Skip to content

Commit

Permalink
chore: Update README
Browse files Browse the repository at this point in the history
  • Loading branch information
js-timbirkett committed Jan 27, 2021
1 parent 2a6b1a2 commit 554a2bf
Showing 1 changed file with 60 additions and 1 deletion.
61 changes: 60 additions & 1 deletion README.md
Original file line number Diff line number Diff line change
@@ -1,2 +1,61 @@
# jenkins-get-library-version
A Jenkins shared library helper function to help load different versions based on job name
A Jenkins shared library helper function to help load different versions based on job name.

## What is this?
On my travels through Jenkins, Groovy Shared Library, and Declarative Pipeline. I read the internet, and watched
many Jenkins World talks, often with conflicting opinions and ideas.

One of the talks had an interesting pattern with some plugin or something that ran code before loading a projects
`Jenkinsfile`. I can't remember which talk it is, but I'll add a link here if I ever find it.

This is an attempt at creating a small library that could be used to do the same thing in a `Jenkinsfile` like this:
```groovy
#!groovy
def libraryVersion = getLibraryVersion()
library "myRealLibrary@${libraryVersion}"
pipeline {
agent any
stages {
stage('Library Dynamically Loaded!') {
steps {
echo "Using library version: ${libraryVersion}"
}
}
}
}
```

## Why is this?
Typically, pipeline library versioning goes one of 2 ways... People either pin to `main` and en(joy|dure) the
exhilaration of innovation, or they pin to a branch, tag, sha of a library and make moving forwards difficult.

I've decided that having different versions based on `JOB_NAME` is sensible. This way you could match jobs beginning
with `pipeline-test` and have them use some "latest" version (tag, branch, whatever...), and all other jobs could
reference a slower moving "stable" version.

## How is this?
Most of the configuration is done with environment variables that should be set at the global configuration level:
### Required:
- `LIBRARY_REPO` - The Github `org/repo` that your main shared library lives in. This is used to query the Github API
for the latest release `tag_name`.

- `LIBRARY_LATEST_JOB_MATCHER` - The pattern to match jobs for the "latest" library version.
Example: `^pipeline-test|^my-cool-jobs`.

### Optional
- `LIBRARY_DEFAULT_STABLE_VERSION` - Override the default "stable" version. Default: `stable`.
- `LIBRARY_DEFAULT_LATEST_VERSION` - Override the default "latest" version. Sometimes, you might just want to use
the `main` branch and not worry about creating Github releases. Defaults to Github "Latest Release" tag.

- `GITHUB_API_URL` - Override the Github API URL for GIthub Enterprise users. Default: `api.github.com`.
- `GITHUB_CREDENTIALS_ID` - The ID of a credential within Jenkins that holds the username and personal access token to use
when querying the Github API. Required if `LIBRARY_REPO` is private.

## Library Configuration
1. Add this library to the Global Configuration and make sure that it is implicitly loaded. Meaning that it will be
available in any pipeline.

2. Add your main library without implicitly loading it. Your `Jenkinsfile` will be responsible for loading it explicitly.
Sadly, you can't load a library dynamically if it is already implicitly loaded. Jenkins will always use the first library
configuration that it finds.

0 comments on commit 554a2bf

Please sign in to comment.