#3 Use AWS Transcribe service in order to evaluate the speaking/pronunciation

backend/backend/.env.example

@@ -0,0 +1,4 @@
+AWS_ACCESS_KEY_ID=
+AWS_SECRET_ACCESS_KEY=
+AWS_DEFAULT_REGION=
+AWS_BUCKET_NAME=

backend/backend/settings.py

@@ -11,6 +11,10 @@
 """
 
 from pathlib import Path
+import environ
+
+env = environ.Env()
+environ.Env.read_env()
 
 # Build paths inside the project like this: BASE_DIR / 'subdir'.
 BASE_DIR = Path(__file__).resolve().parent.parent
@@ -37,6 +41,9 @@
     'django.contrib.sessions',
     'django.contrib.messages',
     'django.contrib.staticfiles',
+    'rest_framework',
+    'backend.transcribe',
+    'backend.upload',
 ]
 
 MIDDLEWARE = [

backend/backend/transcribe/admin.py

@@ -0,0 +1,3 @@
+from django.contrib import admin
+
+# Register your models here.

backend/backend/transcribe/apps.py

@@ -0,0 +1,6 @@
+from django.apps import AppConfig
+
+
+class TranscribeConfig(AppConfig):
+    default_auto_field = 'django.db.models.BigAutoField'
+    name = 'backend.transcribe'

backend/backend/transcribe/models.py

@@ -0,0 +1,3 @@
+from django.db import models
+
+# Create your models here.

backend/backend/transcribe/tests.py

@@ -0,0 +1,3 @@
+from django.test import TestCase
+
+# Create your tests here.

backend/backend/transcribe/urls.py

@@ -0,0 +1,10 @@
+from rest_framework import routers
+from django.urls import path, include
+from . import views
+
+router = routers.DefaultRouter()
+
+urlpatterns = [
+    path('', views.TranscribeViewSet.as_view(), name='transcribe'),
+    path('/<int:id>', views.TranscribeViewSet.as_view(), name='transcribe'),
+]

backend/backend/transcribe/views.py

@@ -0,0 +1,109 @@
+from django.shortcuts import render
+from rest_framework.views import APIView
+from rest_framework.response import Response
+import environ
+import boto3
+import urllib
+import json
+
+env = environ.Env()
+environ.Env.read_env()
+
+transcribe_client = boto3.client('transcribe', aws_access_key_id=env(
+    'AWS_ACCESS_KEY_ID'), aws_secret_access_key=env('AWS_SECRET_ACCESS_KEY'))
+s3_client = boto3.client('s3', aws_access_key_id=env(
+    'AWS_ACCESS_KEY_ID'), aws_secret_access_key=env('AWS_SECRET_ACCESS_KEY'))
+
+
+class TranscribeViewSet(APIView):
+
+    """
+    The post method should receive the following JSON (example)
+    {
+        "s3_url": "https://s3.amazonaws.com/bucket/file.mp3",
+        "language_code": "en-US",
+    }
+    The API will respond with a JSON object with the following fields:
+    {
+        "message": "Successfully started transcription job" || "Failed to start transcription job",
+        "transcription_job_name": "transcription_job_name",
+        "aws_response": "aws_response"
+    }
+
+    The transcription job name has to be stored to be used in the get method to get the results
+    """
+
+    def post(self, request):
+        s3_url = request.data.get('s3_url')
+        language_code = request.data.get('language_code')
+
+        s3_filename = s3_url.split('/')[-1]
+        s3_uri = "s3://{}/{}".format(env('AWS_BUCKET_NAME'), s3_filename)
+
+        if not s3_url or not language_code:
+            return Response({"message": "Please provide all the required fields."})
+
+        response = transcribe_client.start_transcription_job(
+            TranscriptionJobName=s3_filename+'_transcription',
+            Media={'MediaFileUri': s3_uri},
+            MediaFormat=s3_url.split('.')[-1],
+            LanguageCode=language_code,
+            OutputBucketName=env('AWS_BUCKET_NAME'),
+            OutputKey='transcriptions/',
+        )
+
+        if response:
+            return Response({"message": "Successfully started transcription job", "transcription_job_name": response['TranscriptionJob']['TranscriptionJobName'], "aws_response": response})
+        else:
+            return Response({"message": "Failed to start transcription job", "aws_response": response})
+
+    """
+    This get method should recive the following query parameters:
+    Example: api.url/transcribe/?transcription-job-name=test_transcription
+
+    The API will respond with a JSON object with the following fields:
+    {
+        message: "Successfully fetched transcription" || "Failed to get transcription job" || "Transcription job not completed yet",
+        transcription: [{words}],
+        aws_response: "Response from AWS",
+        status: "success" || "failed" || "pending"
+    }
+    """
+
+    def get(self, request):
+        transcription_job_name = request.query_params.get(
+            'transcription-job-name')
+
+        if not transcription_job_name:
+            return Response({"message": "Please provide the transcription job name."})
+        response = transcribe_client.get_transcription_job(
+            TranscriptionJobName=transcription_job_name)
+
+        if response:
+            s3_url = response['TranscriptionJob']['Transcript']['TranscriptFileUri']
+            if response['TranscriptionJob']['TranscriptionJobStatus'] == 'COMPLETED':
+
+                s3_client.put_object_acl(
+                    ACL='public-read', Bucket=env('AWS_BUCKET_NAME'), Key="transcriptions/"+transcription_job_name+'.json')
+                json_data = urllib.request.urlopen(s3_url)
+                data = json.loads(json_data.read())
+                words = []
+
+                for item in data['results']['items']:
+                    word = item['alternatives'][0]['content']
+                    confidence = float(item['alternatives'][0]['confidence'])
+                    if confidence > 0.7 or confidence == 0:
+                        words.append(
+                            {'word': word, 'confidence': confidence, 'color': 'green'})
+                    elif confidence <= 0.7 and confidence > 0.4:
+                        words.append(
+                            {'word': word, 'confidence': confidence, 'color': 'yellow'})
+                    else:
+                        words.append(
+                            {'word': word, 'confidence': confidence, 'color': 'red'})
+
+                return Response({"message": "Successfully fetched transcription", "transcription": words, "aws_response": response, "status": "success"})
+            else:
+                return Response({"message": "Transcription job not completed yet", "aws_response": response, "status": "pending"})
+        else:
+            return Response({"message": "Failed to get transcription job", "aws_response": response, "status": "failed"})

backend/backend/upload/admin.py

@@ -0,0 +1,3 @@
+from django.contrib import admin
+
+# Register your models here.

backend/backend/upload/apps.py

@@ -0,0 +1,6 @@
+from django.apps import AppConfig
+
+
+class UploadConfig(AppConfig):
+    default_auto_field = 'django.db.models.BigAutoField'
+    name = 'backend.upload'

backend/backend/upload/models.py

@@ -0,0 +1,3 @@
+from django.db import models
+
+# Create your models here.

backend/backend/upload/tests.py

@@ -0,0 +1,3 @@
+from django.test import TestCase
+
+# Create your tests here.

backend/backend/upload/urls.py

@@ -0,0 +1,9 @@
+from rest_framework import routers
+from django.urls import path, include
+from . import views
+
+router = routers.DefaultRouter()
+
+urlpatterns = [
+    path('', views.UploadViewSet.as_view(), name='upload'),
+]

backend/backend/upload/views.py

@@ -0,0 +1,31 @@
+from django.shortcuts import render
+from rest_framework.views import APIView
+from rest_framework.response import Response
+import environ
+import boto3
+
+env = environ.Env()
+environ.Env.read_env()
+
+s3_client = boto3.client('s3', aws_access_key_id=env(
+    'AWS_ACCESS_KEY_ID'), aws_secret_access_key=env('AWS_SECRET_ACCESS_KEY'))
+
+
+class UploadViewSet(APIView):
+    """
+    This put method should receive a file in form-data with the name "audio_file".
+    The API will put the file in the S3 bucket and return the URL of the file together with the response from AWS.
+    """
+
+    def put(self, request):
+        audio_file = request.FILES['audio_file']
+        if audio_file.content_type not in ['audio/mpeg', 'audio/mp3']:
+            return Response({"message": "The file is not a valid audio file."})
+        response = s3_client.put_object(
+            Body=audio_file, Bucket=env('AWS_BUCKET_NAME'), Key=audio_file.name, ACL='public-read')
+        if response:
+            URL = "https://{}.s3.amazonaws.com/{}".format(
+                env('AWS_BUCKET_NAME'), audio_file.name)
+            return Response({"message": "Successfully uploaded file", "s3_url": URL, "aws_response": response})
+        else:
+            return Response({"message": "Failed to upload file", "aws_response": response})

backend/backend/urls.py

@@ -14,8 +14,10 @@
     2. Add a URL to urlpatterns:  path('blog/', include('blog.urls'))
 """
 from django.contrib import admin
-from django.urls import path
+from django.urls import path, include
 
 urlpatterns = [
     path('admin/', admin.site.urls),
+    path('transcribe/', include('backend.transcribe.urls')),
+    path('upload/', include('backend.upload.urls'))
 ]

backend/requirements.txt

@@ -1,13 +1,20 @@
 asgiref==3.5.0
 autopep8==1.6.0
+awscli==1.22.47
 boto3==1.20.46
-botocore==1.23.46
+botocore==1.23.47
+colorama==0.4.3
 Django==4.0.2
+django-environ==0.8.1
 djangorestframework==3.13.1
+docutils==0.15.2
 jmespath==0.10.0
+pyasn1==0.4.8
 pycodestyle==2.8.0
 python-dateutil==2.8.2
 pytz==2021.3
+PyYAML==5.4.1
+rsa==4.7.2
 s3transfer==0.5.0
 six==1.16.0
 sqlparse==0.4.2