Calling business logic in a factory is a code-smell, because it
usually requires the object to be saved into database, and may have
unintended consequences when the business logic is changed.
Also, this allows to just build a published procedure, without saving it
to the database.
Providing a query param ("locale") will enable localization. A language picker will be shown once
localization is activated. Locale is stored in a cookie "locale".
The check for whether the checkbox should be checked or not was made by
matching the whole string. Thus, given two options 'valid' and
'invalid', the check for the presence of 'valid' would succeed even when
only 'invalid' was present in the values (because
`'valid'.includes?('invalid')`.
The code now checks against the list of items in the selected_options.
Currently, deselecting all values from a multiple dropdown rendered as
checkboxes doesn't have any effect when submitting the form (the
previous values are still there, instead of being deselected).
This is because unchecked checkboxes are not sent by the browser – so
the "empty selection" never gets sent.
Rails `form.check_box` usually works around this by inserting an empty
hidden checkbox element, that will be sent even if all others are
de-selected. But the documentation warns that this is not possible when
iterating over an array (rather than a model). Which is our case here.
To fix this, this commit uses `collection_check_boxes` instead. It will
insert the proper hidden checkboxes in all cases, and fix our use case.
See https://api.rubyonrails.org/classes/ActionView/Helpers/FormOptionsHelper.html#method-i-collection_check_boxes
After clicking on a radio button option, it is impossible to revert to
the "None of the values selected" state.
However on non-mandatory fields, reverting to the no-selection value
should be possible.
To fix this, add an explicit "N/A" option.
- Make `champ.dossier` a requirement;
- Move the dossier_id assignation to `before_validation` (otherwise
the record is invalid, and never gets saved);
- Allow specs to only build the champ (instead of saving it to the
database), which bypasses the requirement to have a dossier.
Test helpers are separated between two files: spec_helper and
rails_helper. This separation is meant to allow tests that do not
require Rails (like testing standalone libs) to boot faster.
The spec_helper file is always loaded, through `--require spec_helper`
in the `.rspec` config file. When needed, the rails_helper file is
expected to be required manually.
This is fine, but:
- Many test files have a redundant `require 'spec_helper'` line;
- Many test files should require `rails_helper`, but don't.
Not requiring `rails_helper` will cause the Rails-concerned section of
the test environment not to be configured–which may cause subtle bugs
(like the test database not being properly initialized).
Moreover, Spring loads all the Rails files on preloading anyway. So the
gains from using only `spec_helper` are thin.
To streamline this process, this commit:
- Configures `.rspec` to require `rails_helper` by default;
- Remove all manual requires to spec_helper or rails_helper.
Reference: https://stackoverflow.com/questions/24145329/how-is-spec-rails-helper-rb-different-from-spec-spec-helper-rb-do-i-need-it
On the 22/01/2020, a technical error on the demarches-simplifees.fr
instance made us delete some files attached to some dossiers.
This PR adds a warning when browsing a dossier containing attachments
that were deleted.
Capybara logs several warnings during tests:
> Locator Integer:243 for selector :link must be an instance of String
> or Symbol. This will raise an error in a future version of Capybara.
This is because we use selectors like `have_link(dossier.id)`, where the
locator is a Number (instead of a String or Symbol).
Make it more reusable, by:
- Renaming `attachment/update` to `attachment/edit`
- Refactoring the CSS styles into their own stylesheet
- Allow to specify the 'accept' option
The instructeur, dossier and claimant where created regardless of
wether they already existed or not.
With this commit:
- Associations are declarated in a more idiomatic way
- They are correctly not created if they are provided when creating the
object.
Now that all states have a dropdown, we can simplify the code.
This also brings the "Repasser en instruction" action for all states
(including when no attestation or motivation is present).
- Fix the missing link to see attestations (see #4400)
- Fix the "Repasser en instruction" wording being confusing for
accepted state
- Add lot of tests
