92816644f7
This document describes a abstract methodology for analysing reliability of high-availability OpenStack cluster and it's components. Co-Authored-By: Bogdan Dobrelia <bdobrelia@mirantis.com> Change-Id: I5a08c1a39bab96d90c6f7a873fdc771516ffba48
222 lines
8.8 KiB
Python
222 lines
8.8 KiB
Python
# Licensed under the Apache License, Version 2.0 (the "License"); you may
|
|
# not use this file except in compliance with the License. You may obtain
|
|
# a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
|
|
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
|
|
# License for the specific language governing permissions and limitations
|
|
# under the License.
|
|
|
|
import glob
|
|
import os
|
|
import re
|
|
|
|
import docutils.core
|
|
import six
|
|
import testtools
|
|
|
|
|
|
OPTIONAL_SECTIONS = ("Upper level additional section",)
|
|
OPTIONAL_SUBSECTIONS = ("Some additional section",)
|
|
OPTIONAL_SUBSUBSECTIONS = ("Parameters", "Some additional section",)
|
|
OPTIONAL_FIELDS = ("Conventions",)
|
|
|
|
|
|
class TestTitles(testtools.TestCase):
|
|
def _get_title(self, section_tree, depth=1):
|
|
section = {
|
|
"subtitles": [],
|
|
}
|
|
for node in section_tree:
|
|
if node.tagname == "title":
|
|
section["name"] = node.rawsource
|
|
elif node.tagname == "section":
|
|
subsection = self._get_title(node, depth+1)
|
|
if depth < 2:
|
|
if subsection["subtitles"]:
|
|
section["subtitles"].append(subsection)
|
|
else:
|
|
section["subtitles"].append(subsection["name"])
|
|
elif depth == 2:
|
|
section["subtitles"].append(subsection["name"])
|
|
return section
|
|
|
|
def _get_titles(self, test_plan):
|
|
titles = {}
|
|
for node in test_plan:
|
|
if node.tagname == "section":
|
|
section = self._get_title(node)
|
|
titles[section["name"]] = section["subtitles"]
|
|
return titles
|
|
|
|
@staticmethod
|
|
def _get_docinfo(test_plan):
|
|
fields = []
|
|
for node in test_plan:
|
|
if node.tagname == "field_list":
|
|
for field in node:
|
|
for f_opt in field:
|
|
if f_opt.tagname == "field_name":
|
|
fields.append(f_opt.rawsource)
|
|
|
|
if node.tagname == "docinfo":
|
|
for info in node:
|
|
fields.append(info.tagname)
|
|
|
|
if node.tagname == "topic":
|
|
fields.append("abstract")
|
|
|
|
return fields
|
|
|
|
def _check_fields(self, tmpl, test_plan):
|
|
tmpl_fields = self._get_docinfo(tmpl)
|
|
test_plan_fields = self._get_docinfo(test_plan)
|
|
|
|
missing_fields = [f for f in tmpl_fields
|
|
if f not in test_plan_fields and
|
|
f not in OPTIONAL_FIELDS]
|
|
|
|
if len(missing_fields) > 0:
|
|
self.fail("While checking '%s':\n %s"
|
|
% (test_plan[0].rawsource,
|
|
"Missing fields: %s" % missing_fields))
|
|
|
|
def _check_titles(self, filename, expect, actual):
|
|
missing_sections = [x for x in expect.keys() if (
|
|
x not in actual.keys()) and (x not in OPTIONAL_SECTIONS)]
|
|
|
|
msgs = []
|
|
if len(missing_sections) > 0:
|
|
msgs.append("Missing sections: %s" % missing_sections)
|
|
|
|
for section in expect.keys():
|
|
missing_subsections = [x for x in expect[section]
|
|
if x not in actual.get(section, {}) and
|
|
(x not in OPTIONAL_SUBSECTIONS)]
|
|
extra_subsections = [x for x in actual.get(section, {})
|
|
if x not in expect[section]]
|
|
|
|
for ex_s in extra_subsections:
|
|
s_name = (ex_s if isinstance(ex_s, six.string_types)
|
|
else ex_s["name"])
|
|
if s_name.startswith("Test Case"):
|
|
new_missing_subsections = []
|
|
for m_s in missing_subsections:
|
|
m_s_name = (m_s if isinstance(m_s, six.string_types)
|
|
else m_s["name"])
|
|
if not m_s_name.startswith("Test Case"):
|
|
new_missing_subsections.append(m_s)
|
|
missing_subsections = new_missing_subsections
|
|
break
|
|
|
|
if len(missing_subsections) > 0:
|
|
msgs.append("Section '%s' is missing subsections: %s"
|
|
% (section, missing_subsections))
|
|
|
|
for subsection in expect[section]:
|
|
if type(subsection) is dict:
|
|
missing_subsubsections = []
|
|
actual_section = actual.get(section, {})
|
|
matching_actual_subsections = [
|
|
s for s in actual_section
|
|
if type(s) is dict and (
|
|
s["name"] == subsection["name"] or
|
|
(s["name"].startswith("Test Case") and
|
|
subsection["name"].startswith("Test Case")))
|
|
]
|
|
for actual_subsection in matching_actual_subsections:
|
|
for x in subsection["subtitles"]:
|
|
if (x not in actual_subsection["subtitles"] and
|
|
x not in OPTIONAL_SUBSUBSECTIONS):
|
|
missing_subsubsections.append(x)
|
|
if len(missing_subsubsections) > 0:
|
|
msgs.append("Subsection '%s' is missing "
|
|
"subsubsections: %s"
|
|
% (actual_subsection,
|
|
missing_subsubsections))
|
|
|
|
if len(msgs) > 0:
|
|
self.fail("While checking '%s':\n %s"
|
|
% (filename, "\n ".join(msgs)))
|
|
|
|
def _check_lines_wrapping(self, tpl, raw):
|
|
code_block = False
|
|
text_inside_simple_tables = False
|
|
lines = raw.split("\n")
|
|
for i, line in enumerate(lines):
|
|
# NOTE(ndipanov): Allow code block lines to be longer than 79 ch
|
|
if code_block:
|
|
if not line or line.startswith(" "):
|
|
continue
|
|
else:
|
|
code_block = False
|
|
if "::" in line:
|
|
code_block = True
|
|
# simple style tables also can fit >=80 symbols
|
|
# open simple style table
|
|
if "===" in line and not lines[i - 1]:
|
|
text_inside_simple_tables = True
|
|
if "http://" in line or "https://" in line:
|
|
continue
|
|
# Allow lines which do not contain any whitespace
|
|
if re.match("\s*[^\s]+$", line):
|
|
continue
|
|
if not text_inside_simple_tables:
|
|
self.assertTrue(
|
|
len(line) < 80,
|
|
msg="%s:%d: Line limited to a maximum of 79 characters." %
|
|
(tpl, i + 1))
|
|
# close simple style table
|
|
if "===" in line and not lines[i + 1]:
|
|
text_inside_simple_tables = False
|
|
|
|
def _check_no_cr(self, tpl, raw):
|
|
matches = re.findall("\r", raw)
|
|
self.assertEqual(
|
|
len(matches), 0,
|
|
"Found %s literal carriage returns in file %s" %
|
|
(len(matches), tpl))
|
|
|
|
def _check_trailing_spaces(self, tpl, raw):
|
|
for i, line in enumerate(raw.split("\n")):
|
|
trailing_spaces = re.findall("\s+$", line)
|
|
self.assertEqual(
|
|
len(trailing_spaces), 0,
|
|
"Found trailing spaces on line %s of %s" % (i + 1, tpl))
|
|
|
|
def test_template(self):
|
|
# Global repository template
|
|
with open("doc/source/test_plans/template.rst") as f:
|
|
global_template = f.read()
|
|
|
|
files = glob.glob("doc/source/test_plans/*/plan.rst")
|
|
files = [os.path.abspath(filename) for filename in files]
|
|
|
|
for filename in files:
|
|
with open(filename) as f:
|
|
data = f.read()
|
|
|
|
os.chdir(os.path.dirname(filename))
|
|
# Try to use template in directory where plan.rst is located
|
|
try:
|
|
with open("template.rst") as f:
|
|
# use local template
|
|
template = f.read()
|
|
except Exception:
|
|
# use global template
|
|
template = global_template
|
|
pass
|
|
test_plan_tmpl = docutils.core.publish_doctree(template)
|
|
template_titles = self._get_titles(test_plan_tmpl)
|
|
test_plan = docutils.core.publish_doctree(data)
|
|
self._check_titles(filename,
|
|
template_titles,
|
|
self._get_titles(test_plan))
|
|
self._check_fields(test_plan_tmpl, test_plan)
|
|
self._check_lines_wrapping(filename, data)
|
|
self._check_no_cr(filename, data)
|
|
self._check_trailing_spaces(filename, data)
|