tree-sitter-scala

tree-sitter-scala is a tree-sitter grammar package that teaches the tree-sitter incremental parsing engine how to parse Scala 2 and Scala 3 source code into a concrete syntax tree. It’s consumed as a Rust crate (or via its C, Python, Go, Node, and Swift bindings) by editors, linters, and static-analysis tools that need fast, error-tolerant, incremental parsing of Scala files.

Because tree-sitter parsers are designed for editor-speed reparsing on every keystroke, tree-sitter-scala underpins Scala syntax highlighting, code folding, and structural navigation in tools like Neovim, Helix, and Zed, as well as any custom tooling built on the tree-sitter Rust bindings.

What You Get

• A generated C parser (src/parser.c) plus scanner for Scala 2/3 syntax, compiled via the tree-sitter-language crate binding
• Highlight, tag, and injection query files (queries/) for syntax highlighting and code-navigation integrations
• Bindings for Rust, C, Node.js, Python, Go, and Swift so the same grammar can be embedded in tools written in any of those languages
• A test corpus (test/corpus, test/highlight, test/tags) validating parser correctness against real Scala syntax

Common Use Cases

• Powering Scala syntax highlighting and code folding in editors built on tree-sitter (Neovim, Helix, Zed)
• Building static-analysis, linting, or code-search tools that need a fast incremental Scala parser
• Embedding Scala-aware structural navigation (jump-to-definition-style queries) in custom developer tooling
• Generating syntax trees for Scala source as part of a documentation or refactoring pipeline

Under The Hood

Architecture: The grammar is defined declaratively in grammar.js using tree-sitter’s DSL, then compiled offline into a generated src/parser.c (a table-driven GLR-style parser) plus a hand-written src/scanner.c for context-sensitive tokens (e.g. Scala’s significant indentation and string interpolation); the Rust crate (bindings/rust/lib.rs) links this C code via a build script and exposes the LANGUAGE constant consumed by the tree-sitter crate. Tech Stack: Core grammar logic is C (generated) plus a JS grammar-definition file; the Rust binding depends only on tree-sitter-language and a C build (cc crate) at build time, keeping the runtime dependency surface minimal; parallel bindings exist for Node (npm), Python (setup.py/pyproject.toml), Go (go.mod), and Swift (Package.swift). Code Quality: The test/corpus directory contains extensive example-based grammar tests exercising Scala 2 and Scala 3 syntax edge cases, plus separate test/highlight and test/tags suites verifying the query files produce correct highlighting/tagging output; grammar changes are validated via tree-sitter test in CI across all bindings. API Design: Consumers interact with a single exported LANGUAGE (Rust) or equivalent per-binding export, following tree-sitter’s standard cross-language convention, so switching from one tree-sitter grammar crate to another requires no API relearning.