best-practices

Best Practices

A guide for programming well.

General

• Don't duplicate the functionality of a built-in library.
• Don't swallow exceptions or "fail silently."
• Don't write code that guesses at future functionality.
• Exceptions should be exceptional.
• Keep the code simple.

Object-Oriented Design

• Avoid global variables.
• Avoid long parameter lists.
• Limit collaborators of an object (entities an object depends on).
• Limit an object's dependencies (entities that depend on an object).
• Prefer composition over inheritance.
• Prefer small methods. Between one and five lines is best.
• Prefer small classes with a single, well-defined responsibility. When a class exceeds 100 lines, it may be doing too many things.
• Tell, don't ask.

Ruby

• Avoid optional parameters. Does the method do too much?
• Avoid monkey-patching.
• Generate necessary Bundler binstubs for the project, such as rake and rspec, and add them to version control.
• Prefer classes to modules when designing functionality that is shared by multiple models.
• Prefer private when indicating scope. Use protected only with comparison methods like def ==(other), def <(other), and def >(other).

Ruby Gems

• Declare dependencies in the <PROJECT_NAME>.gemspec file.
• Reference the gemspec in the Gemfile.
• Use Appraisal to test the gem against multiple versions of gem dependencies (such as Rails in a Rails engine).
• Use Bundler to manage the gem's dependencies.
• Use Travis CI for Continuous Integration, indicators showing whether GitHub pull requests can be merged, and to test against multiple Ruby versions.

Rails

• Avoid bypassing validations with methods like save(validate: false), update_attribute, and toggle.
• Avoid instantiating more than one object in controllers.
• Avoid naming methods after database columns in the same class.
• Don't change a migration after it has been merged into master if the desired change can be solved with another migration.
• Don't reference a model class directly from a view.
• Don't return false from ActiveModel callbacks, but instead raise an exception.
• Don't use instance variables in partials. Pass local variables to partials from view templates.
• Don't use SQL or SQL fragments (where('inviter_id IS NOT NULL')) outside of models.
• Generate necessary Spring binstubs for the project, such as rake and rspec, and add them to version control.
• If there are default values, set them in migrations.
• Keep db/schema.rb or db/development_structure.sql under version control.
• Use only one instance variable in each view.
• Use SQL, not ActiveRecord models, in migrations.
• Use the .ruby-version file convention to specify the Ruby version and patch level for a project.
• Use _url suffixes for named routes in mailer views and redirects. Use _path suffixes for named routes everywhere else.
• Validate the associated belongs_to object (user), not the database column (user_id).
• Use db/seeds.rb for data that is required in all environments.
• Use dev:prime rake task for development environment seed data.
• Prefer cookies.signed over cookies to prevent tampering.
• Prefer Time.current over Time.now
• Prefer Date.current over Date.today
• Prefer Time.zone.parse("2014-07-04 16:05:37") over Time.parse("2014-07-04 16:05:37")

Testing

• Avoid any_instance in rspec-mocks and mocha. Prefer dependency injection.
• Avoid its, let, let!, specify, before, and subject in RSpec.
• Avoid using instance variables in tests.
• Disable real HTTP requests to external services with WebMock.disable_net_connect!.
• Don't test private methods.
• Test background jobs with a Delayed::Job matcher.
• Use stubs and spies (not mocks) in isolated tests.
• Use a single level of abstraction within scenarios.
• Use an it example or test method for each execution path through the method.
• Use assertions about state for incoming messages.
• Use stubs and spies to assert you sent outgoing messages.
• Use a Fake to stub requests to external services.
• Use integration tests to execute the entire app.
• Use non-SUT methods in expectations when possible.

Bundler

• Specify the Ruby version to be used on the project in the Gemfile.
• Use a pessimistic version in the Gemfile for gems that follow semantic versioning, such as rspec, factory_girl, and capybara.
• Use a versionless Gemfile declarations for gems that are safe to update often, such as pg, thin, and debugger.
• Use an exact version in the Gemfile for fragile gems, such as Rails.

Postgres

• Avoid multicolumn indexes in Postgres. It combines multiple indexes efficiently. Optimize later with a compound index if needed.
• Consider a partial index for queries on booleans.
• Constrain most columns as NOT NULL.
• Index foreign keys.

Background Jobs

• Store IDs, not ActiveRecord objects for cleaner serialization, then re-find the ActiveRecord object in the perform method.

Email

• Use SendGrid or Amazon SES to deliver email in staging and production environments.
• Use a tool like MailView to look at each created or updated mailer view before merging.

JavaScript

• Use CoffeeScript.

HTML

• Don't use a reset button for forms.
• Prefer cancel links to cancel buttons.

CSS

• Use Sass.

Sass

• Use image-url and font-url, not url, so the asset pipeline will re-write the correct paths to assets.

Browsers

• Don't support clients without Javascript.
• Don't support IE6 or IE7.

Objective-C

• Setup new projects using Liftoff and follow provided directory structure.
• Prefer categories on Foundation classes to helper methods.
• Prefer string constants to literals when providing keys or key paths to methods.

Shell

• Don't parse the output of ls. See here for details and alternatives.
• Don't use cat to provide a file on stdin to a process that accepts file arguments itself.
• Don't use a /bin/sh shebang unless you plan to test and run your script on at least: Actual Sh, Dash in POSIX-compatible mode (as it will be run on Debian), and Bash in POSIX-compatible mode (as it will be run on OSX).
• Don't use any non-POSIX features when using a /bin/sh shebang.
• If calling cd, have code to handle a failure to change directories.
• If calling rm with a variable, ensure the variable is not empty.
• Prefer "$@" over "$*" unless you know exactly what you're doing.
• Prefer awk '/re/ { ... }' to grep re | awk '{ ... }'.
• Prefer find -exec {} + to find -print0 | xargs -0.
• Prefer for loops over while read loops.
• Prefer grep -c to grep | wc -l.
• Prefer mktemp over using $$ to "uniquely" name a temporary file.
• Prefer printf over echo.
• Prefer sed '/re/!d; s//.../' to grep re | sed 's/re/.../'.
• Prefer sed 'cmd; cmd' to sed -e 'cmd' -e 'cmd'.
• Prefer checking exit statuses over output in if statements (if grep -q ...; , not if [ -n "$(grep ...)" ];).
• Prefer reading environment variables over process output ($TTY not $(tty), $PWD not $(pwd), etc).
• Use $( ... ), not backticks for capturing command output.
• Use $(( ... )), not expr for executing arithmetic expressions.
• Use 1 and 0, not true and false to represent boolean variables.
• Use find -print0 | xargs -0, not find | xargs.
• Use quotes around every "$variable" and "$( ... )" expression unless you want them to be word-split and/or interpreted as globs.
• Use the local keyword with function-scoped variables.
• Identify common problems with shellcheck.

Bash

In addition to Shell best practices,

• Prefer ${var,,} and ${var^^} over tr for changing case.
• Prefer ${var//from/to} over sed for simple string replacements.
• Prefer [[ over test or [.
• Prefer process substitution over a pipe in while read loops.
• Use (( or let, not $(( when you don't need the result

Haskell

• Avoid partial functions (head, read, etc).
• Compile code with -Wall -Werror.