing) is disabled for Rev Cloud tenants. This means this library will not work out of the box in the browser. You should open a ticket with Vbrick Support if this feature isn't yet configured.
To verify if CORS is enabled for your account check the headers response from https://YOUR_REV_TENANT_URL/api/v2/accounts/branding-settings
- it doesn't require authentication.
If you need to use this library behind a proxy (and you're using Node.js) then the proxy method to use depends on which fetch
library you're using.
node-fetch
polyfill (@vbrick/rev-client
or @vbrick/rev-client/node-fetch
) then you should use a proxy that sets the https.Agent
(like global-agent
).undici
) then you'll need a proxy implementation that sets the undici global agent dispatcherutils.setPolyfills()
On-Prem note: this library targets the latest version of Rev. Some API endpoints may not be available or work as expected in older versions of Rev On-Prem.
import {RevClient} from '/path/to/rev-client.js';
// create client object
const rev = new RevClient({
url: 'https://my.rev.url',
apiKey: 'my.user.apikey',
secret: 'my.user.secret',
// or can login via username + password
// username: 'my.username',
// password: 'my.password',
logEnabled: true, // turn on debug logging
keepAlive: true // automatically extend session
rateLimits: true // automatically enforce rate limiting (avoid 429 error responses)
});
(async () => {
// call login api and start session. will throw error if invalid login
await rev.connect();
// create a category
const {categoryId} = await rev.category.create({ name: 'Created Via API' });
// get details about this category
const categoryDetails = await rev.category.details(categoryId);
console.log('category details: ', categoryDetails);
// get the account media contributor role
const mediaContributorRole = await rev.admin.getRoleByName('Media Contributor');
// create a new user, with the media contributor role
const userId = await rev.user.create({
username: 'new.user',
firstname: 'new',
lastname: 'user',
roleIds: [mediaContributorRole.id]
});
// upload a video, and assign 'new.user' as the owner of that video, and add to the category created above
// if browser instead of node.js - pass in a File object instead of the filepath
const videoId = await rev.upload.video("/path/to/local/video.mp4", {
uploader: 'new.user',
title: 'video uploaded via the API',
categories: [categoryDetails.name], // ['Created Via API']
unlisted: true,
isActive: true
/// ...any additional metadata
});
console.log('Video uploaded!', videoId);
await rev.disconnect();
})();
NOTE Unless otherwise noted all methods of RevClient
are async (they return a Promise
)
new RevClient(options)
url
: REQUIRED - URL of Rev instance (ex https://company.rev.vbrick.com
)
keepAlive
: true
/false
(Default: true
) - If true then automatically extend the Rev session at regular intervals (until rev.disconnect()
is called). If false then you must manually call extendSession()
to maintain a valid token.
rateLimits
: true
/false
/RateLimitOptions
(Default: false
) - Automatically throttle requests client-side to fit within Vbrick's API Request Rate Limits. Note that the default values (when value is true
) is set to the account maximum - see below Rate Limits
section for how to customize.
logEnabled
: true
/false
(Default: false
) - Enable/disable debug logging
log
: (logLevel, ...args) => void
- Custom log function. Default is to log to console
And one of following login options (apiKey
+secret
, username
+password
, oauthConfig
+code
+codeVerifier
, jwtToken
, guestRegistrationToken
+webcastId
, publicOnly
):
User API Key:
apiKey
: User API Key of Rev Usersecret
: User Secret of Rev UserUsername/Password login:
username
: Username of Rev user.password
: Password of Rev user.Legacy OAuth session (NOTE: This is for OAuth 2.0 browser-redirect flow to create sessions, it's not intended for server-side only login).
oauthConfig
: OAuth configuration objectoauthConfig.oauthApiKey
: API key from Rev Admin -> Security. This is a DIFFERENT value from the User Token used for API login/extend sessionoauthConfig.oauthSecret
: API secret from Rev Admin -> Security. This is a DIFFERENT value from a user's secret
. DEPRECATED - only for Legacy OAuth loginoauthConfig.redirectUri
: The local URL Rev should redirect user to after logging in. This must match EXACTLY what's specified in Rev Admin -> Security for the specified API keyauthCode
: the Auth Code returned from the OAuth redirect response as a query parameteroauthConfig
: OAuth configuration objectoauthConfig.oauthApiKey
: API key from Rev Admin -> Security. This is a DIFFERENT value from the User Token used for API login/extend sessionoauthConfig.redirectUri
: The local URL Rev should redirect user to after logging in. This must match EXACTLY what's specified in Rev Admin -> Security for the specified API keycode
: the Code returned from the OAuth redirect response as a query parametercodeVerifier
: the Code Verifier used when initially generating the OAuth2 authorization URL. use rev.auth.buildOAuth2Authentication()
to generate an OAuth2 authorization URL and corresponding codeVerifier.JWT session:
jwtToken
: The JWT TokenGuest Registration session:
guestRegistrationToken
: The Token returned when creating a guest registration.webcastId
: ID of the webcast in questionAccess Token (existing sessions)
session.token
: The Access Token previously received via some login method (see below)session.expiration
: The expiration time of the session.Public Only usage (no authentication)
publicOnly
: Don't use any authentication. This limits use to endpoints that don't require authentication.You can pass in an existing session
to the constructor to reuse a session token
(assuming it isn't expired). When you include session
the additional credential values aren't necessary, however if not included you won't be able to re-login, just extend the session.
const initialRev = new RevClient({ url, apiKey, secret });
await initialRev.connect();
// store state for use elasewhere (like /tmp/ storage in a serverless environment)
// has { token, expiration }
let savedState = rev.sessionState;
// ... time passes ...
const revWithReusedSession = new RevClient({ url, apiKey, sessionState: savedState })
// or set after initial configuration
revWithReusedSession.sessionState = savedState;
connect()
- Login to Rev using configured credentialsdisconnect()
- Logoff from Rev and clear tokenextendSession()
- Extend current session tokenverifySession()
- Returns true/false on if the current token is validisConnected
- Boolean value that indicates if connect()
has been called and the session isn't expiredtoken
- The Authorization token used to make API callslogEnabled
- Boolean value to enable/disable debug logging.sessionExpires
- Date when token is set to expiresessionState
- Current token/expiration session dataMake a request to a specific API endpoint. get
,post
,put
,patch
and delete
are helpers to set the method
value of the main request call
request(method, endpoint, data?, options?)
- Make a HTTP Requestmethod
- HTTP Verb to useendpoint
- Path for call to make. This will be relative to the url
set in RevClient
data
- Query parameters for get/delete
requests, Body for put/post/patch
requests.options
- Additional request parameters to pass to fetch
.options.responseType
- json|text|blob|stream
- specify how to decode the response. Default is to autodetect based on the response. stream
means pass through without decoding.statusCode
- HTTP Status Codeheaders
- Headers objectbody
- Response payload, already decoded based on options.responseType
response
- raw fetch
Response objectRevError
Custom Error with the following additional properties:
status
- HTTP Status Codeurl
- Request URLcode?
- Rev-specific error codedetail?
- Rev-specific detail message;rev = new RevClient({ url: "https://my.rev.url", ...credentials });
await rev.request('get', '/api/v2/users/my.username', { type: 'username' });
// HTTP GET https://my.rev.url/api/v2/users/my.username?type=username
get(endpoint, data?, options?)
post(endpoint, data?, options?)
put(endpoint, data?, options?)
patch(endpoint, data?, options?)
delete(endpoint, data?, options?)
Make HTTP requests with the specified verb (get/post/patch/etc). Unlike request()
these calls return the body directly.
endpoint
- Path for call to make. This will be relative to the url
set in RevClient
data
- Query parameters for get/delete
requests, Body for put/post/patch
requests.options
- Additional request parameters to pass to fetch
.options.responseType
- json|text|blob|stream
- specify how to decode the response. Default is to autodetect based on the response. stream
means pass through without decoding.The Response payload, already decoded based on options.responseType
See the Vbrick documentation for information on Rate Limit behavior.
If you have multiple users / agents using the Public API for a Vbrick account then you may need to set lower rate limits. These values are set Per Minute, so 30
means "30 calls per minute".
// example using default options
const rev = new RevClient({
url: 'https://my.rev.url',
apiKey: 'key', secret: 'secret',
rateLimits: {
get: 24000,
post: 3600,
searchVideos: 120,
videoDetails: 2000,
uploadVideo: 30,
auditEndpoint: 60,
updateVideo: 30,
loginReport: 10,
attendeesRealtime: 2
}
// rateLimits: true // equivalent option
});
For background usage you may consider using lower values to ensure the service doesn't impact other agents using the API:
// example of overriding the limits for a service account that makes background requests
const rev = new RevClient({
url: 'https://my.rev.url',
apiKey: 'key', secret: 'secret',
rateLimits: {
searchVideos: 10, // only make 10 search calls per minute
videoDetails: 100,
// other values use default
}
})
RevClient
also wraps API functionality into separate namespaces. They mostly follow the API Documentation categorization.
admin.roles()
admin.getRoleByName(name)
- Get a specific Role { id: string, name: string }
based on the Role's name (i.e. 'Media Viewer')admin.customFields()
admin.getCustomFieldByName(name)
- Get a specific Custom Field based on the Custom Field's nameadmin.webcastRegistrationFields()
admin.createWebcastRegistrationField(field)
admin.updateWebcastRegistrationField(fieldId, field)
admin.deleteWebcastRegistrationField(fieldId)
admin.brandingSettings()
admin.listIQCreditsUsage(query, options)
admin.verifySystemHealth()
admin.maintenanceSchedule()
admin.userLocationService()
admin.expirationRules()
admin.featureSettings(videoId?)
audit.accountAccess(accountId, options?)
audit.accountUsers(accountId, options?)
audit.accountGroups(accountId, options?)
audit.accountDevices(accountId, options?)
audit.accountVideos(accountId, options?)
audit.accountWebcasts(accountId, options?)
audit.userAccess(userId, accountId, options?)
audit.user(userId, accountId, options?)
audit.group(groupId, accountId, options?)
audit.device(deviceId, accountId, options?)
audit.video(videoId, accountId, options?)
audit.webcast(eventId, accountId, options?)
audit.principal(userId, accountId, options?)
These calls are called automatically by the RevClient
instance, but they're included here for completeness.
auth.extendSession()
- extend any kind of active session, regardless of login methodauth.verifySession()
- throws error if session is not currently validauth.loginToken(apiKey, secret)
auth.extendSessionToken(apiKey)
- DEPRECATEDauth.logoffToken(apiKey)
auth.loginUser(username, password)
auth.logoffUser(userId)
auth.extendSessionUser(userId)
- DEPRECATEDauth.logoffUser(userId)
auth.logoffToken(apiKey)
auth.buildOAuth2Authentication(config, state?)
- returns Promise<{ url, codeVerifier }>
. Sign and format an OAuth2 Authorication URL. Make sure to store the codeVerifier for making a call to get an Access Token.auth.loginOAuth2(config, code, codeVerifier)
auth.buildOAuthAuthenticateURL(config, state?)
- returns Promise<string>
. DEPRECATED Sign and format an OAuth Authorization URL (for browser login flow)auth.parseOAuthRedirectResponse(url)
- DEPRECATED Synchronous, returns { isSuccess, authCode, state, error}
based on returned informationauth.loginOAuth(config, authCode)
- DEPRECATEDauth.extendSessionOAuth(config, refreshToken)
- DEPRECATEDcategory.create(category)
category.details(categoryId)
category.update(categoryId, category)
category.delete(categoryId)
category.list(parentCategoryId?, includeAllDescendants?)
category.listAssignable()
channel.create(channel)
channel.update(channelId, channel)
channel.delete(channelId)
channel.list(start?, options?)
channel.addMembers(channelId, members)
channel.removeMembers(channelId, members)
channel.uploadLogo(channelId, imageFile, options?)
channel.search(searchText?, {type?, assignable?})
Wrapper around the Search Users,Groups and Channels API. If options.assignable: true
then restrict to only assignable entities. options.type
defaults to Channel
to just return channels
device.listDMEs()
device.listZoneDevices()
device.listPresentationProfiles()
device.add(dme)
device.healthStatus(deviceId)
device.delete(deviceId)
device.rebootDME(deviceId)
environment.getAccountId()
environment.getRevVersion()
environment.getUserLocalIp(timeoutMilliseconds)
Wrapper around the Use the Get User Location Service to get a user's IP Address for zoning purposes.
group.create(group)
group.delete(groupId)
group.search(searchText, options?)
NOTE: The response from this endpoint is remapped from the raw API results - it returns camelCase instead of PascalCase ({id: string, name: string, entityType: string }
instead of {Id: string, Name: string, EntityType: string}
group.list(options?)
group.listUsers(groupId, options?)
group.listUserDetails(groupId, options?)
playlist.create(name, videoIds)
playlist.update(playlistId, actions)
playlist.updateFeatured(actions)
playlist.delete(playlistId)
playlist.list()
recording.startVideoConferenceRecording(sipAddress, sipPin, title?)
recording.getVideoConferenceStatus(videoId)
recording.stopVideoConferenceRecording(videoId)
recording.startPresentationProfileRecording(request)
recording.getPresentationProfileStatus(recordingId)
recording.stopPresentationProfileRecording(recordingId)
All upload functions take in a file
argument and an options
argument.
file
: The actual upload data. The possible input options are:Type | Browser | Node | Notes |
---|---|---|---|
File / Blob |
✔ | ✔ | |
"data:" / "blob:" urls |
Passed to fetch() |
✔ | Treated like a Blob |
"file:" urls |
Passed to fetch() |
Read using fs.createReadStream() |
Follows browser/user agent fetch() security policy |
"http:" / "https:" urls |
❌ | ✔ | On node.js these are passed to fetch() , so could load remote resources |
relative/absolute path (string ) |
❌ | Read using fs.createReadStream() |
Follows filesystem permissions |
Web Stream (Readable Stream ), fetch Response |
converted to Blob |
✔ | On browsers entire stream is read into memory. For node.js the stream is piped |
Node.js Streams (including http.IncomingResponse ) |
N/A | ✔ | Piped through to request. |
options
:
fetch
options (most importantly signal
for passing in an AbortSignal
)filename?
: string
- the filename used in the Content-Disposition
field header.contentType?
: string
- the content type of the filecontentLength?
: number
- the known content length of the file. This is rarely needed, but can be used if passing along a HTTP StreamuseChunkedTransfer?
: boolean
- tell upload to not calculate a content length automatically, and just send as Transfer-Encoding: chunked
disableExternalResources?
: boolean
- if true
then reject any string
argument that would read from the network/disk (only blob:
or data:
URLs allowed). Default is false
for backwards compatibilitySECURITY REMINDER - Passing in strings has the chance to trigger a remote request, and on node.js/deno/bun could read arbitrary files on the disk. Be sure to sanitize your inputs. Optionally, use the options.disableExternalResources
to only allow data:
and blob:
URLs.
Note: Rev expects the file extension and content-type to agree. This library will attempt to automatically fix the filename/content-type as needed.
upload.chapters(videoId, chapters, action?, options?)
videoId
: string, videoId of the video in questionchapters
: array of Chapter objects. At least one of title
or imageFile
must be specified
time
: (required) string
title
: string
imageFile
: string
, stream.Readable
or Blob
if using node.js. File
if browser. See file
in the Shared Options section above.action
: One of two string values (default: replace
):
"append"
: Update Video Chapters)"replace"
: Upload Video Chaptersoptions
: Additional fetch
optionsupload.presentationChapters(videoId, file, options?)
upload.supplementalFile(videoId, file, options?)
upload.transcription(videoId, file, language?, options?)
upload.thumbnail(videoId, file, options?)
upload.video(file, metadata, options?)
- Upload a videofile
: string
, stream.Readable
or Blob
if using node.js. File
if browser. See file
in the Shared Options section above.metadata
: Video metadata - see API docs. Note that at minimum uploader
MUST be included.options
: See Shared Options section above.The ID of the video
upload.transcription(videoId, file, language?, options?)
- Upload a transcription / close captions fileuser.roles()
user.create(user)
user.delete(userId)
user.details(userId)
user.getByUsername(username)
user.getByEmail(email)
user.addToGroup(userId, groupId)
- Add a user to the specified Groupuser.removeFromGroup(userId, groupId)
- Remove a user from the specified Groupuser.suspend(userId)
- Use Patch API to suspend useruser.unsuspend(userId)
- Use Patch API to unsuspend useruser.search(searchText, options?)
NOTE: The response from this endpoint is remapped from the raw API results - it returns camelCase instead of PascalCase ({userId: string, firstname: string, profileImageUri: string, entityType: string }
instead of {Id: string, FirstName: string, ProfileImageUri: string, EntityType: string}
.
NOTE: set options.assignable
to true
to use the "Search assignable Users/Groups/Channels" instead of searching for all users
user.listSubscriptions()
user.subscribe(id, type)
user.unsubscribe(id, type)
user.getNotifications(unread?)
user.markNotificationRead(notificationId?)
user.loginReport(sortField?, sortOrder?)
video.setTitle(videoId, title)
- Use PATCH API to change a video's titlevideo.status(videoId)
video.details(videoId)
video.upload(file, metadata, options?)
- alias to upload.video()
video.delete(videoId)
video.playbackInfo(videoId)
- Get Playback Url (basic info about video)video.playbackUrls(videoId, options?, requestOptions?)
video.comments(videoId)
video.chapters(videoId)
video.supplementalFiles(videoId)
video.transcriptions(videoId)
video.thumbnailConfiguration(videoId, options?)
- Get Thumbnail Configurationvideo.migrate(videoId, migratePayload)
video.download(videoId, options?)
video.downloadTranscription(videoId, language)
video.downloadSupplemental(videoId, fileId)
video.downloadChapter(chapter)
video.downloadThumbnail(query)
video.downloadThumbnailSheet(query)
video.listExternalAccess(videoId, searchText?, searchOptions?)
- Get External Accessvideo.createExternalAccess(videoId, {emails, message?, noEmail?})
- Add External Accessvideo.renewExternalAccess(videoId, {emails, noEmail?})
video.deleteExternalAccess(videoId, {emails})
video.revokeExternalAccess(videoId, {emails})
video.waitTranscode(videoId, options?)
- wait for a video to finish transcodingvideo.trim(videoId, removedSegments)
video.convertDualStreamToSwitched(videoId)
video.patch(videoId, operations)
video.report({ videoIds?, startDate?, endDate?, sortDirection? })
- Get Video ReportNOTE: The API only allows searching for 12 months of data at a time. This wrapper function will split up the requests to allow for a larger range of days.
video.uniqueSessionsReport(videoId, { userId?, startDate?, endDate?, sortDirection? })
video.search(query?, options?)
- Search for videosquery
: See API Docs for available search optionsoptions
: Optional, Additional options for requestoptions.maxResults
: number, set the maximum number of results to return.options.onProgress
: (items: <array>, current, total) => void
- callback for each time a page is queried from Rev.options.onScrollExpired
: (err: ScrollError) => void
- Search results use a scrollID cursor that expires after 1-5 minutes from first request. If the scrollID expires then onScrollExpired will be called with a ScrollError. Default behavior is to throw the error.SearchRequest
This method returns a SearchRequest
object, that includes the following methods:
.exec()
- Get all results as an array.nextPage()
- { current, total, items, done }
- Get the search results one page at a time[async iterator]
- The class also implements AsyncIterator
, so can be used as a Stream
or using for await
// get the 10 most recent videos that match 'puppy' in the tags/keywords field
const request = rev.video.search({ q: 'puppies', searchField: 'tags', sortField: 'whenUploaded', sortDirection: 'desc' }, { maxResults: 10 });
const results = await request.exec();
// get ALL videos in the account and report their title, reporting progress and ignoring errors
const request = rev.video.search(undefined, {
onProgress(items, current, total) {
console.log(`PROGRESS: ${current}-${current + items.length} of ${total}`);
},
onScrollExpired(err) {
console.warn('Error while getting video results, ignoring: ', err);
}
});
for await (let video of request) {
console.log(`Video ${video.id}: ${video.title} | duration: ${video.duration}`);
}
video.listDeleted(query?, options?)
- Get Deleted VideosSearchRequest
webcast.list(options?)
webcast.search(query, options?)
webcast.create(event)
webcast.details(eventId)
webcast.edit(eventId, event)
webcast.delete(eventId)
webcast.editAccess(eventId, entities)
webcast.attendees(eventId, runNumber?, options?)
webcast.questions(eventId, runNumber?)
webcast.pollResults(eventId, runNumber?)
webcast.comments(eventId, runNumber?)
webcast.status(eventId)
webcast.isPublic(eventId)
- returns true
/false
webcast.playbackUrls(eventId, options?, requestOptions?)
webcast.playbackUrl(eventId, options?)
DEPRECATED - use webcast.playbackUrls
webcast.startEvent(eventId, preProduction?)
webcast.stopEvent(eventId, preProduction?)
webcast.startBroadcast(eventId)
webcast.stopBroadcast(eventId)
webcast.startRecord(eventId)
webcast.stopRecord(eventId)
webcast.linkVideo(eventId, videoId, autoRedirect?)
webcast.unlinkVideo(eventId)
webcast.muteAttendee(eventId, userId, runNumber?)
webcast.unmuteAttendee(eventId, userId, runNumber?)
webcast.hideComment(eventId, commentId, runNumber?)
webcast.unhideComment(eventId, commentId, runNumber?)
webcast.guestRegistration(eventId, registrationId)
webcast.createGuestRegistration(eventId, registration)
webcast.updateGuestRegistration(eventId, registrationId, registration)
webcast.patchGuestRegistration(eventId, registrationId, registration)
webcast.deleteGuestRegistration(eventId, registrationId)
webcast.listGuestRegistrations(eventId, query, options)
zones.list()
zones.flatList()
zones.create(zone)
zones.edit(zoneId, zone)
zones.delete(zoneId)
zones.devices()
The library exposes some additional utilities:
utils.rateLimit(fn: async function, options)
or rateLimit(options)
The Rev API includes rate limiting for some API endpoints. If you exceed the limit then you'll receive a 429
error response. This function can help to automatically avoid that threshold.
See the API Documentation on Rate Limiting for current limits.
fn
: function to be rate-limitedoptions
:
fn
: function to be rate-limited (if not set as first argument)perSecond
: set limit to X
executions per secondperMinute
: set limit to X
executions per minuteperHour
: set limit to X
executions per hourlimit
: allow limit
executions per interval
millisecondsinterval
: allow limit
executions per interval
millisecondssignal
: AbortSignal
to cancel all pending executions.Wrapped function with same arguments as passed in fn
, with added:
.abort()
- cancel all pending executionsimport { RevClient, rateLimit } from '@vbrick/rev-client';
const rev = new RevClient(options);
// assumes this tool is only tool in the account using API uploads
const throttledUpload = rateLimit({
fn: (...args) => rev.upload.video(...args),
perMinute: 30
});
const numberOfVideos = 100;
for (let i = 0; i < numberOfVideos; i++) {
// same arguments as rev.upload.video
const videoId = await throttledUpload(file, {title: `video ${i}`, uploader: `my.username` });
}
utils.getExtensionForMime(contentType, defaultExtension?)
Get the expected extension for a mime-type supported by Rev for upload. If none found it will return the defaultExtension
(or .mp4
if none specified)
utils.getMimeForExtension(extension?, defaultType?)
Get the expected mime-type for the specified extension (.ext
). If none found it will return the defaultType
(or video/mp4
if none specified)
utils.setPolyfills(callback: (polyfills) => (void | Promise<void>))
Allows overriding the underlying network primitives used by this library, in particular fetch
. This function expects a callback that is called before the first network request is made, to allow lazy load of any dependencies.
import {RevClient, utils} from '@vbrick/rev-client';
utils.setPolyfills(async (polyfills) => {
console.log('Overriding polyfills');
// dynamic import, which works in ESM or commonjs modules
const {fetch} = await import('undici');
Object.assign(polyfills, {
fetch
});
});
const rev = new RevClient(...args);
await rev.connect(); // setpolyfills triggered here
RevError
Custom error returned for error status code responses
status
: number
- http status codeurl
: string
- original URL of requestcode
: string
- Rev-specified error string, for example MalformedRequest
, RequiredFieldMissing
or InvalidFileType
detail
: string
- Additional details about error (if passed).ScrollError
Custom error returned if a paged search request has expired (usually because more than 1 minute has passed between requests for more pages of data)
This code is distributed "as is", with no warranty expressed or implied, and no guarantee for accuracy or applicability to your purpose.