Move devDependencies from individual packages to root package.json

[ntwb]

Details

This PR moves the devDependencies from individual Node.js packages to the root package.json file, this is because we can run the tests for both packages from the root directory instead of each package directory

Actual version bumps to the devDependencies will follow in further PRs on Monday

Additional PRs will follow for the peerDependencies changes

Ideally my initial plan is to release updated versions as v2.0.0 with as few breaking changes as possible

I originally added a Node.js job, but it fails due to a Travis CI issue, so I removed it,will add this back when switching to GitHub Actions

nvm.install
nvm install 18
node: /lib/x86_64-linux-gnu/libm.so.6: version `GLIBC_2.27' not found (required by node)
node: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.25' not found (required by node)
node: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.28' not found (required by node)
The command "npm config set spin false" failed and exited with 1 during .
Your build has been stopped.

Test Instructions

Install Node.js v16, install dependencies, run tests:

nvm install 16
nvm use 16
npm install --legacy-peer-deps
cd packages/eslint-config-humanmade
npm install --legacy-peer-deps
cd -
cd packages/stylelint-config
npm install --legacy-peer-deps
cd -
npm test

Test Results

❯ npm install --legacy-peer-deps
  cd packages/eslint-config-humanmade
  npm install --legacy-peer-deps
  cd -
  cd packages/stylelint-config
  npm install --legacy-peer-deps
  cd -
  npm test

up to date, audited 898 packages in 5s

107 packages are looking for funding
  run `npm fund` for details

7 vulnerabilities (6 moderate, 1 critical)

To address all issues, run:
  npm audit fix

Run `npm audit` for details.

up to date, audited 1 package in 56ms

found 0 vulnerabilities

up to date, audited 13 packages in 524ms

1 package is looking for funding
  run `npm fund` for details

found 0 vulnerabilities

> @humanmade/[email protected] test
> npm run lint-pkg-json && npm run test:eslint && npm run test:stylelint


> @humanmade/[email protected] lint-pkg-json
> npmPkgJsonLint . 'packages/*/package.json'


> @humanmade/[email protected] test:eslint
> cd packages/eslint-config-humanmade && npm test


> @humanmade/[email protected] test
> node fixtures/test-lint-config

Running ESLint on fixture directories. Use --verbose for a detailed report.

Linting `fixtures/fail/**`...
Warning: React version not specified in eslint-plugin-react settings. See https://github.com/yannickcr/eslint-plugin-react#configuration .
ESLint logs errors as expected.

Linting `fixtures/pass/**`...
7 files pass lint.

> @humanmade/[email protected] test:stylelint
> cd packages/stylelint-config && npm test


> @humanmade/[email protected] test
> node fixtures/test-lint-config

All files that should fail log errors as expected.
No errors detected in files that are expected to pass.
~/Code/humanmade/coding-standards move-dev-dependencies                                                                                                                                                                                  8s
❯ 

[tfrommen]

Why this change? With it, the coding standards should actually fail on this...?

[ntwb]

I suspected it was a bug fix in import/order, as all 3 of these imports are external and should not have a blank line between them, particularly the package scope @ character is what was the false positive

Running the tests when including the blank line results in:

Code

import chalk from 'chalk';
import eslint from 'eslint';

import apiFetch from '@wordpress/api-fetch';

Test

/Users/netweb/Code/humanmade/coding-standards/packages/eslint-config-humanmade/fixtures/pass/import-order.js
  6:1  error  There should be no empty line within import group  import/order

✖ 1 problem (1 error, 0 warnings)
  1 error and 0 warnings potentially fixable with the `--fix` option.

1 unexpected error!

[tfrommen]

Hm, so, with the current coding standards (rule config as per #219), it should require @wordpress/* imports to come separately. Which is something that has recently been discussed, and might get changed eventually (see #297).

Not sure if something changed in the ESLint rule...

[ntwb]

Thanks, I'll take a closer look to try to determine why the rule is failing in this PR, as based on the above it shouldn't be, thanks

[tfrommen]

This is looking good!