eb5423d084
* ci: Also check generated HTML with the "vnu" validator Because HTML Tidy apparently misses a bunch of errors. * thead -> tbody * More fixes * More fixes * Start checker in test runner
416 lines
17 KiB
Python
416 lines
17 KiB
Python
# Copyright The IETF Trust 2009-2020, All Rights Reserved
|
|
# -*- coding: utf-8 -*-
|
|
#
|
|
# Portion Copyright (C) 2009 Nokia Corporation and/or its subsidiary(-ies).
|
|
# All rights reserved. Contact: Pasi Eronen <pasi.eronen@nokia.com>
|
|
#
|
|
# Redistribution and use in source and binary forms, with or without
|
|
# modification, are permitted provided that the following conditions
|
|
# are met:
|
|
#
|
|
# * Redistributions of source code must retain the above copyright
|
|
# notice, this list of conditions and the following disclaimer.
|
|
#
|
|
# * Redistributions in binary form must reproduce the above
|
|
# copyright notice, this list of conditions and the following
|
|
# disclaimer in the documentation and/or other materials provided
|
|
# with the distribution.
|
|
#
|
|
# * Neither the name of the Nokia Corporation and/or its
|
|
# subsidiary(-ies) nor the names of its contributors may be used
|
|
# to endorse or promote products derived from this software
|
|
# without specific prior written permission.
|
|
#
|
|
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
|
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
|
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
|
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
|
|
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
|
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
|
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
|
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
|
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
|
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
|
|
import os
|
|
import re
|
|
import email
|
|
import html5lib
|
|
import requests_mock
|
|
import shutil
|
|
import sys
|
|
import subprocess
|
|
|
|
from urllib.parse import unquote
|
|
from unittest.util import strclass
|
|
from bs4 import BeautifulSoup
|
|
from contextlib import contextmanager
|
|
from pathlib import Path
|
|
from tempfile import NamedTemporaryFile
|
|
from tidylib import tidy_document
|
|
|
|
import django.test
|
|
from django.test.client import Client
|
|
from django.conf import settings
|
|
from django.utils.text import slugify
|
|
|
|
import debug # pyflakes:ignore
|
|
|
|
from ietf.utils.mail import get_payload_text
|
|
|
|
real_database_name = settings.DATABASES["default"]["NAME"]
|
|
|
|
def split_url(url):
|
|
if "?" in url:
|
|
url, args = url.split("?", 1)
|
|
args = dict([ list(map(unquote,arg.split("=", 1))) for arg in args.split("&") if "=" in arg ])
|
|
else:
|
|
args = {}
|
|
return url, args
|
|
|
|
def login_testing_unauthorized(test_case, username, url, password=None, method='get', request_kwargs=None):
|
|
"""Test that a request is refused or redirected for login, then log in as the named user
|
|
|
|
Defaults to making a 'get'. Set method to one of the other django.test.Client request method names
|
|
(e.g., 'post') to change that. If that request needs arguments, pass these in request_kwargs.
|
|
"""
|
|
request_method = getattr(test_case.client, method)
|
|
if request_kwargs is None:
|
|
request_kwargs = dict()
|
|
r = request_method(url, **request_kwargs)
|
|
test_case.assertIn(r.status_code, (302, 403))
|
|
if r.status_code == 302:
|
|
test_case.assertTrue("/accounts/login" in r['Location'])
|
|
if not password:
|
|
password = username + "+password"
|
|
return test_case.client.login(username=username, password=password)
|
|
|
|
def unicontent(r):
|
|
"Return a HttpResponse object's content as unicode"
|
|
return r.content.decode(r.charset)
|
|
|
|
def textcontent(r):
|
|
text = BeautifulSoup(r.content, 'lxml').get_text()
|
|
text = re.sub(r'(\n\s+){2,}', '\n\n', text)
|
|
return text
|
|
|
|
def reload_db_objects(*objects):
|
|
"""Rerequest the given arguments from the database so they're refreshed, to be used like
|
|
|
|
foo, bar = reload_db_objects(foo, bar)"""
|
|
|
|
t = tuple(o.__class__.objects.get(pk=o.pk) for o in objects)
|
|
if len(objects) == 1:
|
|
return t[0]
|
|
else:
|
|
return t
|
|
|
|
@contextmanager
|
|
def name_of_file_containing(contents, mode='w'):
|
|
"""Get a context with the name of an email file"""
|
|
f = NamedTemporaryFile(mode, delete=False)
|
|
f.write(contents)
|
|
f.close()
|
|
yield f.name # hand the filename to the context
|
|
Path(f.name).unlink() # clean up after context exits
|
|
|
|
|
|
def assert_ical_response_is_valid(test_inst, response, expected_event_summaries=None,
|
|
expected_event_uids=None, expected_event_count=None):
|
|
"""Validate an HTTP response containing iCal data
|
|
|
|
Based on RFC2445, but not exhaustive by any means. Assumes a single iCalendar object. Checks that
|
|
expected_event_summaries/_uids are found, but other events are allowed to be present. Specify the
|
|
expected_event_count if you want to reject additional events. If any of these are None,
|
|
the check for that property is skipped.
|
|
"""
|
|
test_inst.assertEqual(response.get('Content-Type'), "text/calendar")
|
|
|
|
# Validate iCalendar object
|
|
test_inst.assertContains(response, 'BEGIN:VCALENDAR', count=1)
|
|
test_inst.assertContains(response, 'END:VCALENDAR', count=1)
|
|
test_inst.assertContains(response, 'PRODID:', count=1)
|
|
test_inst.assertContains(response, 'VERSION', count=1)
|
|
|
|
# Validate event objects
|
|
if expected_event_summaries is not None:
|
|
for summary in expected_event_summaries:
|
|
test_inst.assertContains(response, 'SUMMARY:' + summary)
|
|
|
|
if expected_event_uids is not None:
|
|
for uid in expected_event_uids:
|
|
test_inst.assertContains(response, 'UID:' + uid)
|
|
|
|
if expected_event_count is not None:
|
|
test_inst.assertContains(response, 'BEGIN:VEVENT', count=expected_event_count)
|
|
test_inst.assertContains(response, 'END:VEVENT', count=expected_event_count)
|
|
test_inst.assertContains(response, 'UID', count=expected_event_count)
|
|
|
|
|
|
|
|
class ReverseLazyTest(django.test.TestCase):
|
|
def test_redirect_with_lazy_reverse(self):
|
|
response = self.client.get('/ipr/update/')
|
|
self.assertRedirects(response, "/ipr/", status_code=301)
|
|
|
|
|
|
class VerifyingClient(Client):
|
|
def __init__(self, test):
|
|
super(VerifyingClient, self).__init__()
|
|
self.test = test
|
|
|
|
def handle_error(self, path, source, errors):
|
|
file_name = "error" + re.sub("/", "-", path)
|
|
if not file_name.endswith("-"):
|
|
file_name += "-"
|
|
file_name += "source.html"
|
|
with open(file_name, "w") as src:
|
|
src.write(source)
|
|
print("\nHTML validation error for URL path", path)
|
|
print("HTML source saved to", file_name)
|
|
print("See AssertionError below for error location in HTML source.")
|
|
self.test.maxDiff = None
|
|
self.test.assertEqual("", errors)
|
|
|
|
def get(self, path, *args, skip_verify=False, **extra):
|
|
"""GET request
|
|
|
|
Performs verification of HTML responses unless skip_verify is True.
|
|
"""
|
|
r = super(VerifyingClient, self).get(path, *args, **extra)
|
|
|
|
if (
|
|
skip_verify
|
|
or r.status_code >= 300
|
|
or not r["content-type"].lower().startswith("text/html")
|
|
):
|
|
return r
|
|
source = r.content.decode()
|
|
|
|
if settings.VNU:
|
|
# First, run through https://validator.github.io/validator/
|
|
result = subprocess.run(
|
|
["java", "nu.validator.client.HttpClient", "-"],
|
|
input=r.content,
|
|
stdout=subprocess.PIPE,
|
|
stderr=subprocess.DEVNULL,
|
|
)
|
|
errors = result.stdout.decode()
|
|
|
|
if errors:
|
|
msg = ""
|
|
for err in errors.splitlines():
|
|
# TODO: check if some can be removed after validating database templates
|
|
if (
|
|
re.match(r'.*Attribute "required" not allowed', err)
|
|
or re.match(
|
|
r'.*The "type" attribute is unnecessary for JavaScript', err
|
|
)
|
|
or re.match(
|
|
r'.*Element "option" without attribute "label" must not be empty',
|
|
err,
|
|
)
|
|
or re.match(r".*The character encoding was not declared", err)
|
|
or re.match(r".*Consider avoiding viewport values", err)
|
|
or re.match(
|
|
r'.*The value of the "for" attribute of the "label" element must be the ID of a non-hidden form control',
|
|
err,
|
|
)
|
|
or re.match(r".*is not in Unicode Normalization Form C", err)
|
|
):
|
|
continue
|
|
# ignore some errors about obsolete HTML coming from the database
|
|
if re.match(
|
|
r"/meeting/\d+/proceedings/overview/", path
|
|
) and re.match(
|
|
r'.*The "\w+" attribute on the "\w+" element is obsolete', err
|
|
):
|
|
continue
|
|
# ignore some errors coming from outdated but still-needed iframes
|
|
if re.match(r"/meeting/\d+/week-view", path) and re.match(
|
|
r".*Start tag seen without seeing a doctype first", err
|
|
):
|
|
continue
|
|
pos = re.match(r'".*":((\d+)\.(\d+)-(\d+)\.(\d+):.*)', err)
|
|
if not pos:
|
|
self.handle_error(path, source, err)
|
|
return r
|
|
msg += pos.group(1).strip(" .") + ":\n"
|
|
for line in source.splitlines()[
|
|
int(pos.group(2)) - 1 : int(pos.group(4))
|
|
]:
|
|
msg += line.strip() + "\n"
|
|
|
|
if msg:
|
|
self.handle_error(path, source, msg)
|
|
return r
|
|
|
|
# Next, run through https://www.html-tidy.org/
|
|
document, errors = tidy_document(
|
|
r.content,
|
|
options={
|
|
# this is causing way too many generic warnings:
|
|
# "accessibility-check": 1,
|
|
},
|
|
)
|
|
|
|
errors = "\n".join(
|
|
[
|
|
e
|
|
# TODO: check if some can be removed after validating database templates
|
|
for e in errors.splitlines()
|
|
# FIXME: django-bootstrap5 incorrectly sets a "required"
|
|
# proprietary attribute on some <div>s; remove those errors
|
|
if not re.match(r'.*proprietary attribute "required"', e)
|
|
# FIXME: some secretariat templates have this issue, ignore
|
|
and not re.match(r".*id and name attribute value mismatch", e)
|
|
# FIXME: bootstrap-icons and close buttons render as empty, remove those errors.
|
|
# Also, django seems to generate some empty tags, so remove those, too.
|
|
and not re.match(r".*trimming empty <(i|em|button|span|optgroup)>", e)
|
|
# FIXME: some old pages only work correctly in quirks mode :-(
|
|
and not re.match(r".*missing <!DOCTYPE> declaration", e)
|
|
]
|
|
)
|
|
|
|
if errors:
|
|
self.handle_error(path, source, errors)
|
|
|
|
return r
|
|
|
|
|
|
class TestCase(django.test.TestCase):
|
|
"""IETF TestCase class
|
|
|
|
Based on django.test.TestCase, but adds a few things:
|
|
* asserts for html5 validation.
|
|
* tempdir() convenience method
|
|
* setUp() and tearDown() that override settings paths with temp directories
|
|
* mocking the requests library to prevent dependencies on the outside network
|
|
|
|
The setUp() and tearDown() methods create / remove temporary paths and override
|
|
Django's settings with the temp dir names. Subclasses of this class must
|
|
be sure to call the superclass methods if they are overridden. These are created
|
|
anew for each test to avoid risk of cross-talk between test cases. Overriding
|
|
the settings_temp_path_overrides class value will modify which path settings are
|
|
replaced with temp test dirs.
|
|
|
|
Uses requests-mock to prevent the requests library from making requests to outside
|
|
resources. The requests-mock library allows nested mocks, so individual tests can
|
|
ignore this. Note that the mock set up by this class will intercept any requests
|
|
not handled by a test's inner mock - even if the latter is created with
|
|
real_http=True.
|
|
"""
|
|
# These settings will be overridden with empty temporary directories
|
|
settings_temp_path_overrides = [
|
|
'RFC_PATH',
|
|
'INTERNET_ALL_DRAFTS_ARCHIVE_DIR',
|
|
'INTERNET_DRAFT_ARCHIVE_DIR',
|
|
'INTERNET_DRAFT_PATH',
|
|
]
|
|
|
|
parser = html5lib.HTMLParser(strict=True)
|
|
|
|
def assertValidHTML(self, data):
|
|
try:
|
|
self.parser.parse(data)
|
|
except Exception as e:
|
|
raise self.failureException(str(e))
|
|
|
|
def assertValidHTMLResponse(self, resp):
|
|
self.assertHttpOK(resp)
|
|
self.assertTrue(resp['Content-Type'].startswith('text/html'))
|
|
self.assertValidHTML(resp.content)
|
|
|
|
def assertSameEmail(self, a, b, msg=None):
|
|
def normalize(x):
|
|
if x:
|
|
if not isinstance(x, list):
|
|
x = [ x ]
|
|
x = email.utils.getaddresses(x)
|
|
x.sort()
|
|
return x
|
|
return self.assertEqual(normalize(a), normalize(b), msg)
|
|
|
|
def tempdir(self, label):
|
|
slug = slugify(self.__class__.__name__.replace('.','-'))
|
|
dirname = "tmp-{label}-{slug}-dir".format(**locals())
|
|
if 'VIRTUAL_ENV' in os.environ:
|
|
dirname = os.path.join(os.environ['VIRTUAL_ENV'], dirname)
|
|
path = os.path.abspath(dirname)
|
|
if not os.path.exists(path):
|
|
os.mkdir(path)
|
|
return path
|
|
|
|
def assertNoFormPostErrors(self, response, error_css_selector=".is-invalid"):
|
|
"""Try to fish out form errors, if none found at least check the
|
|
status code to be a redirect.
|
|
|
|
Assumptions:
|
|
- a POST is followed by a 302 redirect
|
|
- form errors can be found with a simple CSS selector
|
|
|
|
"""
|
|
|
|
if response.status_code == 200:
|
|
from pyquery import PyQuery
|
|
from lxml import html
|
|
self.maxDiff = None
|
|
|
|
errors = [html.tostring(n).decode() for n in PyQuery(response.content)(error_css_selector)]
|
|
if errors:
|
|
explanation = "{} != {}\nGot form back with errors:\n----\n".format(response.status_code, 302) + "----\n".join(errors)
|
|
self.assertEqual(response.status_code, 302, explanation)
|
|
|
|
self.assertEqual(response.status_code, 302)
|
|
|
|
def assertMailboxContains(self, mailbox, subject=None, text=None, count=None):
|
|
"""
|
|
Asserts that the given mailbox contains *count* mails with the given
|
|
*subject* and body *text* (if not None). At least one of subject,
|
|
text, and count must be different from None. If count is None, the
|
|
filtered mailbox must be non-empty.
|
|
"""
|
|
if subject is None and text is None and count is None:
|
|
raise self.failureException("No assertion made, both text and count is None")
|
|
mlist = mailbox
|
|
if subject:
|
|
mlist = [ m for m in mlist if subject in m["Subject"] ]
|
|
if text:
|
|
assert isinstance(text, str)
|
|
mlist = [ m for m in mlist if text in get_payload_text(m) ]
|
|
if count and len(mlist) != count:
|
|
sys.stderr.write("Wrong count in assertMailboxContains(). The complete mailbox contains %s emails:\n\n" % len(mailbox))
|
|
for m in mailbox:
|
|
sys.stderr.write(m.as_string())
|
|
sys.stderr.write('\n\n')
|
|
if count:
|
|
self.assertEqual(len(mlist), count)
|
|
else:
|
|
self.assertGreater(len(mlist), 0)
|
|
|
|
def __str__(self):
|
|
return u"%s (%s.%s)" % (self._testMethodName, strclass(self.__class__),self._testMethodName)
|
|
|
|
def setUp(self):
|
|
super().setUp()
|
|
# Prevent the requests library from making live requests during tests
|
|
self.requests_mock = requests_mock.Mocker()
|
|
self.requests_mock.start()
|
|
|
|
self.client = VerifyingClient(self) # Set up the HTML verifier
|
|
|
|
# Replace settings paths with temporary directories.
|
|
self._ietf_temp_dirs = {} # trashed during tearDown, DO NOT put paths you care about in this
|
|
for setting in self.settings_temp_path_overrides:
|
|
self._ietf_temp_dirs[setting] = self.tempdir(slugify(setting))
|
|
self._ietf_saved_context = django.test.utils.override_settings(**self._ietf_temp_dirs)
|
|
self._ietf_saved_context.enable()
|
|
|
|
def tearDown(self):
|
|
self._ietf_saved_context.disable()
|
|
for dir in self._ietf_temp_dirs.values():
|
|
shutil.rmtree(dir)
|
|
self.requests_mock.stop()
|
|
super().tearDown() |