There’s a lot of code in service settings which:
- talks to the API directly through the clients
- passes that information through to the Jinja template
By encapsulating this logic in the service model:
- the Jinja template can access the data directly
- the logic can be reused across multiple methods
The view here is rebuilding a pseudo-service object. Now that service
objects have templates it’s cleaner to use the actual service object.
Requires a small change to the `templates_by_type` method so that it can
filter by one or many template types (a user should be able to copy any
template whose type is enabled for their service, and the service
they’re copying from).
This commit is the first step to disentangling the models from the API
clients. With the models in the same folder as the API clients it makes
it hard to import the API clients within the model without getting a
circular import.
After this commit the user API clients still has this problem, but at
least the service API client doesn’t.
Making people use a property is a sure way to make sure they’re spelling
the name of the property correctly, and allows us to easily swap out
properties that call through to the underlying JSON, and properties
which are implemented as methods.
This means we can have a method on the service model which hits the API
(or Redis) but can be called multiple times (within the context of a
request) without making multiple network requests.
It does this by storing the results of the method on the object’s
internal `__dict__` the first time the method is called.
Inheriting from `dict` has some unexpected side effects that don’t
happen with plain object. The one we want to avoid right now is that
a dict doesn’t seem to implement `__dict__` in a normal way, which
is required by `werkzeug.utils.cached_property`.
We do a lot of logic around choosing which templates to show. This logic
is all inside one view method.
It makes it cleaner to break this logic up into functions. But this
would mean passing around variables from one function to another.
Putting these methods onto a class (the service model) means that
there’s a place to store this data (rather than having to pass it around
a lot).
Making this code more manageable is important so that when we have
templates and folders it’s easy to encapsulate the logic around
combining the two.
`_get_current_service` is a function which gets called every time
`current_service` is referenced in a view method or Jinja template.
Because the service model was getting initialised inside this function
it was being reconstructed many times in one request. On the service
settings page, for example, it was getting initialised 43 times, adding
about 200ms to the response time.
This commit moves its initialisation to the point where we’re getting
the data from the API, which only happens once per request.
Once all our users have upgraded to the latest clients they won’t need
this. The latest clients only use the combined key and service ID.
Discuss: when can we safely remove it?
At the moment we show precompiled letters that have failed the
validation as having been sent. This is confusing.
We should communicate it as having been cancelled (rather than failed),
with the implication being that Notify has come along and cancelled the
letter before printing it. I think this is conceptually what makes the
most sense.
From the user’s point of view any letters that show up as cancelled
probably need to be fixed and resent, so it makes sense to group them
with the same name.
At the moment we are manually cancelling letters for people when they
ask us to. Once’s we’ve done this there is no indication that it’s
happened except for the date going red on the list of letters.
This commit adds some error messaging and styling to show when a letter
is cancelled.
Letting people cancel their own letters will be a future enhancement.
We were passing both dvla_org_id and filename to template-preview
temporarily while we switch to only using filename. Now that
template-preview is set up to use filename, we can stop sending the
dvla_org_id too.
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.
