This changeset removes webauthn from the Notify.gov admin app. We are not using webauthn at all in our implementation and will be looking at an entirely different authentication system in the near future.
Signed-off-by: Carlo Costino <carlo.costino@gsa.gov>
* Updated header and footer
* Moved files around and updated gulpfile to correct the build process when it goes to production
* Updated fonts
* Adjusted grid templating
* Adding images to assets
* Updated account pages, dashboard, and pages in message sending flow
* Updated the styling for the landing pages in the account section once logged in
It looks like, by default, Flask no longer makes full URLs, for example
`https://example.com/path`. Instead it does `/path`. This will still
work fine, and if anything is better because it reduces the number of
bytes of HTML we are sending.
It won’t mean that requests go over `http` instead of `https` without
the protocol because we set the appropriate HSTS header here:
0c57da7781/ansible/roles/paas-proxy/templates/admin.conf.j2 (L11)
This commit changes all our tests to reflect that URLs no longer have
the protocol and domain in them. `_external=True` is Flask’s way of
saying whether a URL should be generated with the domain and protocol
(`True`) or without it (`False`).
Again, I can’t find the changelog or diff where this was introuduced,
but if you’d like to go spelunking then here’s a starting point:
50374e3cfe/src/flask/helpers.py (L192)
We have a `client_request` fixture which does a bunch of useful stuff
like:
- checking the status code of the response
- returning a `BeautifulSoup` object
Lots of our tests still use an older fixture called `client`. This is
not as good because it:
- returns a raw `Response` object
- doesn’t do the additional checks
- means our tests contain a lot of repetetive boilerplate like `page = BeautifulSoup(response.data.decode('utf-8'), 'html.parser')`
This commit converts all the tests which had a `client.get(…)` or
`client.post(…)` statement to use their equivalents on `client_request`
instead.
Subsequent commits will remove uses of `client` in other tests, but
doing it this way means the work can be broken up into more manageable
chunks.
We have a `client_request` fixture which does a bunch of useful stuff
like:
- checking the status code of the response
- returning a `BeautifulSoup` object
Lots of our tests still use an older fixture called `logged_in_client`.
This is not as good because:
- it returns a raw `Response` object
- doesn’t do the additional checks
- means our tests contain a lot of repetetive boilerplate like `page = BeautifulSoup(response.data.decode('utf-8'), 'html.parser')`
This commit converts all the tests using `logged_in_client` to:
use `client_request` instead.
Previously we were passing a flag to the API which handled this. Now
we are doing it at the time of clicking the link, not at the time of
storing the new password. We don’t need to update the timestamp twice,
so this commit removes the code which tells the API to do it.
It should be enough to check the user has it set as their auth type.
Even if a user is no longer eligible to register a security key, it
should still be OK for them to continue using the feature.
Previously this was duplicated between the "two_factor" and the
"webauthn" views, and required more test setup. This DRYs up the
check and tests it once, using mocks to simplify the view tests.
As part of DRYing up the check into a util module, I've also moved
the "is_less_than_days_ago" function it uses.
if user has `webauthn_auth` as their auth type, then redirect them to an
interstitial that prompts them to click on a button which right now just
logs to the JS console, but in a future commit will open up the webauthn
browser prompt
content is unsurprisingly not final.
This naming was introduced in 2016 without explanation [1]. I find it
confusing because:
- It's reminiscent of "_app", which is a Python convention indicating
the variable is internal, so maybe avoid using it.
- It suggests there's some other "app" fixture I should be using (there
isn't, though).
The Python style guide describes using an underscore suffix to avoid
clashes with inbuilt names [1], which is sort of applicable if we need
to import the "app" module [2]. However, we can also avoid clashes by
choosing a different name, without the strange underscore.
[1]: 3b1d521c10
[2]: 78824f54fd/tests/app/main/views/test_forgot_password.py (L5)
I prefer to avoid `assert_not_called`, because if I make a typo like
this, the tests will still pass:
```
app.invite_api_client.create_invite.asset_not_called()
```
It’s harder to have a false positive with the statement written this
way:
```
assert app.invite_api_client.create_invite.called is False
```
Some email clients will pre-fetch links in emails to check whether
they’re safe. This has the unfortunate side effect of claiming the token
that’s in the link.
Long term, we don’t want to let the link be used multiple times, because
this reduces how secure it is (eg someone with access to your browser
history could re-use the link even if you’d signed out).
Instead, this commit adds an extra page which is served when the user
clicks the link from the email. This page includes a form which submits
to the actual URL that uses the token, thereby not claiming the token as
soon as the page is loaded.
For convenience, this page also includes some Javascript which clicks
the link on the user’s behalf. If the user has Javascript turned off
they will see the link and can click it themselves. This is going on the
assumption that whatever the email clients are doing when prefetching
the link doesn’t involve running any Javascript.
This Javascript is inlined so that:
- it is run as fast as possible
- it’s more resilient – even if our assets domain is unreachable or the
connection is interrupted, it will still run
We’re going to add an interstitial page that redirects to this new URL.
But we don’t want that redirect to 404 while the change is deploying,
because some boxes will have the new URL and some won’t. So let’s deploy
the new URL to all the boxes first, then the redirect page can safely
take over the new one.
The new URL is going to be `post` not `get` because that feels more
HTTP-y, so we need to make sure that’s part of this change too.
This page is slow to load which means:
- it’s annoying for us
- it’s potentially causing load on the database
This commit does two things to reduce the amount we’re unnecessarily
looking at this page:
1. Avoid redirecting to it when signing in as a platform admin user
2. Don’t go directly to it when clicking ‘platform admin’ at the top,
but instead show a holding page (there’s a fair chance you’ve clicked
that link in order to go and manage some email branding or find a
user, not wait for stats to load)
Flake8 Bugbear checks for some extra things that aren’t code style
errors, but are likely to introduce bugs or unexpected behaviour. A
good example is having mutable default function arguments, which get
shared between every call to the function and therefore mutating a value
in one place can unexpectedly cause it to change in another.
This commit enables all the extra warnings provided by Flake8 Bugbear,
except for the line length one (because we already lint for that
separately).
It disables:
- _B003: Assigning to os.environ_ because I don’t really understand this
- _B306: BaseException.message is removed in Python 3_ because I think
our exceptions have a custom structure that means the `.message`
attribute is still present
The data flow of other bits of our application looks like this:
```
API (returns JSON)
⬇
API client (returns a built in type, usually `dict`)
⬇
Model (returns an instance, eg of type `Service`)
⬇
View (returns HTML)
```
The user API client was architected weirdly, in that it returned a model
directly, like this:
```
API (returns JSON)
⬇
API client (returns a model, of type `User`, `InvitedUser`, etc)
⬇
View (returns HTML)
```
This mixing of different layers of the application is bad because it
makes it hard to write model code that doesn’t have circular
dependencies. As our application gets more complicated we will be
relying more on models to manage this complexity, so we should make it
easy, not hard to write them.
It also means that most of our mocking was of the User model, not just
the underlying JSON. So it would have been easy to introduce subtle bugs
to the user model, because it wasn’t being comprehensively tested. A lot
of the changed lines of code in this commit mean changing the tests to
mock only the JSON, which means that the model layer gets implicitly
tested.
For those reasons this commit changes the user API client to return
JSON, not an instance of `User` or other models.
This commit stops a new email verification link from being sent to a
user if they click on an email link which has expired or which has
already been used. Instead, they will be see an error message with a
link to the sign in page. This stops the situation where someone could
log in indefinitely (without the needing to enter their password) by
trying to use a used / expired email verification link and receiving a
valid link automatically.
This commit stops a new email verification link from being sent to a
user if they click on an email link which has expired or which has
already been used. Instead, they will be see an error message with a
link to the sign in page. This stops the situation where someone could
log in indefinitely (without the needing to enter their password) by
trying to use a used / expired email verification link and receiving a
valid link automatically.
One of the changes this pulls in is encoding of
periods in the token used for new password
requests.
In real-life the resulting URL is build by
concatenating the base url with the token so it
isn't processed further.
The test for new password requests builds the URL
with url_for. This encodes the result returned so
the periods are encoded twice.
tests weren't patching out create_event (which is invoked every time a
user logs in). This was getting caught by our egress proxy on jenkins.
We didn't notice because the event handler code was swallowing all
exceptions and not re-raising.
This changes that code to no longer swallow exceptions. Since we did
that, we also need to update all the tests that test log-in to mock
the call
we were seeing isort produce different outputs locally and in docker -
this was due to it having different opinions about whether the tests
module (ie all our unit tests) is a first party (local) or third party
(pip installed) import. It's a first party import, so by defining this
in the setup.cfg isort settings, we can force it to be consistent
between environments.
Note: I don't know why it was different in the first place though
show_accounts_or_dashboard has logic about where you should redirect
to. If we let it do this, then that's nicer than duplicating its
logic. We found that it wasn't accounting for orgs in redirects
properly.
Done using isort[1], with the following command:
```
isort -rc ./app ./tests
```
Adds linting to the `run_tests.sh` script to stop badly-sorted imports
getting re-introduced.
Chosen style is ‘Vertical Hanging Indent’ with trailing commas, because
I think it gives the cleanest diffs, eg:
```
from third_party import (
lib1,
lib2,
lib3,
lib4,
)
```
1. https://pypi.python.org/pypi/isort
It’s weird that `/services/<service_id>` returns `404`. The home page
for every service is the dashboard – should be possible to get there
from any other page just by stemming the URL.
Also makes it consistent with organisations, which will have
`/organisations/<org_id>`.
