BOT-2154 SPEECH: Sanitize TTS Text and Download Filenames

Author: Attila Ujj

docker-compose-picotts.yml

@@ -8,6 +8,8 @@ services:
     ports:
       - 80:80
   frontend:
+    build:
+      context: frontend
     image: botium/botium-speech-frontend:latest
     restart: always
     environment:
@@ -16,22 +18,22 @@ services:
       BOTIUM_SPEECH_PROVIDER_STT: kaldi
     volumes:
       - "./frontend/resources:/app/resources"
-  watcher:
-    image: botium/botium-speech-watcher:latest
-    restart: always
-    volumes:
-      - "./watcher:/app/watch"
-      - "./logs/watcher:/app/logs"
-  stt-en:
-    image: botium/botium-speech-kaldi-en:latest
-    restart: always
-    volumes:
-      - "./logs/stt-en:/opt/logs"
-  stt-de:
-    image: botium/botium-speech-kaldi-de:latest
-    restart: always
-    volumes:
-      - "./logs/stt-de:/opt/logs"
-  dictate:
-    image: botium/botium-speech-dictate:latest
-    restart: always
+#  watcher:
+#    image: botium/botium-speech-watcher:latest
+#    restart: always
+#    volumes:
+#      - "./watcher:/app/watch"
+#      - "./logs/watcher:/app/logs"
+#  stt-en:
+#    image: botium/botium-speech-kaldi-en:latest
+#    restart: always
+#    volumes:
+#      - "./logs/stt-en:/opt/logs"
+#  stt-de:
+#    image: botium/botium-speech-kaldi-de:latest
+#    restart: always
+#    volumes:
+#      - "./logs/stt-de:/opt/logs"
+#  dictate:
+#    image: botium/botium-speech-dictate:latest
+#    restart: always

docker-compose-test.yml

@@ -0,0 +1,44 @@
+version: '3'
+services:
+#  watcher:
+#    build:
+#      context: watcher
+#    image: botium/botium-speech-watcher:latest
+#    user: 1000:1000
+#    restart: always
+#    volumes:
+#      - "./watcher:/app/watch"
+#      - "./logs/watcher:/app/logs"
+#  frontend:
+#    build:
+#      context: frontend
+#    image: botium/botium-speech-frontend:latest
+#    user: 1000:1000
+#    restart: always
+#    environment:
+#      BOTIUM_API_TOKENS:
+#      BOTIUM_SPEECH_PROVIDER_TTS: picotts
+#      BOTIUM_SPEECH_PROVIDER_STT: kaldi
+#    volumes:
+#      - "./frontend/resources:/app/resources"
+  stt_en:
+    build:
+      context: stt
+      dockerfile: Dockerfile.kaldi.en
+    image: botium/botium-speech-kaldi-en:latest
+    user: 1000:1000
+    restart: always
+    volumes:
+      - "./logs/stt_en:/opt/logs"
+#  dictate:
+#    build:
+#      context: dictate
+#    image: botium/botium-speech-dictate:latest
+#    user: 1000:1000
+#    restart: always
+#  tts:
+#    build:
+#      context: tts
+#      dockerfile: Dockerfile.marytts
+#    image: botium/botium-speech-marytts:${TAG}
+#    restart: always

docker-compose.yml

@@ -18,22 +18,22 @@ services:
       BOTIUM_SPEECH_GOOGLE_PRIVATE_KEY:
     volumes:
       - "./frontend/resources:/app/resources"
-  watcher:
-    image: botium/botium-speech-watcher:${TAG}
-    restart: always
-    volumes:
-      - "./watcher:/app/watch"
-      - "./logs/watcher:/app/logs"
-  stt-en:
-    image: botium/botium-speech-kaldi-en:${TAG}
-    restart: always
-    volumes:
-      - "./logs/stt-en:/opt/logs"
-  stt-de:
-    image: botium/botium-speech-kaldi-de:${TAG}
-    restart: always
-    volumes:
-      - "./logs/stt-de:/opt/logs"
+#  watcher:
+#    image: botium/botium-speech-watcher:${TAG}
+#    restart: always
+#    volumes:
+#      - "./watcher:/app/watch"
+#      - "./logs/watcher:/app/logs"
+#  stt-en:
+#    image: botium/botium-speech-kaldi-en:${TAG}
+#    restart: always
+#    volumes:
+#      - "./logs/stt-en:/opt/logs"
+#  stt-de:
+#    image: botium/botium-speech-kaldi-de:${TAG}
+#    restart: always
+#    volumes:
+#      - "./logs/stt-de:/opt/logs"
   tts:
     image: botium/botium-speech-marytts:${TAG}
     restart: always

frontend/.gitignore

@@ -1,6 +1,6 @@
 node_modules
 package-lock.json
 *.local
-resources/.cache
+resources/.cache_del
 resources/.tmp
 resources/google.json

frontend/package.json

@@ -17,6 +17,7 @@
     "@google-cloud/text-to-speech": "^3.1.3",
     "body-parser": "^1.19.0",
     "cheerio": "^1.0.0-rc.5",
+    "content-disposition": "^0.5.3",
     "cross-env": "^7.0.3",
     "debug": "^4.3.1",
     "dotenv-flow": "^3.2.0",

frontend/src/routes.js

@@ -4,6 +4,7 @@ const mkdirp = require('mkdirp')
 const crypto = require('crypto')
 const express = require('express')
 const sanitize = require('sanitize-filename')
+const contentDisposition = require('content-disposition')
 const { runconvert } = require('./convert/convert')
 const { wer } = require('./utils')
 const debug = require('debug')('botium-speech-processing-routes')
@@ -317,7 +318,7 @@ router.get('/api/tts/:language', async (req, res, next) => {
           const buffer = fs.readFileSync(cacheFileBuffer)
           debug(`Reading tts result ${cacheFileName} from cache: ${name}`)
           res.writeHead(200, {
-            'Content-disposition': `attachment; filename="${name}"`,
+            'Content-disposition': `${contentDisposition(name)}`,
             'Content-Length': buffer.length
           })
           return res.end(buffer)
@@ -335,7 +336,7 @@ router.get('/api/tts/:language', async (req, res, next) => {
         text: req.query.text
       })
       res.writeHead(200, {
-        'Content-disposition': `attachment; filename="${name}"`,
+        'Content-disposition': `${contentDisposition(name)}`,
         'Content-Length': buffer.length
       })
       res.end(buffer)

frontend/src/utils.js

@@ -12,7 +12,7 @@ const wer = async (text1, text2) => {
 
 const ttsFilename = (text) => {
   const shortenedText = _.truncate(text, { length: 500 })
-  return sanitize(shortenedText)
+  return sanitize(shortenedText) || 'unknown'
 }
 
 const cleanEnv = (envName) => {