def get_metadata_content(base_url, repomd_elem, meta_type):
"""Return the text contents of metadata file.
Provided a url, a repomd root element, and a metadata type, locate the metadata
file's location href, download it from the provided url, un-gzip it, parse it, and
return the root element node.
Don't use this with large repos because it will blow up.
"""
# <ns0:repomd xmlns:ns0="http://linux.duke.edu/metadata/repo">
# <ns0:data type="primary">
# <ns0:checksum type="sha256">[…]</ns0:checksum>
# <ns0:location href="repodata/[…]-primary.xml.gz" />
# …
# </ns0:data>
# …
xpath = "{{{}}}data".format(RPM_NAMESPACES["metadata/repo"])
data_elems = [
elem for elem in repomd_elem.findall(xpath)
if elem.get("type") == meta_type
]
xpath = "{{{}}}location".format(RPM_NAMESPACES["metadata/repo"])
location_href = data_elems[0].find(xpath).get("href")
return read_xml_gz(http_get(os.path.join(base_url, location_href)))
def test_all(self):
"""Sync and publish an RPM repository and verify the checksum."""
# 1. create repo and remote
repo = self.repo_api.create(gen_repo())
self.addCleanup(self.repo_api.delete, repo.pulp_href)
body = gen_rpm_remote()
remote = self.remote_api.create(body)
self.addCleanup(self.remote_api.delete, remote.pulp_href)
# 2. Sync it
repository_sync_data = RpmRepositorySyncURL(remote=remote.pulp_href)
sync_response = self.repo_api.sync(repo.pulp_href,
repository_sync_data)
monitor_task(sync_response.task)
# 3. Publish and distribute
publish_data = RpmRpmPublication(repository=repo.pulp_href)
publish_response = self.publications.create(publish_data)
created_resources = monitor_task(
publish_response.task).created_resources
publication_href = created_resources[0]
self.addCleanup(self.publications.delete, publication_href)
body = gen_distribution()
body["publication"] = publication_href
distribution_response = self.distributions.create(body)
created_resources = monitor_task(
distribution_response.task).created_resources
distribution = self.distributions.read(created_resources[0])
self.addCleanup(self.distributions.delete, distribution.pulp_href)
# 4. check the tag 'sum' is not present in updateinfo.xml
repomd = ElementTree.fromstring(
http_get(os.path.join(distribution.base_url,
"repodata/repomd.xml")))
update_xml_url = self._get_updateinfo_xml_path(repomd)
update_xml_content = http_get(
os.path.join(distribution.base_url, update_xml_url))
update_xml = read_xml_gz(update_xml_content)
update_info_content = ElementTree.fromstring(update_xml)
tags = {elem.tag for elem in update_info_content.iter()}
self.assertNotIn("sum", tags, update_info_content)
def get_checksum_types(self, **kwargs):
"""Sync and publish an RPM repository."""
fixture_url = kwargs.get("fixture_url", RPM_UNSIGNED_FIXTURE_URL)
package_checksum_type = kwargs.get("package_checksum_type")
metadata_checksum_type = kwargs.get("metadata_checksum_type")
policy = kwargs.get("policy", "immediate")
# 1. create repo and remote
repo = self.repo_api.create(gen_repo())
self.addCleanup(self.repo_api.delete, repo.pulp_href)
body = gen_rpm_remote(policy=policy, url=fixture_url)
remote = self.remote_api.create(body)
self.addCleanup(self.remote_api.delete, remote.pulp_href)
# 2. Sync it
repository_sync_data = RpmRepositorySyncURL(remote=remote.pulp_href,
mirror=False)
sync_response = self.repo_api.sync(repo.pulp_href,
repository_sync_data)
monitor_task(sync_response.task)
# 3. Publish and distribute
publish_data = RpmRpmPublication(
repository=repo.pulp_href,
package_checksum_type=package_checksum_type,
metadata_checksum_type=metadata_checksum_type,
)
publish_response = self.publications.create(publish_data)
created_resources = monitor_task(
publish_response.task).created_resources
publication_href = created_resources[0]
self.addCleanup(self.publications.delete, publication_href)
body = gen_distribution()
body["publication"] = publication_href
distribution_response = self.distributions.create(body)
created_resources = monitor_task(
distribution_response.task).created_resources
distribution = self.distributions.read(created_resources[0])
self.addCleanup(self.distributions.delete, distribution.pulp_href)
repomd = ElementTree.fromstring(
http_get(os.path.join(distribution.base_url,
"repodata/repomd.xml")))
data_xpath = "{{{}}}data".format(RPM_NAMESPACES["metadata/repo"])
data_elems = [elem for elem in repomd.findall(data_xpath)]
repomd_checksum_types = {}
primary_checksum_types = {}
checksum_xpath = "{{{}}}checksum".format(
RPM_NAMESPACES["metadata/repo"])
for data_elem in data_elems:
checksum_type = data_elem.find(checksum_xpath).get("type")
repomd_checksum_types[data_elem.get("type")] = checksum_type
if data_elem.get("type") == "primary":
location_xpath = "{{{}}}location".format(
RPM_NAMESPACES["metadata/repo"])
primary_href = data_elem.find(location_xpath).get("href")
primary = ElementTree.fromstring(
read_xml_gz(
http_get(
os.path.join(distribution.base_url,
primary_href))))
package_checksum_xpath = "{{{}}}checksum".format(
RPM_NAMESPACES["metadata/common"])
package_xpath = "{{{}}}package".format(
RPM_NAMESPACES["metadata/common"])
package_elems = [
elem for elem in primary.findall(package_xpath)
]
pkg_checksum_type = package_elems[0].find(
package_checksum_xpath).get("type")
primary_checksum_types[package_elems[0].get(
"type")] = pkg_checksum_type
return repomd_checksum_types, primary_checksum_types