def publish_repo(self, repo, publish_conduit, config):
"""
Publishes the given repository.
While this call may be implemented using multiple threads, its execution
from the Pulp server's standpoint should be synchronous. This call should
not return until the publish is complete.
It is not expected that this call be atomic. Should an error occur, it
is not the responsibility of the distributor to rollback any changes
that have been made.
:param repo: metadata describing the repository
:type repo: pulp.plugins.model.Repository
:param publish_conduit: provides access to relevant Pulp functionality
:type publish_conduit: pulp.plugins.conduits.repo_publish.RepoPublishConduit
:param config: plugin configuration
:type config: pulp.plugins.config.PluginConfiguration
:return: report describing the publish run
:rtype: pulp.plugins.model.PublishReport
"""
_logger.debug('Publishing Python repository: %s' % repo.id)
self._publisher = PythonPublisher(repo, publish_conduit, config)
return self._publisher.process_lifecycle()
def publish_repo(self, repo, publish_conduit, config):
"""
Publishes the given repository.
While this call may be implemented using multiple threads, its execution
from the Pulp server's standpoint should be synchronous. This call should
not return until the publish is complete.
It is not expected that this call be atomic. Should an error occur, it
is not the responsibility of the distributor to rollback any changes
that have been made.
:param repo: metadata describing the repository
:type repo: pulp.plugins.model.Repository
:param publish_conduit: provides access to relevant Pulp functionality
:type publish_conduit: pulp.plugins.conduits.repo_publish.RepoPublishConduit
:param config: plugin configuration
:type config: pulp.plugins.config.PluginConfiguration
:return: report describing the publish run
:rtype: pulp.plugins.model.PublishReport
"""
_logger.debug('Publishing Python repository: %s' % repo.id)
self._publisher = PythonPublisher(repo, publish_conduit, config)
return self._publisher.process_lifecycle()
class PythonDistributor(Distributor):
"""
This distributor is for use with the pip client.
"""
@classmethod
def metadata(cls):
"""
Used by Pulp to classify the capabilities of this distributor. The
following keys must be present in the returned dictionary:
* id - Programmatic way to refer to this distributor. Must be unique
across all distributors. Only letters and underscores are valid.
* display_name - User-friendly identification of the distributor.
* types - List of all content type IDs that may be published using this
distributor.
:return: keys and values listed above
:rtype: dict
"""
return {
'id': constants.DISTRIBUTOR_TYPE_ID,
'display_name': _('Python Python Distributor'),
'types': [constants.PACKAGE_TYPE_ID]
}
def __init__(self):
"""
Initialize the PythonDistributor.
"""
super(PythonDistributor, self).__init__()
self._publisher = None
self.canceled = False
def publish_repo(self, repo, publish_conduit, config):
"""
Publishes the given repository.
While this call may be implemented using multiple threads, its execution
from the Pulp server's standpoint should be synchronous. This call should
not return until the publish is complete.
It is not expected that this call be atomic. Should an error occur, it
is not the responsibility of the distributor to rollback any changes
that have been made.
:param repo: metadata describing the repository
:type repo: pulp.plugins.model.Repository
:param publish_conduit: provides access to relevant Pulp functionality
:type publish_conduit: pulp.plugins.conduits.repo_publish.RepoPublishConduit
:param config: plugin configuration
:type config: pulp.plugins.config.PluginConfiguration
:return: report describing the publish run
:rtype: pulp.plugins.model.PublishReport
"""
_logger.debug('Publishing Python repository: %s' % repo.id)
self._publisher = PythonPublisher(repo, publish_conduit, config)
return self._publisher.process_lifecycle()
def cancel_publish_repo(self):
"""
Call cancellation control hook.
"""
_logger.debug('Canceling Python repository publish')
self.canceled = True
if self._publisher is not None:
self._publisher.cancel()
def distributor_removed(self, repo, config):
"""
Called when a distributor of this type is removed from a repository.
This hook allows the distributor to clean up any files that may have
been created during the actual publishing.
The distributor may use the contents of the working directory in cleanup.
It is not required that the contents of this directory be deleted by
the distributor; Pulp will ensure it is wiped following this call.
If this call raises an exception, the distributor will still be removed
from the repository and the working directory contents will still be
wiped by Pulp.
:param repo: metadata describing the repository
:type repo: pulp.plugins.model.Repository
:param config: plugin configuration
:type config: pulp.plugins.config.PluginCallConfiguration
"""
# remove the directories that might have been created for this repo/distributor
dir_list = [
repo.working_dir,
configuration.get_master_publish_dir(repo, config),
configuration.get_web_publish_dir(repo, config)
]
for repo_dir in dir_list:
# in case repo_dir is None
# ignore_errors set to True does not cover this.
if repo_dir:
shutil.rmtree(repo_dir, ignore_errors=True)
def validate_config(self, repo, config, config_conduit):
"""
Allows the distributor to check the contents of a potential configuration
for the given repository. This call is made both for the addition of
this distributor to a new repository as well as updating the configuration
for this distributor on a previously configured repository. The implementation
should use the given repository data to ensure that updating the
configuration does not put the repository into an inconsistent state.
The return is a tuple of the result of the validation (True for success,
False for failure) and a message. The message may be None and is unused
in the success case. For a failed validation, the message will be
communicated to the caller so the plugin should take i18n into
consideration when generating the message.
The related_repos parameter contains a list of other repositories that
have a configured distributor of this type. The distributor configurations
is found in each repository in the "plugin_configs" field.
:param repo: metadata describing the repository to which the
configuration applies
:type repo: pulp.plugins.model.Repository
:param config: plugin configuration instance; the proposed repo
configuration is found within
:type config: pulp.plugins.config.PluginCallConfiguration
:param config_conduit: Configuration Conduit;
:type config_conduit: pulp.plugins.conduits.repo_config.RepoConfigConduit
:return: tuple of (bool, str) to describe the result
:rtype: tuple
:raises: PulpCodedValidationException if any validations failed
"""
return configuration.validate_config(config)
class PythonDistributor(Distributor):
"""
This distributor is for use with the pip client.
"""
@classmethod
def metadata(cls):
"""
Used by Pulp to classify the capabilities of this distributor. The
following keys must be present in the returned dictionary:
* id - Programmatic way to refer to this distributor. Must be unique
across all distributors. Only letters and underscores are valid.
* display_name - User-friendly identification of the distributor.
* types - List of all content type IDs that may be published using this
distributor.
:return: keys and values listed above
:rtype: dict
"""
return {
'id': constants.DISTRIBUTOR_TYPE_ID,
'display_name': _('Python Python Distributor'),
'types': [constants.PACKAGE_TYPE_ID]
}
def __init__(self):
"""
Initialize the PythonDistributor.
"""
super(PythonDistributor, self).__init__()
self._publisher = None
self.canceled = False
def publish_repo(self, repo, publish_conduit, config):
"""
Publishes the given repository.
While this call may be implemented using multiple threads, its execution
from the Pulp server's standpoint should be synchronous. This call should
not return until the publish is complete.
It is not expected that this call be atomic. Should an error occur, it
is not the responsibility of the distributor to rollback any changes
that have been made.
:param repo: metadata describing the repository
:type repo: pulp.plugins.model.Repository
:param publish_conduit: provides access to relevant Pulp functionality
:type publish_conduit: pulp.plugins.conduits.repo_publish.RepoPublishConduit
:param config: plugin configuration
:type config: pulp.plugins.config.PluginConfiguration
:return: report describing the publish run
:rtype: pulp.plugins.model.PublishReport
"""
_logger.debug('Publishing Python repository: %s' % repo.id)
self._publisher = PythonPublisher(repo, publish_conduit, config)
return self._publisher.process_lifecycle()
def cancel_publish_repo(self):
"""
Call cancellation control hook.
"""
_logger.debug('Canceling Python repository publish')
self.canceled = True
if self._publisher is not None:
self._publisher.cancel()
def distributor_removed(self, repo, config):
"""
Called when a distributor of this type is removed from a repository.
This hook allows the distributor to clean up any files that may have
been created during the actual publishing.
The distributor may use the contents of the working directory in cleanup.
It is not required that the contents of this directory be deleted by
the distributor; Pulp will ensure it is wiped following this call.
If this call raises an exception, the distributor will still be removed
from the repository and the working directory contents will still be
wiped by Pulp.
:param repo: metadata describing the repository
:type repo: pulp.plugins.model.Repository
:param config: plugin configuration
:type config: pulp.plugins.config.PluginCallConfiguration
"""
# remove the directories that might have been created for this repo/distributor
dir_list = [repo.working_dir,
configuration.get_master_publish_dir(repo, config),
configuration.get_web_publish_dir(repo, config)]
for repo_dir in dir_list:
# in case repo_dir is None
# ignore_errors set to True does not cover this.
if repo_dir:
shutil.rmtree(repo_dir, ignore_errors=True)
def validate_config(self, repo, config, config_conduit):
"""
Allows the distributor to check the contents of a potential configuration
for the given repository. This call is made both for the addition of
this distributor to a new repository as well as updating the configuration
for this distributor on a previously configured repository. The implementation
should use the given repository data to ensure that updating the
configuration does not put the repository into an inconsistent state.
The return is a tuple of the result of the validation (True for success,
False for failure) and a message. The message may be None and is unused
in the success case. For a failed validation, the message will be
communicated to the caller so the plugin should take i18n into
consideration when generating the message.
The related_repos parameter contains a list of other repositories that
have a configured distributor of this type. The distributor configurations
is found in each repository in the "plugin_configs" field.
:param repo: metadata describing the repository to which the
configuration applies
:type repo: pulp.plugins.model.Repository
:param config: plugin configuration instance; the proposed repo
configuration is found within
:type config: pulp.plugins.config.PluginCallConfiguration
:param config_conduit: Configuration Conduit;
:type config_conduit: pulp.plugins.conduits.repo_config.RepoConfigConduit
:return: tuple of (bool, str) to describe the result
:rtype: tuple
:raises: PulpCodedValidationException if any validations failed
"""
return configuration.validate_config(config)