node-nlp
Version:
Library for NLU (Natural Language Understanding) done in Node.js
49 lines (43 loc) • 2.1 kB
text/typescript
/**
* @module botbuilder
*/
/**
* Copyright (c) Microsoft Corporation. All rights reserved.
* Licensed under the MIT License.
*/
export class EndorsementsValidator {
/**
* Verify that the set of ChannelIds, which come from the incoming activities,
* all match the endorsements found on the JWT Token.
* For example, if an Activity comes from webchat, that channelId says
* says "webchat" and the jwt token endorsement MUST match that.
* @param channelId The channel name, typically extracted from the activity.ChannelId field, that to which the Activity is affinitized.
* @param endorsements Whoever signed the JWT token is permitted to send activities only for
* some specific channels. That list is the endorsement list, and is validated here against the channelId.
* @returns {boolean} True is the channelId is found in the Endorsement set. False if the channelId is not found.
*/
public static validate(channelId: string, endorsements: string[]): boolean {
// If the Activity came in and doesn't have a Channel ID then it's making no
// assertions as to who endorses it. This means it should pass.
if (channelId === null || channelId.trim() === '') {
return true;
}
if (endorsements === null) {
throw new Error('endorsements required');
}
// The Call path to get here is:
// JwtTokenValidation.AuthenticateRequest
// ->
// JwtTokenValidation.ValidateAuthHeader
// ->
// ChannelValidation.AuthenticateChannelToken
// ->
// JWTTokenExtractor
// Does the set of endorsements match the channelId that was passed in?
// ToDo: Consider moving this to a HashSet instead of a string
// array, to make lookups O(1) instead of O(N). To give a sense
// of scope, tokens from WebChat have about 10 endorsements, and
// tokens coming from Teams have about 20.
return endorsements.some((value: string) => value === channelId);
}
}