python-to-scala-idioms

name	python-to-scala-idioms
description	Idiomatic translation of Python constructs to Scala 2.13, including classes, interfaces, enums, and naming conventions.

When translating Python code to Scala, particularly for distributed environments (like Apache Spark), you must favor immutability and functional abstractions.

1. Enums and Constants

Python Enum should be translated to a sealed trait with case object instances in Scala 2.13.

Python:

from enum import Enum
class TokenType(Enum):
    STRING = 1
    NUMERIC = 2

Scala 2.13:

sealed trait TokenType
object TokenType {
  case object String extends TokenType
  case object Numeric extends TokenType
}

2. Data Classes

Python @dataclass translates perfectly to Scala case class. Case classes provide immutability, value-based equality, and pattern matching out of the box.

Python:

@dataclass
class Token:
    value: str
    token_type: TokenType

Scala:

case class Token(value: String, tokenType: TokenType)

3. Abstract Base Classes (Interfaces)

Python ABC should be converted to Scala trait.

Python:

class BaseTokenizer(ABC):
    @abstractmethod
    def tokenize(self, text: str) -> list[Token]: pass

Scala:

trait BaseTokenizer {
  def tokenize(text: String): Seq[Token]
}

4. Naming Conventions

• Classes/Traits/Objects: PascalCase (e.g., StringTokenizer).
• Methods/Variables: camelCase (e.g., tokenizeBatch, tokenType).
• Packages: lowercase (e.g., com.example.tokenizer).
• Scala avoids snake_case completely.

5. Companion Objects for Static Methods

Python standalone functions or @staticmethod should be placed in a companion object (an object with the same name as a class or trait in the same file).