IvyRepo contains the tool chain to populate the Nepherte Ivy Repository. This includes the packaging instructions for all hosted artifacts. Browse this repository for a current list of build files. Keep on reading for more info on the tool chain.
Usage: IvyRepo.sh [-h] <command> <artifact>
Script to manage the Nepherte Ivy Repository
-h,--help Print out a help message with examples
<command> Possible values are resolve, install and publish
<artifact> The artifact id as organisation:module:revision
This repository is made possible by Ivy's Packager Resolver. This resolver supports downloading, extracting and packaging artifacts from upstream repositories. It allows for a clear separation between artifact meta-data and actually hosting them.
For every artifact in the production repository, there is an Ivy module
definition (ivy.xml) that contains the artifact meta-data, plus instructions
on where to download and how to package them (packager.xml).
The tool chain can read the artifact meta-data and build instructions. The repackaged artifacts are published in an artifact repository, on which projects can rely for resolving their dependencies.
Follow these steps to install the tool chain:
- Install Ant into
$ANT_HOME - Download Ivy to
$USER_HOME\.ant\lib - Use Git to clone this repository to your device
- Configure the Production Resolver accordingly
The tool chain relies on Ant to run the available commands. Make sure to set the
environment variable $ANT_HOME. Ant in turn depends on the presence of a Java
Runtime Environment, available under $JAVA_HOME or on the system $PATH.
The tool chain automatically downloads Ivy and its dependencies. The jars will
be available in $USER_HOME\.ant\lib. The environment variable $IVY_HOME
eventually determines the location in which to cache resolved artifacts.
The executable IvyRepo.sh can be used to manage an online Ivy repository:
from building artifacts to publishing them. It makes upgrading artifacts to
newer versions very easy. Out-of-the-box only artifacts in this git repository
are supported, but new artifacts can be added in the blink of an eye.
To download and build an ivy artifact, use the following command:
./IvyRepo.sh resolve <organisation>:<module>:<revision>
To install an artifact into the staging repo, use the following command:
./IvyRepo.sh install <organisation>:<module>:<revision>
To publish an artifact into the production repo, use the following command:
./IvyRepo.sh publish <organisation>:<module>:<revision>
Changing the behavior of the tool chain boils down to configuring the packager
resolver (packager), the staging resolver (filesystem), or the production
resolver (sftp). Each resolver has a dedicated configuration file.
It is also possible to replace each resolver - say your production repository is
not accessible via sftp - by directly editing the Ivy settings file
(ivysettings.xml).
The packager resolver (packager) reconstructs artifacts based on upstream
sources by reading the Ivy meta-data (ivy.xml) and packaging instructions
(packager.xml) in the source repository, which defaults to src/repo.
The location of the packager repository is specified by the directory
pkg.repo.dir in packager.properties. The exact location of the Ivy
meta-data and packaging instructions are specified by pkg.repo.pattern.
The staging resolver (filesystem) acts as a staging area before publishing
artifacts into production. It is highly recommended to try this repository first
before committing anything to your production repository. The target directory
defaults to target/repo.
The location of the staging repository is specified by the directory
stage.repo.dir in staging.properties. The exact location and names of the
artifacts are determined by stage.repo.pattern and stage.artifact.pattern.
The production resolver (sftp) defines the repository that hosts the published
artifacts. The configuration file production.properties contains the
credentials to connect to the online repository. Make sure to update them.