changesets-gitlab

GitLab CI cli for changesets like its GitHub Action, it creates a merge request with all of the package versions updated and changelogs updated and when there are new changesets on master, the MR will be updated. When you're ready, you can merge the merge request and you can either publish the packages to npm manually or setup the action to do it for you.

Usage

Inputs

Environment valuables starts with INPUT_, case insensitive

• publish - The command to use to build and publish packages
• version - The command to update version, edit CHANGELOG, read and delete changesets. Default to changeset version if not provided
• commit - The commit message to use. Default to Version Packages
• title - The merge request title. Default to Version Packages

Only available in changesets-gitlab

• published - Command executed after published
• only_changesets - Command executed on only changesets detected

Outputs

• published - A boolean value to indicate whether a publishing is happened or not
• publishedPackages - A JSON array to present the published packages. The format is [{"name": "@xx/xx", "version": "1.2.0"}, {"name": "@xx/xy", "version": "0.8.9"}]

Environment Variables

GLOBAL_AGENT_HTTP_PROXY  # optional, if you're using custom GitLab service under proxy
GLOBAL_AGENT_HTTPS_PROXY # As above but for https requests
GLOBAL_AGENT_NO_PROXY    # Like above but for no proxied requests

# http_proxy, https_proxy, no_proxy environment variables are supported at the same time

GITLAB_HOST # optional, if you're using custom GitLab host

GITLAB_TOKEN            # required, token with accessibility to push
GITLAB_TOKEN_TYPE       # optional, type of the provided token in GITLAB_TOKEN. defaults to personal access token. can be `job` if you provide the Gitlab CI_JOB_TOKEN or `oauth` if you use Gitlab Oauth token
GITLAB_CI_USER_NAME     # required, username with accessibility to push, used in pairs of the above token (if it was personal access token)
GITLAB_CI_USER_EMAIL    # optional, default `gitlab[bot]@users.noreply.gitlab.com`
DEBUG_GITLAB_CREDENTIAL # optional, default `false`

Example workflow

Without Publishing

Create a file at .gitlab-ci.yml with the following content.

stages:
  - comment
  - release

before_script: yarn --frozen-lockfile

comment:
  image: node:lts-alpine
  stage: comment
  only: merge_requests
  script: yarn changesets-gitlab comment # comment automatically like https://github.com/changesets/bot

release:
  image: node:lts-alpine
  only: main
  script: yarn changesets-gitlab

With Publishing

Before you can setup this action with publishing, you'll need to have an npm token that can publish the packages in the repo you're setting up the action for and doesn't have 2FA on publish enabled (2FA on auth can be enabled). You'll also need to add it as a custom environment variable on your GitLab repo with the name NPM_TOKEN. Once you've done that, you can create a file at .gitlab-ci.yml with the following content.

stages:
  - comment
  - release

before_script: yarn --frozen-lockfile

comment:
  image: node:lts-alpine
  stage: comment
  only: merge_requests
  script: yarn changesets-gitlab comment

release:
  image: node:lts-alpine
  only: main
  script: yarn changesets-gitlab
  variables:
    INPUT_PUBLISH: yarn release

By default the GitLab CI cli creates a .npmrc file with the following content:

//registry.npmjs.org/:_authToken=${process.env.NPM_TOKEN}

However, if a .npmrc file is found, the GitLab CI cli does not recreate the file. This is useful if you need to configure the .npmrc file on your own. For example, you can add a step before running the Changesets GitLab CI cli:

script: |
  cat << EOF > "$HOME/.npmrc"
    [email protected]
    //registry.npmjs.org/:_authToken=$NPM_TOKEN
  EOF

With version script

If you need to add additional logic to the version command, you can do so by using a version script.

If the version script is present, this action will run that script instead of changeset version, so please make sure that your script calls changeset version at some point. All the changes made by the script will be included in the MR.

stages:
  - comment
  - release

before_script: yarn --frozen-lockfile

comment:
  image: node:lts-alpine
  stage: comment
  only:
    - merge_requests
  script: yarn changesets-gitlab comment

release:
  image: node:lts-alpine
  only: main
  script: yarn changesets-gitlab
  variables:
    INPUT_VERSION: yarn version

With Yarn 2 / Plug'n'Play

If you are using Yarn Plug'n'Play, you should use a custom version command so that the action can resolve the changeset CLI:

stages:
  - comment
  - release

before_script: yarn --frozen-lockfile

comment:
  image: node:lts-alpine
  stage: comment
  only:
    - merge_requests
  script: yarn changesets-gitlab comment

release:
  image: node:lts-alpine
  only: main
  script: yarn changesets-gitlab
  variables:
    INPUT_VERSION: yarn changeset version

You may also want to run yarn install after the changeset verion command to update the yarn.lock in the version MR. You need to disable immutable lock file setting using an env variable:

release:
  image: node:lts-alpine
  only: main
  script: yarn changesets-gitlab
  variables:
    YARN_ENABLE_IMMUTABLE_INSTALLS: 'false'
    INPUT_VERSION: yarn update-versions

And your update-versions script would be:

{
  "update-versions": "changeset version && yarn install"
}