/hello-reason

an example esy-powered Reason project

Primary LanguageJavaScriptMIT LicenseMIT

hello-reason

Build Status

A project which demonstrates a Reason workflow with Esy.

Usage

You need Esy, you can install the beta using npm:

% npm install -g esy@latest

NOTE: Make sure esy --version returns at least 0.5.4 for this project to build.

On Linux you will need m4 package to run esy build.

% sudo apt-get install m4

Then run the esy command from this project root to install and build dependencies.

% esy

Now you can run your editor within the environment (which also includes merlin):

% esy $EDITOR
% esy vim

Alternatively you can try vim-reasonml which loads esy project environments automatically.

After you make some changes to source code, you can re-run project's build again with the same simple esy command.

% esy

And test compiled executable (runs scripts.tests specified in package.json):

% esy test

You can format your code (runs scripts.format specified in package.json):

% esy format

If you plan to add some .ml(i) files and want to enable formatting you can run: esy add "@opam/ocamlformat" and esy format will be automatically extended.

Documentation for the libraries in the project can be generated with (runs scripts.doc specified in package.json):

% esy doc
% esy open '#{self.target_dir}/default/_doc/_html/index.html'

Shell into environment:

% esy shell

Create Prebuilt Release:

esy allows creating prebuilt binary packages for your current platform, with no dependencies.

% esy npm-release
% cd _release
% npm publish

Continuous Integration:

hello-reason includes CI configuration for Azure DevOps pipelines out of the box.

  • Create your Azure DevOps account.
  • Add a new project, and point that new Azure DevOps project to your github repo that includes the CI (./azure-pipelines.yml and the .ci/ directory) from hello-reason.
  • Create a new Pipeline within that project.
    • When asked how to configure the new pipeline, select the option to use existing configuration inside the repo.

The CI is configured to build caches on the master branch, and also any branch named one of (global, release-*, releases-*). That means that pull requests to any branch with those names will be fast, once you have landed at least one commit to that branch. The first time you submit a pull request to one of those branches, the builds will be slow but then subsequent pull requests will be faster once a pull request is merged to it.