MatterAI - Frontier LLM Intelligence at 70% Less Cost

Crystal Anti-Patterns Overview

Crystal, despite being a modern language with many safeguards, has several common anti-patterns that can lead to performance issues, maintainability problems, and bugs. Here are the most important anti-patterns to avoid when writing Crystal code.

Not Leveraging the Type System

# Anti-pattern: Using generic Object type
def process(data : Array(Object))
  data.each do |item|
    if item.is_a?(String)
      puts item.upcase
    elsif item.is_a?(Int32)
      puts item + 1
    end
  end
end

# Better approach: Use union types
def process(data : Array(String | Int32))
  data.each do |item|
    case item
    when String
      puts item.upcase
    when Int32
      puts item + 1
    end
  end
end

# Even better: Use generics with constraints
def process(data : Array(T)) forall T
  data.each do |item|
    puts item
  end
end

Leverage Crystal’s type system instead of using generic Object types. Use union types, generics, and type restrictions to make your code more type-safe and efficient. This helps catch errors at compile time rather than runtime.

Using Exceptions for Control Flow

# Anti-pattern: Using exceptions for control flow
def find_user(id)
  begin
    user = User.find(id)
    return user
  rescue
    return nil
  end
end

# Better approach: Use nil or Result types
def find_user(id) : User?
  User.find?(id)  # Returns nil if not found
end

# Alternative: Use a Result type
record Success(T), value : T
record Failure, error : String

alias Result(T) = Success(T) | Failure

def find_user(id) : Result(User)
  user = User.find?(id)
  if user
    Success.new(user)
  else
    Failure.new("User not found")
  end
end

Avoid using exceptions for normal control flow. Exceptions should be reserved for exceptional conditions. Instead, use nil or Result types to represent the possibility of failure in a type-safe way.

Not Using Proper Nil Handling

# Anti-pattern: Unsafe nil handling
def get_username(user_id)
  user = User.find(user_id)  # Might raise if not found
  return user.username       # Might raise if username is nil
end

# Better approach: Use safe navigation operator
def get_username(user_id)
  user = User.find?(user_id)
  return user.try &.username
end

# Even better: Use explicit nil handling
def get_username(user_id) : String?
  user = User.find?(user_id)
  return nil unless user
  
  user.username
end

Use proper nil handling techniques like the safe navigation operator (&.), nil checks, or the try method. This makes your code more robust and prevents nil-related runtime errors.

Mutable Global State

# Anti-pattern: Using mutable global variables
$config = {"api_key" => "secret"}

def update_config(key, value)
  $config[key] = value
end

def use_config
  puts $config["api_key"]
end

# Better approach: Use class variables with controlled access
class Config
  @@instance = {"api_key" => "secret"}
  
  def self.get(key)
    @@instance[key]?
  end
  
  def self.set(key, value)
    @@instance[key] = value
  end
end

# Usage
Config.set("api_key", "new_secret")
puts Config.get("api_key")

Avoid using mutable global state. Global variables make code harder to reason about, test, and maintain. Instead, use dependency injection, class variables with controlled access, or dedicated configuration objects.

Not Using Named Arguments for Clarity

# Anti-pattern: Using positional arguments for complex methods
def create_user(name, email, age, admin, active)
  # Create user with the given parameters
end

# Usage - unclear what each argument means
create_user("John Doe", "john@example.com", 30, true, false)

# Better approach: Use named arguments
def create_user(name : String, email : String, age : Int32, admin : Bool = false, active : Bool = true)
  # Create user with the given parameters
end

# Usage - much clearer
create_user(
  name: "John Doe",
  email: "john@example.com",
  age: 30,
  admin: true,
  active: false
)

Use named arguments for methods with many parameters or boolean flags. This makes your code more readable and less prone to errors from mixing up argument order.

Inefficient String Concatenation

# Anti-pattern: Inefficient string concatenation
def build_report(items)
  report = ""
  items.each do |item|
    report += "Item: #{item.name}, Price: #{item.price}\n"
  end
  report
end

# Better approach: Use string interpolation or String.build
def build_report(items)
  String.build do |str|
    items.each do |item|
      str << "Item: #{item.name}, Price: #{item.price}\n"
    end
  end
end

# Alternative: Use array join
def build_report(items)
  items.map { |item| "Item: #{item.name}, Price: #{item.price}" }.join("\n")
end

Avoid inefficient string concatenation in loops. Instead, use String.build, array joining, or string interpolation for better performance and readability.

Not Using Crystal's Concurrency Features Properly

# Anti-pattern: Misusing fibers without synchronization
def process_data(items)
  results = [] of String
  
  items.each do |item|
    spawn do
      result = process_item(item)
      results << result  # Unsafe concurrent access
    end
  end
  
  Fiber.yield
  results
end

# Better approach: Use channels for communication
def process_data(items)
  channel = Channel(String).new
  
  items.each do |item|
    spawn do
      result = process_item(item)
      channel.send(result)
    end
  end
  
  results = [] of String
  items.size.times do
    results << channel.receive
  end
  
  results
end

Use Crystal’s concurrency features properly. When using fibers with spawn, ensure proper synchronization for shared resources using channels, mutexes, or other synchronization primitives.

Not Using Proper Error Handling

# Anti-pattern: Ignoring errors
def save_user(user)
  user.save
  puts "User saved"
end

# Better approach: Handle errors explicitly
def save_user(user)
  begin
    user.save
    puts "User saved"
  rescue ex : ValidationError
    puts "Validation error: #{ex.message}"
    false
  rescue ex
    puts "Unexpected error: #{ex.message}"
    false
  else
    true
  end
end

Implement proper error handling in your code. Don’t ignore exceptions; catch and handle them appropriately. Use specific rescue clauses for different error types to provide better error messages and recovery strategies.

Not Using Crystal's Standard Library

# Anti-pattern: Reinventing the wheel
def parse_json(json_string)
  # Custom JSON parsing logic
end

# Better approach: Use the standard library
require "json"

def parse_json(json_string)
  JSON.parse(json_string)
end

Leverage Crystal’s rich standard library instead of reinventing the wheel. The standard library provides efficient, well-tested implementations for common tasks like JSON parsing, HTTP requests, file I/O, and more.

Not Using Type Annotations When Helpful

# Anti-pattern: Missing type annotations where helpful
def process_data(data)
  # What type is data? What does this return?
  data.map { |item| item.to_s.upcase }
end

# Better approach: Add type annotations for clarity
def process_data(data : Array(Int32)) : Array(String)
  data.map { |item| item.to_s.upcase }
end

Add type annotations to method signatures when it helps with clarity, especially for public APIs. While Crystal can often infer types, explicit annotations make your code more self-documenting and can catch type errors earlier.

Not Using Proper Abstractions

# Anti-pattern: Duplicated code without abstractions
def process_users
  users = User.all
  users.each do |user|
    # Complex processing logic
  end
end

def process_orders
  orders = Order.all
  orders.each do |order|
    # Similar complex processing logic
  end
end

# Better approach: Use abstractions
module Processable
  abstract def process
end

class User
  include Processable
  
  def process
    # User-specific processing
  end
end

class Order
  include Processable
  
  def process
    # Order-specific processing
  end
end

def process_items(items : Array(Processable))
  items.each &.process
end

Use proper abstractions like modules, inheritance, and generics to avoid code duplication. This makes your code more maintainable and extensible.

Not Using Proper Immutability

# Anti-pattern: Excessive mutability
class User
  property name : String
  property email : String
  
  def initialize(@name, @email)
  end
  
  def update_email(new_email)
    @email = new_email
  end
end

# Better approach: Use immutability where appropriate
class User
  getter name : String
  getter email : String
  
  def initialize(@name, @email)
  end
  
  def with_email(new_email) : User
    User.new(name, new_email)
  end
end

Consider using immutability for your data structures when appropriate. Immutable objects are easier to reason about, especially in concurrent contexts. Use methods that return new instances instead of modifying existing ones.

Not Using Proper Enums

# Anti-pattern: Using strings for fixed sets of values
def process_status(status : String)
  case status
  when "pending"
    # Handle pending
  when "processing"
    # Handle processing
  when "completed"
    # Handle completed
  else
    raise "Invalid status: #{status}"
  end
end

# Better approach: Use enums
enum Status
  Pending
  Processing
  Completed
end

def process_status(status : Status)
  case status
  when Status::Pending
    # Handle pending
  when Status::Processing
    # Handle processing
  when Status::Completed
    # Handle completed
  end
end

Use enums for fixed sets of related values instead of strings or integers. Enums provide type safety, better documentation, and prevent invalid values.

Not Using Proper Macros

# Anti-pattern: Overusing macros for simple cases
macro define_getter(name, value)
  def {{name.id}}
    {{value}}
  end
end

define_getter :api_url, "https://api.example.com"

# Better approach: Use simpler constructs when possible
API_URL = "https://api.example.com"

def api_url
  API_URL
end

# Appropriate use of macros for complex metaprogramming
macro define_json_mappings(properties)
  {% for key, type in properties %}
    property {{key.id}} : {{type.id}}
  {% end %}
  
  def self.from_json(json : String)
    data = JSON.parse(json)
    instance = new
    {% for key, type in properties %}
      instance.{{key.id}} = data[{{key.stringify}}].as_s
    {% end %}
    instance
  end
end

Use macros judiciously. While Crystal’s macro system is powerful, overusing macros can make code harder to understand and debug. Use simpler constructs like methods, constants, or classes when they suffice, and reserve macros for cases where metaprogramming is truly needed.

Not Using Proper Testing

# Anti-pattern: Insufficient testing
class Calculator
  def add(a, b)
    a + b
  end
end

# Better approach: Write proper tests
require "spec"

describe Calculator do
  describe "#add" do
    it "adds two positive numbers" do
      calc = Calculator.new
      calc.add(2, 3).should eq(5)
    end
    
    it "adds a positive and a negative number" do
      calc = Calculator.new
      calc.add(2, -3).should eq(-1)
    end
    
    it "adds two negative numbers" do
      calc = Calculator.new
      calc.add(-2, -3).should eq(-5)
    end
  end
end

Write comprehensive tests for your code. Crystal’s spec framework makes it easy to write and run tests. Test different scenarios, edge cases, and error conditions to ensure your code works correctly in all situations.

Not Using Proper Documentation

# Anti-pattern: Insufficient documentation
class User
  def initialize(@name, @email)
  end
  
  def valid?
    @email.includes?("@") && @name.size > 0
  end
end

# Better approach: Add proper documentation
# A class representing a user in the system.
class User
  # Creates a new user with the given name and email.
  #
  # ## Parameters
  # * name : The user's full name
  # * email : The user's email address
  def initialize(@name : String, @email : String)
  end
  
  # Checks if the user has valid information.
  #
  # Returns true if the email contains '@' and the name is not empty.
  def valid? : Bool
    @email.includes?("@") && @name.size > 0
  end
end

Document your code properly, especially public APIs. Include descriptions of classes, methods, parameters, return values, and any exceptions that might be raised. Good documentation makes your code more accessible to others and to your future self.

Not Using Proper Resource Management

# Anti-pattern: Not closing resources properly
def read_file(path)
  file = File.open(path)
  content = file.gets_to_end
  # file is never closed
  content
end

# Better approach: Use blocks for automatic resource management
def read_file(path)
  File.open(path) do |file|
    file.gets_to_end
  end
end

Use proper resource management techniques. When working with resources like files, database connections, or network sockets, use blocks that automatically handle closing the resource, or ensure you close them manually in a begin/ensure block.

Getting Started

AI Coding Agents

Axon Models

AI Code Reviews

Integrations

Crystal