multivocal-backend

Description

This is the backend & rest API for the multivocal project. It is based on node with express with mongodb.

Requirements

There are a few requirements for a system to run the API.

• MongoDB - v3 or newer (https://www.mongodb.com/)
• node.js - v4 or newer (https://nodejs.org/en/)
• yarn:
  • Not required, but recommended. Install by following the instructions on https://yarnpkg.com/en/docs/install#linux-tab after installing node.js
• nodemon :
  • Daemon used to run node - requires node to be install. Install by running npm install -g nodemon.

Setup

There are a few step to follow when setting up the project

• AWS credentials should be stored in configs/aws_credentials.json in the following format:

{
    "accessKeyId": "accesskey",
    "secretAccessKey": "secretkey",
    "region": "awsregion",
    "Bucket": "bucketnam"
}

• Mongo configuration should be stored in configs/db.js in the following format:

module.exports = {
    mongo: {
        port: 'mongoport',
        db: 'dbname'
    }
}

• The script populate_transcriptions_to_db.js should be run, after it has been edited so that it has the data variable set to the correct json file containing all trascriptions in the following format:

[
    {
        "transcription_text": "Author of the danger trail, Philip Steels, etc.",
         "transcription_id": "arctic_a0001"
    },
    {
        "transcription_text": "Not at this particular case, Tom, apologized Whittemore.",
         "transcription_id": "arctic_a0002"
    }
    ...
]

Configuration

The server is configured by default to run on either port 80 or 3000. Port 80 should be used in production, while 3000 should be used in development.

To set the port, use the NODE_ENV environment variable, set to either development or production

Usage

Before running the application for the first time, run yarn in the project root. This will install all required dependencies.

The application is configured to run with nodemon as a daemon - so it will automatically restart on errors, and reload on changes.

To start the server, simply run npm start from a terminal in the root of the project.

API Documentation

These Are the API endpoints that are a part of the backend so far.

POST /api/recording

Used for uploading new files and supplying data about the recording. It takes the following form data:

• file: The actual file of a recording
• transcription_id: The id of the transcription the recording is for.
• client_id (optional): The id of the client uploading the recording
• notes (optional): Notes about the specific recording

At this point in time, no authentication is implemented for the endpoint, but will be at a later time.

It will return a status code of 422 if missing data, and 200 on succesful upload.

PUT /api/recording/:file_name/edit

Replaces the current file with an edited version.

It takes a file, just like in the endpoint for uploading, as form data

DELETE /api/recording/:file_name

Deletes the recording with the specified name from the system

GET /api/recordings/:transcription_id

Gets all recordings for the specified transcription_id.

The transcription_id in the url is the id to get recordings for.

GET /api/recordings

Returns a JSON array with all transcriptions and their corresponding recordings.

PUT /api/recordings/:file_name/verify

Marks the recording as verified

PUT /api/recording/:file_name/rate/:rating

Adds a rating for the specified file_name

As of now, there are no spec for the ratings, that will have to decided.