code-gauge 1.13.0 → 2.0.0

package/dist/crossFileDuplication.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"crossFileDuplication.cjs","names":["selectMaximalGroups"],"sources":["../src/crossFileDuplication.ts"],"sourcesContent":["import { selectMaximalGroups } from './duplicateSelection.js';\nimport type { CrossFileDuplicateCandidate } from './duplication.js';\n\nexport interface CrossFileDuplicationSourceFile {\n  file: string;\n  candidates: CrossFileDuplicateCandidate[];\n}\n\nexport interface CrossFileDuplicateOccurrence {\n  endLine: number;\n  file: string;\n  startLine: number;\n}\n\nexport interface CrossFileDuplicateBlockGroup {\n  files: string[];\n  occurrences: CrossFileDuplicateOccurrence[];\n  /** Normalized token count of one occurrence (all occurrences share it). */\n  tokenCount: number;\n}\n\nexport interface CrossFileDuplicationMetrics {\n  /** Number of redundant copies across all groups, i.e. sum of (occurrenceCount - 1). */\n  duplicateBlockCount: number;\n  /** Groups the file participates in, keyed by the file name passed in. */\n  duplicateBlockGroupCountByFile: Record<string, number>;\n  groups: CrossFileDuplicateBlockGroup[];\n}\n\ninterface SelectableCandidate extends CrossFileDuplicateCandidate {\n  regionBucket: number;\n  file: string;\n}\n\n/**\n * Detects code regions duplicated across files: per-file candidates (whole block subtrees and\n * full container runs, fingerprinted with the same normalization as within-file duplication) are\n * grouped by fingerprint, and only maximal, non-overlapping regions whose group spans at least two\n * files are counted. Groups that shrink to a single file during selection are shed — a\n * within-file repeat is already reported by that file's own duplication metrics.\n */\nexport function measureCrossFileDuplication(files: CrossFileDuplicationSourceFile[]): CrossFileDuplicationMetrics {\n  const candidates: SelectableCandidate[] = files.flatMap(({ file, candidates }, fileIndex) =>\n    candidates.map((candidate) => ({ ...candidate, regionBucket: fileIndex, file }))\n  );\n  const counted = selectMaximalGroups(\n    candidates,\n    spansMultipleFiles,\n    // File index and position break coverage ties deterministically.\n    (left, right) => left.regionBucket - right.regionBucket || left.startIndex - right.startIndex\n  );\n  return summarize(counted);\n}\n\nfunction spansMultipleFiles(group: SelectableCandidate[]): boolean {\n  return group.length >= 2 && new Set(group.map((candidate) => candidate.regionBucket)).size >= 2;\n}\n\nfunction summarize(counted: Map<string, SelectableCandidate[]>): CrossFileDuplicationMetrics {\n  const groups: CrossFileDuplicateBlockGroup[] = [];\n  // Accumulated in a Map: file names are arbitrary strings, and a plain object would read\n  // inherited properties for names like \"constructor\".\n  const groupCountByFile = new Map<string, number>();\n  let duplicateBlockCount = 0;\n  for (const group of counted.values()) {\n    duplicateBlockCount += group.length - 1;\n    const occurrences = group\n      .map(({ file, startLine, endLine }) => ({ file, startLine, endLine }))\n      .toSorted((left, right) => left.file.localeCompare(right.file) || left.startLine - right.startLine);\n    const files = [...new Set(occurrences.map(({ file }) => file))];\n    for (const file of files) {\n      groupCountByFile.set(file, (groupCountByFile.get(file) ?? 0) + 1);\n    }\n    groups.push({ files, occurrences, tokenCount: group[0]?.tokenCount ?? 0 });\n  }\n  groups.sort(\n    (left, right) =>\n      right.tokenCount - left.tokenCount ||\n      (left.occurrences[0]?.file ?? '').localeCompare(right.occurrences[0]?.file ?? '') ||\n      (left.occurrences[0]?.startLine ?? 0) - (right.occurrences[0]?.startLine ?? 0)\n  );\n  return { duplicateBlockCount, duplicateBlockGroupCountByFile: Object.fromEntries(groupCountByFile), groups };\n}\n"],"mappings":"yDAyCA,SAAgB,EAA4B,EAAsE,CAChH,IAAM,EAAoC,EAAM,SAAS,CAAE,OAAM,cAAc,IAC7E,EAAW,IAAK,IAAe,CAAE,GAAG,EAAW,aAAc,EAAW,MAAK,EAAE,CACjF,EAOA,OAAO,EANSA,EAAAA,oBACd,EACA,GAEC,EAAM,IAAU,EAAK,aAAe,EAAM,cAAgB,EAAK,WAAa,EAAM,UAE9D,CAAC,CAC1B,CAEA,SAAS,EAAmB,EAAuC,CACjE,OAAO,EAAM,QAAU,GAAK,IAAI,IAAI,EAAM,IAAK,GAAc,EAAU,YAAY,CAAC,CAAC,CAAC,MAAQ,CAChG,CAEA,SAAS,EAAU,EAA0E,CAC3F,IAAM,EAAyC,CAAC,EAG1C,EAAmB,IAAI,IACzB,EAAsB,EAC1B,IAAK,IAAM,KAAS,EAAQ,OAAO,EAAG,CACpC,GAAuB,EAAM,OAAS,EACtC,IAAM,EAAc,EACjB,KAAK,CAAE,OAAM,YAAW,cAAe,CAAE,OAAM,YAAW,SAAQ,EAAE,CAAC,CACrE,UAAU,EAAM,IAAU,EAAK,KAAK,cAAc,EAAM,IAAI,GAAK,EAAK,UAAY,EAAM,SAAS,EAC9F,EAAQ,CAAC,GAAG,IAAI,IAAI,EAAY,KAAK,CAAE,UAAW,CAAI,CAAC,CAAC,EAC9D,IAAK,IAAM,KAAQ,EACjB,EAAiB,IAAI,GAAO,EAAiB,IAAI,CAAI,GAAK,GAAK,CAAC,EAElE,EAAO,KAAK,CAAE,QAAO,cAAa,WAAY,EAAM,EAAE,EAAE,YAAc,CAAE,CAAC,CAC3E,CAOA,OANA,EAAO,MACJ,EAAM,IACL,EAAM,WAAa,EAAK,aACvB,EAAK,YAAY,EAAE,EAAE,MAAQ,GAAA,CAAI,cAAc,EAAM,YAAY,EAAE,EAAE,MAAQ,EAAE,IAC/E,EAAK,YAAY,EAAE,EAAE,WAAa,IAAM,EAAM,YAAY,EAAE,EAAE,WAAa,EAChF,EACO,CAAE,sBAAqB,+BAAgC,OAAO,YAAY,CAAgB,EAAG,QAAO,CAC7G"}
+{"version":3,"file":"crossFileDuplication.cjs","names":["defaultDuplicationOptions","selectMaximalGroups","buildLiteralCountPrefix","collectSequenceWindowCandidates","mergeAdjacentGroups","countRedundantFragments"],"sources":["../src/crossFileDuplication.ts"],"sourcesContent":["import { selectMaximalGroups } from './duplicateSelection.js';\nimport {\n  buildLiteralCountPrefix,\n  collectSequenceWindowCandidates,\n  countRedundantFragments,\n  defaultDuplicationOptions,\n  mergeAdjacentGroups,\n  type CountedOccurrence,\n  type CrossFileDuplicateCandidate,\n  type CrossFileDuplicationFileData,\n  type SequenceWindowContext,\n} from './duplication.js';\nimport type { DuplicationOptions } from './types.js';\n\nexport interface CrossFileDuplicationSourceFile extends Partial<CrossFileDuplicationFileData> {\n  file: string;\n  candidates: CrossFileDuplicateCandidate[];\n}\n\nexport interface CrossFileDuplicateOccurrence {\n  endLine: number;\n  file: string;\n  startLine: number;\n}\n\nexport interface CrossFileDuplicateBlockGroup {\n  files: string[];\n  occurrences: CrossFileDuplicateOccurrence[];\n  /** Matched token count of one occurrence (all occurrences share it; gaps are not counted). */\n  tokenCount: number;\n}\n\nexport interface CrossFileDuplicationMetrics {\n  /** Number of redundant copies across all groups, counted per matched fragment like within-file. */\n  duplicateBlockCount: number;\n  /** Groups the file participates in, keyed by the file name passed in. */\n  duplicateBlockGroupCountByFile: Record<string, number>;\n  groups: CrossFileDuplicateBlockGroup[];\n}\n\ninterface SelectableCandidate extends CrossFileDuplicateCandidate {\n  regionBucket: number;\n  file: string;\n}\n\n/** A cross-file occurrence: a within-file occurrence in the project-wide token index space. */\ninterface CrossFileOccurrence extends CountedOccurrence {\n  file: string;\n}\n\n/**\n * Detects code regions duplicated across files. Per-file candidates (whole block subtrees and full\n * container runs, fingerprinted with the same normalization as within-file duplication) are joined\n * by a project-level window index over per-statement fingerprint sequences (CPD-style), so a\n * copy-pasted partial statement run embedded in different surrounding code is matched even though\n * no single file can know it repeats elsewhere. Candidates are grouped by fingerprint, and only\n * maximal, non-overlapping regions whose group spans at least two files are counted. Groups that\n * shrink to a single file during selection are shed — a within-file repeat is already reported by\n * that file's own duplication metrics. Groups separated by a small token gap within each file then\n * merge into gapped (Type-3) clone groups under `maxGapTokens`, exactly like within-file merging.\n */\nexport function measureCrossFileDuplication(\n  files: CrossFileDuplicationSourceFile[],\n  options?: DuplicationOptions\n): CrossFileDuplicationMetrics {\n  const minTokens = options?.minTokens ?? defaultDuplicationOptions.minTokens;\n  const maxGapTokens = options?.maxGapTokens ?? defaultDuplicationOptions.maxGapTokens;\n  const candidates: SelectableCandidate[] = files.flatMap(({ file, candidates }, fileIndex) =>\n    candidates.map((candidate) => ({ ...candidate, regionBucket: fileIndex, file }))\n  );\n  // Pushed one by one: spreading the project-scale window-candidate array as call arguments\n  // overflows V8's argument limit (~124k) and crashes on Node, though Bun/JSC tolerates it.\n  for (const candidate of collectWindowCandidates(files, minTokens)) {\n    candidates.push(candidate);\n  }\n  const counted = selectMaximalGroups(\n    candidates,\n    spansMultipleFiles,\n    // File index and position break coverage ties deterministically.\n    (left, right) => left.regionBucket - right.regionBucket || left.startIndex - right.startIndex\n  );\n  return summarize(mergeGapAdjacentGroups([...counted.values()], files, maxGapTokens));\n}\n\n/** Repeated sub-windows of sibling statements matched across the whole project's files. */\nfunction collectWindowCandidates(files: CrossFileDuplicationSourceFile[], minTokens: number): SelectableCandidate[] {\n  const fileIndexByContext: number[] = [];\n  const contexts: SequenceWindowContext[] = [];\n  for (const [fileIndex, { tokens, containerStatements }] of files.entries()) {\n    if (tokens && containerStatements) {\n      fileIndexByContext.push(fileIndex);\n      contexts.push({ tokens, literalCountPrefix: buildLiteralCountPrefix(tokens), containers: containerStatements });\n    }\n  }\n  if (contexts.length < 2) {\n    return [];\n  }\n  return collectSequenceWindowCandidates(contexts, minTokens, true).flatMap(({ candidate, contextIndex }) => {\n    const fileIndex = fileIndexByContext[contextIndex];\n    const file = fileIndex === undefined ? undefined : files[fileIndex];\n    return fileIndex === undefined || file === undefined\n      ? []\n      : [{ ...candidate, regionBucket: fileIndex, file: file.file }];\n  });\n}\n\nfunction spansMultipleFiles(group: SelectableCandidate[]): boolean {\n  return group.length >= 2 && new Set(group.map((candidate) => candidate.regionBucket)).size >= 2;\n}\n\n/**\n * Reuses the within-file gapped (Type-3) merging by mapping every occurrence into one project-wide\n * token index space: each file's tokens are offset by more than `maxGapTokens` past the previous\n * file's, so occurrences in different files are never gap-adjacent and pairs always stay within\n * one file.\n */\nfunction mergeGapAdjacentGroups(\n  groups: SelectableCandidate[][],\n  files: CrossFileDuplicationSourceFile[],\n  maxGapTokens: number\n): CrossFileOccurrence[][] {\n  const tokenOffsets: number[] = [];\n  let offset = 0;\n  for (const { tokens, candidates } of files) {\n    tokenOffsets.push(offset);\n    // Accumulated in a loop: spreading a project-scale candidate array as call arguments would\n    // overflow V8's argument limit (~124k) and crash on Node.\n    let tokenCount = tokens?.length ?? 0;\n    if (!tokens) {\n      for (const candidate of candidates) {\n        tokenCount = Math.max(tokenCount, candidate.endTokenIndex);\n      }\n    }\n    offset += tokenCount + maxGapTokens + 1;\n  }\n  const occurrenceGroups = groups.map((group) =>\n    group\n      .map((candidate): CrossFileOccurrence => {\n        const start = candidate.startTokenIndex + (tokenOffsets[candidate.regionBucket] ?? 0);\n        const end = candidate.endTokenIndex + (tokenOffsets[candidate.regionBucket] ?? 0);\n        return {\n          file: candidate.file,\n          segments: [{ startTokenIndex: start, endTokenIndex: end }],\n          tokenCount: candidate.tokenCount,\n          startTokenIndex: start,\n          endTokenIndex: end,\n          startIndex: candidate.startIndex,\n          endIndex: candidate.endIndex,\n          startLine: candidate.startLine,\n          endLine: candidate.endLine,\n        };\n      })\n      .toSorted((left, right) => left.startTokenIndex - right.startTokenIndex)\n  );\n  return mergeAdjacentGroups(occurrenceGroups, maxGapTokens);\n}\n\nfunction summarize(groups: CrossFileOccurrence[][]): CrossFileDuplicationMetrics {\n  const reported: CrossFileDuplicateBlockGroup[] = [];\n  // Accumulated in a Map: file names are arbitrary strings, and a plain object would read\n  // inherited properties for names like \"constructor\".\n  const groupCountByFile = new Map<string, number>();\n  let duplicateBlockCount = 0;\n  for (const group of groups) {\n    // Mirrors within-file counting: each redundant occurrence contributes one count per matched\n    // fragment, gapped merging consolidates the grouping without halving the count, and spans a\n    // partial merge shares between a retained group and the merged group count once.\n    duplicateBlockCount += countRedundantFragments(group);\n    const occurrences = group\n      .map(({ file, startLine, endLine }) => ({ file, startLine, endLine }))\n      .toSorted((left, right) => left.file.localeCompare(right.file) || left.startLine - right.startLine);\n    const files = [...new Set(occurrences.map(({ file }) => file))];\n    for (const file of files) {\n      groupCountByFile.set(file, (groupCountByFile.get(file) ?? 0) + 1);\n    }\n    reported.push({ files, occurrences, tokenCount: group[0]?.tokenCount ?? 0 });\n  }\n  reported.sort(\n    (left, right) =>\n      right.tokenCount - left.tokenCount ||\n      (left.occurrences[0]?.file ?? '').localeCompare(right.occurrences[0]?.file ?? '') ||\n      (left.occurrences[0]?.startLine ?? 0) - (right.occurrences[0]?.startLine ?? 0)\n  );\n  return {\n    duplicateBlockCount,\n    duplicateBlockGroupCountByFile: Object.fromEntries(groupCountByFile),\n    groups: reported,\n  };\n}\n"],"mappings":"wFA6DA,SAAgB,EACd,EACA,EAC6B,CAC7B,IAAM,EAAY,GAAS,WAAaA,EAAAA,0BAA0B,UAC5D,EAAe,GAAS,cAAgBA,EAAAA,0BAA0B,aAClE,EAAoC,EAAM,SAAS,CAAE,OAAM,cAAc,IAC7E,EAAW,IAAK,IAAe,CAAE,GAAG,EAAW,aAAc,EAAW,MAAK,EAAE,CACjF,EAGA,IAAK,IAAM,KAAa,EAAwB,EAAO,CAAS,EAC9D,EAAW,KAAK,CAAS,EAQ3B,OAAO,EAAU,EAAuB,CAAC,GANzBC,EAAAA,oBACd,EACA,GAEC,EAAM,IAAU,EAAK,aAAe,EAAM,cAAgB,EAAK,WAAa,EAAM,UAEnC,CAAC,CAAC,OAAO,CAAC,EAAG,EAAO,CAAY,CAAC,CACrF,CAGA,SAAS,EAAwB,EAAyC,EAA0C,CAClH,IAAM,EAA+B,CAAC,EAChC,EAAoC,CAAC,EAC3C,IAAK,GAAM,CAAC,EAAW,CAAE,SAAQ,0BAA0B,EAAM,QAAQ,EACnE,GAAU,IACZ,EAAmB,KAAK,CAAS,EACjC,EAAS,KAAK,CAAE,SAAQ,mBAAoBC,EAAAA,wBAAwB,CAAM,EAAG,WAAY,CAAoB,CAAC,GAMlH,OAHI,EAAS,OAAS,EACb,CAAC,EAEHC,EAAAA,gCAAgC,EAAU,EAAW,EAAI,CAAC,CAAC,SAAS,CAAE,YAAW,kBAAmB,CACzG,IAAM,EAAY,EAAmB,GAC/B,EAAO,IAAc,IAAA,GAAY,IAAA,GAAY,EAAM,GACzD,OAAO,IAAc,IAAA,IAAa,IAAS,IAAA,GACvC,CAAC,EACD,CAAC,CAAE,GAAG,EAAW,aAAc,EAAW,KAAM,EAAK,IAAK,CAAC,CACjE,CAAC,CACH,CAEA,SAAS,EAAmB,EAAuC,CACjE,OAAO,EAAM,QAAU,GAAK,IAAI,IAAI,EAAM,IAAK,GAAc,EAAU,YAAY,CAAC,CAAC,CAAC,MAAQ,CAChG,CAQA,SAAS,EACP,EACA,EACA,EACyB,CACzB,IAAM,EAAyB,CAAC,EAC5B,EAAS,EACb,IAAK,GAAM,CAAE,SAAQ,gBAAgB,EAAO,CAC1C,EAAa,KAAK,CAAM,EAGxB,IAAI,EAAa,GAAQ,QAAU,EACnC,GAAI,CAAC,EACH,IAAK,IAAM,KAAa,EACtB,EAAa,KAAK,IAAI,EAAY,EAAU,aAAa,EAG7D,GAAU,EAAa,EAAe,CACxC,CACA,IAAM,EAAmB,EAAO,IAAK,GACnC,EACG,IAAK,GAAmC,CACvC,IAAM,EAAQ,EAAU,iBAAmB,EAAa,EAAU,eAAiB,GAC7E,EAAM,EAAU,eAAiB,EAAa,EAAU,eAAiB,GAC/E,MAAO,CACL,KAAM,EAAU,KAChB,SAAU,CAAC,CAAE,gBAAiB,EAAO,cAAe,CAAI,CAAC,EACzD,WAAY,EAAU,WACtB,gBAAiB,EACjB,cAAe,EACf,WAAY,EAAU,WACtB,SAAU,EAAU,SACpB,UAAW,EAAU,UACrB,QAAS,EAAU,OACrB,CACF,CAAC,CAAC,CACD,UAAU,EAAM,IAAU,EAAK,gBAAkB,EAAM,eAAe,CAC3E,EACA,OAAOC,EAAAA,oBAAoB,EAAkB,CAAY,CAC3D,CAEA,SAAS,EAAU,EAA8D,CAC/E,IAAM,EAA2C,CAAC,EAG5C,EAAmB,IAAI,IACzB,EAAsB,EAC1B,IAAK,IAAM,KAAS,EAAQ,CAI1B,GAAuBC,EAAAA,wBAAwB,CAAK,EACpD,IAAM,EAAc,EACjB,KAAK,CAAE,OAAM,YAAW,cAAe,CAAE,OAAM,YAAW,SAAQ,EAAE,CAAC,CACrE,UAAU,EAAM,IAAU,EAAK,KAAK,cAAc,EAAM,IAAI,GAAK,EAAK,UAAY,EAAM,SAAS,EAC9F,EAAQ,CAAC,GAAG,IAAI,IAAI,EAAY,KAAK,CAAE,UAAW,CAAI,CAAC,CAAC,EAC9D,IAAK,IAAM,KAAQ,EACjB,EAAiB,IAAI,GAAO,EAAiB,IAAI,CAAI,GAAK,GAAK,CAAC,EAElE,EAAS,KAAK,CAAE,QAAO,cAAa,WAAY,EAAM,EAAE,EAAE,YAAc,CAAE,CAAC,CAC7E,CAOA,OANA,EAAS,MACN,EAAM,IACL,EAAM,WAAa,EAAK,aACvB,EAAK,YAAY,EAAE,EAAE,MAAQ,GAAA,CAAI,cAAc,EAAM,YAAY,EAAE,EAAE,MAAQ,EAAE,IAC/E,EAAK,YAAY,EAAE,EAAE,WAAa,IAAM,EAAM,YAAY,EAAE,EAAE,WAAa,EAChF,EACO,CACL,sBACA,+BAAgC,OAAO,YAAY,CAAgB,EACnE,OAAQ,CACV,CACF"}

package/dist/crossFileDuplication.d.ts

@@ -1,5 +1,6 @@
-import type { CrossFileDuplicateCandidate } from './duplication.js';
-export interface CrossFileDuplicationSourceFile {
+import { type CrossFileDuplicateCandidate, type CrossFileDuplicationFileData } from './duplication.js';
+import type { DuplicationOptions } from './types.js';
+export interface CrossFileDuplicationSourceFile extends Partial<CrossFileDuplicationFileData> {
     file: string;
     candidates: CrossFileDuplicateCandidate[];
 }
@@ -11,21 +12,25 @@ export interface CrossFileDuplicateOccurrence {
 export interface CrossFileDuplicateBlockGroup {
     files: string[];
     occurrences: CrossFileDuplicateOccurrence[];
-    /** Normalized token count of one occurrence (all occurrences share it). */
+    /** Matched token count of one occurrence (all occurrences share it; gaps are not counted). */
     tokenCount: number;
 }
 export interface CrossFileDuplicationMetrics {
-    /** Number of redundant copies across all groups, i.e. sum of (occurrenceCount - 1). */
+    /** Number of redundant copies across all groups, counted per matched fragment like within-file. */
     duplicateBlockCount: number;
     /** Groups the file participates in, keyed by the file name passed in. */
     duplicateBlockGroupCountByFile: Record<string, number>;
     groups: CrossFileDuplicateBlockGroup[];
 }
 /**
- * Detects code regions duplicated across files: per-file candidates (whole block subtrees and
- * full container runs, fingerprinted with the same normalization as within-file duplication) are
- * grouped by fingerprint, and only maximal, non-overlapping regions whose group spans at least two
- * files are counted. Groups that shrink to a single file during selection are shed — a
- * within-file repeat is already reported by that file's own duplication metrics.
+ * Detects code regions duplicated across files. Per-file candidates (whole block subtrees and full
+ * container runs, fingerprinted with the same normalization as within-file duplication) are joined
+ * by a project-level window index over per-statement fingerprint sequences (CPD-style), so a
+ * copy-pasted partial statement run embedded in different surrounding code is matched even though
+ * no single file can know it repeats elsewhere. Candidates are grouped by fingerprint, and only
+ * maximal, non-overlapping regions whose group spans at least two files are counted. Groups that
+ * shrink to a single file during selection are shed — a within-file repeat is already reported by
+ * that file's own duplication metrics. Groups separated by a small token gap within each file then
+ * merge into gapped (Type-3) clone groups under `maxGapTokens`, exactly like within-file merging.
  */
-export declare function measureCrossFileDuplication(files: CrossFileDuplicationSourceFile[]): CrossFileDuplicationMetrics;
+export declare function measureCrossFileDuplication(files: CrossFileDuplicationSourceFile[], options?: DuplicationOptions): CrossFileDuplicationMetrics;

package/dist/crossFileDuplication.js

@@ -1,2 +1,2 @@
-import{selectMaximalGroups as e}from"./duplicateSelection.js";function t(t){let i=t.flatMap(({file:e,candidates:t},n)=>t.map(t=>({...t,regionBucket:n,file:e})));return r(e(i,n,(e,t)=>e.regionBucket-t.regionBucket||e.startIndex-t.startIndex))}function n(e){return e.length>=2&&new Set(e.map(e=>e.regionBucket)).size>=2}function r(e){let t=[],n=new Map,r=0;for(let i of e.values()){r+=i.length-1;let e=i.map(({file:e,startLine:t,endLine:n})=>({file:e,startLine:t,endLine:n})).toSorted((e,t)=>e.file.localeCompare(t.file)||e.startLine-t.startLine),a=[...new Set(e.map(({file:e})=>e))];for(let e of a)n.set(e,(n.get(e)??0)+1);t.push({files:a,occurrences:e,tokenCount:i[0]?.tokenCount??0})}return t.sort((e,t)=>t.tokenCount-e.tokenCount||(e.occurrences[0]?.file??``).localeCompare(t.occurrences[0]?.file??``)||(e.occurrences[0]?.startLine??0)-(t.occurrences[0]?.startLine??0)),{duplicateBlockCount:r,duplicateBlockGroupCountByFile:Object.fromEntries(n),groups:t}}export{t as measureCrossFileDuplication};
+import{selectMaximalGroups as e}from"./duplicateSelection.js";import{buildLiteralCountPrefix as t,collectSequenceWindowCandidates as n,countRedundantFragments as r,defaultDuplicationOptions as i,mergeAdjacentGroups as a}from"./duplication.js";function o(t,n){let r=n?.minTokens??i.minTokens,a=n?.maxGapTokens??i.maxGapTokens,o=t.flatMap(({file:e,candidates:t},n)=>t.map(t=>({...t,regionBucket:n,file:e})));for(let e of s(t,r))o.push(e);return u(l([...e(o,c,(e,t)=>e.regionBucket-t.regionBucket||e.startIndex-t.startIndex).values()],t,a))}function s(e,r){let i=[],a=[];for(let[n,{tokens:r,containerStatements:o}]of e.entries())r&&o&&(i.push(n),a.push({tokens:r,literalCountPrefix:t(r),containers:o}));return a.length<2?[]:n(a,r,!0).flatMap(({candidate:t,contextIndex:n})=>{let r=i[n],a=r===void 0?void 0:e[r];return r===void 0||a===void 0?[]:[{...t,regionBucket:r,file:a.file}]})}function c(e){return e.length>=2&&new Set(e.map(e=>e.regionBucket)).size>=2}function l(e,t,n){let r=[],i=0;for(let{tokens:e,candidates:a}of t){r.push(i);let t=e?.length??0;if(!e)for(let e of a)t=Math.max(t,e.endTokenIndex);i+=t+n+1}let o=e.map(e=>e.map(e=>{let t=e.startTokenIndex+(r[e.regionBucket]??0),n=e.endTokenIndex+(r[e.regionBucket]??0);return{file:e.file,segments:[{startTokenIndex:t,endTokenIndex:n}],tokenCount:e.tokenCount,startTokenIndex:t,endTokenIndex:n,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startLine,endLine:e.endLine}}).toSorted((e,t)=>e.startTokenIndex-t.startTokenIndex));return a(o,n)}function u(e){let t=[],n=new Map,i=0;for(let a of e){i+=r(a);let e=a.map(({file:e,startLine:t,endLine:n})=>({file:e,startLine:t,endLine:n})).toSorted((e,t)=>e.file.localeCompare(t.file)||e.startLine-t.startLine),o=[...new Set(e.map(({file:e})=>e))];for(let e of o)n.set(e,(n.get(e)??0)+1);t.push({files:o,occurrences:e,tokenCount:a[0]?.tokenCount??0})}return t.sort((e,t)=>t.tokenCount-e.tokenCount||(e.occurrences[0]?.file??``).localeCompare(t.occurrences[0]?.file??``)||(e.occurrences[0]?.startLine??0)-(t.occurrences[0]?.startLine??0)),{duplicateBlockCount:i,duplicateBlockGroupCountByFile:Object.fromEntries(n),groups:t}}export{o as measureCrossFileDuplication};
 //# sourceMappingURL=crossFileDuplication.js.map

package/dist/crossFileDuplication.js.map

@@ -1 +1 @@
-{"version":3,"file":"crossFileDuplication.js","names":[],"sources":["../src/crossFileDuplication.ts"],"sourcesContent":["import { selectMaximalGroups } from './duplicateSelection.js';\nimport type { CrossFileDuplicateCandidate } from './duplication.js';\n\nexport interface CrossFileDuplicationSourceFile {\n  file: string;\n  candidates: CrossFileDuplicateCandidate[];\n}\n\nexport interface CrossFileDuplicateOccurrence {\n  endLine: number;\n  file: string;\n  startLine: number;\n}\n\nexport interface CrossFileDuplicateBlockGroup {\n  files: string[];\n  occurrences: CrossFileDuplicateOccurrence[];\n  /** Normalized token count of one occurrence (all occurrences share it). */\n  tokenCount: number;\n}\n\nexport interface CrossFileDuplicationMetrics {\n  /** Number of redundant copies across all groups, i.e. sum of (occurrenceCount - 1). */\n  duplicateBlockCount: number;\n  /** Groups the file participates in, keyed by the file name passed in. */\n  duplicateBlockGroupCountByFile: Record<string, number>;\n  groups: CrossFileDuplicateBlockGroup[];\n}\n\ninterface SelectableCandidate extends CrossFileDuplicateCandidate {\n  regionBucket: number;\n  file: string;\n}\n\n/**\n * Detects code regions duplicated across files: per-file candidates (whole block subtrees and\n * full container runs, fingerprinted with the same normalization as within-file duplication) are\n * grouped by fingerprint, and only maximal, non-overlapping regions whose group spans at least two\n * files are counted. Groups that shrink to a single file during selection are shed — a\n * within-file repeat is already reported by that file's own duplication metrics.\n */\nexport function measureCrossFileDuplication(files: CrossFileDuplicationSourceFile[]): CrossFileDuplicationMetrics {\n  const candidates: SelectableCandidate[] = files.flatMap(({ file, candidates }, fileIndex) =>\n    candidates.map((candidate) => ({ ...candidate, regionBucket: fileIndex, file }))\n  );\n  const counted = selectMaximalGroups(\n    candidates,\n    spansMultipleFiles,\n    // File index and position break coverage ties deterministically.\n    (left, right) => left.regionBucket - right.regionBucket || left.startIndex - right.startIndex\n  );\n  return summarize(counted);\n}\n\nfunction spansMultipleFiles(group: SelectableCandidate[]): boolean {\n  return group.length >= 2 && new Set(group.map((candidate) => candidate.regionBucket)).size >= 2;\n}\n\nfunction summarize(counted: Map<string, SelectableCandidate[]>): CrossFileDuplicationMetrics {\n  const groups: CrossFileDuplicateBlockGroup[] = [];\n  // Accumulated in a Map: file names are arbitrary strings, and a plain object would read\n  // inherited properties for names like \"constructor\".\n  const groupCountByFile = new Map<string, number>();\n  let duplicateBlockCount = 0;\n  for (const group of counted.values()) {\n    duplicateBlockCount += group.length - 1;\n    const occurrences = group\n      .map(({ file, startLine, endLine }) => ({ file, startLine, endLine }))\n      .toSorted((left, right) => left.file.localeCompare(right.file) || left.startLine - right.startLine);\n    const files = [...new Set(occurrences.map(({ file }) => file))];\n    for (const file of files) {\n      groupCountByFile.set(file, (groupCountByFile.get(file) ?? 0) + 1);\n    }\n    groups.push({ files, occurrences, tokenCount: group[0]?.tokenCount ?? 0 });\n  }\n  groups.sort(\n    (left, right) =>\n      right.tokenCount - left.tokenCount ||\n      (left.occurrences[0]?.file ?? '').localeCompare(right.occurrences[0]?.file ?? '') ||\n      (left.occurrences[0]?.startLine ?? 0) - (right.occurrences[0]?.startLine ?? 0)\n  );\n  return { duplicateBlockCount, duplicateBlockGroupCountByFile: Object.fromEntries(groupCountByFile), groups };\n}\n"],"mappings":"8DAyCA,SAAgB,EAA4B,EAAsE,CAChH,IAAM,EAAoC,EAAM,SAAS,CAAE,OAAM,cAAc,IAC7E,EAAW,IAAK,IAAe,CAAE,GAAG,EAAW,aAAc,EAAW,MAAK,EAAE,CACjF,EAOA,OAAO,EANS,EACd,EACA,GAEC,EAAM,IAAU,EAAK,aAAe,EAAM,cAAgB,EAAK,WAAa,EAAM,UAE9D,CAAC,CAC1B,CAEA,SAAS,EAAmB,EAAuC,CACjE,OAAO,EAAM,QAAU,GAAK,IAAI,IAAI,EAAM,IAAK,GAAc,EAAU,YAAY,CAAC,CAAC,CAAC,MAAQ,CAChG,CAEA,SAAS,EAAU,EAA0E,CAC3F,IAAM,EAAyC,CAAC,EAG1C,EAAmB,IAAI,IACzB,EAAsB,EAC1B,IAAK,IAAM,KAAS,EAAQ,OAAO,EAAG,CACpC,GAAuB,EAAM,OAAS,EACtC,IAAM,EAAc,EACjB,KAAK,CAAE,OAAM,YAAW,cAAe,CAAE,OAAM,YAAW,SAAQ,EAAE,CAAC,CACrE,UAAU,EAAM,IAAU,EAAK,KAAK,cAAc,EAAM,IAAI,GAAK,EAAK,UAAY,EAAM,SAAS,EAC9F,EAAQ,CAAC,GAAG,IAAI,IAAI,EAAY,KAAK,CAAE,UAAW,CAAI,CAAC,CAAC,EAC9D,IAAK,IAAM,KAAQ,EACjB,EAAiB,IAAI,GAAO,EAAiB,IAAI,CAAI,GAAK,GAAK,CAAC,EAElE,EAAO,KAAK,CAAE,QAAO,cAAa,WAAY,EAAM,EAAE,EAAE,YAAc,CAAE,CAAC,CAC3E,CAOA,OANA,EAAO,MACJ,EAAM,IACL,EAAM,WAAa,EAAK,aACvB,EAAK,YAAY,EAAE,EAAE,MAAQ,GAAA,CAAI,cAAc,EAAM,YAAY,EAAE,EAAE,MAAQ,EAAE,IAC/E,EAAK,YAAY,EAAE,EAAE,WAAa,IAAM,EAAM,YAAY,EAAE,EAAE,WAAa,EAChF,EACO,CAAE,sBAAqB,+BAAgC,OAAO,YAAY,CAAgB,EAAG,QAAO,CAC7G"}
+{"version":3,"file":"crossFileDuplication.js","names":[],"sources":["../src/crossFileDuplication.ts"],"sourcesContent":["import { selectMaximalGroups } from './duplicateSelection.js';\nimport {\n  buildLiteralCountPrefix,\n  collectSequenceWindowCandidates,\n  countRedundantFragments,\n  defaultDuplicationOptions,\n  mergeAdjacentGroups,\n  type CountedOccurrence,\n  type CrossFileDuplicateCandidate,\n  type CrossFileDuplicationFileData,\n  type SequenceWindowContext,\n} from './duplication.js';\nimport type { DuplicationOptions } from './types.js';\n\nexport interface CrossFileDuplicationSourceFile extends Partial<CrossFileDuplicationFileData> {\n  file: string;\n  candidates: CrossFileDuplicateCandidate[];\n}\n\nexport interface CrossFileDuplicateOccurrence {\n  endLine: number;\n  file: string;\n  startLine: number;\n}\n\nexport interface CrossFileDuplicateBlockGroup {\n  files: string[];\n  occurrences: CrossFileDuplicateOccurrence[];\n  /** Matched token count of one occurrence (all occurrences share it; gaps are not counted). */\n  tokenCount: number;\n}\n\nexport interface CrossFileDuplicationMetrics {\n  /** Number of redundant copies across all groups, counted per matched fragment like within-file. */\n  duplicateBlockCount: number;\n  /** Groups the file participates in, keyed by the file name passed in. */\n  duplicateBlockGroupCountByFile: Record<string, number>;\n  groups: CrossFileDuplicateBlockGroup[];\n}\n\ninterface SelectableCandidate extends CrossFileDuplicateCandidate {\n  regionBucket: number;\n  file: string;\n}\n\n/** A cross-file occurrence: a within-file occurrence in the project-wide token index space. */\ninterface CrossFileOccurrence extends CountedOccurrence {\n  file: string;\n}\n\n/**\n * Detects code regions duplicated across files. Per-file candidates (whole block subtrees and full\n * container runs, fingerprinted with the same normalization as within-file duplication) are joined\n * by a project-level window index over per-statement fingerprint sequences (CPD-style), so a\n * copy-pasted partial statement run embedded in different surrounding code is matched even though\n * no single file can know it repeats elsewhere. Candidates are grouped by fingerprint, and only\n * maximal, non-overlapping regions whose group spans at least two files are counted. Groups that\n * shrink to a single file during selection are shed — a within-file repeat is already reported by\n * that file's own duplication metrics. Groups separated by a small token gap within each file then\n * merge into gapped (Type-3) clone groups under `maxGapTokens`, exactly like within-file merging.\n */\nexport function measureCrossFileDuplication(\n  files: CrossFileDuplicationSourceFile[],\n  options?: DuplicationOptions\n): CrossFileDuplicationMetrics {\n  const minTokens = options?.minTokens ?? defaultDuplicationOptions.minTokens;\n  const maxGapTokens = options?.maxGapTokens ?? defaultDuplicationOptions.maxGapTokens;\n  const candidates: SelectableCandidate[] = files.flatMap(({ file, candidates }, fileIndex) =>\n    candidates.map((candidate) => ({ ...candidate, regionBucket: fileIndex, file }))\n  );\n  // Pushed one by one: spreading the project-scale window-candidate array as call arguments\n  // overflows V8's argument limit (~124k) and crashes on Node, though Bun/JSC tolerates it.\n  for (const candidate of collectWindowCandidates(files, minTokens)) {\n    candidates.push(candidate);\n  }\n  const counted = selectMaximalGroups(\n    candidates,\n    spansMultipleFiles,\n    // File index and position break coverage ties deterministically.\n    (left, right) => left.regionBucket - right.regionBucket || left.startIndex - right.startIndex\n  );\n  return summarize(mergeGapAdjacentGroups([...counted.values()], files, maxGapTokens));\n}\n\n/** Repeated sub-windows of sibling statements matched across the whole project's files. */\nfunction collectWindowCandidates(files: CrossFileDuplicationSourceFile[], minTokens: number): SelectableCandidate[] {\n  const fileIndexByContext: number[] = [];\n  const contexts: SequenceWindowContext[] = [];\n  for (const [fileIndex, { tokens, containerStatements }] of files.entries()) {\n    if (tokens && containerStatements) {\n      fileIndexByContext.push(fileIndex);\n      contexts.push({ tokens, literalCountPrefix: buildLiteralCountPrefix(tokens), containers: containerStatements });\n    }\n  }\n  if (contexts.length < 2) {\n    return [];\n  }\n  return collectSequenceWindowCandidates(contexts, minTokens, true).flatMap(({ candidate, contextIndex }) => {\n    const fileIndex = fileIndexByContext[contextIndex];\n    const file = fileIndex === undefined ? undefined : files[fileIndex];\n    return fileIndex === undefined || file === undefined\n      ? []\n      : [{ ...candidate, regionBucket: fileIndex, file: file.file }];\n  });\n}\n\nfunction spansMultipleFiles(group: SelectableCandidate[]): boolean {\n  return group.length >= 2 && new Set(group.map((candidate) => candidate.regionBucket)).size >= 2;\n}\n\n/**\n * Reuses the within-file gapped (Type-3) merging by mapping every occurrence into one project-wide\n * token index space: each file's tokens are offset by more than `maxGapTokens` past the previous\n * file's, so occurrences in different files are never gap-adjacent and pairs always stay within\n * one file.\n */\nfunction mergeGapAdjacentGroups(\n  groups: SelectableCandidate[][],\n  files: CrossFileDuplicationSourceFile[],\n  maxGapTokens: number\n): CrossFileOccurrence[][] {\n  const tokenOffsets: number[] = [];\n  let offset = 0;\n  for (const { tokens, candidates } of files) {\n    tokenOffsets.push(offset);\n    // Accumulated in a loop: spreading a project-scale candidate array as call arguments would\n    // overflow V8's argument limit (~124k) and crash on Node.\n    let tokenCount = tokens?.length ?? 0;\n    if (!tokens) {\n      for (const candidate of candidates) {\n        tokenCount = Math.max(tokenCount, candidate.endTokenIndex);\n      }\n    }\n    offset += tokenCount + maxGapTokens + 1;\n  }\n  const occurrenceGroups = groups.map((group) =>\n    group\n      .map((candidate): CrossFileOccurrence => {\n        const start = candidate.startTokenIndex + (tokenOffsets[candidate.regionBucket] ?? 0);\n        const end = candidate.endTokenIndex + (tokenOffsets[candidate.regionBucket] ?? 0);\n        return {\n          file: candidate.file,\n          segments: [{ startTokenIndex: start, endTokenIndex: end }],\n          tokenCount: candidate.tokenCount,\n          startTokenIndex: start,\n          endTokenIndex: end,\n          startIndex: candidate.startIndex,\n          endIndex: candidate.endIndex,\n          startLine: candidate.startLine,\n          endLine: candidate.endLine,\n        };\n      })\n      .toSorted((left, right) => left.startTokenIndex - right.startTokenIndex)\n  );\n  return mergeAdjacentGroups(occurrenceGroups, maxGapTokens);\n}\n\nfunction summarize(groups: CrossFileOccurrence[][]): CrossFileDuplicationMetrics {\n  const reported: CrossFileDuplicateBlockGroup[] = [];\n  // Accumulated in a Map: file names are arbitrary strings, and a plain object would read\n  // inherited properties for names like \"constructor\".\n  const groupCountByFile = new Map<string, number>();\n  let duplicateBlockCount = 0;\n  for (const group of groups) {\n    // Mirrors within-file counting: each redundant occurrence contributes one count per matched\n    // fragment, gapped merging consolidates the grouping without halving the count, and spans a\n    // partial merge shares between a retained group and the merged group count once.\n    duplicateBlockCount += countRedundantFragments(group);\n    const occurrences = group\n      .map(({ file, startLine, endLine }) => ({ file, startLine, endLine }))\n      .toSorted((left, right) => left.file.localeCompare(right.file) || left.startLine - right.startLine);\n    const files = [...new Set(occurrences.map(({ file }) => file))];\n    for (const file of files) {\n      groupCountByFile.set(file, (groupCountByFile.get(file) ?? 0) + 1);\n    }\n    reported.push({ files, occurrences, tokenCount: group[0]?.tokenCount ?? 0 });\n  }\n  reported.sort(\n    (left, right) =>\n      right.tokenCount - left.tokenCount ||\n      (left.occurrences[0]?.file ?? '').localeCompare(right.occurrences[0]?.file ?? '') ||\n      (left.occurrences[0]?.startLine ?? 0) - (right.occurrences[0]?.startLine ?? 0)\n  );\n  return {\n    duplicateBlockCount,\n    duplicateBlockGroupCountByFile: Object.fromEntries(groupCountByFile),\n    groups: reported,\n  };\n}\n"],"mappings":"mPA6DA,SAAgB,EACd,EACA,EAC6B,CAC7B,IAAM,EAAY,GAAS,WAAa,EAA0B,UAC5D,EAAe,GAAS,cAAgB,EAA0B,aAClE,EAAoC,EAAM,SAAS,CAAE,OAAM,cAAc,IAC7E,EAAW,IAAK,IAAe,CAAE,GAAG,EAAW,aAAc,EAAW,MAAK,EAAE,CACjF,EAGA,IAAK,IAAM,KAAa,EAAwB,EAAO,CAAS,EAC9D,EAAW,KAAK,CAAS,EAQ3B,OAAO,EAAU,EAAuB,CAAC,GANzB,EACd,EACA,GAEC,EAAM,IAAU,EAAK,aAAe,EAAM,cAAgB,EAAK,WAAa,EAAM,UAEnC,CAAC,CAAC,OAAO,CAAC,EAAG,EAAO,CAAY,CAAC,CACrF,CAGA,SAAS,EAAwB,EAAyC,EAA0C,CAClH,IAAM,EAA+B,CAAC,EAChC,EAAoC,CAAC,EAC3C,IAAK,GAAM,CAAC,EAAW,CAAE,SAAQ,0BAA0B,EAAM,QAAQ,EACnE,GAAU,IACZ,EAAmB,KAAK,CAAS,EACjC,EAAS,KAAK,CAAE,SAAQ,mBAAoB,EAAwB,CAAM,EAAG,WAAY,CAAoB,CAAC,GAMlH,OAHI,EAAS,OAAS,EACb,CAAC,EAEH,EAAgC,EAAU,EAAW,EAAI,CAAC,CAAC,SAAS,CAAE,YAAW,kBAAmB,CACzG,IAAM,EAAY,EAAmB,GAC/B,EAAO,IAAc,IAAA,GAAY,IAAA,GAAY,EAAM,GACzD,OAAO,IAAc,IAAA,IAAa,IAAS,IAAA,GACvC,CAAC,EACD,CAAC,CAAE,GAAG,EAAW,aAAc,EAAW,KAAM,EAAK,IAAK,CAAC,CACjE,CAAC,CACH,CAEA,SAAS,EAAmB,EAAuC,CACjE,OAAO,EAAM,QAAU,GAAK,IAAI,IAAI,EAAM,IAAK,GAAc,EAAU,YAAY,CAAC,CAAC,CAAC,MAAQ,CAChG,CAQA,SAAS,EACP,EACA,EACA,EACyB,CACzB,IAAM,EAAyB,CAAC,EAC5B,EAAS,EACb,IAAK,GAAM,CAAE,SAAQ,gBAAgB,EAAO,CAC1C,EAAa,KAAK,CAAM,EAGxB,IAAI,EAAa,GAAQ,QAAU,EACnC,GAAI,CAAC,EACH,IAAK,IAAM,KAAa,EACtB,EAAa,KAAK,IAAI,EAAY,EAAU,aAAa,EAG7D,GAAU,EAAa,EAAe,CACxC,CACA,IAAM,EAAmB,EAAO,IAAK,GACnC,EACG,IAAK,GAAmC,CACvC,IAAM,EAAQ,EAAU,iBAAmB,EAAa,EAAU,eAAiB,GAC7E,EAAM,EAAU,eAAiB,EAAa,EAAU,eAAiB,GAC/E,MAAO,CACL,KAAM,EAAU,KAChB,SAAU,CAAC,CAAE,gBAAiB,EAAO,cAAe,CAAI,CAAC,EACzD,WAAY,EAAU,WACtB,gBAAiB,EACjB,cAAe,EACf,WAAY,EAAU,WACtB,SAAU,EAAU,SACpB,UAAW,EAAU,UACrB,QAAS,EAAU,OACrB,CACF,CAAC,CAAC,CACD,UAAU,EAAM,IAAU,EAAK,gBAAkB,EAAM,eAAe,CAC3E,EACA,OAAO,EAAoB,EAAkB,CAAY,CAC3D,CAEA,SAAS,EAAU,EAA8D,CAC/E,IAAM,EAA2C,CAAC,EAG5C,EAAmB,IAAI,IACzB,EAAsB,EAC1B,IAAK,IAAM,KAAS,EAAQ,CAI1B,GAAuB,EAAwB,CAAK,EACpD,IAAM,EAAc,EACjB,KAAK,CAAE,OAAM,YAAW,cAAe,CAAE,OAAM,YAAW,SAAQ,EAAE,CAAC,CACrE,UAAU,EAAM,IAAU,EAAK,KAAK,cAAc,EAAM,IAAI,GAAK,EAAK,UAAY,EAAM,SAAS,EAC9F,EAAQ,CAAC,GAAG,IAAI,IAAI,EAAY,KAAK,CAAE,UAAW,CAAI,CAAC,CAAC,EAC9D,IAAK,IAAM,KAAQ,EACjB,EAAiB,IAAI,GAAO,EAAiB,IAAI,CAAI,GAAK,GAAK,CAAC,EAElE,EAAS,KAAK,CAAE,QAAO,cAAa,WAAY,EAAM,EAAE,EAAE,YAAc,CAAE,CAAC,CAC7E,CAOA,OANA,EAAS,MACN,EAAM,IACL,EAAM,WAAa,EAAK,aACvB,EAAK,YAAY,EAAE,EAAE,MAAQ,GAAA,CAAI,cAAc,EAAM,YAAY,EAAE,EAAE,MAAQ,EAAE,IAC/E,EAAK,YAAY,EAAE,EAAE,WAAa,IAAM,EAAM,YAAY,EAAE,EAAE,WAAa,EAChF,EACO,CACL,sBACA,+BAAgC,OAAO,YAAY,CAAgB,EACnE,OAAQ,CACV,CACF"}

package/dist/duplication.cjs

@@ -1,2 +1,2 @@
-"use strict";const e=require("./duplicateSelection.cjs"),t=new Set(`statement_block.block.compound_statement.body_statement.constructor_body.do_block.if_statement.for_statement.for_in_statement.enhanced_for_statement.for_range_loop.while_statement.do_statement.try_statement.try_with_resources_statement.with_statement.switch_statement.switch_expression.switch_case.switch_block_statement_group.switch_rule.case_clause.case_statement.match_statement.match_arm.except_clause.catch_clause.finally_clause.elif_clause.ensure.expression_statement.return_statement.return_expression.if_expression.for_expression.while_expression.loop_expression.match_expression.jsx_element.jsx_self_closing_element.if.unless.case.case_match.while.until.for.begin.when`.split(`.`)),n=new Set([`program`,`source_file`,`translation_unit`,`module`,`statement_block`,`block`,`compound_statement`,`body_statement`,`constructor_body`,`class_body`,`block_body`,`do_block`,`do`,`ensure`,`then`,`else`,`case_statement`,`switch_block_statement_group`,`switch_rule`,`expression_case`,`type_case`,`communication_case`,`default_case`]),r=new Set([`identifier`,`constant`,`instance_variable`,`class_variable`,`global_variable`]),i=new Set([`shorthand_property_identifier`,`shorthand_property_identifier_pattern`]),a=new Map([[`number`,`#num`],[`number_literal`,`#num`],[`integer`,`#num`],[`float`,`#num`],[`integer_literal`,`#num`],[`float_literal`,`#num`],[`int_literal`,`#num`],[`rune_literal`,`#char`],[`imaginary_literal`,`#num`],[`decimal_integer_literal`,`#num`],[`hex_integer_literal`,`#num`],[`octal_integer_literal`,`#num`],[`binary_integer_literal`,`#num`],[`decimal_floating_point_literal`,`#num`],[`hex_floating_point_literal`,`#num`],[`string_fragment`,`#str`],[`multiline_string_fragment`,`#str`],[`string_content`,`#str`],[`raw_string_content`,`#str`],[`heredoc_content`,`#str`],[`heredoc_beginning`,`#heredoc`],[`heredoc_end`,`#heredoc`],[`string`,`#str`],[`template_string`,`#str`],[`string_literal`,`#str`],[`interpreted_string_literal`,`#str`],[`raw_string_literal`,`#str`],[`raw_string`,`#str`],[`escape_sequence`,`#str`],[`char_literal`,`#char`],[`character_literal`,`#char`],[`character`,`#char`],[`regex_pattern`,`#regex`]]),o=new Set([`#num`,`#str`,`#char`,`#regex`]),s=new Set([`comment`,`line_comment`,`block_comment`]),c=new Set([`string_fragment`,`multiline_string_fragment`,`string_content`,`raw_string_content`,`escape_sequence`,`heredoc_content`]),l=new Set([`string_fragment`,`multiline_string_fragment`,`string_content`,`raw_string_content`,`escape_sequence`,`heredoc_content`,`string_start`,`string_end`]),u=new Map([[`call_expression`,`function`],[`method_invocation`,`name`],[`call`,`method`],[`attribute`,`attribute`],[`macro_invocation`,`macro`],[`field_access`,`field`],[`new_expression`,`constructor`],[`keyword_argument`,`name`],[`element_value_pair`,`key`],[`generic_function`,`function`],[`template_function`,`name`]]),d={minTokens:40,maxGapTokens:30};function f(e,t){return e*5>=t}function p(t,n,r){let i=r?.minTokens??d.minTokens,a=r?.maxGapTokens??d.maxGapTokens,o=[],s=[],c=[];h(t,o,s,c);let l=S(o),u=[...w(o,l,s,i),...T(o,l,c,i)];return U(B(z(e.selectMaximalGroups(u,e=>e.length>=2)),a),n,o)}function m(t,n){let r=n?.minTokens??d.minTokens,i=[],a=[],o=[];h(t,i,a,o);let s=S(i),c=w(i,s,a,r);for(let e of o){let t=e[0],n=e.at(-1);!t||!n||n.endTokenIndex-t.startTokenIndex<r||c.push(O(`s:${N(i,s,t.startTokenIndex,n.endTokenIndex)}`,t.startTokenIndex,n.endTokenIndex,t.node,n.node))}return e.dedupeByRegion(c).map(({fingerprint:e,tokenCount:t,startIndex:n,endIndex:r,startLine:i,endLine:a})=>({fingerprint:e,tokenCount:t,startIndex:n,endIndex:r,startLine:i,endLine:a}))}function h(e,r,i,a){function o(e){let c=r.length,l=e.childCount===0?void 0:g(e);if(e.childCount===0)_(e,r);else if(l!==void 0)r.push(v(l,y(e,l),e.startPosition.row,e.endPosition.row));else if(!s.has(e.type)){let t=[],r=e.isNamed&&n.has(e.type);for(let n of e.children){let e=o(n);r&&n.isNamed&&!s.has(n.type)&&t.push(e)}r&&t.length>0&&a.push(t)}let u={startTokenIndex:c,endTokenIndex:r.length,node:e};return e.isNamed&&t.has(e.type)&&i.push(u),u}o(e)}function g(e){let t=e.isNamed?a.get(e.type):void 0;if(t!==void 0)return e.namedChildren.every(e=>l.has(e.type))?t:void 0}function _(e,t){if(s.has(e.type))return;let n=e.startPosition.row,o=e.endPosition.row;if(e.isNamed&&i.has(e.type)){t.push(v(e.text,void 0,n,o),v(`:`,void 0,n,o),{kind:`id`,text:e.text,textHash:0,textHash2:0,startRow:n,endRow:o});return}if(e.isNamed&&r.has(e.type)&&!C(e)){t.push({kind:`id`,text:e.text,textHash:0,textHash2:0,startRow:n,endRow:o});return}let c=e.isNamed?a.get(e.type):void 0;c===void 0?t.push(v(e.text,void 0,n,o)):t.push(v(c,y(e,c),n,o))}function v(e,t,n,r){let i={kind:`text`,text:e,textHash:I(e),textHash2:L(e),startRow:n,endRow:r};return t!==void 0&&o.has(e)&&(i.literalHash=I(t),i.literalHash2=L(t)),i}function y(e,t){if(t!==`#str`&&t!==`#char`||c.has(e.type))return e.text;let n=e.namedChildren.filter(e=>c.has(e.type));return n.length>0?n.map(e=>e.text).join(``):x(e.text)}const b=new Set([`"`,`'`,"`"]);function x(e){let t=e[0];return e.length>=2&&t!==void 0&&b.has(t)&&e.endsWith(t)?e.slice(1,-1):e}function S(e){let t=new Int32Array(e.length+1);for(let[n,r]of e.entries())t[n+1]=(t[n]??0)+(r.literalHash===void 0?0:1);return t}function C(e){let t=e.parent;if(!t)return!1;if(t.type===`method_reference`||t.type===`call`&&t.childForFieldName(`function`)?.id===e.id||e.type===`constant`&&t.type===`call`&&t.childForFieldName(`receiver`)?.id===e.id||t.type===`method_invocation`&&t.childForFieldName(`object`)?.id===e.id&&/^\p{Lu}/u.test(e.text))return!0;if((t.type===`scoped_identifier`||t.type===`qualified_identifier`)&&(t.childForFieldName(`name`)?.id===e.id||t.childForFieldName(`path`)?.id===e.id)){let e=t;for(;e.parent&&(e.parent.type===`scoped_identifier`||e.parent.type===`qualified_identifier`||e.parent.type===`generic_function`||e.parent.type===`template_function`);)e=e.parent;if(e.parent?.type===`call_expression`&&e.parent.childForFieldName(`function`)?.id===e.id)return!0}if(t.type===`literal_element`&&t.parent?.type===`keyed_element`&&t.parent.namedChild(0)?.id===t.id)return!0;let n=u.get(t.type);return n!==void 0&&t.childForFieldName(n)?.id===e.id}function w(e,t,n,r){let i=[];for(let a of n)a.endTokenIndex-a.startTokenIndex<r||i.push(O(`b:${N(e,t,a.startTokenIndex,a.endTokenIndex)}`,a.startTokenIndex,a.endTokenIndex,a.node,a.node));return i}function T(e,t,n,r){let i=[],a=new Map,o=n.map(t=>D(e,t,r));for(let[e,t]of o.entries())for(let[n,r]of t.windowKeysByStart.entries())for(let t of r){if(t===void 0)continue;let r=a.get(t);r?(r.count+=1,r.containerIndex!==e&&(r.containerIndex=-1),r.minStart=Math.min(r.minStart,n),r.maxStart=Math.max(r.maxStart,n)):a.set(t,{count:1,containerIndex:e,minStart:n,maxStart:n})}let s=(e,t)=>{if(e===void 0)return!1;let n=a.get(e);return n!==void 0&&n.count>=2&&(n.containerIndex===-1||n.maxStart-n.minStart>=t)},c=e=>{let t=o[e.containerIndex]?.statementHashes??[],n=t[e.start];for(let r=e.start+1;r<e.start+e.length;r+=1)if(t[r]!==n)return!0;return!1},l=[];for(let[e,t]of o.entries())for(let[n,r]of t.windowKeysByStart.entries())for(let[i,a]of r.entries()){if(!s(a,i)||!c({containerIndex:e,start:n,length:i}))continue;let r=t.windowKeysByStart[n]?.[i+1],o=t.windowKeysByStart[n-1]?.[i+1];s(r,i+1)||s(o,i+1)||l.push({containerIndex:e,start:n,length:i})}let u=new Set(l.map(E)),d=l;for(;d.length>0;){let r=[];for(let a of d){let o=n[a.containerIndex],s=o?.[a.start],c=o?.[a.start+a.length-1];if(!s||!c)continue;let l=`s:${N(e,t,s.startTokenIndex,c.endTokenIndex)}`;i.push(O(l,s.startTokenIndex,c.endTokenIndex,s.node,c.node)),r.push(a)}d=[];for(let e of r)for(let t of[e.start,e.start+1]){let n={containerIndex:e.containerIndex,start:t,length:e.length-1},r=o[e.containerIndex]?.windowKeysByStart[t]?.[n.length];u.has(E(n))||!s(r,n.length)||!c(n)||(u.add(E(n)),d.push(n))}}return i}function E(e){return`${e.containerIndex}:${e.start}:${e.length}`}function D(e,t,n){let r=t.map(t=>P(e,t.startTokenIndex,t.endTokenIndex)),i=[];for(let e=0;e<t.length;e+=1){let a=[],o=5381,s=0,c=Math.min(t.length,e+100);for(let i=e;i<c;i+=1){let c=t[i],l=r[i];if(!c||l===void 0)break;o=R(o,l),s+=c.endTokenIndex-c.startTokenIndex;let u=i-e+1;a[u]=u>=2&&s>=n?R(o,u):void 0}i.push(a)}return{windowKeysByStart:i,statementHashes:r}}function O(e,t,n,r,i){return{fingerprint:e,tokenCount:n-t,startTokenIndex:t,endTokenIndex:n,startIndex:r.startIndex,endIndex:i.endIndex,startLine:r.startPosition.row+1,endLine:i.endPosition.row+1}}const k=[],A=[];function j(e){let t=k[e];return t===void 0&&(t=I(`$${e}`),k[e]=t),t}function M(e){let t=A[e];return t===void 0&&(t=L(`$${e}`),A[e]=t),t}function N(e,t,n,r){let[i,a]=F(e,n,r,f((t[r]??0)-(t[n]??0),r-n));return`${i}:${a}:${r-n}`}function P(e,t,n){let[r,i]=F(e,t,n,!1);return r^Math.imul(i,31)}function F(e,t,n,r){let i=new Map,a=5381,o=52711;for(let s=t;s<n;s+=1){let t=e[s];if(!t)continue;let n,c;if(t.kind===`id`){let e=i.get(t.text);e===void 0&&(e=i.size,i.set(t.text,e)),n=j(e),c=M(e)}else n=t.textHash,c=t.textHash2;a=Math.imul(a,31)+n|0,o=Math.imul(o,37)^c,r&&t.literalHash!==void 0&&t.literalHash2!==void 0&&(a=Math.imul(a,31)+t.literalHash|0,o=Math.imul(o,37)^t.literalHash2)}return[a,o]}function I(e){let t=5381;for(let n=0;n<e.length;n+=1)t=Math.imul(t,33)^e.charCodeAt(n);return t}function L(e){let t=-2128831035;for(let n=0;n<e.length;n+=1)t=Math.imul(t^e.charCodeAt(n),16777619);return t}function R(e,t){return Math.imul(e,31)+t}function z(e){let t=[];for(let n of e.values()){let e=n.map(e=>({segments:[{startTokenIndex:e.startTokenIndex,endTokenIndex:e.endTokenIndex}],tokenCount:e.tokenCount,startTokenIndex:e.startTokenIndex,endTokenIndex:e.endTokenIndex,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startLine,endLine:e.endLine}));e.sort((e,t)=>e.startTokenIndex-t.startTokenIndex||e.endTokenIndex-t.endTokenIndex),t.push(e)}return t}function B(e,t){if(t<=0||e.length<2)return e;e.sort(V);for(let n=!0;n;){n=!1;for(let r=0;r<e.length&&!n;r+=1)for(let i=r+1;i<e.length;i+=1){let a=e[r],o=e[i];if(!a||!o)continue;let s=H(a,o,t)??H(o,a,t);if(s){e[r]=s,e.splice(i,1),e.sort(V),n=!0;break}}}return e}function V(e,t){let n=e[0],r=t[0];return(n?.startTokenIndex??0)-(r?.startTokenIndex??0)||(n?.endTokenIndex??0)-(r?.endTokenIndex??0)}function H(e,t,n){if(e.length===t.length){for(let[r,i]of e.entries()){let a=t[r];if(!a)return;let o=a.startTokenIndex-i.endTokenIndex;if(o<0||o>n)return;let s=e[r+1];if(s&&a.endTokenIndex>s.startTokenIndex)return}return e.map((e,n)=>{let r=t[n];return r?{segments:[...e.segments,...r.segments],tokenCount:e.tokenCount+r.tokenCount,startTokenIndex:e.startTokenIndex,endTokenIndex:r.endTokenIndex,startIndex:e.startIndex,endIndex:r.endIndex,startLine:e.startLine,endLine:r.endLine}:e})}}function U(e,t,n){let r=0,i=0,a=[],o=new Set;for(let s of e){r+=(s.length-1)*(s[0]?.segments.length??1);for(let e of s){i=Math.max(i,e.tokenCount);for(let r of e.segments)for(let e=r.startTokenIndex;e<r.endTokenIndex;e+=1){let r=n[e];for(let e=r?.startRow??0;e<=(r?.endRow??-1);e+=1)t.has(e+1)&&o.add(e+1)}}a.push(s.map(({startLine:e,endLine:t})=>({startLine:e,endLine:t})).toSorted((e,t)=>e.startLine-t.startLine))}return a.sort((e,t)=>(e[0]?.startLine??0)-(t[0]?.startLine??0)),{duplicateBlockCount:r,duplicateBlockGroupCount:e.length,duplicateBlockGroups:a,duplicateLineCount:o.size,duplicationRatio:t.size===0?0:o.size/t.size,maxDuplicateBlockSize:i}}exports.collectCrossFileDuplicateCandidates=m,exports.defaultDuplicationOptions=d,exports.measureDuplication=p;
+"use strict";const e=require("./duplicateSelection.cjs"),t=new Set(`statement_block.block.compound_statement.body_statement.constructor_body.do_block.if_statement.for_statement.for_in_statement.enhanced_for_statement.for_range_loop.while_statement.do_statement.try_statement.try_with_resources_statement.with_statement.switch_statement.switch_expression.switch_case.switch_block_statement_group.switch_rule.case_clause.case_statement.match_statement.match_arm.except_clause.catch_clause.finally_clause.elif_clause.ensure.expression_statement.return_statement.return_expression.if_expression.for_expression.while_expression.loop_expression.match_expression.jsx_element.jsx_self_closing_element.if.unless.case.case_match.while.until.for.begin.when`.split(`.`)),n=new Set([`program`,`source_file`,`translation_unit`,`module`,`statement_block`,`block`,`compound_statement`,`body_statement`,`constructor_body`,`class_body`,`block_body`,`do_block`,`do`,`ensure`,`then`,`else`,`case_statement`,`switch_block_statement_group`,`switch_rule`,`expression_case`,`type_case`,`communication_case`,`default_case`]),r=new Set([`identifier`,`constant`,`instance_variable`,`class_variable`,`global_variable`]),i=new Set([`shorthand_property_identifier`,`shorthand_property_identifier_pattern`]),a=new Map([[`number`,`#num`],[`number_literal`,`#num`],[`integer`,`#num`],[`float`,`#num`],[`integer_literal`,`#num`],[`float_literal`,`#num`],[`int_literal`,`#num`],[`rune_literal`,`#char`],[`imaginary_literal`,`#num`],[`decimal_integer_literal`,`#num`],[`hex_integer_literal`,`#num`],[`octal_integer_literal`,`#num`],[`binary_integer_literal`,`#num`],[`decimal_floating_point_literal`,`#num`],[`hex_floating_point_literal`,`#num`],[`string_fragment`,`#str`],[`multiline_string_fragment`,`#str`],[`string_content`,`#str`],[`raw_string_content`,`#str`],[`heredoc_content`,`#str`],[`heredoc_beginning`,`#heredoc`],[`heredoc_end`,`#heredoc`],[`string`,`#str`],[`template_string`,`#str`],[`string_literal`,`#str`],[`interpreted_string_literal`,`#str`],[`raw_string_literal`,`#str`],[`raw_string`,`#str`],[`escape_sequence`,`#str`],[`char_literal`,`#char`],[`character_literal`,`#char`],[`character`,`#char`],[`regex_pattern`,`#regex`]]),o=new Set([`#num`,`#str`,`#char`,`#regex`]),s=new Set([`comment`,`line_comment`,`block_comment`]),c=new Set([`string_fragment`,`multiline_string_fragment`,`string_content`,`raw_string_content`,`escape_sequence`,`heredoc_content`]),l=new Set([`string_fragment`,`multiline_string_fragment`,`string_content`,`raw_string_content`,`escape_sequence`,`heredoc_content`,`string_start`,`string_end`]),u=new Map([[`call_expression`,`function`],[`method_invocation`,`name`],[`call`,`method`],[`attribute`,`attribute`],[`macro_invocation`,`macro`],[`field_access`,`field`],[`new_expression`,`constructor`],[`keyword_argument`,`name`],[`element_value_pair`,`key`],[`generic_function`,`function`],[`template_function`,`name`]]),d={minTokens:40,maxGapTokens:30,minSimilarityPercent:70};function f(e,t){return e*5>=t}function p(t,n,r){let i=r?.minTokens??d.minTokens,a=r?.maxGapTokens??d.maxGapTokens,o=r?.minSimilarityPercent??d.minSimilarityPercent,s=[],c=[],l=[];h(t,s,c,l);let u=S(s),f=[...w(s,u,c,i),...T(s,u,l,i)],p=X(Y(e.selectMaximalGroups(f,e=>e.length>=2)),a),m=k(s,u,c,i,o,p);return re([...p.filter(e=>e.length>0),...m],n,s)}function m(t,n){let r=n?.minTokens??d.minTokens,i=[],a=[],o=[];h(t,i,a,o);let s=S(i),c=w(i,s,a,r);for(let e of o){let t=e[0],n=e.at(-1);!t||!n||n.endTokenIndex-t.startTokenIndex<r||c.push(R(`s:${U(i,s,t.startTokenIndex,n.endTokenIndex)}`,t.startTokenIndex,n.endTokenIndex,t,n))}return{candidates:e.dedupeByRegion(c),tokens:i,containerStatements:o}}function h(e,r,i,a){function o(e){let c=r.length,l=e.childCount===0?void 0:g(e);if(e.childCount===0)_(e,r);else if(l!==void 0)r.push(v(l,y(e,l),e.startPosition.row,e.endPosition.row));else if(!s.has(e.type)){let t=[],r=e.isNamed&&n.has(e.type);for(let n of e.children){let e=o(n);r&&n.isNamed&&!s.has(n.type)&&t.push(e)}r&&t.length>0&&a.push(t)}let u={startTokenIndex:c,endTokenIndex:r.length,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startPosition.row+1,endLine:e.endPosition.row+1};return e.isNamed&&t.has(e.type)&&i.push(u),u}o(e)}function g(e){let t=e.isNamed?a.get(e.type):void 0;if(t!==void 0)return e.namedChildren.every(e=>l.has(e.type))?t:void 0}function _(e,t){if(s.has(e.type))return;let n=e.startPosition.row,o=e.endPosition.row;if(e.isNamed&&i.has(e.type)){t.push(v(e.text,void 0,n,o,!0),v(`:`,void 0,n,o),{kind:`id`,text:e.text,textHash:0,textHash2:0,startRow:n,endRow:o});return}if(e.isNamed&&r.has(e.type)&&!C(e)){t.push({kind:`id`,text:e.text,textHash:0,textHash2:0,startRow:n,endRow:o});return}let c=e.isNamed?a.get(e.type):void 0;c===void 0?t.push(v(e.text,void 0,n,o,e.isNamed)):t.push(v(c,y(e,c),n,o))}function v(e,t,n,r,i=!1){let a={kind:`text`,text:e,textHash:K(e),textHash2:q(e),startRow:n,endRow:r};return t!==void 0&&o.has(e)&&(a.literalHash=K(t),a.literalHash2=q(t)),i&&(a.isName=!0),a}function y(e,t){if(t!==`#str`&&t!==`#char`||c.has(e.type))return e.text;let n=e.namedChildren.filter(e=>c.has(e.type));return n.length>0?n.map(e=>e.text).join(``):x(e.text)}const b=new Set([`"`,`'`,"`"]);function x(e){let t=e[0];return e.length>=2&&t!==void 0&&b.has(t)&&e.endsWith(t)?e.slice(1,-1):e}function S(e){let t=new Int32Array(e.length+1);for(let[n,r]of e.entries())t[n+1]=(t[n]??0)+(r.literalHash===void 0?0:1);return t}function C(e){let t=e.parent;if(!t)return!1;if(t.type===`method_reference`||t.type===`call`&&t.childForFieldName(`function`)?.id===e.id||e.type===`constant`&&t.type===`call`&&t.childForFieldName(`receiver`)?.id===e.id||t.type===`method_invocation`&&t.childForFieldName(`object`)?.id===e.id&&/^\p{Lu}/u.test(e.text))return!0;if((t.type===`scoped_identifier`||t.type===`qualified_identifier`)&&(t.childForFieldName(`name`)?.id===e.id||t.childForFieldName(`path`)?.id===e.id)){let e=t;for(;e.parent&&(e.parent.type===`scoped_identifier`||e.parent.type===`qualified_identifier`||e.parent.type===`generic_function`||e.parent.type===`template_function`);)e=e.parent;if(e.parent?.type===`call_expression`&&e.parent.childForFieldName(`function`)?.id===e.id)return!0}if(t.type===`literal_element`&&t.parent?.type===`keyed_element`&&t.parent.namedChild(0)?.id===t.id)return!0;let n=u.get(t.type);return n!==void 0&&t.childForFieldName(n)?.id===e.id}function w(e,t,n,r){let i=[];for(let a of n)a.endTokenIndex-a.startTokenIndex<r||i.push(R(`b:${U(e,t,a.startTokenIndex,a.endTokenIndex)}`,a.startTokenIndex,a.endTokenIndex,a,a));return i}function T(e,t,n,r){return E([{tokens:e,literalCountPrefix:t,containers:n}],r,!1).map(({candidate:e})=>e)}function E(e,t,n){let r=[],i=[],a=[];for(let[t,n]of e.entries())for(let e of n.containers)i.push(t),a.push(e);let o=t=>e[i[t]??0],s=new Map,c=a.map((e,n)=>O(o(n)?.tokens??[],e,t));for(let[e,t]of c.entries()){let n=i[e]??0;for(let[r,i]of t.windowKeysByStart.entries())for(let t of i){if(t===void 0)continue;let i=s.get(t);i?(i.count+=1,i.containerIndex!==e&&(i.containerIndex=-1),i.contextIndex!==n&&(i.contextIndex=-1),i.minStart=Math.min(i.minStart,r),i.maxStart=Math.max(i.maxStart,r)):s.set(t,{count:1,containerIndex:e,contextIndex:n,minStart:r,maxStart:r})}}let l=(e,t)=>{if(e===void 0)return!1;let r=s.get(e);return r===void 0||r.count<2?!1:n?r.contextIndex===-1:r.containerIndex===-1||r.maxStart-r.minStart>=t},u=e=>{let t=c[e.containerIndex]?.statementHashes??[],n=t[e.start];for(let r=e.start+1;r<e.start+e.length;r+=1)if(t[r]!==n)return!0;return!1},d=[];for(let[e,t]of c.entries())for(let[n,r]of t.windowKeysByStart.entries())for(let[i,a]of r.entries()){if(!l(a,i)||!u({containerIndex:e,start:n,length:i}))continue;let r=t.windowKeysByStart[n]?.[i+1],o=t.windowKeysByStart[n-1]?.[i+1];l(r,i+1)||l(o,i+1)||d.push({containerIndex:e,start:n,length:i})}let f=new Set(d.map(D)),p=d;for(;p.length>0;){let e=[];for(let t of p){let n=a[t.containerIndex],s=n?.[t.start],c=n?.[t.start+t.length-1],l=o(t.containerIndex);if(!s||!c||!l)continue;let u=`s:${U(l.tokens,l.literalCountPrefix,s.startTokenIndex,c.endTokenIndex)}`;r.push({candidate:R(u,s.startTokenIndex,c.endTokenIndex,s,c),contextIndex:i[t.containerIndex]??0}),e.push(t)}p=[];for(let t of e)for(let e of[t.start,t.start+1]){let n={containerIndex:t.containerIndex,start:e,length:t.length-1},r=c[t.containerIndex]?.windowKeysByStart[e]?.[n.length];f.has(D(n))||!l(r,n.length)||!u(n)||(f.add(D(n)),p.push(n))}}return r}function D(e){return`${e.containerIndex}:${e.start}:${e.length}`}function O(e,t,n){let r=t.map(t=>W(e,t.startTokenIndex,t.endTokenIndex)),i=[];for(let e=0;e<t.length;e+=1){let a=[],o=5381,s=0,c=Math.min(t.length,e+100);for(let i=e;i<c;i+=1){let c=t[i],l=r[i];if(!c||l===void 0)break;o=J(o,l),s+=c.endTokenIndex-c.startTokenIndex;let u=i-e+1;a[u]=u>=2&&s>=n?J(o,u):void 0}i.push(a)}return{windowKeysByStart:i,statementHashes:r}}function k(e,t,n,r,i,a){if(i>=100)return[];let o=A(n.filter(e=>{let n=e.endTokenIndex-e.startTokenIndex,i=(t[e.endTokenIndex]??0)-(t[e.startTokenIndex]??0);return n>=r&&!f(i,n)}).toSorted((e,t)=>e.startTokenIndex-t.startTokenIndex||t.endTokenIndex-e.endTokenIndex));if(o.length<2)return[];let s=o.map(e=>{let t=[];for(let[n,r]of a.entries())r.some(t=>t.startTokenIndex<e.endTokenIndex&&e.startTokenIndex<t.endTokenIndex)&&t.push(n);return t}),c=new Map,l=o.map(t=>F(e,t,c)),u=l.map(({sequence:e})=>ee(e)),d=te(u),p=o.map((e,t)=>t),m=e=>{let t=e;for(;p[t]!==t;)t=p[t]??t;for(;p[e]!==t;){let n=p[e]??t;p[e]=t,e=n}return t};for(let[e,t]of d){let n=Math.floor(e/u.length),r=e%u.length,a=l[n],o=l[r],c=u[n],d=u[r];if(!(!a||!o||!c||!d)&&!((s[n]?.length??0)>0&&(s[r]?.length??0)>0)&&!(t*100<10*Math.min(c.size,d.size))&&!(I(a,o)*100<=50*Math.max(a.contentTotal,o.contentTotal))&&ne(a.sequence,o.sequence)*100>=i*Math.max(a.sequence.length,o.sequence.length)){let e=m(n),t=m(r);p[Math.max(e,t)]=Math.min(e,t)}}let h=new Map;for(let e of o.keys()){let t=m(e),n=h.get(t)??[];n.push(e),h.set(t,n)}let g=[];for(let e of h.values()){if(e.length<2)continue;let t=e.filter(e=>(s[e]?.length??0)===0),n=e.filter(e=>(s[e]?.length??0)>0);if(n.length===0){g.push(e.flatMap(e=>o[e]?[P(o[e])]:[]));continue}if(t.length===0)continue;let r=t=>e.some(e=>{let n=o[e];return n!==void 0&&t.startTokenIndex<n.endTokenIndex&&n.startTokenIndex<t.endTokenIndex}),i=[...new Set(n.flatMap(e=>s[e]??[]))].filter(e=>{let t=a[e];return t!==void 0&&t.length>0&&t.every(r)}).toSorted((e,t)=>e-t),[c,...l]=i;if(c!==void 0){let t=new Set,n=[];for(let r of e){let e=o[r];if(!e)continue;let c=[];for(let n of i)for(let r of a[n]??[])!t.has(r)&&r.startTokenIndex<e.endTokenIndex&&e.startTokenIndex<r.endTokenIndex&&(t.add(r),c.push({occurrence:r,groupIndex:n}));c.sort((e,t)=>e.occurrence.startTokenIndex-t.occurrence.startTokenIndex||e.occurrence.endTokenIndex-t.occurrence.endTokenIndex);let l=[],u=new Set;for(let{occurrence:e,groupIndex:t}of c)u.has(t)&&(n.push(M(l)),l=[],u.clear()),l.push(e),u.add(t);l.length>0&&n.push(M(l)),c.length===0&&(s[r]?.length??0)===0&&n.push(P(e))}n.sort((e,t)=>e.startTokenIndex-t.startTokenIndex||e.endTokenIndex-t.endTokenIndex);for(let e of n)e.sharedWithMergedGroup=void 0;a[c]=n;for(let e of l)a[e]=[]}else t.length>=2&&g.push(t.flatMap(e=>o[e]?[P(o[e])]:[]))}return g.sort(Z),g}function A(e){let t=[],n=[];for(let r of e){for(;n.length>0&&n.at(-1).range.endTokenIndex<=r.startTokenIndex;)n.pop();let e=n.at(-1);if(e&&e.range.startTokenIndex===r.startTokenIndex&&e.range.endTokenIndex===r.endTokenIndex)continue;let i={range:r,children:[]};e?e.children.push(i):t.push(i),n.push(i)}let r=[],i=e=>{if(j(e))for(let t of e.children)i(t);else r.push(e.range)};for(let e of t)i(e);return r}function j(e){let[t]=e.children;return e.children.length>=2||t!==void 0&&j(t)}function M(e){let t=e[0];if(!t||e.length===1)return t??N();let n=[];for(let t of e.flatMap(e=>e.segments).toSorted((e,t)=>e.startTokenIndex-t.startTokenIndex||e.endTokenIndex-t.endTokenIndex)){let e=n.at(-1);e&&t.startTokenIndex<e.endTokenIndex?e.endTokenIndex=Math.max(e.endTokenIndex,t.endTokenIndex):n.push({...t})}return{segments:n,tokenCount:n.reduce((e,t)=>e+t.endTokenIndex-t.startTokenIndex,0),startTokenIndex:Math.min(...e.map(e=>e.startTokenIndex)),endTokenIndex:Math.max(...e.map(e=>e.endTokenIndex)),startIndex:Math.min(...e.map(e=>e.startIndex)),endIndex:Math.max(...e.map(e=>e.endIndex)),startLine:Math.min(...e.map(e=>e.startLine)),endLine:Math.max(...e.map(e=>e.endLine))}}function N(){return{segments:[],tokenCount:0,startTokenIndex:0,endTokenIndex:0,startIndex:0,endIndex:0,startLine:0,endLine:0}}function P(e){return{segments:[{startTokenIndex:e.startTokenIndex,endTokenIndex:e.endTokenIndex}],tokenCount:e.endTokenIndex-e.startTokenIndex,startTokenIndex:e.startTokenIndex,endTokenIndex:e.endTokenIndex,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startLine,endLine:e.endLine}}function F(e,t,n){let r=new Int32Array(t.endTokenIndex-t.startTokenIndex),i=new Map,a=new Map,o=0;for(let s=t.startTokenIndex;s<t.endTokenIndex;s+=1){let c=e[s];if(!c)continue;let l;if(c.kind===`id`){let e=i.get(c.text);e===void 0&&(e=i.size,i.set(c.text,e)),l=-(e+1)}else{let e=`${c.textHash}:${c.textHash2}:${c.literalHash??0}:${c.literalHash2??0}`,t=n.get(e);t===void 0&&(t=n.size,n.set(e,t)),l=t,(c.isName===!0||c.literalHash!==void 0)&&(a.set(l,(a.get(l)??0)+1),o+=1)}r[s-t.startTokenIndex]=l}return{sequence:r,contentCountBySymbol:a,contentTotal:o}}function I(e,t){let[n,r]=e.contentCountBySymbol.size<=t.contentCountBySymbol.size?[e,t]:[t,e],i=0;for(let[e,t]of n.contentCountBySymbol)i+=Math.min(t,r.contentCountBySymbol.get(e)??0);return i}function ee(e){let t=new Set;for(let n=0;n+5<=e.length;n+=1){let r=5381;for(let t=0;t<5;t+=1)r=Math.imul(r,31)+(e[n+t]??0)|0;t.add(r)}return t}function te(e){let t=new Map;for(let[n,r]of e.entries())for(let e of r){let r=t.get(e);r||(r=[],t.set(e,r)),r.push(n)}let n=new Map;for(let r of t.values())for(let t=0;t<r.length;t+=1){let i=r[t]??0;for(let a=t+1;a<r.length;a+=1){let t=i*e.length+(r[a]??0);n.set(t,(n.get(t)??0)+1)}}return n}function ne(e,t){let n=e.length+31>>>5,r=new Map;for(let[t,i]of e.entries()){let e=r.get(i);e||(e=new Uint32Array(n),r.set(i,e));let a=t>>>5;e[a]=(e[a]??0)|1<<(t&31)}let i=new Uint32Array(n);for(let e of t){let t=r.get(e),a=1,o=0;for(let e=0;e<n;e+=1){let n=i[e]??0,r=((t?.[e]??0)|n)>>>0,s=(n<<1|a)>>>0;a=n>>>31;let c=r-s-o;o=+(c<0),i[e]=r&~c}}let a=0;for(let e of i)a+=L(e);return a}function L(e){let t=e-(e>>>1&1431655765);return t=(t&858993459)+(t>>>2&858993459),Math.imul(t+(t>>>4)&252645135,16843009)>>>24&255}function R(e,t,n,r,i){return{fingerprint:e,tokenCount:n-t,startTokenIndex:t,endTokenIndex:n,startIndex:r.startIndex,endIndex:i.endIndex,startLine:r.startLine,endLine:i.endLine}}const z=[],B=[];function V(e){let t=z[e];return t===void 0&&(t=K(`$${e}`),z[e]=t),t}function H(e){let t=B[e];return t===void 0&&(t=q(`$${e}`),B[e]=t),t}function U(e,t,n,r){let[i,a]=G(e,n,r,f((t[r]??0)-(t[n]??0),r-n));return`${i}:${a}:${r-n}`}function W(e,t,n){let[r,i]=G(e,t,n,!1);return r^Math.imul(i,31)}function G(e,t,n,r){let i=new Map,a=5381,o=52711;for(let s=t;s<n;s+=1){let t=e[s];if(!t)continue;let n,c;if(t.kind===`id`){let e=i.get(t.text);e===void 0&&(e=i.size,i.set(t.text,e)),n=V(e),c=H(e)}else n=t.textHash,c=t.textHash2;a=Math.imul(a,31)+n|0,o=Math.imul(o,37)^c,r&&t.literalHash!==void 0&&t.literalHash2!==void 0&&(a=Math.imul(a,31)+t.literalHash|0,o=Math.imul(o,37)^t.literalHash2)}return[a,o]}function K(e){let t=5381;for(let n=0;n<e.length;n+=1)t=Math.imul(t,33)^e.charCodeAt(n);return t}function q(e){let t=-2128831035;for(let n=0;n<e.length;n+=1)t=Math.imul(t^e.charCodeAt(n),16777619);return t}function J(e,t){return Math.imul(e,31)+t}function Y(e){let t=[];for(let n of e.values()){let e=n.map(e=>({segments:[{startTokenIndex:e.startTokenIndex,endTokenIndex:e.endTokenIndex}],tokenCount:e.tokenCount,startTokenIndex:e.startTokenIndex,endTokenIndex:e.endTokenIndex,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startLine,endLine:e.endLine}));e.sort((e,t)=>e.startTokenIndex-t.startTokenIndex||e.endTokenIndex-t.endTokenIndex),t.push(e)}return t}function X(e,t){if(t<=0||e.length<2)return e;e.sort(Z);for(let n=!0;n;){n=!1;for(let r=0;r<e.length&&!n;r+=1)for(let i=r+1;i<e.length;i+=1){let a=e[r],o=e[i];if(!a||!o)continue;let s=Q(a,o,t),c=s??Q(o,a,t);if(!c)continue;let l=s?c.firstConsumed:c.secondConsumed,u=s?c.secondConsumed:c.firstConsumed;l&&u?(e[r]=c.merged,e.splice(i,1)):u?e[i]=c.merged:e[r]=c.merged;for(let e of c.pairedRetained)e.sharedWithMergedGroup=!0;e.sort(Z),n=!0;break}}return e}function Z(e,t){let n=e[0],r=t[0];return(n?.startTokenIndex??0)-(r?.startTokenIndex??0)||(n?.endTokenIndex??0)-(r?.endTokenIndex??0)}function Q(e,t,n){let r=e.filter(e=>!e.sharedWithMergedGroup),i=t.filter(e=>!e.sharedWithMergedGroup),a=[],o=0,s=-1;for(let e of i){for(;o<r.length;){let t=r[o];if(t&&t.endTokenIndex+n<e.startTokenIndex)o+=1;else break}let t=r[o];t&&t.endTokenIndex<=e.startTokenIndex&&t.startTokenIndex>=s&&(a.push([t,e]),s=e.endTokenIndex,o+=1)}let c=a.length===e.length,l=a.length===t.length;if(!(a.length<2||!c&&!l))return{merged:a.map(([e,t])=>({...e,sharedWithMergedGroup:void 0,segments:[...e.segments,...t.segments],tokenCount:e.tokenCount+t.tokenCount,endTokenIndex:t.endTokenIndex,endIndex:t.endIndex,endLine:t.endLine})),firstConsumed:c,secondConsumed:l,pairedRetained:c===l?[]:a.map(([e,t])=>c?t:e)}}function $(e){let t=0,n=0,r=!1;for(let i of e){if(i.sharedWithMergedGroup){r=!0;continue}t+=i.segments.length,n=Math.max(n,i.segments.length)}return r?t:t-n}function re(e,t,n){let r=0,i=0,a=[],o=new Set;for(let s of e){r+=$(s);for(let e of s){i=Math.max(i,e.tokenCount);for(let r of e.segments)for(let e=r.startTokenIndex;e<r.endTokenIndex;e+=1){let r=n[e];for(let e=r?.startRow??0;e<=(r?.endRow??-1);e+=1)t.has(e+1)&&o.add(e+1)}}a.push(s.map(({startLine:e,endLine:t})=>({startLine:e,endLine:t})).toSorted((e,t)=>e.startLine-t.startLine))}return a.sort((e,t)=>(e[0]?.startLine??0)-(t[0]?.startLine??0)),{duplicateBlockCount:r,duplicateBlockGroupCount:e.length,duplicateBlockGroups:a,duplicateLineCount:o.size,duplicationRatio:t.size===0?0:o.size/t.size,maxDuplicateBlockSize:i}}exports.buildLiteralCountPrefix=S,exports.collectCrossFileDuplicateCandidates=m,exports.collectSequenceWindowCandidates=E,exports.countRedundantFragments=$,exports.defaultDuplicationOptions=d,exports.measureDuplication=p,exports.mergeAdjacentGroups=X;
 //# sourceMappingURL=duplication.cjs.map