meeting-to-text's Introduction

meeting-to-text

speech to text

to run flask app:

python3 app.py

you will need a valid Aws account, and a user with the necessary privileges: aws transcribe full access, aws s3 full access.

set these as environment variables AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY

you will also need a database setup with the only table in the database.py file

endpoint accepts an audio file, sends it to amazon transcribe (aws speech to text service), and returns response to user

typical response:

{
  "transcriptions": [
    {
      "index": 0, 
      "text": "Testing testing. This is a, this is a test."
    }
  ]
}

Test Generation

rm -r request_logs/request-*
<- browse around with app running, requests are captured ->
python generate-test-suite.py

Test are generated into the generated_test_suites folder. Refactor those folders into the test_suites folder if the test is useful.

Run Single Test Suite

python auto-test.py

Run All Tests

cd test_suites
python run.py

$ python run.py
['delete_clips_and_uploads', 'upload_10s_audio_sample', 'upload_bad_filetype', '__pycache__']
============ delete_clips_and_uploads ============
sending request GET http://localhost:5000/delete_all_sound_clips
200
============ upload_10s_audio_sample ============
sending request GET http://localhost:5000/
200
sending request POST http://localhost:5000/transcribe
handled as file upload
200
============ upload_bad_filetype ============
sending request POST http://localhost:5000/transcribe
handled as file upload
400
============ __pycache__ ============

Recommend Projects