Gipetto / comics

A dumb comics inventory system

Comics

Getting Started

This project uses Yarn Workspaces. Each of server and client define their own package.json, but dependencies are shared between the two.

Install and start all dev servers:

$ yarn install
$ yarn dev:all

Run Cypress tests against dev servers in headless mode:

$ yarn test

Server

The server is an Express/GraphQL app.

To Debug generated DB queries set the debug environment variable to DEBUG=knex:query.

Client

The client is a Svelte app.

GraphQL Schema

Schemas and queries located in ./schemas. Files are prefixed with their purpose.

To update services after changing the schema just run yarn generate in the root workspace. Schemas are published to:

• server/src/graphql/schema.generated.graphql: This is the schema that the GraphQL server will be generated from. It is standard GraphQL schema.
• server/src/types/schemas.generated.ts: This is all GraphQL types converted to TypeScript types for import.
• client/src/graphql/schemas.generated.ts: This is both type and query definitions of the GraphQL schema. Separating them would be great, but the generator isn't very good at doing imports, so a single file is much easier to manage.

TODO

• Run yarn generate before every dev, test and build
• num_titles property for Publisher
• Just why do we need rollup/plugin-replace in the client? It feels like a half-assed replacement for good environment management.
• Optimize queries on Comic Series lookup
  • not a huge deal since it only gets called on comic instance, but on comic list it could blow up
• Inspect requested params to omit subqueries when not needed