A data-mover-and-shaker for Disco Elysium.

The goal of Disco-Courier is to normalize and condense data for easy reading or transfer to the format of your choice. It drives data around in a SSSSOUPED UP MOTOR CARRIAGE!

It is the first of three repos aimed at a larger project.

Things you can use Disco-Courier for:

• quickly look up game elements to plan a walkthrough.
• generate a markdown-friendly table of some interesting dialog.
• quickly format your own custom views of the data for display.
• normalize data into a SQL db for use in your personal project.
• slice-and-dice results as pre-prep for putting into noSQL docs.
• isolate Garte-only dialog and train an AI model on it.*

*(please, please let me know if you do this).

• You'll need a purchased copy of the game and access to its data.
• Assumes you have Node.js and npm installed.

• clone or download this repo
• run npm install in the directory created.
• place a copy of your exported data in /data/dialog.json
• note: make sure your extracted game data is in the JSON format!
• run a health check: npm run courier:health

If everything is setup correctly, you'll get a json response with summary details of the game's "FYS" attribute (...get it? ...Health check? ...GET IT?!? Yeah me either).

Here's a bunch of sample commands you can try to give you an idea of what's possibe (skip to next section for the full options/arguments reference):

npm run...

Important: Make sure you include the empty double-dash (--) between the courier command and the arguments.

Note this is not raw output: each command passes through an extensive templating system that can be customized to taste.

###Arguments

Arguments take the form of options for paging and output, and a list of entities and sub-groups you want to export.

Use these as last argument to specify output. Can use multiple args. Omitting a sub-item exports all groups for the entity (e.g. "actors" on its own producs all rows in actors)

• All output is controlled by the template functions defined in /templates.
• Helper and utility methods are in the /search directory.
• Add a json-querying utility for extra flexibility if you like, or use the utilities that are in /search.

• options are defined in /lib/args.ts
• routing to templates is done in /templates/index.ts
• if stuck, refer to /lib/migration.ts for details on how entities and groups are managed.
• if using the group with the --output=db flag, you'll need to generate a new schema using the sequelize-cli.

• use npm run dev -- --output=read <rest-of-options> if you want to use nodemon. Useful when making edits to the template system.

• If unfamiliar with SQL databases or the Sequelize ORM, this section probably isn't for you.
• The output of the db flag will place a seed in /data/seeders/<timestamp-entity-group.js>.
  • The resulting seed will reference Entity and Group names for the table name, with the first letter Capitalized, followed by an underscore and group name (if a group was specified):
    Example: --output=db actors.skill will generate File: /data/seeders/202107271226-add-actors.skill.js targeting Table: Actors_skill. See above section for more on Entities and groups.
    
• after you've generated some seed files, you can do npm run db:up to populate the database of your choice.
• SQLite should be ready out-of-the-box, you'll need to follow the standard Sequelize setup for Postgres, etc.

Sample models and migration scripts have been provided for nearly every sample template in the app. To import:

• Have a database that Sequelize supports, and make sure to install the proper Sequelize items so you can talk to it.
• Run at least one --output=db command on an entity you wish to transfer to a db.
• Run npm run db:up to create your tables and run the seeders.
• Run npm run db:down to roll back.

• Sample sequelize models and migrations can be found in /data/models and /data/migrations respectively.
• using the --results flag will greatly improve speed if limiting the number of results, but --start will need to go through the full dataset until it reaches the starting index. Further optimization and work on paging needs to be done.
• Dialogs within conversations are still a bit work-in-progress.
• USE CAUTION when working with a large amount of conversations.
• There are lengthy TODO comments in the source for contributors to check out. I don't have a lot of time, so contributions welcome.

• Tidy up dialogueEntry formatting and make the generic model for it.
• Finish up checks and modifiers (see above) (Done).
• Add support for foreign key relationships for sequelize.
• While you can do a lot with the json export option for getting items into noSQL databases, it could use some better convenience options.
• CSV and spreadsheets would be nice (for easy dumps for things like checks).
• Markdown tables? If there's something easy, why not.
• Currently supports the version released 6th April 2021. Build on initial work to support additional versions.