A page should have only one `<h1>` element. So if there’s an error
message, which contains a `<h1>`, it should replace the page’s normal
`<h1>` element, rather than sit above it.
This makes errors on all pages have a `<h1>` element, which is important
for accessibility. It means a bit of rewriting the messages, but I think
they’re better for it.
Currently revoking an API key takes you to a separate page. It should
work the same way as other destructive actions, ie staying on the same
page but with a banner asking you to confirm the action.
The green bordered banner feels too much like ‘success’ or
‘confirmation’. Doesn’t feel like it’s something which just gives you
the status of a thing, or here’s a thing you should be aware of.
We use panels with a blue banner to indicate something that’s clickable.
So we should move away from this style for things that are just
notifications. We can’t use teal like other bits of GOV.UK because it
doesn’t pass colour contrast.
Pay are using a box with a green border, similar to the error validation
box (which has a red border). So let’s do the same for now.
Because we’re setting the API key and service ID after calling the
`__init__` method of the client it wasn’t doing the thing where it
splits the combined key into the two individual UUIDs. So we still need
to set them directly, individually on the client.
There’s some weird interaction between the message attribute of the
exception and mocking.
Luckily there is an internal attribute – `_message` which doesn’t go
through all the magic.
Looks like `radio_button.text` no longer works. Probably a BeautifulSoup
change. More robust to look at the text of the label and the value of
the input anyway.
The Notify API client changed in version 4 to take two arguments, not
three (service ID was removed in favour of the combined API key).
This gets a bit gnarly because the API key has to be at least a certain
length so it can be substringed internally.
We have two new pages for live and trial services that:
- are faster loading
- are now linked to
So the list of services doesn’t need to be on the platform admin index
page any more.
There are way too many links in the top nav now (when you’re using the
site as a platform admin).
And how often are you going to look at the letter jobs page really?
This commit adds some internal navigation for the platform admin area of
the platform, copying the style that works for navigating a single
service.
We have a lot of services now. Mostly we want to look at what live
services are doing. So loading up the trial mode services every time
slows things – generating, rendering and using the page – right down.
This commit adds two new pages, one to view only live services, and one
to view trial mode services.
When I split up the error messages on the check CSV page into multiple
templates, I also reduced the repetition of wrapping `<div>`s and macro
calls by moving them up outside the conditional blocks (see
8e947f315d).
Unfortunately I didn’t make the same adjustments for the one-off flow,
which meant that errors on these pages lost their styling. This commit
re-adds the styling for these error messages.
This was causing a 500 because of a bug in utils. It only occurred when
there was missing data for a cell in the recipient column. I’ve also
added a test for missing data in a non-recipient column just in case.
Previously we used AWS which meant that we could create wheels
from our requirements and then install them offline which made
deployments quicker.
We're no longer using AWS so let's remove that.
Although CloudFoundry supports installing dependencies in an offline
environment as documented here:
http://docs.cloudfoundry.org/buildpacks/python/#vendoring
To achieve this we create a vendor/ directory which will contain
the packages to install. This uses --no-index and --find-links so will
not resolve for any dependencies from pypi. For this reason there is
assumed confidence that the vendor/ directory will contain all
of the dependencies we need.
