mirror of https://github.com/zulip/zulip.git
101 lines
3.3 KiB
Python
101 lines
3.3 KiB
Python
|
# -*- coding: utf-8 -*-
|
||
|
|
from __future__ import absolute_import
|
||
|
|
|
||
|
|
import logging
|
||
|
|
|
||
|
|
from django.test import TestCase
|
||
|
|
from django.template.loader import get_template
|
||
|
|
|
||
|
|
from zerver.models import get_user_profile_by_email
|
||
|
|
from zerver.lib.test_helpers import get_all_templates
|
||
|
|
|
||
|
|
|
||
|
|
class get_form_value(object):
|
||
|
|
def __init__(self, value):
|
||
|
|
self._value = value
|
||
|
|
|
||
|
|
def value(self):
|
||
|
|
return self._value
|
||
|
|
|
||
|
|
|
||
|
|
class DummyForm(dict):
|
||
|
|
pass
|
||
|
|
|
||
|
|
|
||
|
|
class TemplateTestCase(TestCase):
|
||
|
|
"""
|
||
|
|
Tests that backend template rendering doesn't crash.
|
||
|
|
|
||
|
|
This renders all the Zulip backend templates, passing dummy data
|
||
|
|
as the context, which allows us to verify whether any of the
|
||
|
|
templates are broken enough to not render at all (no verification
|
||
|
|
is done that the output looks right). Please see `get_context`
|
||
|
|
function documentation for more information.
|
||
|
|
"""
|
||
|
|
def test_templates(self):
|
||
|
|
# Just add the templates whose context has a conflict with other
|
||
|
|
# templates' context in `exclude`.
|
||
|
|
exclude = ['analytics/activity.html']
|
||
|
|
templates = [t for t in get_all_templates() if t not in exclude]
|
||
|
|
self.render_templates(templates, self.get_context())
|
||
|
|
|
||
|
|
# Test the excluded templates with updated context.
|
||
|
|
update = {'data': [('one', 'two')]}
|
||
|
|
self.render_templates(exclude, self.get_context(**update))
|
||
|
|
|
||
|
|
def render_templates(self, templates, context):
|
||
|
|
for template in templates:
|
||
|
|
template = get_template(template)
|
||
|
|
try:
|
||
|
|
template.render(context)
|
||
|
|
except Exception:
|
||
|
|
logging.exception("Exception while rendering '{}'".format(template.template.name))
|
||
|
|
|
||
|
|
def get_context(self, **kwargs):
|
||
|
|
"""Get the dummy context for shallow testing.
|
||
|
|
|
||
|
|
The context returned will always contain a parameter called
|
||
|
|
`shallow_tested`, which tells the signal receiver that the
|
||
|
|
test was not rendered in an actual logical test (so we can
|
||
|
|
still do coverage reporting on which templates have a logical
|
||
|
|
test).
|
||
|
|
|
||
|
|
Note: `context` just holds dummy values used to make the test
|
||
|
|
pass. This context only ensures that the templates do not
|
||
|
|
throw a 500 error when rendered using dummy data. If new
|
||
|
|
required parameters are added to a template, this test will
|
||
|
|
fail; the usual fix is to just update the context below to add
|
||
|
|
the new parameter to the dummy data.
|
||
|
|
|
||
|
|
:param kwargs: Keyword arguments can be used to update the base
|
||
|
|
context.
|
||
|
|
|
||
|
|
"""
|
||
|
|
email = "hamlet@zulip.com"
|
||
|
|
user_profile = get_user_profile_by_email(email)
|
||
|
|
|
||
|
|
context = dict(
|
||
|
|
shallow_tested=True,
|
||
|
|
user_profile=user_profile,
|
||
|
|
product_name='testing',
|
||
|
|
form=DummyForm(
|
||
|
|
full_name=get_form_value('John Doe'),
|
||
|
|
terms=get_form_value(True),
|
||
|
|
email=get_form_value(email),
|
||
|
|
),
|
||
|
|
current_url=lambda: 'www.zulip.com',
|
||
|
|
referrer=dict(
|
||
|
|
full_name='John Doe',
|
||
|
|
realm=dict(name='zulip.com'),
|
||
|
|
),
|
||
|
|
uid='uid',
|
||
|
|
token='token',
|
||
|
|
message_count=0,
|
||
|
|
messages=[dict(header='Header')],
|
||
|
|
new_streams=dict(html=''),
|
||
|
|
data=dict(title='Title'),
|
||
|
|
)
|
||
|
|
|
||
|
|
context.update(kwargs)
|
||
|
|
return context
|