Zuul includes a simple command line client that may be used to clone repositories with Zuul references applied.
By default, Zuul cloner will clone the project under
basepath which would create sub directories whenever a project name contains slashes. Since you might want to finely tweak the final destination, a clone map lets you change the destination on a per project basis. The configuration is done using a YAML file passed with
With a project hierarchy such as:
You might want to get
project straight in the base path, the clone map would be:
clonemap: - name: 'project' dest: '.'
Then to strip out
thirdparty such that the plugins land under the
/plugins directory of the basepath, you can use regex and capturing groups:
clonemap: - name: 'project' dest: '.' - name: 'thirdparty/(plugins/.*)' dest: '\1'
The resulting workspace will contains:
project -> ./ thirdparty/plugins/plugin1 -> ./plugins/plugin1
The Zuul cloner reuses Zuul parameters such as ZUUL_BRANCH, ZUUL_REF or ZUUL_PROJECT. It will attempt to load them from the environment variables or you can pass them as parameters (in which case it will override the environment variable if it is set). The matching command line parameters use the
zuul prefix hence ZUUL_REF can be passed to the cloner using
The general options that apply are:
When cloning repositories, the destination folder should not exist or
git clone will complain. This happens whenever cloning a sub project before its parent project. For example:
zuul-cloner project/plugins/plugin1 project
Will create the directory
project when cloning the plugin. The cloner processes the clones in the order supplied, so you should swap the projects:
zuul-cloner project project/plugins/plugin1
--cache-dir option can be used to reduce network traffic by cloning from a local repository which may not be up to date.
--cache-dir option is supplied, zuul-cloner will start by cloning any projects it processes from those found in that directory. The URL of origin remote of the resulting clone will be reset to use the
git_base_url and then the remote will be updated so that the repository has all the information in the upstream repository.