Truepill Inventory API

Typescript Express Server API over MySQL DB

Implements web socket in order to push updated inventory detail notifications.

DDL for database, plus sample insert data in the /ddl folder

Endpoints

• GET /api/products - Fetch Products

  • returns array of products

    [
      {
        product_name: string,
        product_current_count: integer,
        product_manual_count: integer
      },
      ...
    ]
    

• PATCH /api/products - Bulk Update

  • requires array of product names and quantities to increment (optional forceReset flag to reset current and manual counts for a product)

      {
        productUpdates: [
          {
            product_name: string,
            product_qty: integer
          },
          ...
        ],
        forceReset: boolean
      }
    

  • returns count of updated rows

      {
        affectedRows: integer
      }
    

• PATCH /api/products/:product_name - Remove Stock

  • requires product name in URL path, and product_qty to decrement stock by

      {
        product_qty: integer
      }
    

  • returns count of updated rows

      {
        affectedRows: integer
      }
    

Environment Variables

Required Build environment variables :

• DB_HOST e.g. localhost
• DB_USER e.g. root
• DB_PWORD e.g. ********
• DB_NAME e.g. truepill

Required Test environment variables :

• TEST_DB_NAME e.g. truepill_test

Install

yarn install

Run

Development

yarn run dev

Production

Builds in to the /build directory

yarn run prod

Test

My TDD was based around whole endpoint integration tests, using supertest, against a test MySQL database (see environment variables). These tests were mostly happy path tests and need extending.

I would add some unit tests but have not done so now - but the integration tests still run through Mocha and will contribute to coverage.

yarn test

NOTES

Production code needs process management, potentially via a node process manager, or via docker orchestration - not implemented.

Production code should include a proper logger - not chosen, and dependent on log digestion strategies.

Production database should probably include auto-incremented ID Primary Keys - but implmemented without as specified in requirements.

Production code would require security such as authorization tokens, and HTTPS