iw.recipe.subversion

This product is used to download packages from svn and tar them in a directory.

We need some buildout vars:

>>> data_dir = join(test_dir, 'data')
>>> parts_dir = join(data_dir, 'parts')
>>> buildout = {'instance': {'location': test_dir},
...             'buildout': {'directory': test_dir,
...             'install-from-cache': False,
...             'download-cache':False,
...             'parts-directory': parts_dir}}
>>> name = 'svn-packages'

The recipe will checkout the package from his repository and put it in the section. For testing, we need a local repository:

>>> repository = join(test_dir, 'test_repos')
>>> if isdir(repository): rmtree(repository)
>>> copytree(join(test_dir, 'repos'), repository)
>>> if isdir(os.path.join(parts_dir, 'svn-packages')):
...     rmtree(os.path.join(parts_dir, 'svn-packages'))

Then, the recipe should work:

>>> options = {
...   'urls':'file:///%s/my_package/trunk my_package' % repository}
>>> from iw.recipe.subversion import Recipe
>>> recipe = Recipe(buildout, name, options)
>>> os.path.join(parts_dir, 'svn-packages') == recipe.install()
True

Ok, we got it:

>>> ls(parts_dir, 'svn-packages')
my_package

If a download-cache directory is given, then an archive is created in the specified path:

>>> cache = join(data_dir, 'cache')
>>> buildout['buildout']['download-cache'] = cache
>>> recipe = Recipe(buildout, name, options)

>>> if isdir(cache): rmtree(cache)

>>> files = recipe.install()

>>> ls(cache)
my_package-r4.tar.gz

Ok, now we can work offline. The package will be installed from the cache directory:

>>> rmtree(repository)
>>> rmtree(parts_dir)
>>> buildout['buildout']['install-from-cache'] = 'true'
>>> buildout['buildout']['download-cache'] = cache
>>> recipe = Recipe(buildout, name, options)
>>> files = recipe.install()

>>> ls(parts_dir, 'svn-packages')
my_package

>>> ls(parts_dir, 'svn-packages', 'my_package')
__init__.py
README.txt
sub

Well, it’s ok for a trunk url. But when working with tagged version, we don’t want to use subversion if we already have an archive:

>>> copytree(join(test_dir, 'repos'), repository)
>>> rmtree(parts_dir)
>>> options = {
...   'urls':'file:///%s/my_package/tags/v1_0_0 my_package' % repository}
>>> buildout['buildout']['install-from-cache'] = 'false'
>>> buildout['buildout']['download-cache'] = cache
>>> recipe = Recipe(buildout, name, options)

At the first time, we need to update and built the archive:

>>> recipe.need_update
True

>>> files = recipe.install()

But when we have an archive, we don’t need to:

>>> recipe.need_update
False

Then we work offline in an online mode:

>>> rmtree(repository)
>>> files = recipe.install()

Test update:

>>> recipe.update() is not None
True