update-content-types.js

require('dotenv').config()
const AWS = require('aws-sdk')
const fs = require('fs')
const async = require('async')
const Bucket = process.env.AWS_BUCKET

const s3 = new AWS.S3({
  accessKeyId: process.env.ACCESS_KEY_ID,
  secretAccessKey: process.env.SECRET_ACCESS_KEY
})

async function updateObject({ Key }) {
  if (Key === 'uploads/') return
  if (Key.indexOf('pdf') !== Key.length - 3) return

  const objectData = await s3.headObject({
    Key,
    Bucket
  }).promise()

  if (objectData.ContentType === 'application/pdf') return

  const { Body } = await s3.getObject({
    Key,
    Bucket
  }).promise()

  // backup in case
  fs.writeFileSync(`${__dirname}/${Key}`, Body)

  return s3.putObject({
    Bucket,
    Key,
    ContentType: 'application/pdf',
    Body: Body
  }).promise()
}

async function go(Marker) {
  const objects = await s3.listObjects({
    Bucket,
    Marker,
    Prefix: 'uploads/'
  }).promise()

  let x = 0

  const process = (object, cb) => {
    console.log(x++, object.Key)
    updateObject(object).then(() => cb()).catch(cb)
  }

  async.eachLimit(objects.Contents, 10, process, err => {
    if (err) throw err

    if (objects.Contents.length === 1000) {
      go(objects.Contents[objects.Contents.length - 1].Key)
    } else {
      console.log('done')
      process.exit()
    }
  })
}

go()