code-gauge 5.0.0 → 6.0.0

package/dist/thresholds.d.ts

@@ -0,0 +1,38 @@
+import type { CodeMetrics, FunctionMetrics } from './types.js';
+/**
+ * One limit of `code-gauge check`. Its `key` is the config setting, the per-language override
+ * setting, and (kebab-cased) the command-line option, so a new entry here is all a new thresholded
+ * metric needs.
+ */
+export interface Threshold<Subject = never> {
+    key: string;
+    /** How the report names the measured value. */
+    label: string;
+    defaultLimit: number;
+    /** What to change in the code when the limit is exceeded. */
+    hint: string;
+    measure: (subject: Subject) => number;
+}
+/** Limits by threshold key; `Infinity` disables a limit. */
+export type Limits = Readonly<Record<string, number>>;
+export declare const functionThresholds: readonly Threshold<FunctionMetrics>[];
+export declare const fileThresholds: readonly Threshold<CodeMetrics>[];
+/** A duplicated block violates when it spans at least this many lines. */
+export declare const duplicationThreshold: Threshold<{
+    startLine: number;
+    endLine: number;
+}>;
+export declare const thresholds: readonly Threshold[];
+/** The name a report gives a threshold's measured value: its key without the `max`/`min` prefix. */
+export declare function metricNameOf(threshold: Threshold): string;
+/** The `thresholds` config section; a missing key falls back to the less specific layer. */
+export interface ThresholdsConfig {
+    limits: Limits;
+    /** Overrides for the files of one language, keyed by language name. */
+    languages: Readonly<Record<string, Limits>>;
+}
+/**
+ * The limits applying to files of each language, with precedence command-line options > the
+ * config's per-language overrides > the config's limits > built-in defaults.
+ */
+export declare function resolveLimits(cliLimits: Limits, config: ThresholdsConfig | undefined): (language: string) => Limits;

package/dist/thresholds.js

@@ -0,0 +1,2 @@
+const e=[{key:`maxFunctionCognitiveComplexity`,label:`cognitive complexity`,defaultLimit:15,hint:`flatten nested branching with early returns and extract nested blocks into named functions.`,measure:e=>e.cognitiveComplexity},{key:`maxFunctionCyclomaticComplexity`,label:`cyclomatic complexity`,defaultLimit:10,hint:`the function has too many independent paths; split it by decision or replace condition chains with a lookup table.`,measure:e=>e.cyclomaticComplexity},{key:`maxFunctionNcss`,label:`NCSS`,defaultLimit:60,hint:`split the function into smaller functions that each do one step.`,measure:e=>e.ncss},{key:`maxFunctionNestingDepth`,label:`nesting depth`,defaultLimit:4,hint:`replace nested conditions with guard clauses or move inner blocks into functions.`,measure:e=>e.nestingDepth},{key:`maxFunctionParameterCount`,label:`parameters`,defaultLimit:7,hint:`group related parameters into one object or split the function.`,measure:e=>e.parameterCount},{key:`maxFunctionHalsteadVolume`,label:`Halstead volume`,defaultLimit:2e3,hint:`the function holds too many operators and operands; move cohesive parts into helper functions.`,measure:e=>e.halstead.volume},{key:`maxFunctionHalsteadDifficulty`,label:`Halstead difficulty`,defaultLimit:20,hint:`the function reuses the same operands many times; split it so each part handles fewer values.`,measure:e=>e.halstead.difficulty},{key:`maxFunctionHalsteadEffort`,label:`Halstead effort`,defaultLimit:3e4,hint:`the function is both large and dense; split it into smaller functions with fewer values each.`,measure:e=>e.halstead.effort},{key:`maxFunctionDepDegree`,label:`DepDegree`,defaultLimit:50,hint:`too many values flow between the variables of the function; split it so each part works on fewer variables.`,measure:e=>e.depDegree}],t=[{key:`maxFileNcss`,label:`file NCSS`,defaultLimit:500,hint:`split the file into smaller modules with one responsibility each.`,measure:e=>e.ncssCount}],n={key:`minDuplicateLines`,label:`duplicated lines`,defaultLimit:10,hint:`extract the repeated code into one shared function or module and call it from every location.`,measure:e=>e.endLine-e.startLine+1},r=[...e,...t,n];function i(e){let t=e.key.slice(3);return t.charAt(0).toLowerCase()+t.slice(1)}function a(e,t){let n={...Object.fromEntries(r.map(({key:e,defaultLimit:t})=>[e,t])),...t?.limits};return r=>({...n,...t&&Object.hasOwn(t.languages,r)?t.languages[r]:void 0,...e})}export{n as duplicationThreshold,t as fileThresholds,e as functionThresholds,i as metricNameOf,a as resolveLimits,r as thresholds};
+//# sourceMappingURL=thresholds.js.map

package/dist/thresholds.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"thresholds.js","names":[],"sources":["../src/thresholds.ts"],"sourcesContent":["import type { CodeMetrics, FunctionMetrics } from './types.js';\n\n/**\n * One limit of `code-gauge check`. Its `key` is the config setting, the per-language override\n * setting, and (kebab-cased) the command-line option, so a new entry here is all a new thresholded\n * metric needs.\n */\nexport interface Threshold<Subject = never> {\n  key: string;\n  /** How the report names the measured value. */\n  label: string;\n  defaultLimit: number;\n  /** What to change in the code when the limit is exceeded. */\n  hint: string;\n  measure: (subject: Subject) => number;\n}\n\n/** Limits by threshold key; `Infinity` disables a limit. */\nexport type Limits = Readonly<Record<string, number>>;\n\nexport const functionThresholds: readonly Threshold<FunctionMetrics>[] = [\n  {\n    key: 'maxFunctionCognitiveComplexity',\n    label: 'cognitive complexity',\n    defaultLimit: 15,\n    hint: 'flatten nested branching with early returns and extract nested blocks into named functions.',\n    measure: (fn) => fn.cognitiveComplexity,\n  },\n  {\n    key: 'maxFunctionCyclomaticComplexity',\n    label: 'cyclomatic complexity',\n    defaultLimit: 10,\n    hint: 'the function has too many independent paths; split it by decision or replace condition chains with a lookup table.',\n    measure: (fn) => fn.cyclomaticComplexity,\n  },\n  {\n    key: 'maxFunctionNcss',\n    label: 'NCSS',\n    defaultLimit: 60,\n    hint: 'split the function into smaller functions that each do one step.',\n    measure: (fn) => fn.ncss,\n  },\n  {\n    key: 'maxFunctionNestingDepth',\n    label: 'nesting depth',\n    defaultLimit: 4,\n    hint: 'replace nested conditions with guard clauses or move inner blocks into functions.',\n    measure: (fn) => fn.nestingDepth,\n  },\n  {\n    key: 'maxFunctionParameterCount',\n    label: 'parameters',\n    defaultLimit: 7,\n    hint: 'group related parameters into one object or split the function.',\n    measure: (fn) => fn.parameterCount,\n  },\n  {\n    key: 'maxFunctionHalsteadVolume',\n    label: 'Halstead volume',\n    defaultLimit: 2000,\n    hint: 'the function holds too many operators and operands; move cohesive parts into helper functions.',\n    measure: (fn) => fn.halstead.volume,\n  },\n  {\n    key: 'maxFunctionHalsteadDifficulty',\n    label: 'Halstead difficulty',\n    defaultLimit: 20,\n    hint: 'the function reuses the same operands many times; split it so each part handles fewer values.',\n    measure: (fn) => fn.halstead.difficulty,\n  },\n  {\n    key: 'maxFunctionHalsteadEffort',\n    label: 'Halstead effort',\n    defaultLimit: 30_000,\n    hint: 'the function is both large and dense; split it into smaller functions with fewer values each.',\n    measure: (fn) => fn.halstead.effort,\n  },\n  {\n    key: 'maxFunctionDepDegree',\n    label: 'DepDegree',\n    defaultLimit: 50,\n    hint: 'too many values flow between the variables of the function; split it so each part works on fewer variables.',\n    measure: (fn) => fn.depDegree,\n  },\n];\n\nexport const fileThresholds: readonly Threshold<CodeMetrics>[] = [\n  {\n    key: 'maxFileNcss',\n    label: 'file NCSS',\n    defaultLimit: 500,\n    hint: 'split the file into smaller modules with one responsibility each.',\n    measure: (file) => file.ncssCount,\n  },\n];\n\n/** A duplicated block violates when it spans at least this many lines. */\nexport const duplicationThreshold: Threshold<{ startLine: number; endLine: number }> = {\n  key: 'minDuplicateLines',\n  label: 'duplicated lines',\n  defaultLimit: 10,\n  hint: 'extract the repeated code into one shared function or module and call it from every location.',\n  measure: (block) => block.endLine - block.startLine + 1,\n};\n\nexport const thresholds: readonly Threshold[] = [...functionThresholds, ...fileThresholds, duplicationThreshold];\n\n/** The name a report gives a threshold's measured value: its key without the `max`/`min` prefix. */\nexport function metricNameOf(threshold: Threshold): string {\n  const name = threshold.key.slice(3);\n  return name.charAt(0).toLowerCase() + name.slice(1);\n}\n\n/** The `thresholds` config section; a missing key falls back to the less specific layer. */\nexport interface ThresholdsConfig {\n  limits: Limits;\n  /** Overrides for the files of one language, keyed by language name. */\n  languages: Readonly<Record<string, Limits>>;\n}\n\n/**\n * The limits applying to files of each language, with precedence command-line options > the\n * config's per-language overrides > the config's limits > built-in defaults.\n */\nexport function resolveLimits(cliLimits: Limits, config: ThresholdsConfig | undefined): (language: string) => Limits {\n  const defaults = Object.fromEntries(thresholds.map(({ key, defaultLimit }) => [key, defaultLimit]));\n  const shared = { ...defaults, ...config?.limits };\n  return (language) => ({\n    ...shared,\n    // Object.hasOwn: a language named like an Object.prototype member must not read an inherited value.\n    ...(config && Object.hasOwn(config.languages, language) ? config.languages[language] : undefined),\n    ...cliLimits,\n  });\n}\n"],"mappings":"AAoBA,MAAa,EAA4D,CACvE,CACE,IAAK,iCACL,MAAO,uBACP,aAAc,GACd,KAAM,8FACN,QAAU,GAAO,EAAG,mBACtB,EACA,CACE,IAAK,kCACL,MAAO,wBACP,aAAc,GACd,KAAM,qHACN,QAAU,GAAO,EAAG,oBACtB,EACA,CACE,IAAK,kBACL,MAAO,OACP,aAAc,GACd,KAAM,mEACN,QAAU,GAAO,EAAG,IACtB,EACA,CACE,IAAK,0BACL,MAAO,gBACP,aAAc,EACd,KAAM,oFACN,QAAU,GAAO,EAAG,YACtB,EACA,CACE,IAAK,4BACL,MAAO,aACP,aAAc,EACd,KAAM,kEACN,QAAU,GAAO,EAAG,cACtB,EACA,CACE,IAAK,4BACL,MAAO,kBACP,aAAc,IACd,KAAM,iGACN,QAAU,GAAO,EAAG,SAAS,MAC/B,EACA,CACE,IAAK,gCACL,MAAO,sBACP,aAAc,GACd,KAAM,gGACN,QAAU,GAAO,EAAG,SAAS,UAC/B,EACA,CACE,IAAK,4BACL,MAAO,kBACP,aAAc,IACd,KAAM,gGACN,QAAU,GAAO,EAAG,SAAS,MAC/B,EACA,CACE,IAAK,uBACL,MAAO,YACP,aAAc,GACd,KAAM,8GACN,QAAU,GAAO,EAAG,SACtB,CACF,EAEa,EAAoD,CAC/D,CACE,IAAK,cACL,MAAO,YACP,aAAc,IACd,KAAM,oEACN,QAAU,GAAS,EAAK,SAC1B,CACF,EAGa,EAA0E,CACrF,IAAK,oBACL,MAAO,mBACP,aAAc,GACd,KAAM,gGACN,QAAU,GAAU,EAAM,QAAU,EAAM,UAAY,CACxD,EAEa,EAAmC,CAAC,GAAG,EAAoB,GAAG,EAAgB,CAAoB,EAG/G,SAAgB,EAAa,EAA8B,CACzD,IAAM,EAAO,EAAU,IAAI,MAAM,CAAC,EAClC,OAAO,EAAK,OAAO,CAAC,CAAC,CAAC,YAAY,EAAI,EAAK,MAAM,CAAC,CACpD,CAaA,SAAgB,EAAc,EAAmB,EAAoE,CAEnH,IAAM,EAAS,CAAE,GADA,OAAO,YAAY,EAAW,KAAK,CAAE,MAAK,kBAAmB,CAAC,EAAK,CAAY,CAAC,CACtE,EAAG,GAAG,GAAQ,MAAO,EAChD,MAAQ,KAAc,CACpB,GAAG,EAEH,GAAI,GAAU,OAAO,OAAO,EAAO,UAAW,CAAQ,EAAI,EAAO,UAAU,GAAY,IAAA,GACvF,GAAG,CACL,EACF"}

package/dist/types.d.ts

@@ -49,6 +49,11 @@ export interface HalsteadMetrics {
     vocabulary: number;
     length: number;
     volume: number;
+    /**
+     * (distinctOperators / 2) * (totalOperands / distinctOperands), i.e. effort / volume; 0 when
+     * there are no operands.
+     */
+    difficulty: number;
     effort: number;
 }
 export interface FunctionMetrics {
@@ -72,7 +77,7 @@ export interface FunctionMetrics {
      * short-circuit operator (`&&`, `||`, `and`, `or`) wherever it appears, and per case-labelled
      * statement or match arm (stacked labels share one; default and catch-all arms add none).
      * Full-evaluation operators and `throw` add nothing. Nested function bodies are excluded, so
-     * summing over functions counts each decision once. Not used by ranking or the regression gate.
+     * summing over functions counts each decision once.
      */
     cyclomaticComplexity: number;
     cognitiveComplexity: number;
@@ -92,9 +97,8 @@ export interface FunctionMetrics {
     /**
      * Approximate def-use dependency degree (DepDegree, Beyer & Fararooy 2010): the number of
      * variable reads whose name has a preceding definition (declaration, assignment, or parameter)
-     * within the same function. A file-local single-assignment approximation — each read is charged
-     * one reaching definition — which is stable enough for regression ratcheting where only the
-     * delta matters.
+     * within the same function. A file-local single-assignment approximation: each read is charged
+     * one reaching definition.
      */
     depDegree: number;
 }

package/dist/worker.cjs

@@ -1,2 +1,2 @@
-"use strict";Object.defineProperty(exports,Symbol.toStringTag,{value:`Module`});const e=require("./_virtual/_rolldown/runtime.cjs"),t=require("./crossFileDuplication.cjs"),n=require("./languages.cjs"),r=require("./nativeMetrics.cjs"),i=require("./metrics.cjs"),a=require("./newDuplication.cjs"),o=require("./regressionGate.cjs");require("./index.cjs");const s=require("./wasmBinding.cjs");let c=require("../native/code-gauge.wasm");c=e.__toESM(c,1),r.setNativeBinding(s.createWasmBinding(c.default)),exports.TreeMeasurer=i.TreeMeasurer,exports.collectCrossFileDuplicationFileData=i.collectCrossFileDuplicationFileData,exports.collectDuplicationCandidates=i.collectDuplicationCandidates,exports.collectFunctionTokenSequences=i.collectFunctionTokenSequences,exports.defaultGateOptions=o.defaultGateOptions,exports.defaultLanguages=n.defaultLanguages,exports.defaultMeasurer=i.defaultMeasurer,exports.detectLanguage=n.detectLanguage,exports.evaluateRegressionGate=o.evaluateRegressionGate,exports.findNewlyDuplicatedLines=a.findNewlyDuplicatedLines,exports.measureCode=i.measureCode,exports.measureCrossFileDuplication=t.measureCrossFileDuplication,exports.supportedLanguages=n.supportedLanguages;
+"use strict";Object.defineProperty(exports,Symbol.toStringTag,{value:`Module`});const e=require("./_virtual/_rolldown/runtime.cjs"),t=require("./crossFileDuplication.cjs"),n=require("./languages.cjs"),r=require("./nativeMetrics.cjs"),i=require("./metrics.cjs");require("./index.cjs");const a=require("./wasmBinding.cjs");let o=require("../native/code-gauge.wasm");o=e.__toESM(o,1),r.setNativeBinding(a.createWasmBinding(o.default)),exports.TreeMeasurer=i.TreeMeasurer,exports.collectCrossFileDuplicationFileData=i.collectCrossFileDuplicationFileData,exports.collectDuplicationCandidates=i.collectDuplicationCandidates,exports.collectFunctionTokenSequences=i.collectFunctionTokenSequences,exports.defaultLanguages=n.defaultLanguages,exports.defaultMeasurer=i.defaultMeasurer,exports.detectLanguage=n.detectLanguage,exports.measureCode=i.measureCode,exports.measureCrossFileDuplication=t.measureCrossFileDuplication,exports.supportedLanguages=n.supportedLanguages;
 //# sourceMappingURL=worker.cjs.map

package/dist/worker.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"worker.cjs","names":["setNativeBinding","createWasmBinding","wasmModule"],"sources":["../src/worker.ts"],"sourcesContent":["// The entry point for Cloudflare Workers (the `workerd` export condition), which cannot load N-API\n// addons: the same API runs on the WebAssembly build of the native addon.\nimport wasmModule from '../native/code-gauge.wasm';\nimport { setNativeBinding } from './nativeMetrics.js';\nimport { createWasmBinding } from './wasmBinding.js';\n\nsetNativeBinding(createWasmBinding(wasmModule));\n\nexport * from './index.js';\n"],"mappings":"icAMAA,EAAAA,iBAAiBC,EAAAA,kBAAkBC,EAAAA,OAAU,CAAC"}
+{"version":3,"file":"worker.cjs","names":["setNativeBinding","createWasmBinding","wasmModule"],"sources":["../src/worker.ts"],"sourcesContent":["// The entry point for Cloudflare Workers (the `workerd` export condition), which cannot load N-API\n// addons: the same API runs on the WebAssembly build of the native addon.\nimport wasmModule from '../native/code-gauge.wasm';\nimport { setNativeBinding } from './nativeMetrics.js';\nimport { createWasmBinding } from './wasmBinding.js';\n\nsetNativeBinding(createWasmBinding(wasmModule));\n\nexport * from './index.js';\n"],"mappings":"6XAMAA,EAAAA,iBAAiBC,EAAAA,kBAAkBC,EAAAA,OAAU,CAAC"}

package/dist/worker.js

@@ -1,2 +1,2 @@
-import{measureCrossFileDuplication as e}from"./crossFileDuplication.js";import{defaultLanguages as t,detectLanguage as n,supportedLanguages as r}from"./languages.js";import{setNativeBinding as i}from"./nativeMetrics.js";import{TreeMeasurer as a,collectCrossFileDuplicationFileData as o,collectDuplicationCandidates as s,collectFunctionTokenSequences as c,defaultMeasurer as l,measureCode as u}from"./metrics.js";import{findNewlyDuplicatedLines as d}from"./newDuplication.js";import{defaultGateOptions as f,evaluateRegressionGate as p}from"./regressionGate.js";import"./index.js";import{createWasmBinding as m}from"./wasmBinding.js";import h from"../native/code-gauge.wasm";i(m(h));export{a as TreeMeasurer,o as collectCrossFileDuplicationFileData,s as collectDuplicationCandidates,c as collectFunctionTokenSequences,f as defaultGateOptions,t as defaultLanguages,l as defaultMeasurer,n as detectLanguage,p as evaluateRegressionGate,d as findNewlyDuplicatedLines,u as measureCode,e as measureCrossFileDuplication,r as supportedLanguages};
+import{measureCrossFileDuplication as e}from"./crossFileDuplication.js";import{defaultLanguages as t,detectLanguage as n,supportedLanguages as r}from"./languages.js";import{setNativeBinding as i}from"./nativeMetrics.js";import{TreeMeasurer as a,collectCrossFileDuplicationFileData as o,collectDuplicationCandidates as s,collectFunctionTokenSequences as c,defaultMeasurer as l,measureCode as u}from"./metrics.js";import"./index.js";import{createWasmBinding as d}from"./wasmBinding.js";import f from"../native/code-gauge.wasm";i(d(f));export{a as TreeMeasurer,o as collectCrossFileDuplicationFileData,s as collectDuplicationCandidates,c as collectFunctionTokenSequences,t as defaultLanguages,l as defaultMeasurer,n as detectLanguage,u as measureCode,e as measureCrossFileDuplication,r as supportedLanguages};
 //# sourceMappingURL=worker.js.map

package/dist/worker.js.map

@@ -1 +1 @@
-{"version":3,"file":"worker.js","names":[],"sources":["../src/worker.ts"],"sourcesContent":["// The entry point for Cloudflare Workers (the `workerd` export condition), which cannot load N-API\n// addons: the same API runs on the WebAssembly build of the native addon.\nimport wasmModule from '../native/code-gauge.wasm';\nimport { setNativeBinding } from './nativeMetrics.js';\nimport { createWasmBinding } from './wasmBinding.js';\n\nsetNativeBinding(createWasmBinding(wasmModule));\n\nexport * from './index.js';\n"],"mappings":"iqBAMA,EAAiB,EAAkB,CAAU,CAAC"}
+{"version":3,"file":"worker.js","names":[],"sources":["../src/worker.ts"],"sourcesContent":["// The entry point for Cloudflare Workers (the `workerd` export condition), which cannot load N-API\n// addons: the same API runs on the WebAssembly build of the native addon.\nimport wasmModule from '../native/code-gauge.wasm';\nimport { setNativeBinding } from './nativeMetrics.js';\nimport { createWasmBinding } from './wasmBinding.js';\n\nsetNativeBinding(createWasmBinding(wasmModule));\n\nexport * from './index.js';\n"],"mappings":"6gBAMA,EAAiB,EAAkB,CAAU,CAAC"}

package/native/Cargo.lock

@@ -45,18 +45,18 @@ dependencies = [
  "rustc-hash",
  "serde",
  "serde_json",
- "tree-sitter",
- "tree-sitter-c",
- "tree-sitter-c-sharp",
- "tree-sitter-cpp",
  "tree-sitter-go",
  "tree-sitter-java",
- "tree-sitter-javascript",
  "tree-sitter-python",
  "tree-sitter-ruby",
- "tree-sitter-rust",
- "tree-sitter-typescript",
+ "willbooster-tree-sitter",
+ "willbooster-tree-sitter-c",
+ "willbooster-tree-sitter-c-sharp",
+ "willbooster-tree-sitter-cpp",
+ "willbooster-tree-sitter-javascript",
  "willbooster-tree-sitter-kotlin",
+ "willbooster-tree-sitter-rust",
+ "willbooster-tree-sitter-typescript",
 ]
 
 [[package]]
@@ -425,141 +425,141 @@ dependencies = [
 ]
 
 [[package]]
-name = "tree-sitter"
-version = "0.25.10"
+name = "tree-sitter-go"
+version = "0.25.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "78f873475d258561b06f1c595d93308a7ed124d9977cb26b148c2084a4a3cc87"
+checksum = "c8560a4d2f835cc0d4d2c2e03cbd0dde2f6114b43bc491164238d333e28b16ea"
 dependencies = [
  "cc",
- "regex",
- "regex-syntax",
- "serde_json",
- "streaming-iterator",
  "tree-sitter-language",
 ]
 
 [[package]]
-name = "tree-sitter-c"
-version = "0.21.4"
+name = "tree-sitter-java"
+version = "0.23.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f956d5351d62652864a4ff3ae861747e7a1940dc96c9998ae400ac0d3ce30427"
+checksum = "0aa6cbcdc8c679b214e616fd3300da67da0e492e066df01bcf5a5921a71e90d6"
 dependencies = [
  "cc",
- "tree-sitter",
+ "tree-sitter-language",
 ]
 
 [[package]]
-name = "tree-sitter-c-sharp"
-version = "0.21.3"
+name = "tree-sitter-language"
+version = "0.1.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8997ad04502208449025114e434c9024a33a74e700513c702a9d2cac6522a771"
-dependencies = [
- "cc",
- "tree-sitter",
-]
+checksum = "ca0d1bf6fdd806e43ae5198f82f527056d359def39e54e67a0f478ac09dac081"
 
 [[package]]
-name = "tree-sitter-cpp"
-version = "0.22.3"
+name = "tree-sitter-python"
+version = "0.25.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1d509a22a992790d38f2c291961ff8a1ff016c437c7ec6befc9220b8eec8918c"
+checksum = "6bf85fd39652e740bf60f46f4cda9492c3a9ad75880575bf14960f775cb74a1c"
 dependencies = [
  "cc",
- "tree-sitter",
+ "tree-sitter-language",
 ]
 
 [[package]]
-name = "tree-sitter-go"
-version = "0.21.2"
+name = "tree-sitter-ruby"
+version = "0.23.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b8d702a98d3c7e70e466456e58ff2b1ac550bf1e29b97e5770676d2fdabec00d"
+checksum = "be0484ea4ef6bb9c575b4fdabde7e31340a8d2dbc7d52b321ac83da703249f95"
 dependencies = [
  "cc",
- "tree-sitter",
+ "tree-sitter-language",
 ]
 
 [[package]]
-name = "tree-sitter-java"
-version = "0.21.0"
+name = "unicode-ident"
+version = "1.0.24"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "33bc21adf831a773c075d9d00107ab43965e6a6ea7607b47fd9ec6f3db4b481b"
-dependencies = [
- "cc",
- "tree-sitter",
-]
+checksum = "e6e4313cd5fcd3dad5cafa179702e2b244f760991f45397d14d4ebf38247da75"
 
 [[package]]
-name = "tree-sitter-javascript"
-version = "0.25.0"
+name = "unicode-segmentation"
+version = "1.13.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "68204f2abc0627a90bdf06e605f5c470aa26fdcb2081ea553a04bdad756693f5"
+checksum = "c6f5d3c3b1bf09027a88a6bc961fc00497d651009560b5463668dc81b0fa87a8"
+
+[[package]]
+name = "willbooster-tree-sitter"
+version = "1.0.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7c6ca3007235ade028b793d6090a5ea16618ea3ab9b53427d1f782b15bf64064"
 dependencies = [
  "cc",
+ "regex",
+ "serde_json",
+ "streaming-iterator",
  "tree-sitter-language",
 ]
 
 [[package]]
-name = "tree-sitter-language"
-version = "0.1.8"
+name = "willbooster-tree-sitter-c"
+version = "1.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ca0d1bf6fdd806e43ae5198f82f527056d359def39e54e67a0f478ac09dac081"
+checksum = "5cee6f1a4e0f64f337d28aeaed73879d7e8a87d77c287afb989ce29adea54700"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
 
 [[package]]
-name = "tree-sitter-python"
-version = "0.21.0"
+name = "willbooster-tree-sitter-c-sharp"
+version = "1.0.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b4066c6cf678f962f8c2c4561f205945c84834cce73d981e71392624fdc390a9"
+checksum = "99b224dcd59cda9b24b588557cca3827abaccaf10da72d125c31e7cdb2e6f7fb"
 dependencies = [
  "cc",
- "tree-sitter",
+ "tree-sitter-language",
 ]
 
 [[package]]
-name = "tree-sitter-ruby"
-version = "0.21.0"
+name = "willbooster-tree-sitter-cpp"
+version = "1.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c0031f687c0772f2dad7b77104c43428611099a1804c81244ada21560f41f0b1"
+checksum = "72b92a9fcb0dc27bd4af18e06fe6c085dd14d91f009f60ecd92c0308b9cbd068"
 dependencies = [
  "cc",
- "tree-sitter",
+ "tree-sitter-language",
 ]
 
 [[package]]
-name = "tree-sitter-rust"
-version = "0.21.2"
+name = "willbooster-tree-sitter-javascript"
+version = "1.0.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "277690f420bf90741dea984f3da038ace46c4fe6047cba57a66822226cde1c93"
+checksum = "e6c40d41b303b525194daa503a71ec94d8269fe8d1648b2d9138fe902623beff"
 dependencies = [
  "cc",
- "tree-sitter",
+ "tree-sitter-language",
 ]
 
 [[package]]
-name = "tree-sitter-typescript"
-version = "0.21.2"
+name = "willbooster-tree-sitter-kotlin"
+version = "1.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ecb35d98a688378e56c18c9c159824fd16f730ccbea19aacf4f206e5d5438ed9"
+checksum = "529ae914939b559754e6dd8b34d2705f04d6dcb73f09cb52f024b49eda9b5b88"
 dependencies = [
  "cc",
- "tree-sitter",
+ "tree-sitter-language",
 ]
 
 [[package]]
-name = "unicode-ident"
-version = "1.0.24"
+name = "willbooster-tree-sitter-rust"
+version = "1.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e6e4313cd5fcd3dad5cafa179702e2b244f760991f45397d14d4ebf38247da75"
+checksum = "324c9f18cfc29e4da7335ad8cb76beb6ac1ae91a8037a44724bc55fce04f26f7"
+dependencies = [
+ "cc",
+ "tree-sitter-language",
+]
 
 [[package]]
-name = "unicode-segmentation"
-version = "1.13.3"
+name = "willbooster-tree-sitter-typescript"
+version = "1.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c6f5d3c3b1bf09027a88a6bc961fc00497d651009560b5463668dc81b0fa87a8"
-
-[[package]]
-name = "willbooster-tree-sitter-kotlin"
-version = "0.0.0"
-source = "git+https://github.com/WillBooster/tree-sitter-kotlin?tag=v1.1.0#a05912e6f07e2cedd44bfb117940361089d4c3e3"
+checksum = "f8c51b6c9e10d6ae34c2f84c4d44fff79a0210890ede63fda558a6a0f042eb28"
 dependencies = [
  "cc",
  "tree-sitter-language",

package/native/Cargo.toml

@@ -16,27 +16,22 @@ regex = "1"
 rustc-hash = "2"
 
 # Grammar crates are pinned exactly: every metric (and the golden-based test suite) is calibrated
-# against these grammar versions, so upgrades are deliberate, test-verified changes. The Kotlin
-# grammar is WillBooster's fork, which is not on crates.io, pinned to a release tag. Core 0.25 is
-# the first to load that grammar's ABI 15 parser, and tree-sitter-javascript 0.21.x pins
-# `cc ~1.0`, which conflicts with the core's `cc ^1.2`, hence 0.25.0. The core reports no field for
-# extra (error-recovery) children in ts_node_field_name_for_child, which
-# util::find_children_by_field_name compensates for.
-tree-sitter = "=0.25.10"
-tree-sitter-c = "=0.21.4"
-tree-sitter-c-sharp = "=0.21.3"
-tree-sitter-cpp = "=0.22.3"
-tree-sitter-go = "=0.21.2"
-tree-sitter-java = "=0.21.0"
-tree-sitter-kotlin = { package = "willbooster-tree-sitter-kotlin", git = "https://github.com/WillBooster/tree-sitter-kotlin", tag = "v1.1.0" }
-tree-sitter-javascript = "=0.25.0"
-tree-sitter-python = "=0.21.0"
-tree-sitter-ruby = "=0.21.0"
-# 0.21.0 breaks MSVC (its build script passes the GCC-only -Wno-unused-parameter
-# unconditionally; cl.exe fails with D8021); 0.21.2 gates the flag and the golden-based test
-# suite verifies its grammar changes leave every measured metric unchanged.
-tree-sitter-rust = "=0.21.2"
-tree-sitter-typescript = "=0.21.2"
+# against these grammar versions, so upgrades are deliberate, test-verified changes. The
+# `willbooster-*` crates are WillBooster's forks of the runtime and grammars. Every grammar crate
+# depends only on `tree-sitter-language`, never on upstream's `tree-sitter` crate, whose
+# `links = "tree-sitter"` would conflict with the forked runtime's.
+tree-sitter = { package = "willbooster-tree-sitter", version = "1" }
+tree-sitter-c = { package = "willbooster-tree-sitter-c", version = "=1.1.0" }
+tree-sitter-c-sharp = { package = "willbooster-tree-sitter-c-sharp", version = "=1.0.3" }
+tree-sitter-cpp = { package = "willbooster-tree-sitter-cpp", version = "=1.1.1" }
+tree-sitter-go = "=0.25.0"
+tree-sitter-java = "=0.23.5"
+tree-sitter-javascript = { package = "willbooster-tree-sitter-javascript", version = "=1.0.3" }
+tree-sitter-kotlin = { package = "willbooster-tree-sitter-kotlin", version = "=1.4.0" }
+tree-sitter-python = "=0.25.0"
+tree-sitter-ruby = "=0.23.1"
+tree-sitter-rust = { package = "willbooster-tree-sitter-rust", version = "=1.1.1" }
+tree-sitter-typescript = { package = "willbooster-tree-sitter-typescript", version = "=1.1.1" }
 
 # The WebAssembly build (src/wasm.rs) exports a plain C ABI instead of N-API.
 [target.'cfg(not(target_family = "wasm"))'.dependencies]

package/native/src/dep_degree.rs

@@ -93,7 +93,7 @@ const NON_BINDING_PARAMETER_FIELDS: &[&str] = &["type", "value"];
 /// One leaf of the def-use walk: its field in the parent and its function-scope chain.
 struct DepDegreeLeaf<'t> {
     node: Node<'t>,
-    field_name: Option<&'static str>,
+    field_name: Option<&'t str>,
     /// Chain of nested-function scope ids below the measured function: "" for its own body,
     /// then "/0", "/0/1", ...
     scope: String,
@@ -221,7 +221,7 @@ fn implicit_accessor_definitions<'a>(function_node: Node<'_>, code: &Source<'a>)
 #[allow(clippy::too_many_arguments)]
 fn collect_dep_degree_leaves<'t>(
     node: Node<'t>,
-    field_name: Option<&'static str>,
+    field_name: Option<&'t str>,
     scope: &str,
     next_scope_id: &mut usize,
     function_nodes: &FxHashSet<&'static str>,
@@ -234,7 +234,7 @@ fn collect_dep_degree_leaves<'t>(
     ) {
         return;
     }
-    if node.child_count() == 0 {
+    if crate::util::is_token(node) {
         leaves.push(DepDegreeLeaf {
             node,
             field_name,
@@ -341,7 +341,7 @@ fn is_definition_field(holder: Node<'_>, field_name: Option<&str>) -> bool {
 /// fieldless child; a member pointer's `pointer_type_declarator` is the `name` of a
 /// `qualified_identifier`; an `array_declarator` size or a nested parameter has another field and
 /// stops the climb) to the outermost wrapper and its field in the declaration.
-fn unwrap_declarator_wrappers<'t>(leaf: &DepDegreeLeaf<'t>) -> (Node<'t>, Option<&'static str>) {
+fn unwrap_declarator_wrappers<'t>(leaf: &DepDegreeLeaf<'t>) -> (Node<'t>, Option<&'t str>) {
     let mut current = leaf.node;
     let mut field_name = leaf.field_name;
     while let Some(parent) = current
@@ -439,11 +439,11 @@ fn is_kotlin_parameter(node: Node<'_>) -> bool {
 }
 
 /// Only called with small-arity parents (declarator wrappers, definition-list holders).
-fn field_name_in_parent(node: Node<'_>, parent: Node<'_>) -> Option<&'static str> {
+fn field_name_in_parent<'t>(node: Node<'_>, parent: Node<'t>) -> Option<&'t str> {
     for index in 0..parent.child_count() {
         if let Some(child) = parent.child(index) {
             if child.id() == node.id() {
-                return parent.field_name_for_child(index as u32);
+                return parent.field_name_for_child(index);
             }
         }
     }

package/native/src/duplication.rs

@@ -10,7 +10,7 @@ use crate::types::{
     CrossFileCandidate, CrossFileToken, CrossFileTokenRange, DuplicateBlockOccurrence,
     DuplicationMetrics,
 };
-use crate::util::{all_children, named_children, node_text, to_int32, Source};
+use crate::util::{all_children, is_token, named_children, node_text, to_int32, Source};
 
 /// Block-like nodes considered as whole-subtree duplicate candidates.
 const DUPLICATE_BLOCK_TYPES: &[&str] = &[
@@ -103,6 +103,7 @@ const STATEMENT_CONTAINER_TYPES: &[&str] = &[
     "default_case",
     "compilation_unit",
     "switch_section",
+    "extension_body",
     "enum_class_body",
     "function_body",
     "lambda_literal",
@@ -229,6 +230,8 @@ const STRING_FRAGMENT_TYPES: &[&str] = &[
     "string_literal_content",
     "character_literal_content",
     "raw_string_content",
+    "interpreted_string_literal_content",
+    "raw_string_literal_content",
     "raw_string_start",
     "raw_string_end",
     "escape_sequence",
@@ -292,6 +295,8 @@ const STRING_CONTENT_FRAGMENT_TYPES: &[&str] = &[
     "string_literal_content",
     "character_literal_content",
     "raw_string_content",
+    "interpreted_string_literal_content",
+    "raw_string_literal_content",
     "escape_sequence",
     "heredoc_content",
 ];
@@ -548,20 +553,25 @@ fn collect_tokens<'a>(
     block_ranges: &mut Vec<TokenRange>,
     container_statement_ranges: &mut Vec<Vec<TokenRange>>,
 ) {
+    /// `statement_list_statements` receives the statements of a Go `statement_list`, which wraps
+    /// the statements of blocks and case clauses and whose statements belong to the run of the
+    /// enclosing container.
     fn visit<'a>(
         node: Node<'_>,
         code: &Source<'a>,
         tokens: &mut Vec<Token<'a>>,
         block_ranges: &mut Vec<TokenRange>,
         container_statement_ranges: &mut Vec<Vec<TokenRange>>,
+        statement_list_statements: &mut Vec<TokenRange>,
     ) -> TokenRange {
         let start_token_index = tokens.len();
-        let atomic_kind = if node.child_count() == 0 {
+        let is_token = is_token(node);
+        let atomic_kind = if is_token {
             None
         } else {
             atomic_literal_kind(node)
         };
-        if node.child_count() == 0 {
+        if is_token {
             append_leaf_token(node, code, tokens);
         } else if let Some(atomic_kind) = atomic_kind {
             // Interpolation-free strings collapse to their kind tag so copies differing only in
@@ -575,14 +585,17 @@ fn collect_tokens<'a>(
             ));
         } else if !is_tokenless(node, code) {
             let mut statement_ranges: Vec<TokenRange> = Vec::new();
-            let is_container = is_statement_container(node);
+            let is_statement_list = node.kind_name() == "statement_list";
+            let is_container = is_statement_list || is_statement_container(node);
             for child in all_children(node) {
+                let mut child_statements = Vec::new();
                 let child_range = visit(
                     child,
                     code,
                     tokens,
                     block_ranges,
                     container_statement_ranges,
+                    &mut child_statements,
                 );
                 // A Kotlin lambda's parameters sit beside its statements.
                 if is_container
@@ -590,13 +603,19 @@ fn collect_tokens<'a>(
                     && !is_tokenless(child, code)
                     && child.kind_name() != "lambda_parameters"
                 {
-                    statement_ranges.push(child_range);
+                    if child.kind_name() == "statement_list" {
+                        statement_ranges.append(&mut child_statements);
+                    } else {
+                        statement_ranges.push(child_range);
+                    }
                 }
             }
-            // Single-statement containers are recorded too: window enumeration needs two
-            // statements and yields nothing for them, but cross-file matching catalogues each
-            // container's full run.
-            if is_container && !statement_ranges.is_empty() {
+            if is_statement_list {
+                *statement_list_statements = statement_ranges;
+            } else if is_container && !statement_ranges.is_empty() {
+                // Single-statement containers are recorded too: window enumeration needs two
+                // statements and yields nothing for them, but cross-file matching catalogues each
+                // container's full run.
                 container_statement_ranges.push(statement_ranges);
             }
         }
@@ -615,7 +634,14 @@ fn collect_tokens<'a>(
         range
     }
 
-    visit(root, code, tokens, block_ranges, container_statement_ranges);
+    visit(
+        root,
+        code,
+        tokens,
+        block_ranges,
+        container_statement_ranges,
+        &mut Vec::new(),
+    );
 }
 
 /// Comments and dependency declarations contribute no tokens, so no clone consists of them; the
@@ -935,7 +961,7 @@ fn is_semantic_name_leaf(node: Node<'_>, code: &Source<'_>) -> bool {
             .named_child(0)
             .is_some_and(|first| first.id() == node.id());
         let is_last_named = parent
-            .named_child(parent.named_child_count().saturating_sub(1))
+            .named_child(parent.named_child_count().saturating_sub(1) as u32)
             .is_some_and(|last| last.id() == node.id());
         if parent.kind_name() == "navigation_expression" && is_last_named && !is_first_named {
             return true;
@@ -943,7 +969,7 @@ fn is_semantic_name_leaf(node: Node<'_>, code: &Source<'_>) -> bool {
         if parent.kind_name() == "call_expression"
             && is_first_named
             && parent
-                .named_child(parent.named_child_count().saturating_sub(1))
+                .named_child(parent.named_child_count().saturating_sub(1) as u32)
                 .is_some_and(|arguments| {
                     matches!(
                         arguments.kind_name(),