We were inconsistent on this one. I consider it particularly useful when
a method starts with a `return` statement.
In other cases, we probably shouldn't have a guard rule in the middle of
a method in any case, but that's a different refactoring.
We were very inconsistent regarding these rules.
Personally I prefer no empty lines around blocks, clases, etc... as
recommended by the Ruby style guide [1], and they're the default values
in rubocop, so those are the settings I'm applying.
The exception is the `private` access modifier, since we were leaving
empty lines around it most of the time. That's the default rubocop rule
as well. Personally I don't have a strong preference about this one.
[1] https://rubystyle.guide/#empty-lines-around-bodies
We were already using `find_by` most of the time.
Since there are false positives related to our `find_by_slug_or_id!` and
`find_by_manger_login` methods, which cannot be replaced with `find_by`,
I'm adding it indicating the "refactor" severity.
It looks like we get this warning if we check the dialog message. Using
`accept_confirm` the same way we do in the rest of the application
solves the problem.
Having exceptions is better than having silent bugs.
There are a few methods I've kept the same way they were.
The `RelatedContentScore#score_with_opposite` method is a bit peculiar:
it creates scores for both itself and the opposite related content,
which means the opposite related content will try to create the same
scores as well.
We've already got a test to check `Budget::Ballot#add_investment` when
creating a line fails ("Edge case voting a non-elegible investment").
Finally, the method `User#send_oauth_confirmation_instructions` doesn't
update the record when the email address isn't already present, leading
to the test "Try to register with the email of an already existing user,
when an unconfirmed email was provided by oauth" fo fail if we raise an
exception for an invalid user. That's because updating a user's email
doesn't update the database automatically, but instead a confirmation
email is sent.
There are also a few false positives for classes which don't have bang
methods (like the GraphQL classes) or destroying attachments.
For these reasons, I'm adding the rule with a "Refactor" severity,
meaning it's a rule we can break if necessary.
The main reason to use it was the `rel` attribute for previous/next
pages not being indexed correctly by certain search engines when using a
relative URL. However, AFAIK that only applied to `<link>` tags, not to
`<a>` tags, and only if a `<base>` tag was defined.
In any case, it looks like the same search engines don't use the `rel`
attribute for previous/next to index pages anymore.
This way we can simplify the way we generate form fields. In some cases,
we also use the human attribute in table headers, which IMHO makes
sense.
I haven't moved all of them: for example, sometimes a label is
different depending on whether it's shown to administrators, valuators,
or users. And I haven't touched the ones related to devise, since I
wasn't sure about possible side effects.
Note I've also removed placeholders when they had the same text as their
labels, since they weren't helpful. On the contrary, the added redundant
text to the form, potentially distracting users.
This is a very subtle behaviour: `match /attachment/i` could represent a
regular expression, but it could also represent a division like
`match / attachment / i`. So we need to make an exception to the usual
way we omit parenthesis in RSpec expectations.
Naming two variables the same way is confusing at the very least, and
can lead to hard to debug errors. That's why the Ruby interpreter issues
a warning when we do so.
It could be argued that the following lines use single quotes to escape
double quotes, but on the other hand, using a single quote isn't a
great benefit.
I haven't found an elegant way to remove them, but since they were the
only three variables left out of 383 we used to have, I can live with
this low percentage of inelegant solutions.
There's a very common pattern in our test, where the setup only has two
lines:
variable = create(:something)
unused_variable = create(:something_else, something: variable)
In this case, since there's a blank line below these ones and then we'll
get to the body of the test, and the second variable is going to be
created based on the first variable, we can remove the useless
assignment and the readability is still OK.
Another option we almost unanimously discarded was:
variable = create(:something)
_unused_variable = create(:something_else, something: variable)
We don't use it anywhere else, either.
One more option we considered but found a bit too much for simple tests:
variable = create(:something) do |something|
create(:something_else, something: variable)
end
Then of course we could move the setup to `let` and `before` blocks, but
the tests could get over-structured really quickly.
These variables can be considered a block, and so removing them doesn't
make the test much harder to undestand.
Sometimes these variables formed the setup, sometimes they formed an
isolated part of the setup, and sometimes they were the part of the test
that made the test different from other tests.
We were creating records with a title we manually set, so to be
consistent with the rest of the code, in the test we check the title is
present using a string literal.
This way we can also remove useless assignments while keeping the code
vertically aligned.
This way we write the tests from the user's point of view: users can see
(for example) a proposal with the title "Make everything awesome", but
they don't see a proposal with a certain ID.
There are probably dozens, if not hundreds, of places where we could
write tests this way. However, it's very hard to filter which ones are
safe to edit, since not many of them have an HTML class we can use in
the tests, and adding a class might generate conflicts with CSS styles.
So, for now, I'm only changing the ones allowing us to cleanly remove
useless assignements while maintaining the code vertically aligned.
Note we usually cannot make it simple because officer assignments are
usually assigned to both a poll and a booth, and on a certain date.
However, in the few cases where the booth nor the date don't matter, we
can make the code a bit easier to read.
While in theory we wouldn't need to use the `transient` nor the
`after(:create)` because there's already a `has_many :through`
association with followers, Factory Bot / ActiveRecord don't
automatically associate the followable, resulting in an invalid record
exception.
The group is automatically assigned when we assign the heading. The
budget isn't needed either, except for a special case related to the
reason to be rejected.
While it could be argued we're hiding the real way we've defined
associations in our models, the tests are so much easier to read when we
don't have so many lines just creating data.
Furthermore, developers who care about vertically aligning the code will
be glad to see some variables disrupting this alignment are now gone.
We barely use this trait. In the votation type spec we're probably using
it wrong, and in the answer spec we assume one of the answers is going
to be "Yes".
The name `yes_no` is more expressive, since it makes it clear what the
answers are.
