Self harm words (#1320)

f2/server.js

@@ -3,6 +3,7 @@ const bodyParser = require('body-parser')
 const path = require('path')
 const formidable = require('formidable')
 const { getAllCerts, encryptAndSend } = require('./src/utils/encryptedEmail')
+const { selfHarmWordsScan } = require('./utils/selfHarmWordsScan')
 
 require('dotenv').config()
 
@@ -65,6 +66,12 @@ const uploadData = (req, res) => {
     // Extract the JSON from the "JSON" form element
     const data = JSON.parse(fields['json'])
     console.log('Parsed JSON:', data)
+
+    const selfHarmWords = selfHarmWordsScan(data)
+    if (selfHarmWords) {
+      console.warn(`Self harm words detected: ${selfHarmWords}`)
+    }
+    data.selfHarmWords = selfHarmWords
     data.submissionTime = new Date().toISOString()
     data.contactInfo.email = randomizeString(data.contactInfo.email)
 

f2/utils/__tests__/selfHarmWordsScan.test.js

@@ -0,0 +1,25 @@
+import { selfHarmWordsScan } from '../selfHarmWordsScan'
+
+describe('selfHarmWordsScan', () => {
+  it('returns empty list when no words match', () => {
+    const data = { a: { b: 'hi' } }
+    expect(selfHarmWordsScan(data)).toEqual([])
+  })
+
+  it('returns 2 words when they are nested in data', () => {
+    const data = {
+      a: { b: 'hi' },
+      aa: ['this', 'agile', 'thing'],
+      cc: { ccc: { dd: 'hi', ee: 'scrum' } },
+    }
+    expect(selfHarmWordsScan(data)).toEqual(['agile', 'scrum'])
+  })
+
+  it('catches words with accents', () => {
+    const data = {
+      a: 'agilé',
+      aa: 'this sçrum',
+    }
+    expect(selfHarmWordsScan(data)).toEqual(['agile', 'scrum'])
+  })
+})

f2/utils/selfHarmWordsScan.js

@@ -0,0 +1,14 @@
+// 'use strict'
+
+const selfHarmWords = 'agile, lean, mvp, scrum'.split(',').map(w => w.trim())
+
+// from https://stackoverflow.com/questions/990904/remove-accents-diacritics-in-a-string-in-javascript
+const removeAccents = s => s.normalize('NFD').replace(/[\u0300-\u036f]/g, '')
+
+const selfHarmWordsScan = data => {
+  const json = removeAccents(JSON.stringify(data).toLowerCase())
+  const wordsUsed = selfHarmWords.filter(w => json.includes(w))
+  return wordsUsed
+}
+
+module.exports = { selfHarmWordsScan }