Work Order Schedule Viewer

Running

This project uses Ruby v3.3.5 and Rails v8.0.1, the latest version available to me on my system. It uses the default SQLite database, so no database setup is required beyond running:

rails db:migrate

To run the server in development mode, run:

./bin/dev

To run it in production mode, run:

rails assets:precompile
rails server

Make sure to migrate and import the data in the production environment first. This project may also be run in production mode using Docker.

Importing Data

To import data from the CSV files into the database, place the files into the root directory of the project, and run:

rake import:csv

If this task is run after making changes to the CSV files, it will update the rows with the corresponding ID in the database with the new values. It will add rows with IDs that are not already present, but will not delete records that are not also in the CSV files.

Development Approach

Some of my design choices are listed below:

The frontend works like most calendar applications, where the blocks in each column have the position: absolute style attribute, and are freely positioned and sized within the column using additional CSS.

I designed the application so that the page does not have to fetch data from the server once loaded. Therefore, data that is not immediately shown to the user are in the data-* attributes of the block and column components for use by Stimulus controllers.

Since all time values are all known to be within the same day, I represent these values as seconds since midnight, and translate them into formatted time representations by treating them as UNIX timestamps in the UTC time zone.

The largest challenge I faced while designing this application was handling overlapping time blocks due to one technician having overlapping work orders. To solve it, I had to ensure that each time block had its start and end times embedded into it as data-* attributes. Each column then gets its own Stimulus controller which creates a list of sets of overlapping blocks, then resizes them horizontally until they no longer overlap.

Potential future improvements

• Allow for the import:csv to take custom file paths.
• Make the schedule UI look better (i.e. add horizontal rules and make the block contents more responsive).
• Add the ability to upload the CSV files from the frontend.