It contains a fix which makes function helpers thread-safe. Without this
fix, our CI fails to compile assets with sprockets 4.x, so we need this
change in order to update sprockets.
Until now, when institutions made custom changes which made certain
tests obsolete, they didn't have a clear way to deal with the failing
tests. They would either:
1. Stop running the test suite
2. Run the test suite and get test failures 100% of the time
3. Comment the failing tests
4. Modify the failing tests in order to keep them in sync with the code
Solution 1 would be suicide from a maintenance perspective, although it
could work if only a couple of small custom changes were done.
Solution 2 would make it really hard to differenciate between "false
failures" and "real failures" when running the test suite.
Solution 3 would cause many conflicts when updating to a newer version
of Consul.
Solution 4 could make sense sometimes (big tests where only one line
needs to be changed), but it would also cause conflicts when updating
Consul.
So now, we're giving an alternative to solution 3 by making it easier to
exclude a test.
For tests that still need to be changed, when to use this solution
combined with a custom test and when to use solution 4 will have to be
decided on a per-case basis.
In the past, users had permission to edit their own legislation
proposals. However, that changed in commit ebfa3fb01, where we replaced
the `can` method with `cannot`.
An easier way to remove this permission is to simply remove the whole
statement, since by default users don't have permissions to do anything.
We're also adding a test checking users can't edit their own legislation
proposals, since it was missing.
The `edit` action is automatically authorized with the rules used for
`:update`, the same way the `new` action is authorized with the rules
used for `:create`.
So we don't need to authorize the edit and new actions.
These changes make it easier for institutions customizing Consul to
notice what they need to change if (for instance) they want users to be
able to edit investments under certain conditions.
See foundation/foundation-sites issue 12167, which was fixed in
Foundation 6.7.1. However, since Foundation 6.7 requires Node 12 and the
foundation-rails gem hasn't been maintained for a year and a half, we're
not updating for now.
We forgot to remove it in commit f28a5cc49.
The generated HTML was invalid, with the error:
> Element meta is missing one or more of the following attributes:
> content, property.
It was a bit confusing to press the "hide" button and then see the user
listed as "blocked". Some moderators might think they accidentally
pressed the wrong button.
In the moderation section there's no clear indicator as to what the
"Hide" and "Block" buttons do and the difference between them.
Since we're using confirmation dialogs in all moderation actions except
these ones, we're adding them here as well, so the difference will
appear in the dialog.
This isn't a very good solution, though, since the confirmation dialog
comes after clicking the button and users have already been wondering
whether clicking that button will be the right choice. A better solution
would be making the purpose clear before the button is clicked, although
that's something we don't do anywhere in the admin/moderation sections.
The test "Action links remember the pagination setting and the filter"
was failing sometimes because it assumed the third user created was
going to appear in the third place, but that wasn't always the case.
So we're using the same order we use in the rest of the sections dealing
with hidden content.
In the past, whenever we hid users, we also hid their comments.
However, we've now implemented an action to hide users without hiding
their comments. In this case, we still want to show the comment, but we
weren't doing so.
This is useful for people using screen readers, since the character used
as a separator won't be read aloud.
Since many screen readers also read content generated via CSS
pseudoelements, we aren't using `content: "|";` or similar but using
elements with a very small width instead.
We're continuing to replace links with buttons, for the reasons
explained in commit 5311daadf.
Since we're using the admin action component, we can also simplify the
logic handling the confirmation message.
In order to avoid duplicate IDs when generating buttons to block the
same author more than once in a page, we're including the record dom_id
in the ID of the button to block an author.
This HTML wasn't valid because it was a <span> containing the <div
class="reply"> element and it wasn't needed because there aren't any
flag actions in proposal notifications.
The `hide` action was calling the `block` method while the `soft_block`
action was calling the `hide` method.
Combined with the fact that we also have a `block` permission which is
used in `ModerateActions` the logic was hard to follow.
Note that in proposal notifications we're writing the call to
render the component in the same line as the <div class="reply">
definition in order to be able to use the `:empty` selector when the
component renders nothing. No browser matches whitespace with the
`:empty` selector, so we can't add newline characters inside the tag. A
more elegant solution would be extracting the proposal notification
actions to a component and only rendering it if the moderation actions
component is rendered.
Other than removing a redundant action, we're fixing two bugs when
blocking an author using the links in the public views:
* We were always redirecting to the debates index, even if we blocked
the author of a proposal or an investment
* We weren't showing any kind of success message
Having proper headers makes it more accessible.
We're also using the table actions component because we're going to add
another action. Since table actions use a flex layout, we have to tweak
the styles a little bit. For that, I'm adding a <main> element which
will make it possible to style just this table while also providing an
extra shortcut for people using screen readers.
We're continuing to replace links with buttons, for the reasons
explained in commit 5311daadf.
We're also adding an ARIA label since on the same page there might be
several links to block different users.
We can use `polymorphic_url` instead of manually setting the domain
every time.
This is a bit of a hack in order to comply with the validation rule
which says related content must start with the URL defined in
`Setting["url"]`.
The `parameterize` method uses the `I18n.transliterate` method, whose
documentation says:
```
I18n.transliterate("Ærøskøbing")
=> "AEroskobing"
I18n.transliterate("日本語")
=> "???"
```
That means we can't use it for dictionaries where characters don't have
a transliteration to the latin alphabet.
So we're changing the code in order to only transliterate characters
with a transliteration to the latin alphabet.
Note the first example ("Česká republika") already worked with the
previous code; the test has been added to make sure accented characters
are handled properly.
We were testing the creation of newsletters and admin notifications for
each existing segment, which IMHO is a bit overkill, considering how
slow system tests are.
So far we don't have any reasons to believe creating newsletters and
admin notifications will only work for some user segments, so we're
testing a random one instead.
Running these tests on my machine is now about 15 seconds faster.
This way we don't have to use the `send` method in other places, like
the AdminNotification class, and we can change the internal
implementation at any point.
We're going to add geozones as user segments, so it's handy to have the
method in the UserSegments class.
We're also changing the `user_segment_emails` parameter name for
consistency and simplicity.
We're going to make it dynamic using the geozones. Besides, class
methods can be overwritten using custom models, while constants can't be
overwritten without getting a warning [1].
Makes the definition of segments with geozones a little cleaner. I
think it’s worth it, compared to the slight memory gain of using a
constant [2].
[1] warning: already initialized constant UserSegments::SEGMENTS
[2] https://stackoverflow.com/questions/15903835/class-method-vs-constant-in-ruby-rails#answer-15903970
Programmers can take advantage of this feature when defining custom
default settings. And, since many CONSUL installations had custom
changes in the `custom/verification/residence.rb` model and those
changes might use regular expressions, we're making it easier to migrate
that code to the new system to define valid postal codes.
We aren't documenting this feature in the description in the admin
section because most administrators don't know what regular expressions
are.
Note that, in order to simplify the setting, we already define the `/\A`
and `\Z/` characters. So, if the custom code had something like
`postal_code =~ /^280/`, the setting would have to be "280*" (without
the quotes) or, in order to comply with a length validation,
"280[0-9]{2}" (without the quotes).
We were using the word "registered" in English as an equivalent of the
Spanish word "empadronado". However, the term "registered" is very
confusing because it might be understood as being registered in the
CONSUL website.
In the message, we're saying "cannot participate" in order to make the
message consistent with the message regarding the required age.
Due to the way Madrid handled postal code validations (see issue 533),
by default we were requiring everyone to validate against the local
census *and* to specify valid postal codes.
This could be useful in some cases, but in other cases, the census
validation will be enough and there'll be no need to manually define the
valid postal codes. Besides, some CONSUL installations are used in
organizations or political parties where the postal code validation
doesn't make sense.
