[Wallaby-Only] Ansible lint custom rule expansion
Expands the existing ansible lint rule to aling
with proposed metadata changes.
Categories and products metadata keys were added to the cep-ansible-installed
playbook in order to make it compliant.
Depends-On: https://review.opendev.org/c/openstack/validations-libs/+/800721
Conflicts:
.ansible-lint
Resolved in favor of Iff0f37b1a76d81df7e5b20d76cefd0c879a150f6
Signed-off-by: Jiri Podivin <jpodivin@redhat.com>
Change-Id: Ib5fc841e23001816612dcdfd8c3335ae99ba2722
(cherry picked from commit 84c91cfcf4)
This commit is contained in:
Jiri Podivin
parent
4c3d77cbf3
commit
c18200cc8b
|
|
@ -1,6 +1,5 @@
|
|||
exclude_paths:
|
||||
- releasenotes/
|
||||
- playbooks/ceph-pg.yaml
|
||||
- ci/playbooks/
|
||||
parseable: true
|
||||
quiet: false
|
||||
|
|
|
|||
|
|
@ -1,6 +1,6 @@
|
|||
import os
|
||||
import yaml
|
||||
|
||||
from ansiblelint.errors import MatchError
|
||||
from ansiblelint.rules import AnsibleLintRule
|
||||
|
||||
|
||||
|
|
@ -20,6 +20,14 @@ class ValidationHasMetadataRule(AnsibleLintRule):
|
|||
- group1
|
||||
- group2
|
||||
- group3
|
||||
categories:
|
||||
- category1
|
||||
- category2
|
||||
- category3
|
||||
products:
|
||||
- product1
|
||||
- product2
|
||||
- product3
|
||||
"""
|
||||
|
||||
description = (
|
||||
|
|
@ -35,102 +43,155 @@ class ValidationHasMetadataRule(AnsibleLintRule):
|
|||
"The validation playbook must contain "
|
||||
"a 'metadata' dictionary under vars"
|
||||
)
|
||||
no_groups_found = \
|
||||
"*metadata* should contain a list of group (groups)"
|
||||
no_classification_found = \
|
||||
"*metadata* should contain a list of {classification}"
|
||||
|
||||
unknown_groups_found = (
|
||||
"Unkown group(s) '{}' found! "
|
||||
"The official list of groups are '{}'. "
|
||||
"To add a new validation group, please add it in the groups.yaml "
|
||||
"file at the root of the tripleo-validations project."
|
||||
unknown_classifications_found = (
|
||||
"Unkown {classification_key}(s) '{unknown_classification}' found! "
|
||||
"The official list of {classification_key} are '{known_classification}'. "
|
||||
)
|
||||
|
||||
def get_groups(self):
|
||||
"""Returns a list of group names supported by
|
||||
tripleo-validations by reading 'groups.yaml'
|
||||
file located in the base direcotry.
|
||||
how_to_add_classification = {
|
||||
'groups': (
|
||||
"To add a new validation group, please add it in the groups.yaml "
|
||||
"file at the root of the tripleo-validations project."
|
||||
)
|
||||
}
|
||||
|
||||
def get_classifications(self, classification='groups'):
|
||||
"""Returns a list classification names
|
||||
defined for tripleo-validations in the '{classification}.yaml' file
|
||||
located in the base repo directory.
|
||||
"""
|
||||
results = []
|
||||
file_path = os.path.abspath(classification + '.yaml')
|
||||
|
||||
grp_file_path = os.path.abspath('groups.yaml')
|
||||
try:
|
||||
with open(file_path, "r") as definitions:
|
||||
contents = yaml.safe_load(definitions)
|
||||
except (PermissionError, OSError):
|
||||
raise RuntimeError(
|
||||
"{}.yaml file at '{}' inacessible.".format(
|
||||
classification,
|
||||
file_path))
|
||||
|
||||
with open(grp_file_path, "r") as grps:
|
||||
contents = yaml.safe_load(grps)
|
||||
|
||||
for grp_name, grp_desc in sorted(contents.items()):
|
||||
results.append(grp_name)
|
||||
results = [name for name, _ in contents.items()]
|
||||
|
||||
return results
|
||||
|
||||
def check_classification(self, metadata, path,
|
||||
classification_key, strict=False):
|
||||
"""Check validatity of validation classifications,
|
||||
such as groups, categories and products.
|
||||
This one is tricky.
|
||||
Empty lists in python evaluate as false
|
||||
So we can't just check for truth value of returned list.
|
||||
Instead we have to compare the returned value with `None`.
|
||||
"""
|
||||
classification = metadata.get(classification_key, None)
|
||||
|
||||
if classification is None:
|
||||
return MatchError(
|
||||
message=self.no_classification_found.format(
|
||||
classification=classification_key
|
||||
),
|
||||
filename=path,
|
||||
details=str(metadata))
|
||||
else:
|
||||
if not isinstance(classification, list):
|
||||
return MatchError(
|
||||
message="*{}* must be a list".format(classification_key),
|
||||
filename=path,
|
||||
details=str(metadata))
|
||||
elif strict:
|
||||
classifications = self.get_classifications(classification_key)
|
||||
unknown_classifications = list(
|
||||
set(classification) - set(classifications))
|
||||
if unknown_classifications:
|
||||
message = self.unknown_classifications_found.format(
|
||||
unknown_classification=unknown_classifications,
|
||||
known_classification=classifications,
|
||||
classification_key=classification_key)
|
||||
message += self.how_to_add_classification.get(classification_key, "")
|
||||
return MatchError(
|
||||
message=message,
|
||||
filename=path,
|
||||
details=str(metadata))
|
||||
|
||||
def matchplay(self, file, data):
|
||||
results = []
|
||||
path = file['path']
|
||||
|
||||
if file['type'] == 'playbook':
|
||||
if path.startswith("playbooks/") or \
|
||||
path.find("tripleo-validations/playbooks/") > 0:
|
||||
if path.startswith("playbooks/") \
|
||||
or "tripleo-validations/playbooks/" in path:
|
||||
|
||||
# *hosts* line check
|
||||
hosts = data.get('hosts', None)
|
||||
if not hosts:
|
||||
return [({
|
||||
path: data
|
||||
}, "No *hosts* key found in the playbook")]
|
||||
results.append(
|
||||
MatchError(
|
||||
message="No *hosts* key found in the playbook",
|
||||
filename=path,
|
||||
details=str(data)))
|
||||
|
||||
# *vars* lines check
|
||||
vars = data.get('vars', None)
|
||||
if not vars:
|
||||
return [({
|
||||
path: data
|
||||
}, self.no_vars_found)]
|
||||
results.append(
|
||||
MatchError(
|
||||
message=self.no_vars_found,
|
||||
filename=path,
|
||||
details=str(data)))
|
||||
else:
|
||||
if not isinstance(vars, dict):
|
||||
return [({path: data}, '*vars* should be a dictionary')]
|
||||
results.append(
|
||||
MatchError(
|
||||
message='*vars* must be a dictionary',
|
||||
filename=path,
|
||||
details=str(data)))
|
||||
|
||||
# *metadata* lines check
|
||||
metadata = data['vars'].get('metadata', None)
|
||||
if metadata:
|
||||
if not isinstance(metadata, dict):
|
||||
return [(
|
||||
{path: data},
|
||||
'*metadata* should be a dictionary')]
|
||||
results.append(
|
||||
MatchError(
|
||||
message='*metadata* must be a dictionary',
|
||||
filename=path,
|
||||
details=str(data)))
|
||||
else:
|
||||
return [({path: data}, self.no_meta_found)]
|
||||
results.append(
|
||||
MatchError(
|
||||
message=self.no_meta_found,
|
||||
filename=path,
|
||||
details=str(data)))
|
||||
|
||||
# *metadata>[name|description] lines check
|
||||
for info in ['name', 'description']:
|
||||
if not metadata.get(info, None):
|
||||
results.append((
|
||||
{path: data},
|
||||
'*metadata* should contain a %s key' % info))
|
||||
results.append(
|
||||
MatchError(
|
||||
message='*metadata* must contain a %s key' % info,
|
||||
filename=path,
|
||||
details=str(data)))
|
||||
continue
|
||||
if not isinstance(metadata.get(info), str):
|
||||
results.append((
|
||||
{path: data},
|
||||
'*%s* should be a string' % info))
|
||||
results.append(
|
||||
MatchError(
|
||||
message='*%s* should be a string' % info,
|
||||
filename=path,
|
||||
details=str(data)))
|
||||
|
||||
# *metadata>groups* lines check
|
||||
if not metadata.get('groups', None):
|
||||
results.append((
|
||||
{path: data},
|
||||
self.no_groups_found))
|
||||
else:
|
||||
if not isinstance(metadata.get('groups'), list):
|
||||
results.append((
|
||||
{path: data},
|
||||
'*groups* should be a list'))
|
||||
else:
|
||||
groups = metadata.get('groups')
|
||||
group_list = self.get_groups()
|
||||
unknown_groups_list = list(
|
||||
set(groups) - set(group_list))
|
||||
if unknown_groups_list:
|
||||
results.append((
|
||||
{path: data},
|
||||
self.unknown_groups_found.format(
|
||||
unknown_groups_list,
|
||||
group_list)
|
||||
))
|
||||
return results
|
||||
#Checks for metadata we use to classify validations.
|
||||
#Groups, categories and products
|
||||
for classification in ['categories', 'products', 'groups']:
|
||||
classification_error = self.check_classification(
|
||||
metadata,
|
||||
path,
|
||||
classification,
|
||||
strict=(classification == 'groups'))
|
||||
|
||||
if classification_error:
|
||||
results.append(classification_error)
|
||||
|
||||
return results
|
||||
|
|
|
|||
|
|
@ -8,6 +8,11 @@
|
|||
groups:
|
||||
- pre-deployment
|
||||
- pre-ceph
|
||||
categories:
|
||||
- storage
|
||||
products:
|
||||
- tripleo
|
||||
- ceph
|
||||
fail_without_ceph_ansible: false
|
||||
ceph_ansible_repo: "centos-ceph-nautilus"
|
||||
tasks:
|
||||
|
|
|
|||
Loading…
Reference in New Issue