TL;DR When using Cache-Control on a Rails application, make sure the Vary: Accept header is set.

Proxy caching is a good feature to serve your publicly visible application content faster and reduce load on your servers. It is e.g. available in nginx, but also affects proxies delivered by ISPs.

Unfortunately, there is a little problem in Rails < 6.1 when delivering responses for different MIME-types. Say you have an arbitrary route in your Rails application that is able to respond with regular HTML and JSON. By sending the specific ...

To reverse lookup a fixture by its table name and id, use the following approach on ActiveRecord::FixtureSet:

table = 'users'       # Specify the fixture table name
id = 123122           # Specify the ID to look for

# Find the fixture that matches the given ID
ActiveRecord::FixtureSet.all_loaded_fixtures[table].fixtures.find { |key, value| value['id'] == id }

Result Example:

[
  "one", # Fixture name
  #<ActiveRecord::Fixture:0x00007e79990234c8>, # ActiveRecord::Fixture object
  @fixture= { ... }, # The raw fixtu...

Updated for selenium-webdriver 4.27 which requires options.add_option(:page_load_strategy, 'none').

Your package-lock.json should always match and resolve all packages from your package.json.
Coming from Yarn, I was looking for an option like Yarn's --frozen-lockfile which validates that. Here is what seems to be the way to do it.

Using npm clean-install

Running npm clean-install instead of npm install will actually validate that your package-lock.json matches your package.json.
If your package-lock.json cannot be fulfilled, npm install would install other versions while npm clean-install will complain.

You can ...

TL;DR: Rails ships two methods to convert strings to constants, constantize and safe_constantize. Neither is safe for untrusted user input. Before you call either method you must validate the input string against an allowlist. The only difference between the two methods is that unresolvable constants raise an error with constantize, but return nil with safe_constantize. If you validate the input string against an allowlist, an error should never happen.

Preventing Dangerous Lookups

Suppose an application uses eit...

In development, we store files using ActiveStorage's disk service. This means that stored files are served by your Rails application, and every request to a file results in (at least!) one non-trivial log entry which can be annoying. Here is how to disable those log entries.

Example

Here is an example of what loading a single <img> in an example application writes to the Rails log.

Started GET "/rails/active_storage/blobs/redirect/..." for ::1 at ...
Processing by ActiveStorage::Blobs::RedirectController#show as SVG
  Parameter...

If you update Selenium regularly, you'll run into deprecation warnings similar to:

WARN Selenium [:clear_local_storage] [DEPRECATION] clear_local_storage is deprecated and will be removed in a future release

You can ignore this deprecation warning and clean your logs with logfilters:

Selenium::WebDriver.logger.ignore(:clear_local_storage)

Just specify the tag (:clear_local_storage) you want to ignore.

When you have a string containing umlauts which don't behave as expected (are not matched with a regexp, can't be found with an SQL query, do not print correctly on LaTeX documents, etc), you may be encountering umlauts which are not actually umlaut characters.

They look, depending on the font, like their "real" umlaut counterpart:

• ä ↔ ä
• ö ↔ ö
• ü ↔ ü

However, they are not the same:

'ä' == 'ä' # false
'ä'.size # 1
'ä'.size # 2

Looking at how those strings are constructed reveals what is going on:

'ä'.unpack('U*...

Rails migrations allow you to use a change method whose calls are automatically inverted for the down path. However, if you need to some path-specific logic (like SQL UPDATE statements) you can not define up and down methods at the same time.

If you were to define define all 3 of them, Rails would only run change and ignore up and down. However, Rails 4+ features a helper method called reversible:

class MyMigration < ActiveRecord::Migration

  def cha...

RubyMine allows bookmarking lines of code. This is super-helpful when working on a complex problem.
I've been using this feature for a few years now, and so should you! :)

Here are the default Linux/Windows keystrokes. See the documentation for other keybindings.

Add an anonymous bookmark

F11

A gray checkmark will be shown in the gutter on the left.
If you press F11 again on a bookmarked line, the bookmark will be removed.

Add a named bookmark ("mnemonic")

Ctrl ...

tl;dr

When a cookie includes an Expires attribute or an HTTP response includes caching headers like Expires or Cache-Control, their validity depends on the server's Date header if present. Otherwise, the browser uses its local time. This can lead to issues in tests with mocked time or inconsistent cache behavior.

Cookie Expires depends on the Date header or browser time

When a cookie includes an Expires attribute, the browser evaluates the expiration date relative to a reference time:

1. If the HTTP response ...

In FactoryBot factories, Rails' file_fixture is not available by default. To enable it, include a support module from rspec-rails:

FactoryBot::SyntaxRunner.include(RSpec::Rails::FileFixtureSupport)

That includes ActiveSupport::Testing::FileFixtures, where file_fixture is defined, but also configures the file_fixture_path so that you can actually use file_fixture.

When you find yourself in the situation that you would like to define a custom matcher in your specs, but you do not want to define a global matcher since you need it only for your specific test, there are two ways to do it:

Custom matcher for a single group

If you're only going to include a matcher once, you can also use the matcher macro within an example group:

describe "group" do
  
  matcher :be_just_like do |expected|
    match {|ac...

Sometimes huge refactorings or refactoring of core concepts of your application are necessary for being able to meet new requirements or to keep your application maintainable on the long run. Here are some thoughts about how to approach such challenges.

Break it down

Try to break your refactoring down in different parts. Try to make tests green for each part of your refactoring as soon as possible and only move to the next big part if your tests are fixed. It's not a good idea to work for weeks or months and wait for all puzzle pieces ...

Please don't simply copy line number links from Github. The URL usually contains a branch name like master which will change over time:

https://github.com/makandra/upjs/blob/master/lib/assets/javascripts/up/link.js.coffee#L76

If someone now posts an insertion or deletion to that file into master your link points to the wrong line!

A better way is to press the Y key after clicking on a line number. This will transform the URL to another URL that points to the particular commit:

https://github.com/makandra/upjs/blob/b3b14...

Some key highlights and points from the linked article TestProf II: Factory therapy for your Ruby tests.

The Problem with Factories in Ruby Tests

• Factories are used to easily generate test data.
• However, they can unintentionally slow down test suites by creating unnecessary or excessive associated data (factory cascades).

Understanding Factory-Induced Slowdowns

• Factories often create additional data (e.g., associated records) th...

Rails slightly changed the fragment cache implementation from Rails 7.0 to Rails 7.1. Unfortunately, this is incompatible with how Haml 5 patches Action View buffers. I tried turning a String buffer into an ActionView::OutputBuffer, but this brought up...

Text fragments allow linking directly to a specific portion of text in a web document, without requiring the author to annotate it with an ID, using particular syntax in the URL fragment. Supporting browsers are free to choose how to draw attention to the linked text, e.g. with a color highlight and/or scrolling to the content on the page. This is useful because it allows web content authors to deep-link to other content they don't control, without relying on the presence of IDs to make that possible. Building on top of that, it could be u...

When Ruby objects are inspected in any modern IRB, some objects (like ActiveRecord instances) are rendered with neat colors and line breaks.
You will not get that for custom classes by default -- which can be annoying if your inspection contains lots of meaningful information.

Here is what you need to do if you want your objects to be inspected nicely.

Implement a pretty_print method

As an example, consider the following class.

class MyClass

  # ...

  def inspect
    "#<#{self.class} attr1: #{attr1.inspect}, attr2: #{attr2...

This card explains how to generate an entity relationship diagram for your Rails application.
We also show how to limit your ERD to a subset of models, e.g. models inside a namespace.

Generating a full ERD

Option A: RubyMine

1. Right-click anywhere in your project tree
2. In the context menu, find the "Diagrams" menu item at/near the bottom
3. Inside, choose "Show diagram" → "Rails Model Dependency Diagram"
4. A new tab will open with the diagram inside. You can modify it there, and export it as an image.

Option B: Use rails-e...

Sometimes you have a test expectation but actually want a better error message in case of a failure. Here is how to do that.

Background

Consider this test:

expect(User.last).to be_present

In case of an error, it will fail with a not-so-helpful error message:

expected present? to return true, got false (Spec::Expectations::ExpectationNotMetError)

Solution

That can be fixed easily. RSpec expectations allow you to pass an error message like this:

expect(User.last).to be_present, 'Could not find a user!'

...

The linked article points out that COUNT queries might be unexpectedly slow in PostgreSQL.

If you just need to know "are there any records" use any?. This uses SELECT 1 AS one FROM ... LIMIT 1 under the hood.
If you just need to know "are there no records" use empty? or none?. This uses SELECT 1 AS one FROM ... LIMIT 1 under the hood.

In short: Replace foo.count > 0 with foo.any? or foo.count == 0 with foo.none?

If you require quick counts and can tolerate some level of imprecision, consider exploring the ...

Capistrano 3 has a doctor task that will print information about

• Environment: Ruby, Rubygems and Bundler versions
• List of Capistrano gems and whether an update is available
• All config variables and their values
• Capistrano server config

$ bundle exec cap staging doctor

Rails' fragment caching caches subtrees of an HTML document tree. While constructing that tree though, it can be really hard to keep track of whether some code is run in a caching context. Fortunately, Rails 7 brings two helpers that simplify this.

Note that these helpers are all about Rails' fragment caching and not about downstream caching (i.e. Cache-Control).

uncacheable!

Invoke this helper in a partial or another helper that should never be cached. Used outside of fragment caches, the helper does just nothing. But should it ...