unified: auto-generate parser files

Uses the `tree-sitter-generate` crate to generate these files on the
fly.

Author: tausbn

Cargo.lock

@@ -150,6 +150,7 @@ use_repo(
     "vendor_ts__tracing-subscriber-0.3.20",
     "vendor_ts__tree-sitter-0.26.8",
     "vendor_ts__tree-sitter-embedded-template-0.25.0",
+    "vendor_ts__tree-sitter-generate-0.26.8",
     "vendor_ts__tree-sitter-json-0.24.8",
     "vendor_ts__tree-sitter-language-0.1.5",
     "vendor_ts__tree-sitter-python-0.23.6",

MODULE.bazel

@@ -0,0 +1,9 @@
+# Generated by tree-sitter from grammar.js. The Cargo build script
+# (bindings/rust/build.rs) and Bazel's cargo_build_script regenerate them into
+# OUT_DIR. The tree-sitter CLI (parse, test, playground, etc.) expects them in
+# src/, so contributors can run `tree-sitter generate` locally to populate
+# these — they are intentionally untracked.
+src/parser.c
+src/grammar.json
+src/node-types.json
+src/tree_sitter/

unified/extractor/tree-sitter-swift/.gitignore

@@ -10,8 +10,10 @@ cargo_build_script(
     name = "tree-sitter-swift-build",
     srcs = ["bindings/rust/build.rs"],
     data = glob([
-        "src/**",
-    ]),
+        "src/scanner.c",
+    ]) + [
+        "grammar.js",
+    ],
     deps = all_crate_deps(
         build = True,
     ),

unified/extractor/tree-sitter-swift/BUILD.bazel

@@ -19,3 +19,4 @@ tree-sitter-language = "0.1"
 
 [build-dependencies]
 cc = "1.2"
+tree-sitter-generate = "0.26.8"

unified/extractor/tree-sitter-swift/Cargo.toml

@@ -1,19 +1,49 @@
+use std::env;
+use std::path::PathBuf;
+
 fn main() {
-    let src_dir = std::path::Path::new("src");
+    // tree-sitter-generate produces parser.c, grammar.json, node-types.json,
+    // and src/tree_sitter/*.h headers from grammar.js. We write them into
+    // OUT_DIR so the build is sandbox-friendly and we don't litter the source
+    // tree.
+    let crate_dir: PathBuf = env::var("CARGO_MANIFEST_DIR").unwrap().into();
+    let out_dir: PathBuf = env::var("OUT_DIR").unwrap().into();
+    let grammar_js = crate_dir.join("grammar.js");
+
+    tree_sitter_generate::generate_parser_in_directory(
+        &crate_dir,
+        Some(&out_dir),
+        Some(&grammar_js),
+        tree_sitter_generate::ABI_VERSION_MAX,
+        None,
+        None,
+        true,
+        tree_sitter_generate::OptLevel::default(),
+    )
+    .expect("failed to generate tree-sitter-swift parser");
 
     let mut c_config = cc::Build::new();
-    c_config.std("c11").include(src_dir);
+    c_config
+        .std("c11")
+        .include(&out_dir)
+        .include(out_dir.join("tree_sitter"));
 
     #[cfg(target_env = "msvc")]
     c_config.flag("-utf-8");
 
-    let parser_path = src_dir.join("parser.c");
-    c_config.file(&parser_path);
-    println!("cargo:rerun-if-changed={}", parser_path.to_str().unwrap());
+    c_config.file(out_dir.join("parser.c"));
+
+    // scanner.c is hand-written and lives in the source tree.
+    let scanner_path = crate_dir.join("src").join("scanner.c");
+    c_config.include(crate_dir.join("src")).file(&scanner_path);
 
-    let scanner_path = src_dir.join("scanner.c");
-    c_config.file(&scanner_path);
+    println!("cargo:rerun-if-changed={}", grammar_js.to_str().unwrap());
     println!("cargo:rerun-if-changed={}", scanner_path.to_str().unwrap());
+    // Re-export OUT_DIR so consumers can include_str! the generated files.
+    println!(
+        "cargo:rustc-env=TREE_SITTER_SWIFT_OUT_DIR={}",
+        out_dir.to_str().unwrap()
+    );
 
     c_config.compile("tree-sitter-swift");
 }

unified/extractor/tree-sitter-swift/bindings/rust/build.rs

@@ -32,7 +32,7 @@ pub const LANGUAGE: LanguageFn = unsafe { LanguageFn::from_raw(tree_sitter_swift
 /// The content of the [`node-types.json`][] file for this grammar.
 ///
 /// [`node-types.json`]: https://tree-sitter.github.io/tree-sitter/using-parsers#static-node-types
-pub const NODE_TYPES: &str = include_str!("../../src/node-types.json");
+pub const NODE_TYPES: &str = include_str!(concat!(env!("TREE_SITTER_SWIFT_OUT_DIR"), "/node-types.json"));
 
 pub const HIGHLIGHTS_QUERY: &str = include_str!("../../queries/highlights.scm");
 pub const INJECTIONS_QUERY: &str = include_str!("../../queries/injections.scm");