Skip to content

Latest commit

 

History

History
50 lines (32 loc) · 2.4 KB

CONTRIBUTING.md

File metadata and controls

50 lines (32 loc) · 2.4 KB
Raw

Contributing

Inlang is set up as monorepo with Nx (with caching some compile process for better dev/build performance).

Getting started

  1. Clone/Fork inlang's repository.
  2. Open the repository folder in Visual Studio Code.
  3. Run doppler setup to select the development config.
  4. doppler run -- pnpm install to install dependencies.
  5. doppler run -- pnpm dev to run the development environment.
  6. doppler run -- pnpm test to run the tests.
  7. doppler run -- pnpm build to compile a production build.

For Windows users

To work on Windows, you will need to change the shell, which is used to execute NPM scripts to bash, before running doppler run -- pnpm dev. A Git installation usually comes with a bash binary.

  • pnpm config set script-shell "C:\\Program Files\\git\\bin\\bash.exe" to set the shell to bash provided by your Git installation

Debugging

  1. Run doppler run -- pnpm dev.
  2. Several debug launch configs can be found in the Visual Studio Code side menu. Select the correct one depending on the subject of your contribution and launch Debug mode.

Tips for running tests locally

Commands below assume that you are in the root of the monorepo. You may find it helpful to alias p='doppler run -- pnpm' for brevity.

Before running tests or debugging locally, first run doppler run -- pnpm install and doppler run -- pnpm build

If you don't run a build, there is a chance that tests will use builds from a previous branch.

To rebuild everything without the nx cache, do NX_SKIP_NX_CACHE=true doppler run -- pnpm build. This is useful if you're building locally with a slow network.

Most tests will work without a local server, however the machine translation test (cli/src/commands/machine/translate.test.ts) requires the @inlang/server to be listening on localhost:3000 for rpc fetch() calls to the google translate api.

To make this test pass locally, start (all) the dev servers with doppler run -- pnpm dev or
start just the @inlang/server with doppler run -- pnpm --filter @inlang/server dev

To run tests for just one package e.g. cli doppler run -- pnpm --filter @inlang/cli test or
doppler run -- pnpm nx run @inlang/cli:test

To run the full test suite: doppler run -- pnpm test

To re-run all the tests without the nx cache, do NX_SKIP_NX_CACHE=true doppler run -- pnpm test