Fourmolu vs Taplo

Fourmolu is a configurable formatter for Haskell source code and a fork of Ormolu. It uses GHC’s parser and aims to keep diffs small, produce idempotent output, and handle large projects. Set formatting options in `fourmolu.yaml` or pass them on the command line. Controls include indentation, line length, import grouping, record layout, Haddock comments, Unicode output, and operator fixities. Fourmolu can find the relevant Cabal file to read default extensions and dependencies that help determine fixities. Magic comments can disable formatting for selected code, and command-line options can target a line region. There is an interactive browser demo with configurable formatting and program options. Installation instructions cover GHCup, Cabal, Stack, and dotslash; editor integration is documented through Haskell Language Server. The project recommends a GitHub Action to keep projects formatted. Its listed current release is v0.21.0.0. CPP handling is experimental and limited to simple cases where conditionals surround top-level declarations. Output always uses LF line endings, which can cause check mode to fail on CRLF files.

Taplo is a free toolkit for validating, formatting, and querying TOML documents, with support for TOML 1.0.0. Its validator checks syntax and semantic issues such as duplicate keys, and can apply JSON Schema Draft 4 validation. The formatter can edit files in place or use standard input and output; by default, it stops when syntax errors occur to avoid destructive edits. The CLI also converts TOML to JSON and retrieves values through query expressions. Taplo’s language server communicates through standard input/output or TCP, though multiple TCP clients are not supported. It is excluded from default builds and is unavailable in the NPM installation. The Even Better TOML Visual Studio Code extension adds highlighting, validation, folding, navigation, renaming, formatting, and JSON Schema completion; its listing calls it a preview that may contain bugs or crash. Taplo can retrieve schemas from catalogs such as JSON Schema Store and match them to files by name. Install options include precompiled releases, Cargo, Homebrew, NPM, and Docker; precompiled releases are listed for Linux, macOS, and Windows.