Apollo Server Boilerplate

dir structure

/
  .env <-- create this file (see below)
  .gitignore
  package.json <-- eslint / nodemon config, customize with your info
  package-lock.json
  src/
    index.js <-- API entrypoint (only customize context as needed)
    db/
      index.js <-- exports all db related bits
      models/ <-- define models in here
        index.js <-- exports models
    api/
      resolvers/
        index.js <-- bundles and exports all resolvers, dont modify
        mutations.js <-- define root mutation resolvers
        queries.js <-- define root query resolvers
        scalars.js <-- use @okgrow/graphql-scalar or define scalar resolvers
        types.js <-- define type resolvers
      type-defs/
        index.js <-- bundles and exports all type defs, dont modify
        index.graphql <-- bundles and exports all SDL type defs, dont modify
        mutations.graphql <-- define mutations
        queries.graphql <-- define queries
        scalars.graphql <-- define additional scalars
        enums.graphql <-- define enums
        inputs.graphql <-- define inputs
        types.graphql <-- define types

includes

eslint for linting with the AirBnB and Prettier rules
- tip: add the eslint and prettier extensions in vscode to make use of these during development
jest for testing
- eslint jest config for linting test best practices
- type defs for autocomplete in test files
graphql-import for writing type defs in .graphql SDL files
nodemon for local development (auto restarts on change)
- includes custom config so it watches your .graphql SDL files for changes as well
GraphQL scalars from @okgrow/graphql-scalars
- boilerplate includes EmailAddress, URL, and DateTime scalars
- add additional scalar defs to the src/api/type-defs/scalars.graphql file
- add additional scalar resolvers to the src/api/resolvers/scalars.js file (the same way the boilerplate ones are used)
pre-configured winston logger from @vampiire/node-logger
- colorized console logging during development
- writes rotating log files during production

setup

clone the boilerplate repo

git clone https://github.com/the-vampiire/apollo-server-boilerplate

create a new repo on your account and leave it empty
- do not initialize with a readme, license, gitignore etc
copy your new repo URL
navigate to the cloned boilerplate repo in your terminal
replace the git remote with your own

git remote remove origin

git remote add origin <your new repo url>

update package.json with your own information
- name, author, description, license, repository, bugs, homepage
create your .env file in the root directory

PORT=8008
ERROR_LOGS_DIR=logs/errors
COMBINED_LOGS_DIR=logs/combined

PORT: the port the API will listen on (set it to anything you'd like)
ERROR_LOGS_DIR: where error log files should be written
COMBINED_LOGS_DIR: where combined log files should be written
- you dont have to set either of these locally
- they must be supplied if you deploy the project and set NODE_ENV=production

running

start production with npm start (dont forget your env vars)
run tests with npm test (runs linter and jest tests)
start dev mode with npm run dev (loads .env then watches and reloads on change)

customize

add your type defs in src/api/type-defs
- you don't need to modify src/api/type-defs/ index.js or index.graphql they are already set up to bundle and export your schema
add your resolvers in src/api/resolvers
- you don't need to modify src/api/resolvers/index.js it is already set up to bundle and export your resolvers
add and export your database models in src/db/models
customize the API context in src/index.js but do not modify anything else (it is all wired up)
override eslint rules in package.json eslintConfig.rules section
customize the logger see options at @vampiire/node-logger