Revert "Mitigate document loading issues caused by Google Drive API #…

…export method's unreliable size limit"

This reverts commit 5da0762.

package.json

@@ -5,7 +5,6 @@
   "main": "server/index.js",
   "dependencies": {
     "@google-cloud/datastore": "^5.1.0",
-    "axios": "^1.4.0",
     "cache-manager": "^3.3.0",
     "chai": "^4.2.0",
     "cheerio": "^1.0.0-rc.3",

server/auth.js

@@ -19,25 +19,12 @@ if (!process.env.GOOGLE_APPLICATION_CREDENTIALS && process.env.NODE_ENV === 'dev
   process.env.GOOGLE_APPLICATION_CREDENTIALS = path.join(__dirname, '.auth.json')
 }
 
-// returns the authClient that can be used for making other requests
+// only public method, returns the authClient that can be used for making other requests
 exports.getAuth = async () => {
   if (authClient && process.env.NODE_ENV !== 'test') return authClient
   return setAuthClient()
 }
 
-// This is only used when data is fetched manually via a download link vs using the Google Drive library
-exports.getAccessToken = async () => {
-  const accessToken = await exports.getAuth()
-    .then((client) => {
-      return client.getAccessToken()
-    })
-    .catch((err) => {
-      console.error('Error getting access token:', err)
-    })
-
-  return accessToken.token
-}
-
 // configures the auth client if we don't already have one
 async function setAuthClient() {
   return inflight('auth', async () => {

server/docs.js

@@ -1,6 +1,5 @@
 'use strict'
 
-const axios = require('axios')
 const {google} = require('googleapis')
 const cheerio = require('cheerio')
 const slugify = require('limax')
@@ -9,7 +8,7 @@ const xlsx = require('xlsx')
 const cache = require('./cache')
 const formatter = require('./formatter')
 const log = require('./logger')
-const {getAuth, getAccessToken} = require('./auth')
+const {getAuth} = require('./auth')
 
 const supportedTypes = new Set(['document', 'spreadsheet', 'text/html'])
 
@@ -30,16 +29,18 @@ exports.slugify = (text = '') => {
   return slugify(text)
 }
 
-exports.fetchDoc = async (id, resourceType, exportLinks, req) => {
+exports.fetchDoc = async (id, resourceType, req) => {
   const data = await cache.get(id)
   if (data && data.content) {
     log.info(`CACHE HIT ${req.path}`)
     return data.content
   }
 
   const auth = await getAuth()
-  const driveDoc = await fetch({id, resourceType, exportLinks, req}, auth)
+
+  const driveDoc = await fetch({id, resourceType, req}, auth)
   const originalRevision = driveDoc[1]
+
   const {html, byline, createdBy, sections} = formatter.getProcessedDocAttributes(driveDoc, req.path)
   const payload = {html, byline, createdBy, sections}
 
@@ -52,7 +53,7 @@ exports.fetchDoc = async (id, resourceType, exportLinks, req) => {
   return payload
 }
 
-async function fetchHTMLForId(id, resourceType, exportLinks, req, drive) {
+async function fetchHTMLForId(id, resourceType, req, drive) {
   if (!supportedTypes.has(resourceType)) {
     return `Library does not support viewing ${resourceType}s yet.`
   }
@@ -65,45 +66,12 @@ async function fetchHTMLForId(id, resourceType, exportLinks, req, drive) {
     return fetchHTML(drive, id)
   }
 
-  try {
-    const {data} = await drive.files.export({
-      fileId: id,
-      // text/html exports are not suupported for slideshows
-      mimeType: resourceType === 'presentation' ? 'text/plain' : 'text/html'
-    })
-
-    return data
-  } catch (e) {
-    const errorResponse = e.response.data.error
-    // If the Google Drive API returns 403, we fall back to using the export link directly
-    if (errorResponse.code === 403 && errorResponse.message === "This file is too large to be exported.") {
-      console.log("falling back to using the export link...")
-      const manuallyFetchedData = await fetchManually(resourceType, exportLinks)
-      return manuallyFetchedData
-    } else {
-      throw e
-    }
-  }
-}
-async function fetchManually(resourceType, exportLinks) {
-  const accessToken = await getAccessToken()
-  const exportLink = exportLinks['text/html']
-  const headers = {Authorization: `Bearer ${accessToken}`}
-
-  const fetchedData = await axios({
-    url: exportLink,
-    method: 'GET',
-    responseType: resourceType === 'presentation' ? 'text/plain' : 'text/html',
-    headers: headers
+  const {data} = await drive.files.export({
+    fileId: id,
+    // text/html exports are not suupported for slideshows
+    mimeType: resourceType === 'presentation' ? 'text/plain' : 'text/html'
   })
-    .then((response) => {
-      const fileContents = response.data
-      return fileContents
-    })
-    .catch((err) => {
-      console.error('Error downloading file:', err)
-    })
-  return fetchedData
+  return data
 }
 
 async function fetchOriginalRevisions(id, resourceType, req, drive) {
@@ -121,13 +89,12 @@ async function fetchOriginalRevisions(id, resourceType, req, drive) {
   })
 }
 
-async function fetch({id, resourceType, exportLinks, req}, authClient) {
+async function fetch({id, resourceType, req}, authClient) {
   const drive = google.drive({version: 'v3', auth: authClient})
   const documentData = await Promise.all([
-    fetchHTMLForId(id, resourceType, exportLinks, req, drive),
+    fetchHTMLForId(id, resourceType, req, drive),
     fetchOriginalRevisions(id, resourceType, req, drive)
   ])
-
   return documentData
 }
 

server/list.js

@@ -107,7 +107,7 @@ async function updateTree() {
 }
 
 function getOptions(driveType, id) {
-  const fields = 'nextPageToken,files(id,name,mimeType,parents,webViewLink,createdTime,modifiedTime,lastModifyingUser,exportLinks)'
+  const fields = 'nextPageToken,files(id,name,mimeType,parents,webViewLink,createdTime,modifiedTime,lastModifyingUser)'
 
   if (driveType === 'folder') {
     return {

server/routes/categories.js

@@ -21,8 +21,9 @@ async function handleCategory(req, res) {
 
   if (!meta || !data) return 'next'
 
-  const {resourceType, tags, id, exportLinks} = meta
+  const {resourceType, tags, id} = meta
   const {breadcrumb, duplicates} = data
+
   const layout = categories.has(root) ? root : 'default'
   const template = `categories/${layout}`
 
@@ -67,7 +68,7 @@ async function handleCategory(req, res) {
 
   res.locals.docId = data.id // we need this for history later
   // for docs, fetch the html and then combine with the base data
-  const {html, byline, createdBy, sections} = await fetchDoc(id, resourceType, exportLinks, req)
+  const {html, byline, createdBy, sections} = await fetchDoc(id, resourceType, req)
 
   const renderData = Object.assign({}, baseRenderData, {
     content: html,

test/unit/docs.test.js

@@ -9,12 +9,12 @@ const PAYLOAD_KEYS = ['html', 'byline', 'createdBy', 'sections']
 describe('Docs', () => {
   describe('Fetching Docs', () => {
     it('should fetch document data with expected structure', async () => {
-      const doc = await fetchDoc('id-doc', 'document', {}, {})
+      const doc = await fetchDoc('id-doc', 'document', {})
       expect(doc).to.include.keys('html', 'byline', 'createdBy', 'sections')
     })
 
     it('should parse sections correctly', async () => {
-      const doc = await fetchDoc('mulitsection', 'document', {}, {})
+      const doc = await fetchDoc('mulitsection', 'document', {})
       expect(doc).to.include.keys('html', 'sections')
       const {sections} = doc
       expect(sections.length).equals(2)
@@ -24,31 +24,31 @@ describe('Docs', () => {
 
   describe('Fetching Sheets', () => {
     it('should fetch sheet data with expected structure', async () => {
-      const sheet = await fetchDoc('id-sheet', 'spreadsheet', {}, {})
+      const sheet = await fetchDoc('id-sheet', 'spreadsheet', {})
       expect(sheet).to.include.keys(PAYLOAD_KEYS)
     })
 
     it('should successully parse the sheet to a html table', async () => {
-      const {html} = await fetchDoc('id-sheet', 'spreadsheet', {}, {})
+      const {html} = await fetchDoc('id-sheet', 'spreadsheet', {})
       expect(html).includes('<table>')
       expect(html).includes('</table>')
     })
   })
 
   describe('Fetching html', () => {
     it('should fetch html data with expected structure', async () => {
-      const sheet = await fetchDoc('id-html', 'text/html', {}, {})
+      const sheet = await fetchDoc('id-html', 'text/html', {})
       expect(sheet).to.include.keys(PAYLOAD_KEYS)
     })
 
     it('should not modify html', async () => {
-      const {html} = await fetchDoc('id-html', 'text/html', {}, {})
+      const {html} = await fetchDoc('id-html', 'text/html', {})
       expect(html).equals('<h1>This is a raw HTML document</h1>')
     })
   })
 
   it('should identify bad resource types', async () => {
-    const {html} = await fetchDoc('id-html', 'badtype', {}, {})
+    const {html} = await fetchDoc('id-html', 'badtype', {})
     expect(html).equals('Library does not support viewing badtypes yet.')
   })
 })