server/lib/activitypub/crawl.ts

   1 import { ACTIVITY_PUB, JOB_REQUEST_TIMEOUT, WEBSERVER } from '../../initializers/constants'
   2 import { doRequest } from '../../helpers/requests'
   3 import { logger } from '../../helpers/logger'
   4 import * as Bluebird from 'bluebird'
   5 import { ActivityPubOrderedCollection } from '../../../shared/models/activitypub'
   6 import { checkUrlsSameHost } from '../../helpers/activitypub'
   7 import { parse } from "url"
   8
   9 type HandlerFunction<T> = (items: T[]) => (Promise<any> | Bluebird<any>)
  10 type CleanerFunction = (startedDate: Date) => (Promise<any> | Bluebird<any>)
  11
  12 async function crawlCollectionPage <T> (uri: string, handler: HandlerFunction<T>, cleaner?: CleanerFunction) {
  13   logger.info('Crawling ActivityPub data on %s.', uri)
  14
  15   const options = {
  16     method: 'GET',
  17     uri,
  18     json: true,
  19     activityPub: true,
  20     timeout: JOB_REQUEST_TIMEOUT
  21   }
  22
  23   const startDate = new Date()
  24
  25   const response = await doRequest<ActivityPubOrderedCollection<T>>(options)
  26   const firstBody = response.body
  27
  28   let limit = ACTIVITY_PUB.FETCH_PAGE_LIMIT
  29   let i = 0
  30   let nextLink = firstBody.first
  31   while (nextLink && i < limit) {
  32     // Don't crawl ourselves
  33     const remoteHost = parse(nextLink).host
  34     if (remoteHost === WEBSERVER.HOST) continue
  35
  36     options.uri = nextLink
  37
  38     const { body } = await doRequest<ActivityPubOrderedCollection<T>>(options)
  39     nextLink = body.next
  40     i++
  41
  42     if (Array.isArray(body.orderedItems)) {
  43       const items = body.orderedItems
  44       logger.info('Processing %i ActivityPub items for %s.', items.length, options.uri)
  45
  46       await handler(items)
  47     }
  48   }
  49
  50   if (cleaner) await cleaner(startDate)
  51 }
  52
  53 export {
  54   crawlCollectionPage
  55 }