@willbooster/tree-sitter-kotlin 1.4.1 → 1.5.0

package/README.md

@@ -60,12 +60,15 @@ parser.setLanguage(await Language.load(kotlin));
 
 The package also ships the node types in `src/node-types.json`.
 
-In Rust, depend on the [crate](https://crates.io/crates/willbooster-tree-sitter-kotlin):
+In Rust, depend on the [crate](https://crates.io/crates/willbooster-tree-sitter-kotlin) and on
+[willbooster-tree-sitter](https://crates.io/crates/willbooster-tree-sitter), the runtime this package is tested and
+fuzzed with (the grammar also loads in the upstream `tree-sitter` crate 0.27, whose error recovery never ends on some
+malformed input):
 
 ```toml
 [dependencies]
-tree-sitter = "0.27"
-tree-sitter-kotlin = { package = "willbooster-tree-sitter-kotlin", version = "1.2" }
+tree-sitter = { package = "willbooster-tree-sitter", version = "1" }
+tree-sitter-kotlin = { package = "willbooster-tree-sitter-kotlin", version = "1" }
 ```
 
 ```rust

package/grammar.js

@@ -86,6 +86,11 @@ module.exports = grammar({
     $._in,
     $._q_dot,
     $._multiline_string_content,
+    $._multi_dollar_string_start,
+    $._multi_dollar_multiline_string_start,
+    $._multi_dollar_string_content,
+    $._multi_dollar_interpolation_start,
+    $._multi_dollar_string_end,
     // used to check if a preceding annoation should not have an automatic _semi inserted
     'constructor',
     'get',
@@ -94,6 +99,8 @@ module.exports = grammar({
     '$',
     // used to check if a modifier alone on its line belongs to a declaration on the next line
     'val',
+    // never scanned: it tells the scanner where a primary constructor may start
+    $._primary_constructor_position,
   ],
 
   inline: ($) => [$._statements, $._identifier, $._control_structure_body],
@@ -143,7 +150,7 @@ module.exports = grammar({
           choice('class', seq(optional('fun'), 'interface')),
           field('name', $.identifier),
           optional($.type_parameters),
-          optional($.primary_constructor),
+          optional(seq(optional($._primary_constructor_position), $.primary_constructor)),
           optional(seq(':', $.delegation_specifiers)),
           optional($.type_constraints),
           optional(choice($.class_body, $.enum_class_body))
@@ -681,26 +688,51 @@ module.exports = grammar({
     object_literal: ($) => seq('object', optional(seq(':', $.delegation_specifiers)), $.class_body),
 
     string_literal: ($) =>
-      seq(
-        '"',
-        repeat(
-          choice(
-            alias(
-              choice(token.immediate(prec(2, seq('\\', /[^bnrt'"\\$]/))), token.immediate(prec(1, /[^"\\$]+/)), '$'),
-              $.string_content
-            ),
-            $.escape_sequence,
-            $.interpolation
-          )
+      choice(
+        seq(
+          '"',
+          repeat(
+            choice(
+              alias(
+                choice(token.immediate(prec(2, seq('\\', /[^bnrt'"\\$]/))), token.immediate(prec(1, /[^"\\$]+/)), '$'),
+                $.string_content
+              ),
+              $.escape_sequence,
+              $.interpolation
+            )
+          ),
+          '"'
         ),
-        '"'
+        seq(
+          $._multi_dollar_string_start,
+          repeat(
+            choice(
+              alias($._multi_dollar_string_content, $.string_content),
+              $.escape_sequence,
+              alias($._multi_dollar_interpolation, $.interpolation)
+            )
+          ),
+          $._multi_dollar_string_end
+        )
       ),
 
     multiline_string_literal: ($) =>
-      seq(
-        '"""',
-        repeat(choice(alias($._multiline_string_content, $.string_content), $.interpolation)),
-        choice('"""', '""""')
+      choice(
+        seq(
+          '"""',
+          repeat(choice(alias($._multiline_string_content, $.string_content), $.interpolation)),
+          choice('"""', '""""')
+        ),
+        seq(
+          $._multi_dollar_multiline_string_start,
+          repeat(
+            choice(
+              alias($._multi_dollar_string_content, $.string_content),
+              alias($._multi_dollar_interpolation, $.interpolation)
+            )
+          ),
+          $._multi_dollar_string_end
+        )
       ),
 
     interpolation: ($) =>
@@ -710,6 +742,17 @@ module.exports = grammar({
         seq('${', $.expression, '}')
       ),
 
+    // In a string prefixed with n dollars (Kotlin 2.1), exactly n dollars start an interpolation; the scanner
+    // returns them as one token only where they do.
+    _multi_dollar_interpolation: ($) =>
+      seq(
+        $._multi_dollar_interpolation_start,
+        choice(
+          alias(token.immediate(prec(2, /[\p{L}_][\p{L}_\p{Nd}]*/u)), $.identifier),
+          seq(token.immediate('{'), $.expression, '}')
+        )
+      ),
+
     character_literal: ($) => seq("'", choice(token.immediate(prec(1, /[^'\\\r\n]/)), $.escape_sequence), "'"),
 
     escape_sequence: () => token.immediate(prec(1, seq('\\', choice(/[^xu0-7]/, /u[0-9a-fA-F]{4}/)))),

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@willbooster/tree-sitter-kotlin",
-  "version": "1.4.1",
+  "version": "1.5.0",
   "description": "Kotlin grammar for tree-sitter",
   "keywords": [
     "incremental",
@@ -22,7 +22,7 @@
   ],
   "scripts": {
     "build-wasm": "tree-sitter build --wasm",
-    "build/ci": "tree-sitter generate && tree-sitter build --wasm",
+    "build/ci": "tree-sitter generate && script/generate-letter-ranges && tree-sitter build --wasm",
     "cleanup": "bun wb lint --fix --format",
     "format": "bun wb lint --format",
     "lint": "bun wb lint",

package/tree-sitter.json

@@ -13,7 +13,7 @@
     }
   ],
   "metadata": {
-    "version": "1.4.1",
+    "version": "1.5.0",
     "license": "MIT",
     "description": "Kotlin grammar for tree-sitter",
     "authors": [