Cookbooks
Haskell
Chapter 10. Data Formats and Serialization
Working with TOML

Haskell:
Working with TOML

How to:

First, ensure you have a TOML parsing library. For Haskell, htoml is a popular choice. You’ll need to add it to your project’s dependencies.

-- Import the TOML parsing library
import qualified Text.Toml as Toml

-- Define your config data structure
data Config = Config {
  title :: String,
  owner :: Owner
} deriving (Show)

data Owner = Owner {
  name :: String,
  dob :: Maybe Day -- Optional date
} deriving (Show)

-- Parsing a TOML string
main :: IO ()
main = do
  let tomlData = "[owner]\nname = \"Tom Preston-Werner\"\ndob = 1979-05-27T07:32:00Z"
  case Toml.parseTomlDoc "" tomlData of
    Left err -> putStrLn $ "Error: " ++ show err
    Right toml -> print toml -- Or further process the parsed TOML

Sample output can be structured and accessed like any Haskell data type.

Deep Dive

Historically, TOML was created by Tom Preston-Werner, co-founder of GitHub, as a reaction to the complexities of YAML and JSON for configuration files. It emphasizes being more readable and easier to write than JSON, and more strict and simple than YAML.

Alternatives to TOML include JSON and YAML, with each format having its own strengths. JSON is ubiquitous and language-agnostic, while YAML offers a more human-readable format. TOML is valued for its simplicity and consistency, avoiding some of the pitfalls of its relatives.

Implementation in Haskell typically involves a library that parses TOML into a Haskell data type, often leveraging Haskell’s advanced type system to ensure correctness. Parsing can be done through recursive descent or combinator parsing, which balances efficiency with readability and maintainability of the code.

See Also

Working with JSONWorking with XML