code-gauge 10.0.0 → 10.1.0

package/dist/crossFileDuplication.js

@@ -1,2 +1,2 @@
-import{buildLiteralCountPrefix as e,collectSegmentLines as t,collectSequenceWindowCandidates as n,countRedundantFragments as r,mergeAdjacentGroups as i,resolveDuplicationOptions as a}from"./duplication.js";import{collectCrossFileNearMissGroups as o}from"./crossFileNearMiss.js";import{selectMaximalGroups as s}from"./duplicateSelection.js";function c(e,t){let{minTokens:n,maxGapTokens:r,minSimilarityPercent:i}=a(t),o=e.flatMap(({file:e,candidates:t},n)=>t.map(t=>({...t,regionBucket:n,file:e})));for(let t of u(e,n))o.push(t);let c=s(o,f,(e,t)=>e.regionBucket-t.regionBucket||e.startIndex-t.startIndex),d=p(e,r),g=m([...c.values()],d,r);for(let t of l(e,g,d,n,i))g.push(t);return h(g,e,d)}function l(e,t,n,r,i){let a=e.map(()=>[]);for(let{fileIndex:e,startTokenIndex:r,endTokenIndex:i}of t.flat()){let t=n[e]??0;a[e]?.push({startTokenIndex:r-t,endTokenIndex:i-t})}return o(e,a,r,i).map(t=>t.map(t=>{let r=n[t.fileIndex]??0;return{...t,file:e[t.fileIndex]?.file??``,segments:t.segments.map(e=>({startTokenIndex:e.startTokenIndex+r,endTokenIndex:e.endTokenIndex+r})),startTokenIndex:t.startTokenIndex+r,endTokenIndex:t.endTokenIndex+r}}))}function u(t,r){let i=[],a=[];for(let[n,{tokens:r,containerStatements:o}]of t.entries())r&&o&&(i.push(n),a.push({tokens:r,literalCountPrefix:e(r),containers:o}));return a.length<2?[]:n(a,r,!0).flatMap(({candidate:e,contextIndex:n})=>{let r=i[n],a=r===void 0?void 0:t[r];return r===void 0||a===void 0?[]:[{...e,regionBucket:r,file:a.file}]})}function d(e){return new Set(e.map(e=>e.file)).size>=2}function f(e){return e.length>=2&&new Set(e.map(e=>e.regionBucket)).size>=2}function p(e,t){let n=[],r=0;for(let{tokens:i,candidates:a}of e){n.push(r);let e=i?.length??0;if(!i)for(let t of a)e=Math.max(e,t.endTokenIndex);r+=e+t+1}return n}function m(e,t,n){let r=e.map(e=>e.map(e=>{let n=e.startTokenIndex+(t[e.regionBucket]??0),r=e.endTokenIndex+(t[e.regionBucket]??0);return{file:e.file,fileIndex:e.regionBucket,spanCountedElsewhere:e.nestedInLargerGroup,nestedInLargerGroup:e.nestedInLargerGroup,segments:[{startTokenIndex:n,endTokenIndex:r}],tokenCount:e.tokenCount,startTokenIndex:n,endTokenIndex:r,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startLine,endLine:e.endLine}}).toSorted((e,t)=>e.startTokenIndex-t.startTokenIndex));return i(r,n,d)}function h(e,t,n){let i=[],a=/* @__PURE__ */ new Map,o=new Map(t.map((e,t)=>[e.file,{tokens:e.tokens,codeLineNumbers:e.codeLineNumbers,offset:n[t]??0}])),s=/* @__PURE__ */ new Map,c=0;for(let t of e){c+=r(t);for(let e of t)g(e,o,s);let e=t.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),n=[...new Set(e.map(({file:e})=>e))];for(let e of n)a.set(e,(a.get(e)??0)+1);i.push({files:n,occurrences:e,tokenCount:Math.min(...t.map(({tokenCount:e})=>e))})}return i.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:c,duplicateBlockGroupCountByFile:Object.fromEntries(a),duplicateLineNumbersByFile:Object.fromEntries([...s].map(([e,t])=>[e,[...t].toSorted((e,t)=>e-t)])),groups:i}}function g(e,n,r){let i=n.get(e.file);if(!i?.tokens)return;let a=r.get(e.file);a||(a=/* @__PURE__ */ new Set,r.set(e.file,a));for(let n of e.segments)t({startTokenIndex:n.startTokenIndex-i.offset,endTokenIndex:n.endTokenIndex-i.offset},i.tokens,i.codeLineNumbers,a)}export{c as measureCrossFileDuplication};
+import{buildLiteralCountPrefix as e,collectSegmentLines as t,collectSequenceWindowCandidates as n,countRedundantFragments as r,mergeAdjacentGroups as i,resolveDuplicationOptions as a}from"./duplication.js";import{collectCrossFileNearMissGroups as o}from"./crossFileNearMiss.js";import{selectMaximalGroups as s}from"./duplicateSelection.js";function c(e,t){let{minTokens:n,maxGapTokens:r,minSimilarityPercent:i}=a(t),o=e.flatMap(({file:e,candidates:t},n)=>t.map(t=>({...t,regionBucket:n,file:e})));for(let t of u(e,n))o.push(t);let c=s(o,f,(e,t)=>e.regionBucket-t.regionBucket||e.startIndex-t.startIndex),d=p(e,r),g=m([...c.values()],d,r);for(let t of l(e,g,d,n,i))g.push(t);return h(g,e,d)}function l(e,t,n,r,i){let a=e.map(()=>[]);for(let{fileIndex:e,startTokenIndex:r,endTokenIndex:i}of t.flat()){let t=n[e]??0;a[e]?.push({startTokenIndex:r-t,endTokenIndex:i-t})}return o(e,a,r,i).map(t=>t.map(t=>{let r=n[t.fileIndex]??0;return{...t,file:e[t.fileIndex]?.file??``,segments:t.segments.map(e=>({startTokenIndex:e.startTokenIndex+r,endTokenIndex:e.endTokenIndex+r})),startTokenIndex:t.startTokenIndex+r,endTokenIndex:t.endTokenIndex+r}}))}function u(t,r){let i=[],a=[];for(let[n,{tokens:r,containerStatements:o}]of t.entries())r&&o&&(i.push(n),a.push({tokens:r,literalCountPrefix:e(r),containers:o}));return a.length<2?[]:n(a,r,!0).flatMap(({candidate:e,contextIndex:n})=>{let r=i[n],a=r===void 0?void 0:t[r];return r===void 0||a===void 0?[]:[{...e,regionBucket:r,file:a.file}]})}function d(e){return new Set(e.map(e=>e.file)).size>=2}function f(e){return e.length>=2&&new Set(e.map(e=>e.regionBucket)).size>=2}function p(e,t){let n=[],r=0;for(let{tokens:i,candidates:a}of e){n.push(r);let e=i?.length??0;if(!i)for(let t of a)e=Math.max(e,t.endTokenIndex);r+=e+t+1}return n}function m(e,t,n){let r=e.map(e=>e.map(e=>{let n=e.startTokenIndex+(t[e.regionBucket]??0),r=e.endTokenIndex+(t[e.regionBucket]??0);return{file:e.file,fileIndex:e.regionBucket,spanCountedElsewhere:e.nestedInLargerGroup,nestedInLargerGroup:e.nestedInLargerGroup,segments:[{startTokenIndex:n,endTokenIndex:r}],tokenCount:e.tokenCount,startTokenIndex:n,endTokenIndex:r,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startLine,endLine:e.endLine}}).toSorted((e,t)=>e.startTokenIndex-t.startTokenIndex));return i(r,n,d)}function h(e,t,n){let i=[],a=/* @__PURE__ */ new Map,o=new Map(t.map((e,t)=>[e.file,{tokens:e.tokens,codeLineNumbers:e.codeLineNumbers,offset:n[t]??0}])),s=/* @__PURE__ */ new Map,c=0;for(let t of e){c+=r(t);let e=t.map(e=>{let{file:t,startLine:n,endLine:r,tokenCount:i}=e,a=g(e,o,s);return{file:t,startLine:n,endLine:r,...a&&{lineNumbers:a},tokenCount:i}}).toSorted((e,t)=>e.file.localeCompare(t.file)||e.startLine-t.startLine),n=[...new Set(e.map(({file:e})=>e))];for(let e of n)a.set(e,(a.get(e)??0)+1);i.push({files:n,occurrences:e,tokenCount:Math.min(...t.map(({tokenCount:e})=>e))})}return i.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:c,duplicateBlockGroupCountByFile:Object.fromEntries(a),duplicateLineNumbersByFile:Object.fromEntries([...s].map(([e,t])=>[e,[...t].toSorted((e,t)=>e-t)])),groups:i}}function g(e,n,r){let i=n.get(e.file);if(!i?.tokens)return;let a=r.get(e.file);a||(a=/* @__PURE__ */ new Set,r.set(e.file,a));let o=/* @__PURE__ */ new Set;for(let n of e.segments)t({startTokenIndex:n.startTokenIndex-i.offset,endTokenIndex:n.endTokenIndex-i.offset},i.tokens,i.codeLineNumbers,o);for(let e of o)a.add(e);return[...o].toSorted((e,t)=>e-t)}export{c 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 { collectCrossFileNearMissGroups } from './crossFileNearMiss.js';\nimport { selectMaximalGroups, type SelectableRegion } from './duplicateSelection.js';\nimport {\n  buildLiteralCountPrefix,\n  collectSegmentLines,\n  collectSequenceWindowCandidates,\n  countRedundantFragments,\n  mergeAdjacentGroups,\n  resolveDuplicationOptions,\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  /**\n   * Token count of the smallest occurrence. For exact and gapped groups it is the matched token\n   * count every occurrence shares (gaps are not counted); for near-miss (Type-3) groups it is the\n   * token count of the smallest whole block or set of matched cores, edited tokens included.\n   */\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  /**\n   * Per file, the 1-based code lines covered by the tokens of its cross-file occurrences, sorted\n   * ascending: the matched tokens of exact and gapped occurrences, and every token of a near-miss\n   * block or its matched cores, edited ones included (like within-file near-miss coverage). The unmatched gap of a merged\n   * clone and comment/blank lines inside an occurrence's bounding range are excluded (blank rows\n   * inside multi-row tokens only when the file supplied codeLineNumbers). A file that supplied\n   * only candidates (no `tokens`) has no entry — without its token stream the covered lines are\n   * unknowable, and an approximate bounding range would break this field's exactness.\n   */\n  duplicateLineNumbersByFile: Record<string, number[]>;\n  groups: CrossFileDuplicateBlockGroup[];\n}\n\ninterface SelectableCandidate extends CrossFileDuplicateCandidate, SelectableRegion {\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  fileIndex: number;\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. A copy nested inside a larger group's region (two files share\n * a whole function, a third file only a block of it) is reported with its group, so the third\n * file's copy still shows what it duplicates. 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 * Finally, blocks of files that supplied `nearMissBlocks` are compared across files for near-miss\n * (Type-3) clones under `minSimilarityPercent` (see crossFileNearMiss.ts).\n */\nexport function measureCrossFileDuplication(\n  files: CrossFileDuplicationSourceFile[],\n  options?: DuplicationOptions\n): CrossFileDuplicationMetrics {\n  const { minTokens, maxGapTokens, minSimilarityPercent } = resolveDuplicationOptions(options);\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  const tokenOffsets = computeTokenOffsets(files, maxGapTokens);\n  const groups = mergeGapAdjacentGroups([...counted.values()], tokenOffsets, maxGapTokens);\n  for (const group of collectNearMissGroups(files, groups, tokenOffsets, minTokens, minSimilarityPercent)) {\n    groups.push(group);\n  }\n  return summarize(groups, files, tokenOffsets);\n}\n\n/** Near-miss groups among the blocks the exact groups leave unreported, in the project token space. */\nfunction collectNearMissGroups(\n  files: CrossFileDuplicationSourceFile[],\n  exactGroups: CrossFileOccurrence[][],\n  tokenOffsets: number[],\n  minTokens: number,\n  minSimilarityPercent: number\n): CrossFileOccurrence[][] {\n  const reportedSpansByFile: { startTokenIndex: number; endTokenIndex: number }[][] = files.map(() => []);\n  for (const { fileIndex, startTokenIndex, endTokenIndex } of exactGroups.flat()) {\n    const offset = tokenOffsets[fileIndex] ?? 0;\n    reportedSpansByFile[fileIndex]?.push({\n      startTokenIndex: startTokenIndex - offset,\n      endTokenIndex: endTokenIndex - offset,\n    });\n  }\n  return collectCrossFileNearMissGroups(files, reportedSpansByFile, minTokens, minSimilarityPercent).map((group) =>\n    group.map((occurrence) => {\n      const offset = tokenOffsets[occurrence.fileIndex] ?? 0;\n      return {\n        ...occurrence,\n        file: files[occurrence.fileIndex]?.file ?? '',\n        segments: occurrence.segments.map((segment) => ({\n          startTokenIndex: segment.startTokenIndex + offset,\n          endTokenIndex: segment.endTokenIndex + offset,\n        })),\n        startTokenIndex: occurrence.startTokenIndex + offset,\n        endTokenIndex: occurrence.endTokenIndex + offset,\n      };\n    })\n  );\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\n/** A merged group is reported only while it still covers more than one file. */\nfunction spansMultipleFilesAfterMerge(group: CrossFileOccurrence[]): boolean {\n  return new Set(group.map((occurrence) => occurrence.file)).size >= 2;\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 * Per-file token offsets that map every file into one project-wide token index space: each file's\n * tokens are offset by more than `maxGapTokens` past the previous file's, so occurrences in\n * different files are never gap-adjacent and merged pairs always stay within one file.\n */\nfunction computeTokenOffsets(files: CrossFileDuplicationSourceFile[], maxGapTokens: number): number[] {\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  return tokenOffsets;\n}\n\n/** Reuses the within-file gapped (Type-3) merging in the project-wide token index space. */\nfunction mergeGapAdjacentGroups(\n  groups: SelectableCandidate[][],\n  tokenOffsets: number[],\n  maxGapTokens: number\n): CrossFileOccurrence[][] {\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          fileIndex: candidate.regionBucket,\n          spanCountedElsewhere: candidate.nestedInLargerGroup,\n          nestedInLargerGroup: candidate.nestedInLargerGroup,\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, spansMultipleFilesAfterMerge);\n}\n\nfunction summarize(\n  groups: CrossFileOccurrence[][],\n  files: CrossFileDuplicationSourceFile[],\n  tokenOffsets: number[]\n): CrossFileDuplicationMetrics {\n  const reported: CrossFileDuplicateBlockGroup[] = [];\n  // Accumulated in Maps: 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  const fileDataByName = new Map(\n    files.map((file, index) => [\n      file.file,\n      { tokens: file.tokens, codeLineNumbers: file.codeLineNumbers, offset: tokenOffsets[index] ?? 0 },\n    ])\n  );\n  const lineNumbersByFile = new Map<string, Set<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    for (const occurrence of group) {\n      collectOccurrenceLines(occurrence, fileDataByName, lineNumbersByFile);\n    }\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: Math.min(...group.map(({ tokenCount }) => tokenCount)) });\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    duplicateLineNumbersByFile: Object.fromEntries(\n      [...lineNumbersByFile].map(([file, lines]) => [file, [...lines].toSorted((left, right) => left - right)])\n    ),\n    groups: reported,\n  };\n}\n\n/**\n * Adds the code lines an occurrence's segment tokens cover (matched tokens of an exact or gapped\n * occurrence, the whole block or matched cores of a near-miss one) to its file's line set, mapping the\n * project-wide token segments back into the file's own token stream. A file that supplied only\n * candidates (no token stream) is skipped rather than approximated from the bounding line range,\n * which would include gap and comment/blank lines and break the field's exactness contract.\n */\nfunction collectOccurrenceLines(\n  occurrence: CrossFileOccurrence,\n  fileDataByName: Map<\n    string,\n    { tokens?: CrossFileDuplicationSourceFile['tokens']; codeLineNumbers?: Set<number>; offset: number }\n  >,\n  lineNumbersByFile: Map<string, Set<number>>\n): void {\n  const fileData = fileDataByName.get(occurrence.file);\n  if (!fileData?.tokens) {\n    return;\n  }\n  let lines = lineNumbersByFile.get(occurrence.file);\n  if (!lines) {\n    lines = new Set();\n    lineNumbersByFile.set(occurrence.file, lines);\n  }\n  for (const segment of occurrence.segments) {\n    collectSegmentLines(\n      {\n        startTokenIndex: segment.startTokenIndex - fileData.offset,\n        endTokenIndex: segment.endTokenIndex - fileData.offset,\n      },\n      fileData.tokens,\n      fileData.codeLineNumbers,\n      lines\n    );\n  }\n}\n"],"mappings":"oVAkFA,SAAgB,EACd,EACA,EAC6B,CAC7B,GAAM,CAAE,YAAW,eAAc,wBAAyB,EAA0B,CAAO,EACrF,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,EAE3B,IAAM,EAAU,EACd,EACA,GAEC,EAAM,IAAU,EAAK,aAAe,EAAM,cAAgB,EAAK,WAAa,EAAM,UACrF,EACM,EAAe,EAAoB,EAAO,CAAY,EACtD,EAAS,EAAuB,CAAC,GAAG,EAAQ,OAAO,CAAC,EAAG,EAAc,CAAY,EACvF,IAAK,IAAM,KAAS,EAAsB,EAAO,EAAQ,EAAc,EAAW,CAAoB,EACpG,EAAO,KAAK,CAAK,EAEnB,OAAO,EAAU,EAAQ,EAAO,CAAY,CAC9C,CAGA,SAAS,EACP,EACA,EACA,EACA,EACA,EACyB,CACzB,IAAM,EAA8E,EAAM,QAAU,CAAC,CAAC,EACtG,IAAK,GAAM,CAAE,YAAW,kBAAiB,mBAAmB,EAAY,KAAK,EAAG,CAC9E,IAAM,EAAS,EAAa,IAAc,EAC1C,EAAoB,EAAU,EAAE,KAAK,CACnC,gBAAiB,EAAkB,EACnC,cAAe,EAAgB,CACjC,CAAC,CACH,CACA,OAAO,EAA+B,EAAO,EAAqB,EAAW,CAAoB,CAAC,CAAC,IAAK,GACtG,EAAM,IAAK,GAAe,CACxB,IAAM,EAAS,EAAa,EAAW,YAAc,EACrD,MAAO,CACL,GAAG,EACH,KAAM,EAAM,EAAW,UAAU,EAAE,MAAQ,GAC3C,SAAU,EAAW,SAAS,IAAK,IAAa,CAC9C,gBAAiB,EAAQ,gBAAkB,EAC3C,cAAe,EAAQ,cAAgB,CACzC,EAAE,EACF,gBAAiB,EAAW,gBAAkB,EAC9C,cAAe,EAAW,cAAgB,CAC5C,CACF,CAAC,CACH,CACF,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,CAGA,SAAS,EAA6B,EAAuC,CAC3E,OAAO,IAAI,IAAI,EAAM,IAAK,GAAe,EAAW,IAAI,CAAC,CAAC,CAAC,MAAQ,CACrE,CAEA,SAAS,EAAmB,EAAuC,CACjE,OAAO,EAAM,QAAU,GAAK,IAAI,IAAI,EAAM,IAAK,GAAc,EAAU,YAAY,CAAC,CAAC,CAAC,MAAQ,CAChG,CAOA,SAAS,EAAoB,EAAyC,EAAgC,CACpG,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,OAAO,CACT,CAGA,SAAS,EACP,EACA,EACA,EACyB,CACzB,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,UAAW,EAAU,aACrB,qBAAsB,EAAU,oBAChC,oBAAqB,EAAU,oBAC/B,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,EAAc,CAA4B,CACzF,CAEA,SAAS,EACP,EACA,EACA,EAC6B,CAC7B,IAAM,EAA2C,CAAC,EAG5C,kBAAmB,IAAI,IACvB,EAAiB,IAAI,IACzB,EAAM,KAAK,EAAM,IAAU,CACzB,EAAK,KACL,CAAE,OAAQ,EAAK,OAAQ,gBAAiB,EAAK,gBAAiB,OAAQ,EAAa,IAAU,CAAE,CACjG,CAAC,CACH,EACM,kBAAoB,IAAI,IAC1B,EAAsB,EAC1B,IAAK,IAAM,KAAS,EAAQ,CAI1B,GAAuB,EAAwB,CAAK,EACpD,IAAK,IAAM,KAAc,EACvB,EAAuB,EAAY,EAAgB,CAAiB,EAEtE,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,KAAK,IAAI,GAAG,EAAM,KAAK,CAAE,gBAAiB,CAAU,CAAC,CAAE,CAAC,CAC1G,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,2BAA4B,OAAO,YACjC,CAAC,GAAG,CAAiB,CAAC,CAAC,KAAK,CAAC,EAAM,KAAW,CAAC,EAAM,CAAC,GAAG,CAAK,CAAC,CAAC,UAAU,EAAM,IAAU,EAAO,CAAK,CAAC,CAAC,CAC1G,EACA,OAAQ,CACV,CACF,CASA,SAAS,EACP,EACA,EAIA,EACM,CACN,IAAM,EAAW,EAAe,IAAI,EAAW,IAAI,EACnD,GAAI,CAAC,GAAU,OACb,OAEF,IAAI,EAAQ,EAAkB,IAAI,EAAW,IAAI,EAC5C,IACH,kBAAQ,IAAI,IACZ,EAAkB,IAAI,EAAW,KAAM,CAAK,GAE9C,IAAK,IAAM,KAAW,EAAW,SAC/B,EACE,CACE,gBAAiB,EAAQ,gBAAkB,EAAS,OACpD,cAAe,EAAQ,cAAgB,EAAS,MAClD,EACA,EAAS,OACT,EAAS,gBACT,CACF,CAEJ"}
+{"version":3,"file":"crossFileDuplication.js","names":[],"sources":["../src/crossFileDuplication.ts"],"sourcesContent":["import { collectCrossFileNearMissGroups } from './crossFileNearMiss.js';\nimport { selectMaximalGroups, type SelectableRegion } from './duplicateSelection.js';\nimport {\n  buildLiteralCountPrefix,\n  collectSegmentLines,\n  collectSequenceWindowCandidates,\n  countRedundantFragments,\n  mergeAdjacentGroups,\n  resolveDuplicationOptions,\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  /**\n   * The code lines carrying the occurrence's matched tokens (see\n   * DuplicateBlockOccurrence.lineNumbers). Absent for a file that supplied only candidates (no\n   * `tokens`), whose matched lines are unknowable.\n   */\n  lineNumbers?: number[];\n  startLine: number;\n  /**\n   * The tokens of its matched runs, or of a near-miss block or its matched cores with the edited\n   * tokens among them; the group's `tokenCount` is the smallest of its occurrences'.\n   */\n  tokenCount: number;\n}\n\nexport interface CrossFileDuplicateBlockGroup {\n  files: string[];\n  occurrences: CrossFileDuplicateOccurrence[];\n  /**\n   * Token count of the smallest occurrence. For exact and gapped groups it is the matched token\n   * count every occurrence shares (gaps are not counted); for near-miss (Type-3) groups it is the\n   * token count of the smallest whole block or set of matched cores, edited tokens included.\n   */\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  /**\n   * Per file, the 1-based code lines covered by the tokens of its cross-file occurrences, sorted\n   * ascending: the matched tokens of exact and gapped occurrences, and every token of a near-miss\n   * block or its matched cores, edited ones included (like within-file near-miss coverage). The unmatched gap of a merged\n   * clone and comment/blank lines inside an occurrence's bounding range are excluded (blank rows\n   * inside multi-row tokens only when the file supplied codeLineNumbers). A file that supplied\n   * only candidates (no `tokens`) has no entry — without its token stream the covered lines are\n   * unknowable, and an approximate bounding range would break this field's exactness.\n   */\n  duplicateLineNumbersByFile: Record<string, number[]>;\n  groups: CrossFileDuplicateBlockGroup[];\n}\n\ninterface SelectableCandidate extends CrossFileDuplicateCandidate, SelectableRegion {\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  fileIndex: number;\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. A copy nested inside a larger group's region (two files share\n * a whole function, a third file only a block of it) is reported with its group, so the third\n * file's copy still shows what it duplicates. 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 * Finally, blocks of files that supplied `nearMissBlocks` are compared across files for near-miss\n * (Type-3) clones under `minSimilarityPercent` (see crossFileNearMiss.ts).\n */\nexport function measureCrossFileDuplication(\n  files: CrossFileDuplicationSourceFile[],\n  options?: DuplicationOptions\n): CrossFileDuplicationMetrics {\n  const { minTokens, maxGapTokens, minSimilarityPercent } = resolveDuplicationOptions(options);\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  const tokenOffsets = computeTokenOffsets(files, maxGapTokens);\n  const groups = mergeGapAdjacentGroups([...counted.values()], tokenOffsets, maxGapTokens);\n  for (const group of collectNearMissGroups(files, groups, tokenOffsets, minTokens, minSimilarityPercent)) {\n    groups.push(group);\n  }\n  return summarize(groups, files, tokenOffsets);\n}\n\n/** Near-miss groups among the blocks the exact groups leave unreported, in the project token space. */\nfunction collectNearMissGroups(\n  files: CrossFileDuplicationSourceFile[],\n  exactGroups: CrossFileOccurrence[][],\n  tokenOffsets: number[],\n  minTokens: number,\n  minSimilarityPercent: number\n): CrossFileOccurrence[][] {\n  const reportedSpansByFile: { startTokenIndex: number; endTokenIndex: number }[][] = files.map(() => []);\n  for (const { fileIndex, startTokenIndex, endTokenIndex } of exactGroups.flat()) {\n    const offset = tokenOffsets[fileIndex] ?? 0;\n    reportedSpansByFile[fileIndex]?.push({\n      startTokenIndex: startTokenIndex - offset,\n      endTokenIndex: endTokenIndex - offset,\n    });\n  }\n  return collectCrossFileNearMissGroups(files, reportedSpansByFile, minTokens, minSimilarityPercent).map((group) =>\n    group.map((occurrence) => {\n      const offset = tokenOffsets[occurrence.fileIndex] ?? 0;\n      return {\n        ...occurrence,\n        file: files[occurrence.fileIndex]?.file ?? '',\n        segments: occurrence.segments.map((segment) => ({\n          startTokenIndex: segment.startTokenIndex + offset,\n          endTokenIndex: segment.endTokenIndex + offset,\n        })),\n        startTokenIndex: occurrence.startTokenIndex + offset,\n        endTokenIndex: occurrence.endTokenIndex + offset,\n      };\n    })\n  );\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\n/** A merged group is reported only while it still covers more than one file. */\nfunction spansMultipleFilesAfterMerge(group: CrossFileOccurrence[]): boolean {\n  return new Set(group.map((occurrence) => occurrence.file)).size >= 2;\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 * Per-file token offsets that map every file into one project-wide token index space: each file's\n * tokens are offset by more than `maxGapTokens` past the previous file's, so occurrences in\n * different files are never gap-adjacent and merged pairs always stay within one file.\n */\nfunction computeTokenOffsets(files: CrossFileDuplicationSourceFile[], maxGapTokens: number): number[] {\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  return tokenOffsets;\n}\n\n/** Reuses the within-file gapped (Type-3) merging in the project-wide token index space. */\nfunction mergeGapAdjacentGroups(\n  groups: SelectableCandidate[][],\n  tokenOffsets: number[],\n  maxGapTokens: number\n): CrossFileOccurrence[][] {\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          fileIndex: candidate.regionBucket,\n          spanCountedElsewhere: candidate.nestedInLargerGroup,\n          nestedInLargerGroup: candidate.nestedInLargerGroup,\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, spansMultipleFilesAfterMerge);\n}\n\nfunction summarize(\n  groups: CrossFileOccurrence[][],\n  files: CrossFileDuplicationSourceFile[],\n  tokenOffsets: number[]\n): CrossFileDuplicationMetrics {\n  const reported: CrossFileDuplicateBlockGroup[] = [];\n  // Accumulated in Maps: 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  const fileDataByName = new Map(\n    files.map((file, index) => [\n      file.file,\n      { tokens: file.tokens, codeLineNumbers: file.codeLineNumbers, offset: tokenOffsets[index] ?? 0 },\n    ])\n  );\n  const lineNumbersByFile = new Map<string, Set<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((occurrence) => {\n        const { file, startLine, endLine, tokenCount } = occurrence;\n        const lineNumbers = collectOccurrenceLines(occurrence, fileDataByName, lineNumbersByFile);\n        return { file, startLine, endLine, ...(lineNumbers && { lineNumbers }), tokenCount };\n      })\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: Math.min(...group.map(({ tokenCount }) => tokenCount)) });\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    duplicateLineNumbersByFile: Object.fromEntries(\n      [...lineNumbersByFile].map(([file, lines]) => [file, [...lines].toSorted((left, right) => left - right)])\n    ),\n    groups: reported,\n  };\n}\n\n/**\n * Adds the code lines an occurrence's segment tokens cover (matched tokens of an exact or gapped\n * occurrence, the whole block or matched cores of a near-miss one) to its file's line set, mapping the\n * project-wide token segments back into the file's own token stream, and returns them in order.\n * A file that supplied only candidates (no token stream) is skipped rather than approximated from\n * the bounding line range, which would include gap and comment/blank lines and break the field's\n * exactness contract.\n */\nfunction collectOccurrenceLines(\n  occurrence: CrossFileOccurrence,\n  fileDataByName: Map<\n    string,\n    { tokens?: CrossFileDuplicationSourceFile['tokens']; codeLineNumbers?: Set<number>; offset: number }\n  >,\n  lineNumbersByFile: Map<string, Set<number>>\n): number[] | undefined {\n  const fileData = fileDataByName.get(occurrence.file);\n  if (!fileData?.tokens) {\n    return undefined;\n  }\n  let fileLines = lineNumbersByFile.get(occurrence.file);\n  if (!fileLines) {\n    fileLines = new Set();\n    lineNumbersByFile.set(occurrence.file, fileLines);\n  }\n  const lines = new Set<number>();\n  for (const segment of occurrence.segments) {\n    collectSegmentLines(\n      {\n        startTokenIndex: segment.startTokenIndex - fileData.offset,\n        endTokenIndex: segment.endTokenIndex - fileData.offset,\n      },\n      fileData.tokens,\n      fileData.codeLineNumbers,\n      lines\n    );\n  }\n  for (const line of lines) {\n    fileLines.add(line);\n  }\n  return [...lines].toSorted((left, right) => left - right);\n}\n"],"mappings":"oVA6FA,SAAgB,EACd,EACA,EAC6B,CAC7B,GAAM,CAAE,YAAW,eAAc,wBAAyB,EAA0B,CAAO,EACrF,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,EAE3B,IAAM,EAAU,EACd,EACA,GAEC,EAAM,IAAU,EAAK,aAAe,EAAM,cAAgB,EAAK,WAAa,EAAM,UACrF,EACM,EAAe,EAAoB,EAAO,CAAY,EACtD,EAAS,EAAuB,CAAC,GAAG,EAAQ,OAAO,CAAC,EAAG,EAAc,CAAY,EACvF,IAAK,IAAM,KAAS,EAAsB,EAAO,EAAQ,EAAc,EAAW,CAAoB,EACpG,EAAO,KAAK,CAAK,EAEnB,OAAO,EAAU,EAAQ,EAAO,CAAY,CAC9C,CAGA,SAAS,EACP,EACA,EACA,EACA,EACA,EACyB,CACzB,IAAM,EAA8E,EAAM,QAAU,CAAC,CAAC,EACtG,IAAK,GAAM,CAAE,YAAW,kBAAiB,mBAAmB,EAAY,KAAK,EAAG,CAC9E,IAAM,EAAS,EAAa,IAAc,EAC1C,EAAoB,EAAU,EAAE,KAAK,CACnC,gBAAiB,EAAkB,EACnC,cAAe,EAAgB,CACjC,CAAC,CACH,CACA,OAAO,EAA+B,EAAO,EAAqB,EAAW,CAAoB,CAAC,CAAC,IAAK,GACtG,EAAM,IAAK,GAAe,CACxB,IAAM,EAAS,EAAa,EAAW,YAAc,EACrD,MAAO,CACL,GAAG,EACH,KAAM,EAAM,EAAW,UAAU,EAAE,MAAQ,GAC3C,SAAU,EAAW,SAAS,IAAK,IAAa,CAC9C,gBAAiB,EAAQ,gBAAkB,EAC3C,cAAe,EAAQ,cAAgB,CACzC,EAAE,EACF,gBAAiB,EAAW,gBAAkB,EAC9C,cAAe,EAAW,cAAgB,CAC5C,CACF,CAAC,CACH,CACF,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,CAGA,SAAS,EAA6B,EAAuC,CAC3E,OAAO,IAAI,IAAI,EAAM,IAAK,GAAe,EAAW,IAAI,CAAC,CAAC,CAAC,MAAQ,CACrE,CAEA,SAAS,EAAmB,EAAuC,CACjE,OAAO,EAAM,QAAU,GAAK,IAAI,IAAI,EAAM,IAAK,GAAc,EAAU,YAAY,CAAC,CAAC,CAAC,MAAQ,CAChG,CAOA,SAAS,EAAoB,EAAyC,EAAgC,CACpG,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,OAAO,CACT,CAGA,SAAS,EACP,EACA,EACA,EACyB,CACzB,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,UAAW,EAAU,aACrB,qBAAsB,EAAU,oBAChC,oBAAqB,EAAU,oBAC/B,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,EAAc,CAA4B,CACzF,CAEA,SAAS,EACP,EACA,EACA,EAC6B,CAC7B,IAAM,EAA2C,CAAC,EAG5C,kBAAmB,IAAI,IACvB,EAAiB,IAAI,IACzB,EAAM,KAAK,EAAM,IAAU,CACzB,EAAK,KACL,CAAE,OAAQ,EAAK,OAAQ,gBAAiB,EAAK,gBAAiB,OAAQ,EAAa,IAAU,CAAE,CACjG,CAAC,CACH,EACM,kBAAoB,IAAI,IAC1B,EAAsB,EAC1B,IAAK,IAAM,KAAS,EAAQ,CAI1B,GAAuB,EAAwB,CAAK,EACpD,IAAM,EAAc,EACjB,IAAK,GAAe,CACnB,GAAM,CAAE,OAAM,YAAW,UAAS,cAAe,EAC3C,EAAc,EAAuB,EAAY,EAAgB,CAAiB,EACxF,MAAO,CAAE,OAAM,YAAW,UAAS,GAAI,GAAe,CAAE,aAAY,EAAI,YAAW,CACrF,CAAC,CAAC,CACD,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,KAAK,IAAI,GAAG,EAAM,KAAK,CAAE,gBAAiB,CAAU,CAAC,CAAE,CAAC,CAC1G,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,2BAA4B,OAAO,YACjC,CAAC,GAAG,CAAiB,CAAC,CAAC,KAAK,CAAC,EAAM,KAAW,CAAC,EAAM,CAAC,GAAG,CAAK,CAAC,CAAC,UAAU,EAAM,IAAU,EAAO,CAAK,CAAC,CAAC,CAC1G,EACA,OAAQ,CACV,CACF,CAUA,SAAS,EACP,EACA,EAIA,EACsB,CACtB,IAAM,EAAW,EAAe,IAAI,EAAW,IAAI,EACnD,GAAI,CAAC,GAAU,OACb,OAEF,IAAI,EAAY,EAAkB,IAAI,EAAW,IAAI,EAChD,IACH,kBAAY,IAAI,IAChB,EAAkB,IAAI,EAAW,KAAM,CAAS,GAElD,IAAM,kBAAQ,IAAI,IAClB,IAAK,IAAM,KAAW,EAAW,SAC/B,EACE,CACE,gBAAiB,EAAQ,gBAAkB,EAAS,OACpD,cAAe,EAAQ,cAAgB,EAAS,MAClD,EACA,EAAS,OACT,EAAS,gBACT,CACF,EAEF,IAAK,IAAM,KAAQ,EACjB,EAAU,IAAI,CAAI,EAEpB,MAAO,CAAC,GAAG,CAAK,CAAC,CAAC,UAAU,EAAM,IAAU,EAAO,CAAK,CAC1D"}

package/dist/nativeMetrics.cjs

@@ -1,4 +1,4 @@
 "use strict";let e=require("@willbooster/shared-lib");function t(e,t,n,i,a=!1){return JSON.parse(f().measureCodeNative(...r(e,t,n,i,a)))}function n(e,t,n,i,a=!1){return f().measureCodeNativeAsync(...r(e,t,n,i,a)).then(e=>JSON.parse(e))}function r(e,t,n,r,i){return[o(e),t,n,s(r?.minTokens),s(r?.maxGapTokens),s(r?.minSimilarityPercent),i]}function i(e,t,n){return JSON.parse(f().collectCrossFileDataNative(o(e),t,s(n)))}function a(e,t){return JSON.parse(f().collectFunctionTokenSequencesNative(o(e),t)).map(e=>Int32Array.from(e))}function o(e){return e.isWellFormed()?e:e.toWellFormed()}function s(t){return t===void 0||Number.isNaN(t)?void 0:(0,e.clamp)(Math.trunc(t),0,4294967295)}var c=class extends Error{};let l,u;function d(e){l=e}function f(){if(l)return l;if(u)throw u;let t=process.getBuiltinModule(`node:module`).createRequire(require("url").pathToFileURL(__filename).href),n=[`code-gauge-${p()}`,`../native/code-gauge.node`],r=[];for(let i of n){let n;try{n=t(i)}catch(t){r.push(`  ${i}: ${(0,e.getErrorMessage)(t).split(`
-`)[0]}`);continue}let a=n.payloadVersion?.();if(a!==10){r.push(`  ${i}: payload version ${a??`unknown`} does not match the expected 10; rebuild the addon with \`bun run build-native\``);continue}return l=n,n}throw u=new c(`The code-gauge native addon is not available for ${p()}. Build it with \`node scripts/buildNative.mjs\` in the code-gauge package directory (requires a Rust toolchain).\n${r.join(`
+`)[0]}`);continue}let a=n.payloadVersion?.();if(a!==12){r.push(`  ${i}: payload version ${a??`unknown`} does not match the expected 12; rebuild the addon with \`bun run build-native\``);continue}return l=n,n}throw u=new c(`The code-gauge native addon is not available for ${p()}. Build it with \`node scripts/buildNative.mjs\` in the code-gauge package directory (requires a Rust toolchain).\n${r.join(`
 `)}`),u}function p(){let e=`${process.platform}-${process.arch}`;return process.platform===`win32`?`${e}-msvc`:process.platform===`linux`?(process.report?.getReport())?.header?.glibcVersionRuntime?`${e}-gnu`:`${e}-musl`:e}exports.NativeAddonError=c,exports.collectCrossFileDataNative=i,exports.collectFunctionTokenSequencesNative=a,exports.measureCodeNative=t,exports.measureCodeNativeAsync=n,exports.setNativeBinding=d;
 //# sourceMappingURL=nativeMetrics.cjs.map

package/dist/nativeMetrics.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"nativeMetrics.cjs","names":["clamp","getErrorMessage"],"sources":["../src/nativeMetrics.ts"],"sourcesContent":["import { clamp, getErrorMessage } from '@willbooster/shared-lib';\nimport type { CrossFileDuplicateCandidate, Token, TokenRange } from './duplication.js';\nimport type { CodeMetrics, DuplicationOptions, FunctionMetrics } from './types.js';\n\n/**\n * Halstead counts measured natively; the derived float metrics (volume, effort, ...) are\n * computed in TypeScript because they involve transcendental functions (log2) whose last-bit\n * results can differ between V8 and Rust's libm, and results must not depend on the Rust side's\n * libm build.\n */\nexport interface NativeHalsteadCounts {\n  distinctOperators: number;\n  distinctOperands: number;\n  totalOperators: number;\n  totalOperands: number;\n}\n\nexport interface NativeFunctionMetricsPayload extends Omit<FunctionMetrics, 'halstead' | 'cognitiveBlocks'> {\n  /** Absent when the function has none. */\n  cognitiveBlocks?: FunctionMetrics['cognitiveBlocks'];\n  halsteadCounts: NativeHalsteadCounts;\n}\n\nexport interface NativeMetricsPayload extends Omit<CodeMetrics, 'halstead' | 'functions' | 'syntaxTree'> {\n  functions: NativeFunctionMetricsPayload[];\n  halsteadCounts: NativeHalsteadCounts;\n  syntaxTree?: string;\n  crossFileData?: NativeCrossFileDataPayload;\n}\n\n/** One file's cross-file clone-detection contribution as serialized by the native addon. */\nexport interface NativeCrossFileDataPayload {\n  candidates: CrossFileDuplicateCandidate[];\n  tokens: Token[];\n  containerStatements: TokenRange[][];\n  nearMissBlocks: TokenRange[];\n  /** 1-based lines that are neither blank nor comment-only, sorted ascending. */\n  codeLineNumbers: number[];\n}\n\ntype NativeMeasureArguments = [\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  minTokens: number | undefined,\n  maxGapTokens: number | undefined,\n  minSimilarityPercent: number | undefined,\n  includeCrossFileData: boolean,\n];\n\nexport interface NativeBinding {\n  measureCodeNative(...args: NativeMeasureArguments): string;\n  measureCodeNativeAsync(...args: NativeMeasureArguments): Promise<string>;\n  collectCrossFileDataNative(code: string, language: string, minTokens?: number): string;\n  collectFunctionTokenSequencesNative(code: string, language: string): string;\n  payloadVersion?(): number;\n}\n\n/**\n * Must equal `payload_version` in native/src/lib.rs. A previously built addon survives a\n * `git pull` untouched, so without this handshake it would silently return payloads missing\n * newer fields, or lack newer binding functions, instead of failing with a clear rebuild message.\n */\nexport const expectedPayloadVersion = 10;\n\n/**\n * Measures one file via the native addon, returning the raw payload for assembly in metrics.ts;\n * with `includeCrossFileData`, the payload also carries the file's cross-file contribution.\n */\nexport function measureCodeNative(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication?: DuplicationOptions,\n  includeCrossFileData = false\n): NativeMetricsPayload {\n  return JSON.parse(\n    loadBinding().measureCodeNative(\n      ...toNativeMeasureArguments(code, language, includeSyntaxTree, duplication, includeCrossFileData)\n    )\n  ) as NativeMetricsPayload;\n}\n\n/**\n * measureCodeNative on the addon's worker threads, so several files can be measured in parallel.\n * A missing addon still throws synchronously, like measureCodeNative.\n */\nexport function measureCodeNativeAsync(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication?: DuplicationOptions,\n  includeCrossFileData = false\n): Promise<NativeMetricsPayload> {\n  return loadBinding()\n    .measureCodeNativeAsync(\n      ...toNativeMeasureArguments(code, language, includeSyntaxTree, duplication, includeCrossFileData)\n    )\n    .then((json) => JSON.parse(json) as NativeMetricsPayload);\n}\n\n/** The binding arguments shared by the sync and async measurements, so both measure alike. */\nfunction toNativeMeasureArguments(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication: DuplicationOptions | undefined,\n  includeCrossFileData: boolean\n): NativeMeasureArguments {\n  return [\n    toWellFormed(code),\n    language,\n    includeSyntaxTree,\n    clampToU32(duplication?.minTokens),\n    clampToU32(duplication?.maxGapTokens),\n    clampToU32(duplication?.minSimilarityPercent),\n    includeCrossFileData,\n  ];\n}\n\n/** Collects one file's cross-file clone-detection contribution via the native addon. */\nexport function collectCrossFileDataNative(\n  code: string,\n  language: string,\n  minTokens?: number\n): NativeCrossFileDataPayload {\n  return JSON.parse(\n    loadBinding().collectCrossFileDataNative(toWellFormed(code), language, clampToU32(minTokens))\n  ) as NativeCrossFileDataPayload;\n}\n\n/** Collects normalized token hash sequences of every function via the native addon. */\nexport function collectFunctionTokenSequencesNative(code: string, language: string): Int32Array[] {\n  const sequences = JSON.parse(\n    loadBinding().collectFunctionTokenSequencesNative(toWellFormed(code), language)\n  ) as number[][];\n  return sequences.map((sequence) => Int32Array.from(sequence));\n}\n\n/**\n * Lone surrogates cannot cross the N-API boundary losslessly, so ill-formed strings (invalid\n * UTF-16 occasionally present in real-world files) are measured with U+FFFD replacements — the\n * same code units V8's own UTF-8 conversion would substitute.\n */\nfunction toWellFormed(code: string): string {\n  return code.isWellFormed() ? code : code.toWellFormed();\n}\n\n/**\n * The duplication settings cross the boundary as u32, whose JavaScript conversion wraps modulo\n * 2^32 (2 ** 32 would become 0 and match everything). The public API accepts any safe integer, so\n * out-of-range values clamp to [0, u32::MAX] — no source can hold 2^32 tokens, so a clamped\n * threshold behaves identically to the requested one. NaN (e.g. `Number(unsetEnvVariable)`)\n * survives clamping arithmetic and would also convert to 0, so it is treated as an absent setting\n * instead.\n */\nfunction clampToU32(value: number | undefined): number | undefined {\n  return value === undefined || Number.isNaN(value) ? undefined : clamp(Math.trunc(value), 0, 0xFF_FF_FF_FF);\n}\n\n/**\n * Raised when no usable native addon can be loaded. Every measurement fails identically until the\n * addon is built, so callers measuring many files (the CLI scan) treat it as fatal for the whole\n * run instead of recording one \"skipped\" entry per file.\n */\nexport class NativeAddonError extends Error {}\n\nlet cachedBinding: NativeBinding | undefined;\nlet cachedFailure: NativeAddonError | undefined;\n\n/** Replaces the N-API addon, e.g. with the WebAssembly build on runtimes without N-API. */\nexport function setNativeBinding(binding: NativeBinding): void {\n  cachedBinding = binding;\n}\n\nfunction loadBinding(): NativeBinding {\n  if (cachedBinding) {\n    return cachedBinding;\n  }\n  // The failure is memoized too: resolution (including platformTriplet's diagnostic-report call\n  // on Linux) would otherwise repeat for every measured file of an already-failing run.\n  if (cachedFailure) {\n    throw cachedFailure;\n  }\n  // Resolved relative to this file, so both src/ (tests) and dist/ (build) find the addon.\n  // node:module is loaded lazily so that runtimes without it can bundle this module.\n  const requireNative = process.getBuiltinModule('node:module').createRequire(import.meta.url);\n  const specifiers = [\n    // A prebuilt platform package, when published for this platform.\n    `code-gauge-${platformTriplet()}`,\n    // A locally built addon (`bun run build-native`).\n    '../native/code-gauge.node',\n  ];\n  const failures: string[] = [];\n  for (const specifier of specifiers) {\n    let binding: NativeBinding;\n    try {\n      binding = requireNative(specifier) as NativeBinding;\n    } catch (error) {\n      failures.push(`  ${specifier}: ${getErrorMessage(error).split('\\n')[0]}`);\n      continue;\n    }\n    const version = binding.payloadVersion?.();\n    if (version !== expectedPayloadVersion) {\n      failures.push(\n        `  ${specifier}: payload version ${version ?? 'unknown'} does not match the expected ` +\n          `${expectedPayloadVersion}; rebuild the addon with \\`bun run build-native\\``\n      );\n      continue;\n    }\n    cachedBinding = binding;\n    return binding;\n  }\n  cachedFailure = new NativeAddonError(\n    `The code-gauge native addon is not available for ${platformTriplet()}. Build it with ` +\n      '`node scripts/buildNative.mjs` in the code-gauge package directory (requires a Rust ' +\n      `toolchain).\\n${failures.join('\\n')}`\n  );\n  throw cachedFailure;\n}\n\n/**\n * The platform-package suffix in the napi-rs naming convention: Linux targets are qualified by\n * libc ABI (`linux-x64-gnu` / `linux-x64-musl`) because a glibc-linked addon cannot load on\n * Alpine/musl, and Windows by toolchain ABI (`win32-x64-msvc`), matching what napi-rs tooling\n * generates. Must match scripts/installNative.mjs and the build-native workflow's target list.\n */\nfunction platformTriplet(): string {\n  const base = `${process.platform}-${process.arch}`;\n  if (process.platform === 'win32') {\n    return `${base}-msvc`;\n  }\n  if (process.platform !== 'linux') {\n    return base;\n  }\n  // Musl builds of Node report no glibc runtime version; see napi-rs's isMusl detection.\n  const report = process.report?.getReport() as { header?: { glibcVersionRuntime?: string } } | undefined;\n  return report?.header?.glibcVersionRuntime ? `${base}-gnu` : `${base}-musl`;\n}\n"],"mappings":"sDAqEA,SAAgB,EACd,EACA,EACA,EACA,EACA,EAAuB,GACD,CACtB,OAAO,KAAK,MACV,EAAY,CAAC,CAAC,kBACZ,GAAG,EAAyB,EAAM,EAAU,EAAmB,EAAa,CAAoB,CAClG,CACF,CACF,CAMA,SAAgB,EACd,EACA,EACA,EACA,EACA,EAAuB,GACQ,CAC/B,OAAO,EAAY,CAAC,CACjB,uBACC,GAAG,EAAyB,EAAM,EAAU,EAAmB,EAAa,CAAoB,CAClG,CAAC,CACA,KAAM,GAAS,KAAK,MAAM,CAAI,CAAyB,CAC5D,CAGA,SAAS,EACP,EACA,EACA,EACA,EACA,EACwB,CACxB,MAAO,CACL,EAAa,CAAI,EACjB,EACA,EACA,EAAW,GAAa,SAAS,EACjC,EAAW,GAAa,YAAY,EACpC,EAAW,GAAa,oBAAoB,EAC5C,CACF,CACF,CAGA,SAAgB,EACd,EACA,EACA,EAC4B,CAC5B,OAAO,KAAK,MACV,EAAY,CAAC,CAAC,2BAA2B,EAAa,CAAI,EAAG,EAAU,EAAW,CAAS,CAAC,CAC9F,CACF,CAGA,SAAgB,EAAoC,EAAc,EAAgC,CAIhG,OAHkB,KAAK,MACrB,EAAY,CAAC,CAAC,oCAAoC,EAAa,CAAI,EAAG,CAAQ,CAEjE,CAAC,CAAC,IAAK,GAAa,WAAW,KAAK,CAAQ,CAAC,CAC9D,CAOA,SAAS,EAAa,EAAsB,CAC1C,OAAO,EAAK,aAAa,EAAI,EAAO,EAAK,aAAa,CACxD,CAUA,SAAS,EAAW,EAA+C,CACjE,OAAO,IAAU,IAAA,IAAa,OAAO,MAAM,CAAK,EAAI,IAAA,IAAA,EAAYA,EAAAA,MAAAA,CAAM,KAAK,MAAM,CAAK,EAAG,EAAG,UAAa,CAC3G,CAOA,IAAa,EAAb,cAAsC,KAAM,CAAC,EAE7C,IAAI,EACA,EAGJ,SAAgB,EAAiB,EAA8B,CAC7D,EAAgB,CAClB,CAEA,SAAS,GAA6B,CACpC,GAAI,EACF,OAAO,EAIT,GAAI,EACF,MAAM,EAIR,IAAM,EAAgB,QAAQ,iBAAiB,aAAa,CAAC,CAAC,cAAA,QAAA,KAAA,CAAA,CAAA,cAAA,UAAA,CAAA,CAAA,IAA6B,EACrF,EAAa,CAEjB,cAAc,EAAgB,IAE9B,2BACF,EACM,EAAqB,CAAC,EAC5B,IAAK,IAAM,KAAa,EAAY,CAClC,IAAI,EACJ,GAAI,CACF,EAAU,EAAc,CAAS,CACnC,OAAS,EAAO,CACd,EAAS,KAAK,KAAK,EAAU,KAAA,EAAIC,EAAAA,gBAAAA,CAAgB,CAAK,CAAC,CAAC,MAAM;CAAI,CAAC,CAAC,IAAI,EACxE,QACF,CACA,IAAM,EAAU,EAAQ,iBAAiB,EACzC,GAAI,IAAA,GAAoC,CACtC,EAAS,KACP,KAAK,EAAU,oBAAoB,GAAW,UAAU,iFAE1D,EACA,QACF,CAEA,MADA,GAAgB,EACT,CACT,CAMA,KALA,GAAgB,IAAI,EAClB,oDAAoD,EAAgB,EAAE,qHAEpD,EAAS,KAAK;CAAI,GACtC,EACM,CACR,CAQA,SAAS,GAA0B,CACjC,IAAM,EAAO,GAAG,QAAQ,SAAS,GAAG,QAAQ,OAS5C,OARI,QAAQ,WAAa,QAChB,GAAG,EAAK,OAEb,QAAQ,WAAa,SAIV,QAAQ,QAAQ,UAAU,EAAA,EAC1B,QAAQ,oBAAsB,GAAG,EAAK,MAAQ,GAAG,EAAK,OAJ5D,CAKX"}
+{"version":3,"file":"nativeMetrics.cjs","names":["clamp","getErrorMessage"],"sources":["../src/nativeMetrics.ts"],"sourcesContent":["import { clamp, getErrorMessage } from '@willbooster/shared-lib';\nimport type { CrossFileDuplicateCandidate, Token, TokenRange } from './duplication.js';\nimport type { CodeMetrics, DuplicationOptions, FunctionMetrics } from './types.js';\n\n/**\n * Halstead counts measured natively; the derived float metrics (volume, effort, ...) are\n * computed in TypeScript because they involve transcendental functions (log2) whose last-bit\n * results can differ between V8 and Rust's libm, and results must not depend on the Rust side's\n * libm build.\n */\nexport interface NativeHalsteadCounts {\n  distinctOperators: number;\n  distinctOperands: number;\n  totalOperators: number;\n  totalOperands: number;\n}\n\nexport interface NativeFunctionMetricsPayload extends Omit<FunctionMetrics, 'halstead' | 'cognitiveBlocks'> {\n  /** Absent when the function has none. */\n  cognitiveBlocks?: FunctionMetrics['cognitiveBlocks'];\n  halsteadCounts: NativeHalsteadCounts;\n}\n\nexport interface NativeMetricsPayload extends Omit<CodeMetrics, 'halstead' | 'functions' | 'syntaxTree'> {\n  functions: NativeFunctionMetricsPayload[];\n  halsteadCounts: NativeHalsteadCounts;\n  syntaxTree?: string;\n  crossFileData?: NativeCrossFileDataPayload;\n}\n\n/** One file's cross-file clone-detection contribution as serialized by the native addon. */\nexport interface NativeCrossFileDataPayload {\n  candidates: CrossFileDuplicateCandidate[];\n  tokens: Token[];\n  containerStatements: TokenRange[][];\n  nearMissBlocks: TokenRange[];\n  /** 1-based lines that are neither blank nor comment-only, sorted ascending. */\n  codeLineNumbers: number[];\n}\n\ntype NativeMeasureArguments = [\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  minTokens: number | undefined,\n  maxGapTokens: number | undefined,\n  minSimilarityPercent: number | undefined,\n  includeCrossFileData: boolean,\n];\n\nexport interface NativeBinding {\n  measureCodeNative(...args: NativeMeasureArguments): string;\n  measureCodeNativeAsync(...args: NativeMeasureArguments): Promise<string>;\n  collectCrossFileDataNative(code: string, language: string, minTokens?: number): string;\n  collectFunctionTokenSequencesNative(code: string, language: string): string;\n  payloadVersion?(): number;\n}\n\n/**\n * Must equal `payload_version` in native/src/lib.rs. A previously built addon survives a\n * `git pull` untouched, so without this handshake it would silently return payloads missing\n * newer fields, or lack newer binding functions, instead of failing with a clear rebuild message.\n */\nexport const expectedPayloadVersion = 12;\n\n/**\n * Measures one file via the native addon, returning the raw payload for assembly in metrics.ts;\n * with `includeCrossFileData`, the payload also carries the file's cross-file contribution.\n */\nexport function measureCodeNative(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication?: DuplicationOptions,\n  includeCrossFileData = false\n): NativeMetricsPayload {\n  return JSON.parse(\n    loadBinding().measureCodeNative(\n      ...toNativeMeasureArguments(code, language, includeSyntaxTree, duplication, includeCrossFileData)\n    )\n  ) as NativeMetricsPayload;\n}\n\n/**\n * measureCodeNative on the addon's worker threads, so several files can be measured in parallel.\n * A missing addon still throws synchronously, like measureCodeNative.\n */\nexport function measureCodeNativeAsync(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication?: DuplicationOptions,\n  includeCrossFileData = false\n): Promise<NativeMetricsPayload> {\n  return loadBinding()\n    .measureCodeNativeAsync(\n      ...toNativeMeasureArguments(code, language, includeSyntaxTree, duplication, includeCrossFileData)\n    )\n    .then((json) => JSON.parse(json) as NativeMetricsPayload);\n}\n\n/** The binding arguments shared by the sync and async measurements, so both measure alike. */\nfunction toNativeMeasureArguments(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication: DuplicationOptions | undefined,\n  includeCrossFileData: boolean\n): NativeMeasureArguments {\n  return [\n    toWellFormed(code),\n    language,\n    includeSyntaxTree,\n    clampToU32(duplication?.minTokens),\n    clampToU32(duplication?.maxGapTokens),\n    clampToU32(duplication?.minSimilarityPercent),\n    includeCrossFileData,\n  ];\n}\n\n/** Collects one file's cross-file clone-detection contribution via the native addon. */\nexport function collectCrossFileDataNative(\n  code: string,\n  language: string,\n  minTokens?: number\n): NativeCrossFileDataPayload {\n  return JSON.parse(\n    loadBinding().collectCrossFileDataNative(toWellFormed(code), language, clampToU32(minTokens))\n  ) as NativeCrossFileDataPayload;\n}\n\n/** Collects normalized token hash sequences of every function via the native addon. */\nexport function collectFunctionTokenSequencesNative(code: string, language: string): Int32Array[] {\n  const sequences = JSON.parse(\n    loadBinding().collectFunctionTokenSequencesNative(toWellFormed(code), language)\n  ) as number[][];\n  return sequences.map((sequence) => Int32Array.from(sequence));\n}\n\n/**\n * Lone surrogates cannot cross the N-API boundary losslessly, so ill-formed strings (invalid\n * UTF-16 occasionally present in real-world files) are measured with U+FFFD replacements — the\n * same code units V8's own UTF-8 conversion would substitute.\n */\nfunction toWellFormed(code: string): string {\n  return code.isWellFormed() ? code : code.toWellFormed();\n}\n\n/**\n * The duplication settings cross the boundary as u32, whose JavaScript conversion wraps modulo\n * 2^32 (2 ** 32 would become 0 and match everything). The public API accepts any safe integer, so\n * out-of-range values clamp to [0, u32::MAX] — no source can hold 2^32 tokens, so a clamped\n * threshold behaves identically to the requested one. NaN (e.g. `Number(unsetEnvVariable)`)\n * survives clamping arithmetic and would also convert to 0, so it is treated as an absent setting\n * instead.\n */\nfunction clampToU32(value: number | undefined): number | undefined {\n  return value === undefined || Number.isNaN(value) ? undefined : clamp(Math.trunc(value), 0, 0xFF_FF_FF_FF);\n}\n\n/**\n * Raised when no usable native addon can be loaded. Every measurement fails identically until the\n * addon is built, so callers measuring many files (the CLI scan) treat it as fatal for the whole\n * run instead of recording one \"skipped\" entry per file.\n */\nexport class NativeAddonError extends Error {}\n\nlet cachedBinding: NativeBinding | undefined;\nlet cachedFailure: NativeAddonError | undefined;\n\n/** Replaces the N-API addon, e.g. with the WebAssembly build on runtimes without N-API. */\nexport function setNativeBinding(binding: NativeBinding): void {\n  cachedBinding = binding;\n}\n\nfunction loadBinding(): NativeBinding {\n  if (cachedBinding) {\n    return cachedBinding;\n  }\n  // The failure is memoized too: resolution (including platformTriplet's diagnostic-report call\n  // on Linux) would otherwise repeat for every measured file of an already-failing run.\n  if (cachedFailure) {\n    throw cachedFailure;\n  }\n  // Resolved relative to this file, so both src/ (tests) and dist/ (build) find the addon.\n  // node:module is loaded lazily so that runtimes without it can bundle this module.\n  const requireNative = process.getBuiltinModule('node:module').createRequire(import.meta.url);\n  const specifiers = [\n    // A prebuilt platform package, when published for this platform.\n    `code-gauge-${platformTriplet()}`,\n    // A locally built addon (`bun run build-native`).\n    '../native/code-gauge.node',\n  ];\n  const failures: string[] = [];\n  for (const specifier of specifiers) {\n    let binding: NativeBinding;\n    try {\n      binding = requireNative(specifier) as NativeBinding;\n    } catch (error) {\n      failures.push(`  ${specifier}: ${getErrorMessage(error).split('\\n')[0]}`);\n      continue;\n    }\n    const version = binding.payloadVersion?.();\n    if (version !== expectedPayloadVersion) {\n      failures.push(\n        `  ${specifier}: payload version ${version ?? 'unknown'} does not match the expected ` +\n          `${expectedPayloadVersion}; rebuild the addon with \\`bun run build-native\\``\n      );\n      continue;\n    }\n    cachedBinding = binding;\n    return binding;\n  }\n  cachedFailure = new NativeAddonError(\n    `The code-gauge native addon is not available for ${platformTriplet()}. Build it with ` +\n      '`node scripts/buildNative.mjs` in the code-gauge package directory (requires a Rust ' +\n      `toolchain).\\n${failures.join('\\n')}`\n  );\n  throw cachedFailure;\n}\n\n/**\n * The platform-package suffix in the napi-rs naming convention: Linux targets are qualified by\n * libc ABI (`linux-x64-gnu` / `linux-x64-musl`) because a glibc-linked addon cannot load on\n * Alpine/musl, and Windows by toolchain ABI (`win32-x64-msvc`), matching what napi-rs tooling\n * generates. Must match scripts/installNative.mjs and the build-native workflow's target list.\n */\nfunction platformTriplet(): string {\n  const base = `${process.platform}-${process.arch}`;\n  if (process.platform === 'win32') {\n    return `${base}-msvc`;\n  }\n  if (process.platform !== 'linux') {\n    return base;\n  }\n  // Musl builds of Node report no glibc runtime version; see napi-rs's isMusl detection.\n  const report = process.report?.getReport() as { header?: { glibcVersionRuntime?: string } } | undefined;\n  return report?.header?.glibcVersionRuntime ? `${base}-gnu` : `${base}-musl`;\n}\n"],"mappings":"sDAqEA,SAAgB,EACd,EACA,EACA,EACA,EACA,EAAuB,GACD,CACtB,OAAO,KAAK,MACV,EAAY,CAAC,CAAC,kBACZ,GAAG,EAAyB,EAAM,EAAU,EAAmB,EAAa,CAAoB,CAClG,CACF,CACF,CAMA,SAAgB,EACd,EACA,EACA,EACA,EACA,EAAuB,GACQ,CAC/B,OAAO,EAAY,CAAC,CACjB,uBACC,GAAG,EAAyB,EAAM,EAAU,EAAmB,EAAa,CAAoB,CAClG,CAAC,CACA,KAAM,GAAS,KAAK,MAAM,CAAI,CAAyB,CAC5D,CAGA,SAAS,EACP,EACA,EACA,EACA,EACA,EACwB,CACxB,MAAO,CACL,EAAa,CAAI,EACjB,EACA,EACA,EAAW,GAAa,SAAS,EACjC,EAAW,GAAa,YAAY,EACpC,EAAW,GAAa,oBAAoB,EAC5C,CACF,CACF,CAGA,SAAgB,EACd,EACA,EACA,EAC4B,CAC5B,OAAO,KAAK,MACV,EAAY,CAAC,CAAC,2BAA2B,EAAa,CAAI,EAAG,EAAU,EAAW,CAAS,CAAC,CAC9F,CACF,CAGA,SAAgB,EAAoC,EAAc,EAAgC,CAIhG,OAHkB,KAAK,MACrB,EAAY,CAAC,CAAC,oCAAoC,EAAa,CAAI,EAAG,CAAQ,CAEjE,CAAC,CAAC,IAAK,GAAa,WAAW,KAAK,CAAQ,CAAC,CAC9D,CAOA,SAAS,EAAa,EAAsB,CAC1C,OAAO,EAAK,aAAa,EAAI,EAAO,EAAK,aAAa,CACxD,CAUA,SAAS,EAAW,EAA+C,CACjE,OAAO,IAAU,IAAA,IAAa,OAAO,MAAM,CAAK,EAAI,IAAA,IAAA,EAAYA,EAAAA,MAAAA,CAAM,KAAK,MAAM,CAAK,EAAG,EAAG,UAAa,CAC3G,CAOA,IAAa,EAAb,cAAsC,KAAM,CAAC,EAE7C,IAAI,EACA,EAGJ,SAAgB,EAAiB,EAA8B,CAC7D,EAAgB,CAClB,CAEA,SAAS,GAA6B,CACpC,GAAI,EACF,OAAO,EAIT,GAAI,EACF,MAAM,EAIR,IAAM,EAAgB,QAAQ,iBAAiB,aAAa,CAAC,CAAC,cAAA,QAAA,KAAA,CAAA,CAAA,cAAA,UAAA,CAAA,CAAA,IAA6B,EACrF,EAAa,CAEjB,cAAc,EAAgB,IAE9B,2BACF,EACM,EAAqB,CAAC,EAC5B,IAAK,IAAM,KAAa,EAAY,CAClC,IAAI,EACJ,GAAI,CACF,EAAU,EAAc,CAAS,CACnC,OAAS,EAAO,CACd,EAAS,KAAK,KAAK,EAAU,KAAA,EAAIC,EAAAA,gBAAAA,CAAgB,CAAK,CAAC,CAAC,MAAM;CAAI,CAAC,CAAC,IAAI,EACxE,QACF,CACA,IAAM,EAAU,EAAQ,iBAAiB,EACzC,GAAI,IAAA,GAAoC,CACtC,EAAS,KACP,KAAK,EAAU,oBAAoB,GAAW,UAAU,iFAE1D,EACA,QACF,CAEA,MADA,GAAgB,EACT,CACT,CAMA,KALA,GAAgB,IAAI,EAClB,oDAAoD,EAAgB,EAAE,qHAEpD,EAAS,KAAK;CAAI,GACtC,EACM,CACR,CAQA,SAAS,GAA0B,CACjC,IAAM,EAAO,GAAG,QAAQ,SAAS,GAAG,QAAQ,OAS5C,OARI,QAAQ,WAAa,QAChB,GAAG,EAAK,OAEb,QAAQ,WAAa,SAIV,QAAQ,QAAQ,UAAU,EAAA,EAC1B,QAAQ,oBAAsB,GAAG,EAAK,MAAQ,GAAG,EAAK,OAJ5D,CAKX"}

package/dist/nativeMetrics.d.ts

@@ -53,7 +53,7 @@ export interface NativeBinding {
  * `git pull` untouched, so without this handshake it would silently return payloads missing
  * newer fields, or lack newer binding functions, instead of failing with a clear rebuild message.
  */
-export declare const expectedPayloadVersion = 10;
+export declare const expectedPayloadVersion = 12;
 /**
  * Measures one file via the native addon, returning the raw payload for assembly in metrics.ts;
  * with `includeCrossFileData`, the payload also carries the file's cross-file contribution.

package/dist/nativeMetrics.js

@@ -1,4 +1,4 @@
 import{clamp as e,getErrorMessage as t}from"@willbooster/shared-lib";function n(e,t,n,r,a=!1){return JSON.parse(p().measureCodeNative(...i(e,t,n,r,a)))}function r(e,t,n,r,a=!1){return p().measureCodeNativeAsync(...i(e,t,n,r,a)).then(e=>JSON.parse(e))}function i(e,t,n,r,i){return[s(e),t,n,c(r?.minTokens),c(r?.maxGapTokens),c(r?.minSimilarityPercent),i]}function a(e,t,n){return JSON.parse(p().collectCrossFileDataNative(s(e),t,c(n)))}function o(e,t){return JSON.parse(p().collectFunctionTokenSequencesNative(s(e),t)).map(e=>Int32Array.from(e))}function s(e){return e.isWellFormed()?e:e.toWellFormed()}function c(t){return t===void 0||Number.isNaN(t)?void 0:e(Math.trunc(t),0,4294967295)}var l=class extends Error{};let u,d;function f(e){u=e}function p(){if(u)return u;if(d)throw d;let e=process.getBuiltinModule(`node:module`).createRequire(import.meta.url),n=[`code-gauge-${m()}`,`../native/code-gauge.node`],r=[];for(let i of n){let n;try{n=e(i)}catch(e){r.push(`  ${i}: ${t(e).split(`
-`)[0]}`);continue}let a=n.payloadVersion?.();if(a!==10){r.push(`  ${i}: payload version ${a??`unknown`} does not match the expected 10; rebuild the addon with \`bun run build-native\``);continue}return u=n,n}throw d=new l(`The code-gauge native addon is not available for ${m()}. Build it with \`node scripts/buildNative.mjs\` in the code-gauge package directory (requires a Rust toolchain).\n${r.join(`
+`)[0]}`);continue}let a=n.payloadVersion?.();if(a!==12){r.push(`  ${i}: payload version ${a??`unknown`} does not match the expected 12; rebuild the addon with \`bun run build-native\``);continue}return u=n,n}throw d=new l(`The code-gauge native addon is not available for ${m()}. Build it with \`node scripts/buildNative.mjs\` in the code-gauge package directory (requires a Rust toolchain).\n${r.join(`
 `)}`),d}function m(){let e=`${process.platform}-${process.arch}`;return process.platform===`win32`?`${e}-msvc`:process.platform===`linux`?(process.report?.getReport())?.header?.glibcVersionRuntime?`${e}-gnu`:`${e}-musl`:e}export{l as NativeAddonError,a as collectCrossFileDataNative,o as collectFunctionTokenSequencesNative,n as measureCodeNative,r as measureCodeNativeAsync,f as setNativeBinding};
 //# sourceMappingURL=nativeMetrics.js.map

package/dist/nativeMetrics.js.map

@@ -1 +1 @@
-{"version":3,"file":"nativeMetrics.js","names":[],"sources":["../src/nativeMetrics.ts"],"sourcesContent":["import { clamp, getErrorMessage } from '@willbooster/shared-lib';\nimport type { CrossFileDuplicateCandidate, Token, TokenRange } from './duplication.js';\nimport type { CodeMetrics, DuplicationOptions, FunctionMetrics } from './types.js';\n\n/**\n * Halstead counts measured natively; the derived float metrics (volume, effort, ...) are\n * computed in TypeScript because they involve transcendental functions (log2) whose last-bit\n * results can differ between V8 and Rust's libm, and results must not depend on the Rust side's\n * libm build.\n */\nexport interface NativeHalsteadCounts {\n  distinctOperators: number;\n  distinctOperands: number;\n  totalOperators: number;\n  totalOperands: number;\n}\n\nexport interface NativeFunctionMetricsPayload extends Omit<FunctionMetrics, 'halstead' | 'cognitiveBlocks'> {\n  /** Absent when the function has none. */\n  cognitiveBlocks?: FunctionMetrics['cognitiveBlocks'];\n  halsteadCounts: NativeHalsteadCounts;\n}\n\nexport interface NativeMetricsPayload extends Omit<CodeMetrics, 'halstead' | 'functions' | 'syntaxTree'> {\n  functions: NativeFunctionMetricsPayload[];\n  halsteadCounts: NativeHalsteadCounts;\n  syntaxTree?: string;\n  crossFileData?: NativeCrossFileDataPayload;\n}\n\n/** One file's cross-file clone-detection contribution as serialized by the native addon. */\nexport interface NativeCrossFileDataPayload {\n  candidates: CrossFileDuplicateCandidate[];\n  tokens: Token[];\n  containerStatements: TokenRange[][];\n  nearMissBlocks: TokenRange[];\n  /** 1-based lines that are neither blank nor comment-only, sorted ascending. */\n  codeLineNumbers: number[];\n}\n\ntype NativeMeasureArguments = [\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  minTokens: number | undefined,\n  maxGapTokens: number | undefined,\n  minSimilarityPercent: number | undefined,\n  includeCrossFileData: boolean,\n];\n\nexport interface NativeBinding {\n  measureCodeNative(...args: NativeMeasureArguments): string;\n  measureCodeNativeAsync(...args: NativeMeasureArguments): Promise<string>;\n  collectCrossFileDataNative(code: string, language: string, minTokens?: number): string;\n  collectFunctionTokenSequencesNative(code: string, language: string): string;\n  payloadVersion?(): number;\n}\n\n/**\n * Must equal `payload_version` in native/src/lib.rs. A previously built addon survives a\n * `git pull` untouched, so without this handshake it would silently return payloads missing\n * newer fields, or lack newer binding functions, instead of failing with a clear rebuild message.\n */\nexport const expectedPayloadVersion = 10;\n\n/**\n * Measures one file via the native addon, returning the raw payload for assembly in metrics.ts;\n * with `includeCrossFileData`, the payload also carries the file's cross-file contribution.\n */\nexport function measureCodeNative(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication?: DuplicationOptions,\n  includeCrossFileData = false\n): NativeMetricsPayload {\n  return JSON.parse(\n    loadBinding().measureCodeNative(\n      ...toNativeMeasureArguments(code, language, includeSyntaxTree, duplication, includeCrossFileData)\n    )\n  ) as NativeMetricsPayload;\n}\n\n/**\n * measureCodeNative on the addon's worker threads, so several files can be measured in parallel.\n * A missing addon still throws synchronously, like measureCodeNative.\n */\nexport function measureCodeNativeAsync(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication?: DuplicationOptions,\n  includeCrossFileData = false\n): Promise<NativeMetricsPayload> {\n  return loadBinding()\n    .measureCodeNativeAsync(\n      ...toNativeMeasureArguments(code, language, includeSyntaxTree, duplication, includeCrossFileData)\n    )\n    .then((json) => JSON.parse(json) as NativeMetricsPayload);\n}\n\n/** The binding arguments shared by the sync and async measurements, so both measure alike. */\nfunction toNativeMeasureArguments(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication: DuplicationOptions | undefined,\n  includeCrossFileData: boolean\n): NativeMeasureArguments {\n  return [\n    toWellFormed(code),\n    language,\n    includeSyntaxTree,\n    clampToU32(duplication?.minTokens),\n    clampToU32(duplication?.maxGapTokens),\n    clampToU32(duplication?.minSimilarityPercent),\n    includeCrossFileData,\n  ];\n}\n\n/** Collects one file's cross-file clone-detection contribution via the native addon. */\nexport function collectCrossFileDataNative(\n  code: string,\n  language: string,\n  minTokens?: number\n): NativeCrossFileDataPayload {\n  return JSON.parse(\n    loadBinding().collectCrossFileDataNative(toWellFormed(code), language, clampToU32(minTokens))\n  ) as NativeCrossFileDataPayload;\n}\n\n/** Collects normalized token hash sequences of every function via the native addon. */\nexport function collectFunctionTokenSequencesNative(code: string, language: string): Int32Array[] {\n  const sequences = JSON.parse(\n    loadBinding().collectFunctionTokenSequencesNative(toWellFormed(code), language)\n  ) as number[][];\n  return sequences.map((sequence) => Int32Array.from(sequence));\n}\n\n/**\n * Lone surrogates cannot cross the N-API boundary losslessly, so ill-formed strings (invalid\n * UTF-16 occasionally present in real-world files) are measured with U+FFFD replacements — the\n * same code units V8's own UTF-8 conversion would substitute.\n */\nfunction toWellFormed(code: string): string {\n  return code.isWellFormed() ? code : code.toWellFormed();\n}\n\n/**\n * The duplication settings cross the boundary as u32, whose JavaScript conversion wraps modulo\n * 2^32 (2 ** 32 would become 0 and match everything). The public API accepts any safe integer, so\n * out-of-range values clamp to [0, u32::MAX] — no source can hold 2^32 tokens, so a clamped\n * threshold behaves identically to the requested one. NaN (e.g. `Number(unsetEnvVariable)`)\n * survives clamping arithmetic and would also convert to 0, so it is treated as an absent setting\n * instead.\n */\nfunction clampToU32(value: number | undefined): number | undefined {\n  return value === undefined || Number.isNaN(value) ? undefined : clamp(Math.trunc(value), 0, 0xFF_FF_FF_FF);\n}\n\n/**\n * Raised when no usable native addon can be loaded. Every measurement fails identically until the\n * addon is built, so callers measuring many files (the CLI scan) treat it as fatal for the whole\n * run instead of recording one \"skipped\" entry per file.\n */\nexport class NativeAddonError extends Error {}\n\nlet cachedBinding: NativeBinding | undefined;\nlet cachedFailure: NativeAddonError | undefined;\n\n/** Replaces the N-API addon, e.g. with the WebAssembly build on runtimes without N-API. */\nexport function setNativeBinding(binding: NativeBinding): void {\n  cachedBinding = binding;\n}\n\nfunction loadBinding(): NativeBinding {\n  if (cachedBinding) {\n    return cachedBinding;\n  }\n  // The failure is memoized too: resolution (including platformTriplet's diagnostic-report call\n  // on Linux) would otherwise repeat for every measured file of an already-failing run.\n  if (cachedFailure) {\n    throw cachedFailure;\n  }\n  // Resolved relative to this file, so both src/ (tests) and dist/ (build) find the addon.\n  // node:module is loaded lazily so that runtimes without it can bundle this module.\n  const requireNative = process.getBuiltinModule('node:module').createRequire(import.meta.url);\n  const specifiers = [\n    // A prebuilt platform package, when published for this platform.\n    `code-gauge-${platformTriplet()}`,\n    // A locally built addon (`bun run build-native`).\n    '../native/code-gauge.node',\n  ];\n  const failures: string[] = [];\n  for (const specifier of specifiers) {\n    let binding: NativeBinding;\n    try {\n      binding = requireNative(specifier) as NativeBinding;\n    } catch (error) {\n      failures.push(`  ${specifier}: ${getErrorMessage(error).split('\\n')[0]}`);\n      continue;\n    }\n    const version = binding.payloadVersion?.();\n    if (version !== expectedPayloadVersion) {\n      failures.push(\n        `  ${specifier}: payload version ${version ?? 'unknown'} does not match the expected ` +\n          `${expectedPayloadVersion}; rebuild the addon with \\`bun run build-native\\``\n      );\n      continue;\n    }\n    cachedBinding = binding;\n    return binding;\n  }\n  cachedFailure = new NativeAddonError(\n    `The code-gauge native addon is not available for ${platformTriplet()}. Build it with ` +\n      '`node scripts/buildNative.mjs` in the code-gauge package directory (requires a Rust ' +\n      `toolchain).\\n${failures.join('\\n')}`\n  );\n  throw cachedFailure;\n}\n\n/**\n * The platform-package suffix in the napi-rs naming convention: Linux targets are qualified by\n * libc ABI (`linux-x64-gnu` / `linux-x64-musl`) because a glibc-linked addon cannot load on\n * Alpine/musl, and Windows by toolchain ABI (`win32-x64-msvc`), matching what napi-rs tooling\n * generates. Must match scripts/installNative.mjs and the build-native workflow's target list.\n */\nfunction platformTriplet(): string {\n  const base = `${process.platform}-${process.arch}`;\n  if (process.platform === 'win32') {\n    return `${base}-msvc`;\n  }\n  if (process.platform !== 'linux') {\n    return base;\n  }\n  // Musl builds of Node report no glibc runtime version; see napi-rs's isMusl detection.\n  const report = process.report?.getReport() as { header?: { glibcVersionRuntime?: string } } | undefined;\n  return report?.header?.glibcVersionRuntime ? `${base}-gnu` : `${base}-musl`;\n}\n"],"mappings":"qEAqEA,SAAgB,EACd,EACA,EACA,EACA,EACA,EAAuB,GACD,CACtB,OAAO,KAAK,MACV,EAAY,CAAC,CAAC,kBACZ,GAAG,EAAyB,EAAM,EAAU,EAAmB,EAAa,CAAoB,CAClG,CACF,CACF,CAMA,SAAgB,EACd,EACA,EACA,EACA,EACA,EAAuB,GACQ,CAC/B,OAAO,EAAY,CAAC,CACjB,uBACC,GAAG,EAAyB,EAAM,EAAU,EAAmB,EAAa,CAAoB,CAClG,CAAC,CACA,KAAM,GAAS,KAAK,MAAM,CAAI,CAAyB,CAC5D,CAGA,SAAS,EACP,EACA,EACA,EACA,EACA,EACwB,CACxB,MAAO,CACL,EAAa,CAAI,EACjB,EACA,EACA,EAAW,GAAa,SAAS,EACjC,EAAW,GAAa,YAAY,EACpC,EAAW,GAAa,oBAAoB,EAC5C,CACF,CACF,CAGA,SAAgB,EACd,EACA,EACA,EAC4B,CAC5B,OAAO,KAAK,MACV,EAAY,CAAC,CAAC,2BAA2B,EAAa,CAAI,EAAG,EAAU,EAAW,CAAS,CAAC,CAC9F,CACF,CAGA,SAAgB,EAAoC,EAAc,EAAgC,CAIhG,OAHkB,KAAK,MACrB,EAAY,CAAC,CAAC,oCAAoC,EAAa,CAAI,EAAG,CAAQ,CAEjE,CAAC,CAAC,IAAK,GAAa,WAAW,KAAK,CAAQ,CAAC,CAC9D,CAOA,SAAS,EAAa,EAAsB,CAC1C,OAAO,EAAK,aAAa,EAAI,EAAO,EAAK,aAAa,CACxD,CAUA,SAAS,EAAW,EAA+C,CACjE,OAAO,IAAU,IAAA,IAAa,OAAO,MAAM,CAAK,EAAI,IAAA,GAAY,EAAM,KAAK,MAAM,CAAK,EAAG,EAAG,UAAa,CAC3G,CAOA,IAAa,EAAb,cAAsC,KAAM,CAAC,EAE7C,IAAI,EACA,EAGJ,SAAgB,EAAiB,EAA8B,CAC7D,EAAgB,CAClB,CAEA,SAAS,GAA6B,CACpC,GAAI,EACF,OAAO,EAIT,GAAI,EACF,MAAM,EAIR,IAAM,EAAgB,QAAQ,iBAAiB,aAAa,CAAC,CAAC,cAAc,YAAY,GAAG,EACrF,EAAa,CAEjB,cAAc,EAAgB,IAE9B,2BACF,EACM,EAAqB,CAAC,EAC5B,IAAK,IAAM,KAAa,EAAY,CAClC,IAAI,EACJ,GAAI,CACF,EAAU,EAAc,CAAS,CACnC,OAAS,EAAO,CACd,EAAS,KAAK,KAAK,EAAU,IAAI,EAAgB,CAAK,CAAC,CAAC,MAAM;CAAI,CAAC,CAAC,IAAI,EACxE,QACF,CACA,IAAM,EAAU,EAAQ,iBAAiB,EACzC,GAAI,IAAA,GAAoC,CACtC,EAAS,KACP,KAAK,EAAU,oBAAoB,GAAW,UAAU,iFAE1D,EACA,QACF,CAEA,MADA,GAAgB,EACT,CACT,CAMA,KALA,GAAgB,IAAI,EAClB,oDAAoD,EAAgB,EAAE,qHAEpD,EAAS,KAAK;CAAI,GACtC,EACM,CACR,CAQA,SAAS,GAA0B,CACjC,IAAM,EAAO,GAAG,QAAQ,SAAS,GAAG,QAAQ,OAS5C,OARI,QAAQ,WAAa,QAChB,GAAG,EAAK,OAEb,QAAQ,WAAa,SAIV,QAAQ,QAAQ,UAAU,EAAA,EAC1B,QAAQ,oBAAsB,GAAG,EAAK,MAAQ,GAAG,EAAK,OAJ5D,CAKX"}
+{"version":3,"file":"nativeMetrics.js","names":[],"sources":["../src/nativeMetrics.ts"],"sourcesContent":["import { clamp, getErrorMessage } from '@willbooster/shared-lib';\nimport type { CrossFileDuplicateCandidate, Token, TokenRange } from './duplication.js';\nimport type { CodeMetrics, DuplicationOptions, FunctionMetrics } from './types.js';\n\n/**\n * Halstead counts measured natively; the derived float metrics (volume, effort, ...) are\n * computed in TypeScript because they involve transcendental functions (log2) whose last-bit\n * results can differ between V8 and Rust's libm, and results must not depend on the Rust side's\n * libm build.\n */\nexport interface NativeHalsteadCounts {\n  distinctOperators: number;\n  distinctOperands: number;\n  totalOperators: number;\n  totalOperands: number;\n}\n\nexport interface NativeFunctionMetricsPayload extends Omit<FunctionMetrics, 'halstead' | 'cognitiveBlocks'> {\n  /** Absent when the function has none. */\n  cognitiveBlocks?: FunctionMetrics['cognitiveBlocks'];\n  halsteadCounts: NativeHalsteadCounts;\n}\n\nexport interface NativeMetricsPayload extends Omit<CodeMetrics, 'halstead' | 'functions' | 'syntaxTree'> {\n  functions: NativeFunctionMetricsPayload[];\n  halsteadCounts: NativeHalsteadCounts;\n  syntaxTree?: string;\n  crossFileData?: NativeCrossFileDataPayload;\n}\n\n/** One file's cross-file clone-detection contribution as serialized by the native addon. */\nexport interface NativeCrossFileDataPayload {\n  candidates: CrossFileDuplicateCandidate[];\n  tokens: Token[];\n  containerStatements: TokenRange[][];\n  nearMissBlocks: TokenRange[];\n  /** 1-based lines that are neither blank nor comment-only, sorted ascending. */\n  codeLineNumbers: number[];\n}\n\ntype NativeMeasureArguments = [\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  minTokens: number | undefined,\n  maxGapTokens: number | undefined,\n  minSimilarityPercent: number | undefined,\n  includeCrossFileData: boolean,\n];\n\nexport interface NativeBinding {\n  measureCodeNative(...args: NativeMeasureArguments): string;\n  measureCodeNativeAsync(...args: NativeMeasureArguments): Promise<string>;\n  collectCrossFileDataNative(code: string, language: string, minTokens?: number): string;\n  collectFunctionTokenSequencesNative(code: string, language: string): string;\n  payloadVersion?(): number;\n}\n\n/**\n * Must equal `payload_version` in native/src/lib.rs. A previously built addon survives a\n * `git pull` untouched, so without this handshake it would silently return payloads missing\n * newer fields, or lack newer binding functions, instead of failing with a clear rebuild message.\n */\nexport const expectedPayloadVersion = 12;\n\n/**\n * Measures one file via the native addon, returning the raw payload for assembly in metrics.ts;\n * with `includeCrossFileData`, the payload also carries the file's cross-file contribution.\n */\nexport function measureCodeNative(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication?: DuplicationOptions,\n  includeCrossFileData = false\n): NativeMetricsPayload {\n  return JSON.parse(\n    loadBinding().measureCodeNative(\n      ...toNativeMeasureArguments(code, language, includeSyntaxTree, duplication, includeCrossFileData)\n    )\n  ) as NativeMetricsPayload;\n}\n\n/**\n * measureCodeNative on the addon's worker threads, so several files can be measured in parallel.\n * A missing addon still throws synchronously, like measureCodeNative.\n */\nexport function measureCodeNativeAsync(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication?: DuplicationOptions,\n  includeCrossFileData = false\n): Promise<NativeMetricsPayload> {\n  return loadBinding()\n    .measureCodeNativeAsync(\n      ...toNativeMeasureArguments(code, language, includeSyntaxTree, duplication, includeCrossFileData)\n    )\n    .then((json) => JSON.parse(json) as NativeMetricsPayload);\n}\n\n/** The binding arguments shared by the sync and async measurements, so both measure alike. */\nfunction toNativeMeasureArguments(\n  code: string,\n  language: string,\n  includeSyntaxTree: boolean,\n  duplication: DuplicationOptions | undefined,\n  includeCrossFileData: boolean\n): NativeMeasureArguments {\n  return [\n    toWellFormed(code),\n    language,\n    includeSyntaxTree,\n    clampToU32(duplication?.minTokens),\n    clampToU32(duplication?.maxGapTokens),\n    clampToU32(duplication?.minSimilarityPercent),\n    includeCrossFileData,\n  ];\n}\n\n/** Collects one file's cross-file clone-detection contribution via the native addon. */\nexport function collectCrossFileDataNative(\n  code: string,\n  language: string,\n  minTokens?: number\n): NativeCrossFileDataPayload {\n  return JSON.parse(\n    loadBinding().collectCrossFileDataNative(toWellFormed(code), language, clampToU32(minTokens))\n  ) as NativeCrossFileDataPayload;\n}\n\n/** Collects normalized token hash sequences of every function via the native addon. */\nexport function collectFunctionTokenSequencesNative(code: string, language: string): Int32Array[] {\n  const sequences = JSON.parse(\n    loadBinding().collectFunctionTokenSequencesNative(toWellFormed(code), language)\n  ) as number[][];\n  return sequences.map((sequence) => Int32Array.from(sequence));\n}\n\n/**\n * Lone surrogates cannot cross the N-API boundary losslessly, so ill-formed strings (invalid\n * UTF-16 occasionally present in real-world files) are measured with U+FFFD replacements — the\n * same code units V8's own UTF-8 conversion would substitute.\n */\nfunction toWellFormed(code: string): string {\n  return code.isWellFormed() ? code : code.toWellFormed();\n}\n\n/**\n * The duplication settings cross the boundary as u32, whose JavaScript conversion wraps modulo\n * 2^32 (2 ** 32 would become 0 and match everything). The public API accepts any safe integer, so\n * out-of-range values clamp to [0, u32::MAX] — no source can hold 2^32 tokens, so a clamped\n * threshold behaves identically to the requested one. NaN (e.g. `Number(unsetEnvVariable)`)\n * survives clamping arithmetic and would also convert to 0, so it is treated as an absent setting\n * instead.\n */\nfunction clampToU32(value: number | undefined): number | undefined {\n  return value === undefined || Number.isNaN(value) ? undefined : clamp(Math.trunc(value), 0, 0xFF_FF_FF_FF);\n}\n\n/**\n * Raised when no usable native addon can be loaded. Every measurement fails identically until the\n * addon is built, so callers measuring many files (the CLI scan) treat it as fatal for the whole\n * run instead of recording one \"skipped\" entry per file.\n */\nexport class NativeAddonError extends Error {}\n\nlet cachedBinding: NativeBinding | undefined;\nlet cachedFailure: NativeAddonError | undefined;\n\n/** Replaces the N-API addon, e.g. with the WebAssembly build on runtimes without N-API. */\nexport function setNativeBinding(binding: NativeBinding): void {\n  cachedBinding = binding;\n}\n\nfunction loadBinding(): NativeBinding {\n  if (cachedBinding) {\n    return cachedBinding;\n  }\n  // The failure is memoized too: resolution (including platformTriplet's diagnostic-report call\n  // on Linux) would otherwise repeat for every measured file of an already-failing run.\n  if (cachedFailure) {\n    throw cachedFailure;\n  }\n  // Resolved relative to this file, so both src/ (tests) and dist/ (build) find the addon.\n  // node:module is loaded lazily so that runtimes without it can bundle this module.\n  const requireNative = process.getBuiltinModule('node:module').createRequire(import.meta.url);\n  const specifiers = [\n    // A prebuilt platform package, when published for this platform.\n    `code-gauge-${platformTriplet()}`,\n    // A locally built addon (`bun run build-native`).\n    '../native/code-gauge.node',\n  ];\n  const failures: string[] = [];\n  for (const specifier of specifiers) {\n    let binding: NativeBinding;\n    try {\n      binding = requireNative(specifier) as NativeBinding;\n    } catch (error) {\n      failures.push(`  ${specifier}: ${getErrorMessage(error).split('\\n')[0]}`);\n      continue;\n    }\n    const version = binding.payloadVersion?.();\n    if (version !== expectedPayloadVersion) {\n      failures.push(\n        `  ${specifier}: payload version ${version ?? 'unknown'} does not match the expected ` +\n          `${expectedPayloadVersion}; rebuild the addon with \\`bun run build-native\\``\n      );\n      continue;\n    }\n    cachedBinding = binding;\n    return binding;\n  }\n  cachedFailure = new NativeAddonError(\n    `The code-gauge native addon is not available for ${platformTriplet()}. Build it with ` +\n      '`node scripts/buildNative.mjs` in the code-gauge package directory (requires a Rust ' +\n      `toolchain).\\n${failures.join('\\n')}`\n  );\n  throw cachedFailure;\n}\n\n/**\n * The platform-package suffix in the napi-rs naming convention: Linux targets are qualified by\n * libc ABI (`linux-x64-gnu` / `linux-x64-musl`) because a glibc-linked addon cannot load on\n * Alpine/musl, and Windows by toolchain ABI (`win32-x64-msvc`), matching what napi-rs tooling\n * generates. Must match scripts/installNative.mjs and the build-native workflow's target list.\n */\nfunction platformTriplet(): string {\n  const base = `${process.platform}-${process.arch}`;\n  if (process.platform === 'win32') {\n    return `${base}-msvc`;\n  }\n  if (process.platform !== 'linux') {\n    return base;\n  }\n  // Musl builds of Node report no glibc runtime version; see napi-rs's isMusl detection.\n  const report = process.report?.getReport() as { header?: { glibcVersionRuntime?: string } } | undefined;\n  return report?.header?.glibcVersionRuntime ? `${base}-gnu` : `${base}-musl`;\n}\n"],"mappings":"qEAqEA,SAAgB,EACd,EACA,EACA,EACA,EACA,EAAuB,GACD,CACtB,OAAO,KAAK,MACV,EAAY,CAAC,CAAC,kBACZ,GAAG,EAAyB,EAAM,EAAU,EAAmB,EAAa,CAAoB,CAClG,CACF,CACF,CAMA,SAAgB,EACd,EACA,EACA,EACA,EACA,EAAuB,GACQ,CAC/B,OAAO,EAAY,CAAC,CACjB,uBACC,GAAG,EAAyB,EAAM,EAAU,EAAmB,EAAa,CAAoB,CAClG,CAAC,CACA,KAAM,GAAS,KAAK,MAAM,CAAI,CAAyB,CAC5D,CAGA,SAAS,EACP,EACA,EACA,EACA,EACA,EACwB,CACxB,MAAO,CACL,EAAa,CAAI,EACjB,EACA,EACA,EAAW,GAAa,SAAS,EACjC,EAAW,GAAa,YAAY,EACpC,EAAW,GAAa,oBAAoB,EAC5C,CACF,CACF,CAGA,SAAgB,EACd,EACA,EACA,EAC4B,CAC5B,OAAO,KAAK,MACV,EAAY,CAAC,CAAC,2BAA2B,EAAa,CAAI,EAAG,EAAU,EAAW,CAAS,CAAC,CAC9F,CACF,CAGA,SAAgB,EAAoC,EAAc,EAAgC,CAIhG,OAHkB,KAAK,MACrB,EAAY,CAAC,CAAC,oCAAoC,EAAa,CAAI,EAAG,CAAQ,CAEjE,CAAC,CAAC,IAAK,GAAa,WAAW,KAAK,CAAQ,CAAC,CAC9D,CAOA,SAAS,EAAa,EAAsB,CAC1C,OAAO,EAAK,aAAa,EAAI,EAAO,EAAK,aAAa,CACxD,CAUA,SAAS,EAAW,EAA+C,CACjE,OAAO,IAAU,IAAA,IAAa,OAAO,MAAM,CAAK,EAAI,IAAA,GAAY,EAAM,KAAK,MAAM,CAAK,EAAG,EAAG,UAAa,CAC3G,CAOA,IAAa,EAAb,cAAsC,KAAM,CAAC,EAE7C,IAAI,EACA,EAGJ,SAAgB,EAAiB,EAA8B,CAC7D,EAAgB,CAClB,CAEA,SAAS,GAA6B,CACpC,GAAI,EACF,OAAO,EAIT,GAAI,EACF,MAAM,EAIR,IAAM,EAAgB,QAAQ,iBAAiB,aAAa,CAAC,CAAC,cAAc,YAAY,GAAG,EACrF,EAAa,CAEjB,cAAc,EAAgB,IAE9B,2BACF,EACM,EAAqB,CAAC,EAC5B,IAAK,IAAM,KAAa,EAAY,CAClC,IAAI,EACJ,GAAI,CACF,EAAU,EAAc,CAAS,CACnC,OAAS,EAAO,CACd,EAAS,KAAK,KAAK,EAAU,IAAI,EAAgB,CAAK,CAAC,CAAC,MAAM;CAAI,CAAC,CAAC,IAAI,EACxE,QACF,CACA,IAAM,EAAU,EAAQ,iBAAiB,EACzC,GAAI,IAAA,GAAoC,CACtC,EAAS,KACP,KAAK,EAAU,oBAAoB,GAAW,UAAU,iFAE1D,EACA,QACF,CAEA,MADA,GAAgB,EACT,CACT,CAMA,KALA,GAAgB,IAAI,EAClB,oDAAoD,EAAgB,EAAE,qHAEpD,EAAS,KAAK;CAAI,GACtC,EACM,CACR,CAQA,SAAS,GAA0B,CACjC,IAAM,EAAO,GAAG,QAAQ,SAAS,GAAG,QAAQ,OAS5C,OARI,QAAQ,WAAa,QAChB,GAAG,EAAK,OAEb,QAAQ,WAAa,SAIV,QAAQ,QAAQ,UAAU,EAAA,EAC1B,QAAQ,oBAAsB,GAAG,EAAK,MAAQ,GAAG,EAAK,OAJ5D,CAKX"}

package/dist/thresholdCheck.cjs

@@ -1,2 +1,2 @@
-"use strict";const e=require("./_virtual/_rolldown/runtime.cjs"),t=require("./thresholds.cjs");let n=require("node:assert");n=e.__toESM(n,1);const r=[`file`,`function`,`duplication`],i=new Set(t.functionThresholds.filter(({includesNestedFunctions:e})=>e).map(e=>t.metricNameOf(e)));function a(e,n,r,i){let a=g(n,new Set(e.map(({file:e})=>e))),l=[],d=0;for(let n of e){let{file:e,metrics:o,hunks:m}=n,h=r(o.language),g=m?.length===0?[]:p(t.fileThresholds,o,h);g.length>0&&l.push(f({file:e,startLine:1,endLine:o.lines.total},`file`,g));let v=s(n);d+=v.length;let y=v.map(e=>({fn:e,exceeded:p(t.functionThresholds,e,h)})).filter(({exceeded:e})=>e.length>0);for(let{fn:t,exceeded:n}of y){let r=n.filter(e=>!c(e,t,y));r.length>0&&l.push(u(e,t,r))}l.push(..._(n,a.get(e)??[],h,i))}return{violations:l.toSorted(o),checkedFunctionCount:d}}function o(e,n){return t.levels.indexOf(n.level)-t.levels.indexOf(e.level)||x(e.file,n.file)||e.startLine-n.startLine||r.indexOf(e.kind)-r.indexOf(n.kind)||e.endLine-n.endLine}function s({metrics:e,hunks:t}){let n=e.language===`python`;return t?e.functions.filter(e=>t.some(t=>h(t,e.startLine,e.endLine,n))):e.functions}function c(e,t,n){return i.has(e.metric)?n.some(n=>n.fn!==t&&l(n.fn,t)&&n.exceeded.some(({metric:t})=>t===e.metric)):!1}function l(e,t){let n=e.startLine<t.startLine||e.startLine===t.startLine&&e.startColumn<=t.startColumn,r=e.endLine>t.endLine||e.endLine===t.endLine&&e.endColumn>=t.endColumn;return n&&r}function u(e,t,n){let{startLine:r,endLine:i}=t,a=n.some(({metric:e})=>e===`functionCognitiveComplexity`);return{...f({file:e,startLine:r,endLine:i},`function`,n),name:d(t),...a&&t.cognitiveBlocks.length>0&&{largestBlocks:t.cognitiveBlocks}}}function d({name:e=`<anonymous>`,containerName:t}){return t?`${t}.${e}`:e}function f(e,t,n){return{kind:t,level:n.some(e=>e.level===`error`)?`error`:`warning`,...e,exceeded:n}}function p(e,t,n){return e.flatMap(e=>m(e,t,n,(e,t)=>e>t))}function m(e,n,r,i){let a=e.measure(n),o=t.levels.findLast(t=>i(a,r[t][e.key]));return o?[{metric:t.metricNameOf(e),value:a,level:o,limit:r[o][e.key]}]:[]}function h(e,t,n,r=!1){return e.headCount===0?e.headStart>=t&&(e.headStart<n||r&&e.headStart===n):e.headStart<=n&&e.headStart+e.headCount>t}function g(e,t){let n=/* @__PURE__ */ new Map;for(let r of e?.groups??[])for(let e of r.files)if(t.has(e)){let t=n.get(e)??[];t.push(r),n.set(e,t)}return n}function _({file:e,metrics:n,hunks:r},i,a,o){let s=e=>m(t.duplicationThreshold,{codeLineCount:y(e,o)},a,(e,t)=>e>=t),c=[...n.duplication.duplicateBlockGroups.map(t=>t.map(t=>({...t,file:e}))),...i.map(e=>e.occurrences)],l=t=>t.file===e&&s(t).length>0&&(!r||r.some(e=>e.headCount>0&&h(e,t.startLine,t.endLine))),u=/* @__PURE__ */ new Map;for(let e of c)for(let t of e.filter(e=>l(e)))u.set(t,e);return v([...u.keys()]).map(({merged:e,sources:t})=>({...f(e,`duplication`,s(e)),partners:v(t.flatMap(e=>u.get(e)??[])).map(e=>e.merged).filter(t=>!b(t,e)).toSorted((e,t)=>y(t,o)-y(e,o))}))}function v(e){let t=[],n=e.toSorted((e,t)=>x(e.file,t.file)||e.startLine-t.startLine);for(let e of n){let n=t.at(-1);if(n&&b(n.merged,e))n.merged.endLine=Math.max(n.merged.endLine,e.endLine),n.sources.push(e);else{let{file:n,startLine:r,endLine:i}=e;t.push({merged:{file:n,startLine:r,endLine:i},sources:[e]})}}return t}function y(e,t){let r=t.get(e.file);n.default.ok(r,`No code lines for ${e.file}, which holds a duplicated block.`);let i=0;for(let t=e.startLine;t<=e.endLine;t++)r.has(t)&&(i+=1);return i}function b(e,t){return e.file===t.file&&e.startLine<=t.endLine&&t.startLine<=e.endLine}function x(e,t){return e<t?-1:+(e>t)}exports.checkThresholds=a,exports.qualifyFunctionName=d;
+"use strict";const e=require("./_virtual/_rolldown/runtime.cjs"),t=require("./thresholds.cjs");let n=require("node:assert");n=e.__toESM(n,1);const r=[`file`,`function`,`duplication`],i=new Set(t.functionThresholds.filter(({includesNestedFunctions:e})=>e).map(e=>t.metricNameOf(e)));function a(e,n,r){let i=g(n.crossFile,new Set(e.map(({file:e})=>e))),a=[],l=0;for(let o of e){let{file:e,metrics:d,hunks:m}=o,h=r(d.language),g=m?.length===0?[]:p(t.fileThresholds,d,h);g.length>0&&a.push(f({file:e,startLine:1,endLine:d.lines.total},`file`,g));let v=s(o);l+=v.length;let y=v.map(e=>({fn:e,exceeded:p(t.functionThresholds,e,h)})).filter(({exceeded:e})=>e.length>0);for(let{fn:t,exceeded:n}of y){let r=n.filter(e=>!c(e,t,y));r.length>0&&a.push(u(e,t,r))}a.push(..._(o,i.get(e)??[],h,n))}return{violations:a.toSorted(o),checkedFunctionCount:l}}function o(e,n){return t.levels.indexOf(n.level)-t.levels.indexOf(e.level)||C(e.file,n.file)||e.startLine-n.startLine||r.indexOf(e.kind)-r.indexOf(n.kind)||e.endLine-n.endLine}function s({metrics:e,hunks:t}){let n=e.language===`python`;return t?e.functions.filter(e=>t.some(t=>h(t,e.startLine,e.endLine,n))):e.functions}function c(e,t,n){return i.has(e.metric)?n.some(n=>n.fn!==t&&l(n.fn,t)&&n.exceeded.some(({metric:t})=>t===e.metric)):!1}function l(e,t){let n=e.startLine<t.startLine||e.startLine===t.startLine&&e.startColumn<=t.startColumn,r=e.endLine>t.endLine||e.endLine===t.endLine&&e.endColumn>=t.endColumn;return n&&r}function u(e,t,n){let{startLine:r,endLine:i}=t,a=n.some(({metric:e})=>e===`functionCognitiveComplexity`);return{...f({file:e,startLine:r,endLine:i},`function`,n),name:d(t),...a&&t.cognitiveBlocks.length>0&&{largestBlocks:t.cognitiveBlocks}}}function d({name:e=`<anonymous>`,containerName:t}){return t?`${t}.${e}`:e}function f(e,t,n){return{kind:t,level:n.some(e=>e.level===`error`)?`error`:`warning`,...e,exceeded:n}}function p(e,t,n){return e.flatMap(e=>m(e,t,n,(e,t)=>e>t))}function m(e,n,r,i){let a=e.measure(n),o=t.levels.findLast(t=>i(a,r[t][e.key]));return o?[{metric:t.metricNameOf(e),value:a,level:o,limit:r[o][e.key]}]:[]}function h(e,t,n,r=!1){return e.headCount===0?e.headStart>=t&&(e.headStart<n||r&&e.headStart===n):e.headStart<=n&&e.headStart+e.headCount>t}function g(e,t){let n=/* @__PURE__ */ new Map;for(let r of e?.groups??[])for(let e of r.files)if(t.has(e)){let t=n.get(e)??[];t.push(r),n.set(e,t)}return n}function _({file:e,metrics:n,hunks:r},i,a,{minSimilarityPercent:o}){let s=e=>m(t.duplicationThreshold,{duplicatedLineCount:e},a,(e,t)=>e>=t),c=[...n.duplication.duplicateBlockGroups.map(t=>t.map(t=>({...t,file:e}))),...i.map(e=>e.occurrences.map(e=>v(e)))],l=t=>t.file===e&&s(t.lineNumbers.length).length>0&&(!r||r.some(e=>e.headCount>0&&h(e,t.startLine,t.endLine))),u=/* @__PURE__ */ new Map;for(let e of c)for(let t of e.filter(e=>l(e)))u.set(t,e);return b([...u.keys()]).map(({merged:e,sources:t})=>{let n=n=>b(t.flatMap(e=>{let r=Math.max(...t.filter(t=>x(t,e)).map(e=>e.tokenCount));return(u.get(e)??[]).filter(e=>e.tokenCount*100>=r*n)})).filter(t=>!S(t.merged,e)).toSorted((e,t)=>y(t.sources)-y(e.sources)).map(e=>e.merged),r=n(o);return{...f(e,`duplication`,s(new Set(t.flatMap(e=>e.lineNumbers)).size)),partners:r.length>0?r:n(0)}})}function v(e){let{lineNumbers:t}=e;return n.default.ok(t,`No matched lines for ${e.file}, a scanned file.`),{...e,lineNumbers:t}}function y(e){return e.reduce((e,t)=>e+t.tokenCount,0)}function b(e){let t=[],n=e.toSorted((e,t)=>C(e.file,t.file)||e.startLine-t.startLine);for(let e of n){let n=t.at(-1);if(n&&S(n.merged,e))n.merged.endLine=Math.max(n.merged.endLine,e.endLine),n.sources.push(e);else{let{file:n,startLine:r,endLine:i}=e;t.push({merged:{file:n,startLine:r,endLine:i},sources:[e]})}}return t}function x(e,t){return e.file===t.file&&e.startLine<=t.startLine&&t.endLine<=e.endLine}function S(e,t){return e.file===t.file&&e.startLine<=t.endLine&&t.startLine<=e.endLine}function C(e,t){return e<t?-1:+(e>t)}exports.checkThresholds=a,exports.qualifyFunctionName=d;
 //# sourceMappingURL=thresholdCheck.cjs.map

package/dist/thresholdCheck.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"thresholdCheck.cjs","names":["functionThresholds","metricNameOf","fileThresholds","levels","duplicationThreshold"],"sources":["../src/thresholdCheck.ts"],"sourcesContent":["import assert from 'node:assert';\nimport type { CrossFileDuplicateBlockGroup, CrossFileDuplicationMetrics } from './crossFileDuplication.js';\nimport type { LineHunk } from './git.js';\nimport {\n  duplicationThreshold,\n  fileThresholds,\n  functionThresholds,\n  levels,\n  metricNameOf,\n  type Level,\n  type LimitsByLevel,\n  type Threshold,\n} from './thresholds.js';\nimport type { CodeMetrics, CognitiveBlock, FunctionMetrics } from './types.js';\n\n/** A measured file to check, under the path cross-file duplication knows it by. */\nexport interface CheckedFile {\n  file: string;\n  metrics: CodeMetrics;\n  /** The change to the file; when set, only what the change touches is checked. */\n  hunks?: readonly LineHunk[];\n}\n\n/**\n * By file path, the 1-based lines that are neither blank nor comment-only, by which duplicated\n * blocks are measured and their copies ranked: of the checked files and of every other measured\n * file a copy may lie in.\n */\nexport type CodeLineNumbersByFile = ReadonlyMap<string, ReadonlySet<number>>;\n\nexport interface ExceededLimit {\n  metric: string;\n  value: number;\n  /** The most severe level whose limit the value violates. */\n  level: Level;\n  /** That level's limit: the value is allowed up to it; a duplicated block is allowed below it. */\n  limit: number;\n}\n\nexport interface BlockLocation {\n  file: string;\n  startLine: number;\n  endLine: number;\n}\n\nexport interface Violation extends BlockLocation {\n  kind: 'function' | 'file' | 'duplication';\n  /** The most severe level among `exceeded`. */\n  level: Level;\n  /** The function's name, qualified with the class it is a member of, for kind `function`. */\n  name?: string;\n  exceeded: ExceededLimit[];\n  /**\n   * The parts adding the most to the cognitive complexity, largest first, for a function that\n   * exceeds a limit of it and has such parts: operator sequences and jumps outside any branching\n   * construct add to the complexity without being one.\n   */\n  largestBlocks?: CognitiveBlock[];\n  /** The other copies of the block, most code lines first, for kind `duplication`. */\n  partners?: BlockLocation[];\n}\n\nexport interface CheckResult {\n  violations: Violation[];\n  checkedFunctionCount: number;\n}\n\nconst violationKinds: readonly Violation['kind'][] = ['file', 'function', 'duplication'];\n\nconst cognitiveComplexityMetric = 'functionCognitiveComplexity';\n\nconst nestedInclusiveMetrics = new Set(\n  functionThresholds\n    .filter(({ includesNestedFunctions }) => includesNestedFunctions)\n    .map((threshold) => metricNameOf(threshold))\n);\n\n/** Every limit the files exceed, errors first, then ordered by path, then line. */\nexport function checkThresholds(\n  files: readonly CheckedFile[],\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined,\n  limitsOf: (language: string) => LimitsByLevel,\n  codeLineNumbersByFile: CodeLineNumbersByFile\n): CheckResult {\n  const crossFileGroupsByFile = indexGroupsByFile(crossFileDuplication, new Set(files.map(({ file }) => file)));\n  const violations: Violation[] = [];\n  let checkedFunctionCount = 0;\n  for (const checkedFile of files) {\n    const { file, metrics, hunks } = checkedFile;\n    const limits = limitsOf(metrics.language);\n    // A rename or a mode change lists the file as changed without touching any of its lines.\n    const fileExceeded = hunks?.length === 0 ? [] : collectExceeded(fileThresholds, metrics, limits);\n    if (fileExceeded.length > 0) {\n      violations.push(toViolation({ file, startLine: 1, endLine: metrics.lines.total }, 'file', fileExceeded));\n    }\n    const checkedFunctions = listCheckedFunctions(checkedFile);\n    checkedFunctionCount += checkedFunctions.length;\n    const exceedingFunctions = checkedFunctions\n      .map((fn) => ({ fn, exceeded: collectExceeded(functionThresholds, fn, limits) }))\n      .filter(({ exceeded }) => exceeded.length > 0);\n    for (const { fn, exceeded } of exceedingFunctions) {\n      const ownExceeded = exceeded.filter((limit) => !isReportedByEnclosingFunction(limit, fn, exceedingFunctions));\n      if (ownExceeded.length > 0) {\n        violations.push(toFunctionViolation(file, fn, ownExceeded));\n      }\n    }\n    violations.push(\n      ...collectDuplicationViolations(checkedFile, crossFileGroupsByFile.get(file) ?? [], limits, codeLineNumbersByFile)\n    );\n  }\n  return { violations: violations.toSorted(compareViolations), checkedFunctionCount };\n}\n\n// The kind is compared before the end line because a file violation's end line is corrected after\n// this ordering (to the last line that exists), and the order must not depend on that correction.\nfunction compareViolations(left: Violation, right: Violation): number {\n  return (\n    levels.indexOf(right.level) - levels.indexOf(left.level) ||\n    compareStrings(left.file, right.file) ||\n    left.startLine - right.startLine ||\n    violationKinds.indexOf(left.kind) - violationKinds.indexOf(right.kind) ||\n    left.endLine - right.endLine\n  );\n}\n\n/** The file's functions or, for a changed file, those the change touches. */\nfunction listCheckedFunctions({ metrics, hunks }: CheckedFile): FunctionMetrics[] {\n  const endsByIndentation = metrics.language === 'python';\n  return hunks\n    ? metrics.functions.filter((fn) =>\n        hunks.some((hunk) => touchesSpan(hunk, fn.startLine, fn.endLine, endsByIndentation))\n      )\n    : metrics.functions;\n}\n\n/**\n * Whether a function enclosing `fn` exceeds the same limit with a value that covers `fn`, so\n * reporting it for `fn` too would count the same code twice. The enclosing value is at least that\n * of `fn` and both share the file's limits, so its level is never milder.\n */\nfunction isReportedByEnclosingFunction(\n  limit: ExceededLimit,\n  fn: FunctionMetrics,\n  exceedingFunctions: readonly { fn: FunctionMetrics; exceeded: ExceededLimit[] }[]\n): boolean {\n  if (!nestedInclusiveMetrics.has(limit.metric)) {\n    return false;\n  }\n  return exceedingFunctions.some(\n    (outer) => outer.fn !== fn && encloses(outer.fn, fn) && outer.exceeded.some(({ metric }) => metric === limit.metric)\n  );\n}\n\nfunction encloses(outer: FunctionMetrics, inner: FunctionMetrics): boolean {\n  const startsBefore =\n    outer.startLine < inner.startLine ||\n    (outer.startLine === inner.startLine && outer.startColumn <= inner.startColumn);\n  const endsAfter =\n    outer.endLine > inner.endLine || (outer.endLine === inner.endLine && outer.endColumn >= inner.endColumn);\n  return startsBefore && endsAfter;\n}\n\nfunction toFunctionViolation(file: string, fn: FunctionMetrics, exceeded: ExceededLimit[]): Violation {\n  const { startLine, endLine } = fn;\n  const exceedsCognitiveComplexity = exceeded.some(({ metric }) => metric === cognitiveComplexityMetric);\n  return {\n    ...toViolation({ file, startLine, endLine }, 'function', exceeded),\n    name: qualifyFunctionName(fn),\n    ...(exceedsCognitiveComplexity && fn.cognitiveBlocks.length > 0 && { largestBlocks: fn.cognitiveBlocks }),\n  };\n}\n\n/** How reports name a function: with the type or module it is a member of, so that it can be found. */\nexport function qualifyFunctionName({ name = '<anonymous>', containerName }: FunctionMetrics): string {\n  return containerName ? `${containerName}.${name}` : name;\n}\n\nfunction toViolation(location: BlockLocation, kind: Violation['kind'], exceeded: ExceededLimit[]): Violation {\n  const level = exceeded.some((limit) => limit.level === 'error') ? 'error' : 'warning';\n  return { kind, level, ...location, exceeded };\n}\n\nfunction collectExceeded<Subject>(\n  thresholds: readonly Threshold<Subject>[],\n  subject: Subject,\n  limits: LimitsByLevel\n): ExceededLimit[] {\n  return thresholds.flatMap((threshold) => findExceeded(threshold, subject, limits, (value, limit) => value > limit));\n}\n\n/** The most severe violated limit of the threshold, if any. */\nfunction findExceeded<Subject>(\n  threshold: Threshold<Subject>,\n  subject: Subject,\n  limits: LimitsByLevel,\n  violates: (value: number, limit: number) => boolean\n): ExceededLimit[] {\n  const value = threshold.measure(subject);\n  const level = levels.findLast((candidate) => violates(value, limits[candidate][threshold.key] as number));\n  return level\n    ? [{ metric: metricNameOf(threshold), value, level, limit: limits[level][threshold.key] as number }]\n    : [];\n}\n\n/**\n * Whether the hunk changes the head lines [startLine, endLine]. A hunk without head lines (a pure\n * deletion) sits between its headStart line and the next one. Where indentation ends a span, its\n * last line is a statement rather than a closing delimiter, so a deletion right after that line may\n * have removed the span's tail; `endsByIndentation` counts it as touching, since the head alone\n * cannot tell it from a deletion after the span.\n */\nfunction touchesSpan(hunk: LineHunk, startLine: number, endLine: number, endsByIndentation = false): boolean {\n  return hunk.headCount === 0\n    ? hunk.headStart >= startLine && (hunk.headStart < endLine || (endsByIndentation && hunk.headStart === endLine))\n    : hunk.headStart <= endLine && hunk.headStart + hunk.headCount > startLine;\n}\n\n/** The cross-file groups with an occurrence in each of the given files. */\nfunction indexGroupsByFile(\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined,\n  files: ReadonlySet<string>\n): Map<string, CrossFileDuplicateBlockGroup[]> {\n  const groupsByFile = new Map<string, CrossFileDuplicateBlockGroup[]>();\n  for (const group of crossFileDuplication?.groups ?? []) {\n    for (const file of group.files) {\n      if (files.has(file)) {\n        const groups = groupsByFile.get(file) ?? [];\n        groups.push(group);\n        groupsByFile.set(file, groups);\n      }\n    }\n  }\n  return groupsByFile;\n}\n\n/**\n * One violation per duplicated region of the file: its clone occurrences, within-file and\n * cross-file, that hold at least the code lines of a level (with `hunks`, only those overlapping\n * added lines), with overlapping occurrences merged into one region.\n */\nfunction collectDuplicationViolations(\n  { file, metrics, hunks }: CheckedFile,\n  crossFileGroups: readonly CrossFileDuplicateBlockGroup[],\n  limits: LimitsByLevel,\n  codeLineNumbersByFile: CodeLineNumbersByFile\n): Violation[] {\n  const findExceededLines = (block: BlockLocation): ExceededLimit[] =>\n    findExceeded(\n      duplicationThreshold,\n      { codeLineCount: countCodeLines(block, codeLineNumbersByFile) },\n      limits,\n      (value, limit) => value >= limit\n    );\n  const groups: BlockLocation[][] = [\n    ...metrics.duplication.duplicateBlockGroups.map((group) => group.map((occurrence) => ({ ...occurrence, file }))),\n    ...crossFileGroups.map((group) => group.occurrences),\n  ];\n  const isReported = (block: BlockLocation): boolean =>\n    block.file === file &&\n    findExceededLines(block).length > 0 &&\n    (!hunks || hunks.some((hunk) => hunk.headCount > 0 && touchesSpan(hunk, block.startLine, block.endLine)));\n  const copiesByBlock = new Map<BlockLocation, BlockLocation[]>();\n  for (const group of groups) {\n    for (const block of group.filter((occurrence) => isReported(occurrence))) {\n      copiesByBlock.set(block, group);\n    }\n  }\n  return mergeOverlapping([...copiesByBlock.keys()]).map(({ merged, sources }) => ({\n    ...toViolation(merged, 'duplication', findExceededLines(merged)),\n    // The region's own occurrences are among the copies of the groups it belongs to.\n    // The region merges every clone that overlaps it, so its copies range from whole copies of it\n    // to a few lines matching one fragment; the largest say the most about what to share.\n    partners: mergeOverlapping(sources.flatMap((source) => copiesByBlock.get(source) ?? []))\n      .map((partner) => partner.merged)\n      .filter((partner) => !overlaps(partner, merged))\n      .toSorted(\n        (left, right) => countCodeLines(right, codeLineNumbersByFile) - countCodeLines(left, codeLineNumbersByFile)\n      ),\n  }));\n}\n\n/** The locations with those overlapping in the same file merged, ordered by file, then line. */\nfunction mergeOverlapping(locations: readonly BlockLocation[]): { merged: BlockLocation; sources: BlockLocation[] }[] {\n  const regions: { merged: BlockLocation; sources: BlockLocation[] }[] = [];\n  const ordered = locations.toSorted(\n    (left, right) => compareStrings(left.file, right.file) || left.startLine - right.startLine\n  );\n  for (const location of ordered) {\n    const last = regions.at(-1);\n    if (last && overlaps(last.merged, location)) {\n      last.merged.endLine = Math.max(last.merged.endLine, location.endLine);\n      last.sources.push(location);\n    } else {\n      const { file, startLine, endLine } = location;\n      regions.push({ merged: { file, startLine, endLine }, sources: [location] });\n    }\n  }\n  return regions;\n}\n\n/** The length of a duplicated block. */\nfunction countCodeLines(block: BlockLocation, codeLineNumbersByFile: CodeLineNumbersByFile): number {\n  const codeLineNumbers = codeLineNumbersByFile.get(block.file);\n  assert.ok(codeLineNumbers, `No code lines for ${block.file}, which holds a duplicated block.`);\n  let count = 0;\n  for (let line = block.startLine; line <= block.endLine; line++) {\n    if (codeLineNumbers.has(line)) count += 1;\n  }\n  return count;\n}\n\nfunction overlaps(left: BlockLocation, right: BlockLocation): boolean {\n  return left.file === right.file && left.startLine <= right.endLine && right.startLine <= left.endLine;\n}\n\n/** Code-unit order, so reports do not depend on the locale. */\nfunction compareStrings(left: string, right: string): number {\n  return left < right ? -1 : left > right ? 1 : 0;\n}\n"],"mappings":"6IAmEA,MAAM,EAA+C,CAAC,OAAQ,WAAY,aAAa,EAIjF,EAAyB,IAAI,IACjCA,EAAAA,mBACG,QAAQ,CAAE,6BAA8B,CAAuB,CAAC,CAChE,IAAK,GAAcC,EAAAA,aAAa,CAAS,CAAC,CAC/C,EAGA,SAAgB,EACd,EACA,EACA,EACA,EACa,CACb,IAAM,EAAwB,EAAkB,EAAsB,IAAI,IAAI,EAAM,KAAK,CAAE,UAAW,CAAI,CAAC,CAAC,EACtG,EAA0B,CAAC,EAC7B,EAAuB,EAC3B,IAAK,IAAM,KAAe,EAAO,CAC/B,GAAM,CAAE,OAAM,UAAS,SAAU,EAC3B,EAAS,EAAS,EAAQ,QAAQ,EAElC,EAAe,GAAO,SAAW,EAAI,CAAC,EAAI,EAAgBC,EAAAA,eAAgB,EAAS,CAAM,EAC3F,EAAa,OAAS,GACxB,EAAW,KAAK,EAAY,CAAE,OAAM,UAAW,EAAG,QAAS,EAAQ,MAAM,KAAM,EAAG,OAAQ,CAAY,CAAC,EAEzG,IAAM,EAAmB,EAAqB,CAAW,EACzD,GAAwB,EAAiB,OACzC,IAAM,EAAqB,EACxB,IAAK,IAAQ,CAAE,KAAI,SAAU,EAAgBF,EAAAA,mBAAoB,EAAI,CAAM,CAAE,EAAE,CAAC,CAChF,QAAQ,CAAE,cAAe,EAAS,OAAS,CAAC,EAC/C,IAAK,GAAM,CAAE,KAAI,cAAc,EAAoB,CACjD,IAAM,EAAc,EAAS,OAAQ,GAAU,CAAC,EAA8B,EAAO,EAAI,CAAkB,CAAC,EACxG,EAAY,OAAS,GACvB,EAAW,KAAK,EAAoB,EAAM,EAAI,CAAW,CAAC,CAE9D,CACA,EAAW,KACT,GAAG,EAA6B,EAAa,EAAsB,IAAI,CAAI,GAAK,CAAC,EAAG,EAAQ,CAAqB,CACnH,CACF,CACA,MAAO,CAAE,WAAY,EAAW,SAAS,CAAiB,EAAG,sBAAqB,CACpF,CAIA,SAAS,EAAkB,EAAiB,EAA0B,CACpE,OACEG,EAAAA,OAAO,QAAQ,EAAM,KAAK,EAAIA,EAAAA,OAAO,QAAQ,EAAK,KAAK,GACvD,EAAe,EAAK,KAAM,EAAM,IAAI,GACpC,EAAK,UAAY,EAAM,WACvB,EAAe,QAAQ,EAAK,IAAI,EAAI,EAAe,QAAQ,EAAM,IAAI,GACrE,EAAK,QAAU,EAAM,OAEzB,CAGA,SAAS,EAAqB,CAAE,UAAS,SAAyC,CAChF,IAAM,EAAoB,EAAQ,WAAa,SAC/C,OAAO,EACH,EAAQ,UAAU,OAAQ,GACxB,EAAM,KAAM,GAAS,EAAY,EAAM,EAAG,UAAW,EAAG,QAAS,CAAiB,CAAC,CACrF,EACA,EAAQ,SACd,CAOA,SAAS,EACP,EACA,EACA,EACS,CAIT,OAHK,EAAuB,IAAI,EAAM,MAAM,EAGrC,EAAmB,KACvB,GAAU,EAAM,KAAO,GAAM,EAAS,EAAM,GAAI,CAAE,GAAK,EAAM,SAAS,MAAM,CAAE,YAAa,IAAW,EAAM,MAAM,CACrH,EAJS,EAKX,CAEA,SAAS,EAAS,EAAwB,EAAiC,CACzE,IAAM,EACJ,EAAM,UAAY,EAAM,WACvB,EAAM,YAAc,EAAM,WAAa,EAAM,aAAe,EAAM,YAC/D,EACJ,EAAM,QAAU,EAAM,SAAY,EAAM,UAAY,EAAM,SAAW,EAAM,WAAa,EAAM,UAChG,OAAO,GAAgB,CACzB,CAEA,SAAS,EAAoB,EAAc,EAAqB,EAAsC,CACpG,GAAM,CAAE,YAAW,WAAY,EACzB,EAA6B,EAAS,MAAM,CAAE,YAAa,IAAW,6BAAyB,EACrG,MAAO,CACL,GAAG,EAAY,CAAE,OAAM,YAAW,SAAQ,EAAG,WAAY,CAAQ,EACjE,KAAM,EAAoB,CAAE,EAC5B,GAAI,GAA8B,EAAG,gBAAgB,OAAS,GAAK,CAAE,cAAe,EAAG,eAAgB,CACzG,CACF,CAGA,SAAgB,EAAoB,CAAE,OAAO,cAAe,iBAA0C,CACpG,OAAO,EAAgB,GAAG,EAAc,GAAG,IAAS,CACtD,CAEA,SAAS,EAAY,EAAyB,EAAyB,EAAsC,CAE3G,MAAO,CAAE,OAAM,MADD,EAAS,KAAM,GAAU,EAAM,QAAU,OAAO,EAAI,QAAU,UACtD,GAAG,EAAU,UAAS,CAC9C,CAEA,SAAS,EACP,EACA,EACA,EACiB,CACjB,OAAO,EAAW,QAAS,GAAc,EAAa,EAAW,EAAS,GAAS,EAAO,IAAU,EAAQ,CAAK,CAAC,CACpH,CAGA,SAAS,EACP,EACA,EACA,EACA,EACiB,CACjB,IAAM,EAAQ,EAAU,QAAQ,CAAO,EACjC,EAAQA,EAAAA,OAAO,SAAU,GAAc,EAAS,EAAO,EAAO,EAAU,CAAC,EAAU,IAAc,CAAC,EACxG,OAAO,EACH,CAAC,CAAE,OAAQF,EAAAA,aAAa,CAAS,EAAG,QAAO,QAAO,MAAO,EAAO,EAAM,CAAC,EAAU,IAAe,CAAC,EACjG,CAAC,CACP,CASA,SAAS,EAAY,EAAgB,EAAmB,EAAiB,EAAoB,GAAgB,CAC3G,OAAO,EAAK,YAAc,EACtB,EAAK,WAAa,IAAc,EAAK,UAAY,GAAY,GAAqB,EAAK,YAAc,GACrG,EAAK,WAAa,GAAW,EAAK,UAAY,EAAK,UAAY,CACrE,CAGA,SAAS,EACP,EACA,EAC6C,CAC7C,IAAM,kBAAe,IAAI,IACzB,IAAK,IAAM,KAAS,GAAsB,QAAU,CAAC,EACnD,IAAK,IAAM,KAAQ,EAAM,MACvB,GAAI,EAAM,IAAI,CAAI,EAAG,CACnB,IAAM,EAAS,EAAa,IAAI,CAAI,GAAK,CAAC,EAC1C,EAAO,KAAK,CAAK,EACjB,EAAa,IAAI,EAAM,CAAM,CAC/B,CAGJ,OAAO,CACT,CAOA,SAAS,EACP,CAAE,OAAM,UAAS,SACjB,EACA,EACA,EACa,CACb,IAAM,EAAqB,GACzB,EACEG,EAAAA,qBACA,CAAE,cAAe,EAAe,EAAO,CAAqB,CAAE,EAC9D,GACC,EAAO,IAAU,GAAS,CAC7B,EACI,EAA4B,CAChC,GAAG,EAAQ,YAAY,qBAAqB,IAAK,GAAU,EAAM,IAAK,IAAgB,CAAE,GAAG,EAAY,MAAK,EAAE,CAAC,EAC/G,GAAG,EAAgB,IAAK,GAAU,EAAM,WAAW,CACrD,EACM,EAAc,GAClB,EAAM,OAAS,GACf,EAAkB,CAAK,CAAC,CAAC,OAAS,IACjC,CAAC,GAAS,EAAM,KAAM,GAAS,EAAK,UAAY,GAAK,EAAY,EAAM,EAAM,UAAW,EAAM,OAAO,CAAC,GACnG,kBAAgB,IAAI,IAC1B,IAAK,IAAM,KAAS,EAClB,IAAK,IAAM,KAAS,EAAM,OAAQ,GAAe,EAAW,CAAU,CAAC,EACrE,EAAc,IAAI,EAAO,CAAK,EAGlC,OAAO,EAAiB,CAAC,GAAG,EAAc,KAAK,CAAC,CAAC,CAAC,CAAC,KAAK,CAAE,SAAQ,cAAe,CAC/E,GAAG,EAAY,EAAQ,cAAe,EAAkB,CAAM,CAAC,EAI/D,SAAU,EAAiB,EAAQ,QAAS,GAAW,EAAc,IAAI,CAAM,GAAK,CAAC,CAAC,CAAC,CAAC,CACrF,IAAK,GAAY,EAAQ,MAAM,CAAC,CAChC,OAAQ,GAAY,CAAC,EAAS,EAAS,CAAM,CAAC,CAAC,CAC/C,UACE,EAAM,IAAU,EAAe,EAAO,CAAqB,EAAI,EAAe,EAAM,CAAqB,CAC5G,CACJ,EAAE,CACJ,CAGA,SAAS,EAAiB,EAA4F,CACpH,IAAM,EAAiE,CAAC,EAClE,EAAU,EAAU,UACvB,EAAM,IAAU,EAAe,EAAK,KAAM,EAAM,IAAI,GAAK,EAAK,UAAY,EAAM,SACnF,EACA,IAAK,IAAM,KAAY,EAAS,CAC9B,IAAM,EAAO,EAAQ,GAAG,EAAE,EAC1B,GAAI,GAAQ,EAAS,EAAK,OAAQ,CAAQ,EACxC,EAAK,OAAO,QAAU,KAAK,IAAI,EAAK,OAAO,QAAS,EAAS,OAAO,EACpE,EAAK,QAAQ,KAAK,CAAQ,MACrB,CACL,GAAM,CAAE,OAAM,YAAW,WAAY,EACrC,EAAQ,KAAK,CAAE,OAAQ,CAAE,OAAM,YAAW,SAAQ,EAAG,QAAS,CAAC,CAAQ,CAAE,CAAC,CAC5E,CACF,CACA,OAAO,CACT,CAGA,SAAS,EAAe,EAAsB,EAAsD,CAClG,IAAM,EAAkB,EAAsB,IAAI,EAAM,IAAI,EAC5D,EAAA,QAAO,GAAG,EAAiB,qBAAqB,EAAM,KAAK,kCAAkC,EAC7F,IAAI,EAAQ,EACZ,IAAK,IAAI,EAAO,EAAM,UAAW,GAAQ,EAAM,QAAS,IAClD,EAAgB,IAAI,CAAI,IAAG,GAAS,GAE1C,OAAO,CACT,CAEA,SAAS,EAAS,EAAqB,EAA+B,CACpE,OAAO,EAAK,OAAS,EAAM,MAAQ,EAAK,WAAa,EAAM,SAAW,EAAM,WAAa,EAAK,OAChG,CAGA,SAAS,EAAe,EAAc,EAAuB,CAC3D,OAAO,EAAO,EAAQ,GAAK,IAAO,EACpC"}
+{"version":3,"file":"thresholdCheck.cjs","names":["functionThresholds","metricNameOf","fileThresholds","levels","duplicationThreshold"],"sources":["../src/thresholdCheck.ts"],"sourcesContent":["import assert from 'node:assert';\nimport type {\n  CrossFileDuplicateBlockGroup,\n  CrossFileDuplicateOccurrence,\n  CrossFileDuplicationMetrics,\n} from './crossFileDuplication.js';\nimport type { LineHunk } from './git.js';\nimport {\n  duplicationThreshold,\n  fileThresholds,\n  functionThresholds,\n  levels,\n  metricNameOf,\n  type Level,\n  type LimitsByLevel,\n  type Threshold,\n} from './thresholds.js';\nimport type { CodeMetrics, CognitiveBlock, FunctionMetrics } from './types.js';\n\n/** A measured file to check, under the path cross-file duplication knows it by. */\nexport interface CheckedFile {\n  file: string;\n  metrics: CodeMetrics;\n  /** The change to the file; when set, only what the change touches is checked. */\n  hunks?: readonly LineHunk[];\n}\n\n/** The duplication of the measured files, in which duplicated blocks and their copies are found. */\nexport interface CheckedDuplication {\n  crossFile?: CrossFileDuplicationMetrics;\n  /** The detection's `minSimilarityPercent`, the share of a block a copy of it holds at least. */\n  minSimilarityPercent: number;\n}\n\nexport interface ExceededLimit {\n  metric: string;\n  value: number;\n  /** The most severe level whose limit the value violates. */\n  level: Level;\n  /** That level's limit: the value is allowed up to it; a duplicated block is allowed below it. */\n  limit: number;\n}\n\nexport interface BlockLocation {\n  file: string;\n  startLine: number;\n  endLine: number;\n}\n\n/** One copy of a clone group. */\ninterface CloneOccurrence extends BlockLocation {\n  /**\n   * Its code lines matched in another copy, whose number is the length of a duplicated block. Its\n   * line span would also count the lines between the matched parts of one copy, which the copies\n   * do not share; a block matched as a whole by a near-miss copy holds all its code lines.\n   */\n  lineNumbers: readonly number[];\n  tokenCount: number;\n}\n\nexport interface Violation extends BlockLocation {\n  kind: 'function' | 'file' | 'duplication';\n  /** The most severe level among `exceeded`. */\n  level: Level;\n  /** The function's name, qualified with the class it is a member of, for kind `function`. */\n  name?: string;\n  exceeded: ExceededLimit[];\n  /**\n   * The parts adding the most to the cognitive complexity, largest first, for a function that\n   * exceeds a limit of it and has such parts: operator sequences and jumps outside any branching\n   * construct add to the complexity without being one.\n   */\n  largestBlocks?: CognitiveBlock[];\n  /** The other copies of the block, most matched tokens first, for kind `duplication`. */\n  partners?: BlockLocation[];\n}\n\nexport interface CheckResult {\n  violations: Violation[];\n  checkedFunctionCount: number;\n}\n\nconst violationKinds: readonly Violation['kind'][] = ['file', 'function', 'duplication'];\n\nconst cognitiveComplexityMetric = 'functionCognitiveComplexity';\n\nconst nestedInclusiveMetrics = new Set(\n  functionThresholds\n    .filter(({ includesNestedFunctions }) => includesNestedFunctions)\n    .map((threshold) => metricNameOf(threshold))\n);\n\n/** Every limit the files exceed, errors first, then ordered by path, then line. */\nexport function checkThresholds(\n  files: readonly CheckedFile[],\n  duplication: CheckedDuplication,\n  limitsOf: (language: string) => LimitsByLevel\n): CheckResult {\n  const crossFileGroupsByFile = indexGroupsByFile(duplication.crossFile, new Set(files.map(({ file }) => file)));\n  const violations: Violation[] = [];\n  let checkedFunctionCount = 0;\n  for (const checkedFile of files) {\n    const { file, metrics, hunks } = checkedFile;\n    const limits = limitsOf(metrics.language);\n    // A rename or a mode change lists the file as changed without touching any of its lines.\n    const fileExceeded = hunks?.length === 0 ? [] : collectExceeded(fileThresholds, metrics, limits);\n    if (fileExceeded.length > 0) {\n      violations.push(toViolation({ file, startLine: 1, endLine: metrics.lines.total }, 'file', fileExceeded));\n    }\n    const checkedFunctions = listCheckedFunctions(checkedFile);\n    checkedFunctionCount += checkedFunctions.length;\n    const exceedingFunctions = checkedFunctions\n      .map((fn) => ({ fn, exceeded: collectExceeded(functionThresholds, fn, limits) }))\n      .filter(({ exceeded }) => exceeded.length > 0);\n    for (const { fn, exceeded } of exceedingFunctions) {\n      const ownExceeded = exceeded.filter((limit) => !isReportedByEnclosingFunction(limit, fn, exceedingFunctions));\n      if (ownExceeded.length > 0) {\n        violations.push(toFunctionViolation(file, fn, ownExceeded));\n      }\n    }\n    violations.push(\n      ...collectDuplicationViolations(checkedFile, crossFileGroupsByFile.get(file) ?? [], limits, duplication)\n    );\n  }\n  return { violations: violations.toSorted(compareViolations), checkedFunctionCount };\n}\n\n// The kind is compared before the end line because a file violation's end line is corrected after\n// this ordering (to the last line that exists), and the order must not depend on that correction.\nfunction compareViolations(left: Violation, right: Violation): number {\n  return (\n    levels.indexOf(right.level) - levels.indexOf(left.level) ||\n    compareStrings(left.file, right.file) ||\n    left.startLine - right.startLine ||\n    violationKinds.indexOf(left.kind) - violationKinds.indexOf(right.kind) ||\n    left.endLine - right.endLine\n  );\n}\n\n/** The file's functions or, for a changed file, those the change touches. */\nfunction listCheckedFunctions({ metrics, hunks }: CheckedFile): FunctionMetrics[] {\n  const endsByIndentation = metrics.language === 'python';\n  return hunks\n    ? metrics.functions.filter((fn) =>\n        hunks.some((hunk) => touchesSpan(hunk, fn.startLine, fn.endLine, endsByIndentation))\n      )\n    : metrics.functions;\n}\n\n/**\n * Whether a function enclosing `fn` exceeds the same limit with a value that covers `fn`, so\n * reporting it for `fn` too would count the same code twice. The enclosing value is at least that\n * of `fn` and both share the file's limits, so its level is never milder.\n */\nfunction isReportedByEnclosingFunction(\n  limit: ExceededLimit,\n  fn: FunctionMetrics,\n  exceedingFunctions: readonly { fn: FunctionMetrics; exceeded: ExceededLimit[] }[]\n): boolean {\n  if (!nestedInclusiveMetrics.has(limit.metric)) {\n    return false;\n  }\n  return exceedingFunctions.some(\n    (outer) => outer.fn !== fn && encloses(outer.fn, fn) && outer.exceeded.some(({ metric }) => metric === limit.metric)\n  );\n}\n\nfunction encloses(outer: FunctionMetrics, inner: FunctionMetrics): boolean {\n  const startsBefore =\n    outer.startLine < inner.startLine ||\n    (outer.startLine === inner.startLine && outer.startColumn <= inner.startColumn);\n  const endsAfter =\n    outer.endLine > inner.endLine || (outer.endLine === inner.endLine && outer.endColumn >= inner.endColumn);\n  return startsBefore && endsAfter;\n}\n\nfunction toFunctionViolation(file: string, fn: FunctionMetrics, exceeded: ExceededLimit[]): Violation {\n  const { startLine, endLine } = fn;\n  const exceedsCognitiveComplexity = exceeded.some(({ metric }) => metric === cognitiveComplexityMetric);\n  return {\n    ...toViolation({ file, startLine, endLine }, 'function', exceeded),\n    name: qualifyFunctionName(fn),\n    ...(exceedsCognitiveComplexity && fn.cognitiveBlocks.length > 0 && { largestBlocks: fn.cognitiveBlocks }),\n  };\n}\n\n/** How reports name a function: with the type or module it is a member of, so that it can be found. */\nexport function qualifyFunctionName({ name = '<anonymous>', containerName }: FunctionMetrics): string {\n  return containerName ? `${containerName}.${name}` : name;\n}\n\nfunction toViolation(location: BlockLocation, kind: Violation['kind'], exceeded: ExceededLimit[]): Violation {\n  const level = exceeded.some((limit) => limit.level === 'error') ? 'error' : 'warning';\n  return { kind, level, ...location, exceeded };\n}\n\nfunction collectExceeded<Subject>(\n  thresholds: readonly Threshold<Subject>[],\n  subject: Subject,\n  limits: LimitsByLevel\n): ExceededLimit[] {\n  return thresholds.flatMap((threshold) => findExceeded(threshold, subject, limits, (value, limit) => value > limit));\n}\n\n/** The most severe violated limit of the threshold, if any. */\nfunction findExceeded<Subject>(\n  threshold: Threshold<Subject>,\n  subject: Subject,\n  limits: LimitsByLevel,\n  violates: (value: number, limit: number) => boolean\n): ExceededLimit[] {\n  const value = threshold.measure(subject);\n  const level = levels.findLast((candidate) => violates(value, limits[candidate][threshold.key] as number));\n  return level\n    ? [{ metric: metricNameOf(threshold), value, level, limit: limits[level][threshold.key] as number }]\n    : [];\n}\n\n/**\n * Whether the hunk changes the head lines [startLine, endLine]. A hunk without head lines (a pure\n * deletion) sits between its headStart line and the next one. Where indentation ends a span, its\n * last line is a statement rather than a closing delimiter, so a deletion right after that line may\n * have removed the span's tail; `endsByIndentation` counts it as touching, since the head alone\n * cannot tell it from a deletion after the span.\n */\nfunction touchesSpan(hunk: LineHunk, startLine: number, endLine: number, endsByIndentation = false): boolean {\n  return hunk.headCount === 0\n    ? hunk.headStart >= startLine && (hunk.headStart < endLine || (endsByIndentation && hunk.headStart === endLine))\n    : hunk.headStart <= endLine && hunk.headStart + hunk.headCount > startLine;\n}\n\n/** The cross-file groups with an occurrence in each of the given files. */\nfunction indexGroupsByFile(\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined,\n  files: ReadonlySet<string>\n): Map<string, CrossFileDuplicateBlockGroup[]> {\n  const groupsByFile = new Map<string, CrossFileDuplicateBlockGroup[]>();\n  for (const group of crossFileDuplication?.groups ?? []) {\n    for (const file of group.files) {\n      if (files.has(file)) {\n        const groups = groupsByFile.get(file) ?? [];\n        groups.push(group);\n        groupsByFile.set(file, groups);\n      }\n    }\n  }\n  return groupsByFile;\n}\n\n/**\n * One violation per duplicated region of the file: its clone occurrences, within-file and\n * cross-file, that hold at least the duplicated lines of a level (with `hunks`, only those\n * overlapping added lines), with overlapping occurrences merged into one region.\n */\nfunction collectDuplicationViolations(\n  { file, metrics, hunks }: CheckedFile,\n  crossFileGroups: readonly CrossFileDuplicateBlockGroup[],\n  limits: LimitsByLevel,\n  { minSimilarityPercent }: CheckedDuplication\n): Violation[] {\n  const findExceededLines = (duplicatedLineCount: number): ExceededLimit[] =>\n    findExceeded(duplicationThreshold, { duplicatedLineCount }, limits, (value, limit) => value >= limit);\n  const groups: CloneOccurrence[][] = [\n    ...metrics.duplication.duplicateBlockGroups.map((group) => group.map((occurrence) => ({ ...occurrence, file }))),\n    ...crossFileGroups.map((group) => group.occurrences.map((occurrence) => toCloneOccurrence(occurrence))),\n  ];\n  const isReported = (block: CloneOccurrence): boolean =>\n    block.file === file &&\n    findExceededLines(block.lineNumbers.length).length > 0 &&\n    (!hunks || hunks.some((hunk) => hunk.headCount > 0 && touchesSpan(hunk, block.startLine, block.endLine)));\n  const groupByBlock = new Map<CloneOccurrence, CloneOccurrence[]>();\n  for (const group of groups) {\n    for (const block of group.filter((occurrence) => isReported(occurrence))) {\n      groupByBlock.set(block, group);\n    }\n  }\n  return mergeOverlapping([...groupByBlock.keys()]).map(({ merged, sources }) => {\n    const listCopies = (minTokenPercent: number): BlockLocation[] =>\n      mergeOverlapping(\n        sources.flatMap((source) => {\n          // The region merges every clone that overlaps it, so a copy of an occurrence lying\n          // within a larger one is measured against the larger one.\n          const tokenCount = Math.max(\n            ...sources.filter((outer) => contains(outer, source)).map((outer) => outer.tokenCount)\n          );\n          return (groupByBlock.get(source) ?? []).filter(\n            (copy) => copy.tokenCount * 100 >= tokenCount * minTokenPercent\n          );\n        })\n      )\n        // The region's own occurrences are among the copies of the groups it belongs to.\n        .filter((partner) => !overlaps(partner.merged, merged))\n        // The largest say the most about what to share.\n        .toSorted((left, right) => sumTokens(right.sources) - sumTokens(left.sources))\n        .map((partner) => partner.merged);\n    // A group joins a block with every place repeating a part of it: a function another file\n    // copies whole is grouped with each file holding a few of its lines. Such a fragment is not a\n    // copy of the block: sharing it would not remove the block, and a report listing it reads as\n    // one more copy. Sizes are compared in tokens, since the same code may be laid out over any\n    // number of lines.\n    const copies = listCopies(minSimilarityPercent);\n    return {\n      ...toViolation(\n        merged,\n        'duplication',\n        // Occurrences that overlap share lines, which count once.\n        findExceededLines(new Set(sources.flatMap((source) => source.lineNumbers)).size)\n      ),\n      // A group does not always hold the copy a block was matched with as a whole: within a file,\n      // a copy that already belongs to an exact group is listed as its part in that group. The\n      // fragments then are all that locates the block's copies.\n      partners: copies.length > 0 ? copies : listCopies(0),\n    };\n  });\n}\n\nfunction toCloneOccurrence(occurrence: CrossFileDuplicateOccurrence): CloneOccurrence {\n  const { lineNumbers } = occurrence;\n  assert.ok(lineNumbers, `No matched lines for ${occurrence.file}, a scanned file.`);\n  return { ...occurrence, lineNumbers };\n}\n\nfunction sumTokens(occurrences: readonly CloneOccurrence[]): number {\n  return occurrences.reduce((sum, occurrence) => sum + occurrence.tokenCount, 0);\n}\n\n/** The locations with those overlapping in the same file merged, ordered by file, then line. */\nfunction mergeOverlapping(\n  locations: readonly CloneOccurrence[]\n): { merged: BlockLocation; sources: CloneOccurrence[] }[] {\n  const regions: { merged: BlockLocation; sources: CloneOccurrence[] }[] = [];\n  const ordered = locations.toSorted(\n    (left, right) => compareStrings(left.file, right.file) || left.startLine - right.startLine\n  );\n  for (const location of ordered) {\n    const last = regions.at(-1);\n    if (last && overlaps(last.merged, location)) {\n      last.merged.endLine = Math.max(last.merged.endLine, location.endLine);\n      last.sources.push(location);\n    } else {\n      const { file, startLine, endLine } = location;\n      regions.push({ merged: { file, startLine, endLine }, sources: [location] });\n    }\n  }\n  return regions;\n}\n\n/** Whether `outer` spans every line of `inner`; a block contains itself. */\nfunction contains(outer: BlockLocation, inner: BlockLocation): boolean {\n  return outer.file === inner.file && outer.startLine <= inner.startLine && inner.endLine <= outer.endLine;\n}\n\nfunction overlaps(left: BlockLocation, right: BlockLocation): boolean {\n  return left.file === right.file && left.startLine <= right.endLine && right.startLine <= left.endLine;\n}\n\n/** Code-unit order, so reports do not depend on the locale. */\nfunction compareStrings(left: string, right: string): number {\n  return left < right ? -1 : left > right ? 1 : 0;\n}\n"],"mappings":"6IAkFA,MAAM,EAA+C,CAAC,OAAQ,WAAY,aAAa,EAIjF,EAAyB,IAAI,IACjCA,EAAAA,mBACG,QAAQ,CAAE,6BAA8B,CAAuB,CAAC,CAChE,IAAK,GAAcC,EAAAA,aAAa,CAAS,CAAC,CAC/C,EAGA,SAAgB,EACd,EACA,EACA,EACa,CACb,IAAM,EAAwB,EAAkB,EAAY,UAAW,IAAI,IAAI,EAAM,KAAK,CAAE,UAAW,CAAI,CAAC,CAAC,EACvG,EAA0B,CAAC,EAC7B,EAAuB,EAC3B,IAAK,IAAM,KAAe,EAAO,CAC/B,GAAM,CAAE,OAAM,UAAS,SAAU,EAC3B,EAAS,EAAS,EAAQ,QAAQ,EAElC,EAAe,GAAO,SAAW,EAAI,CAAC,EAAI,EAAgBC,EAAAA,eAAgB,EAAS,CAAM,EAC3F,EAAa,OAAS,GACxB,EAAW,KAAK,EAAY,CAAE,OAAM,UAAW,EAAG,QAAS,EAAQ,MAAM,KAAM,EAAG,OAAQ,CAAY,CAAC,EAEzG,IAAM,EAAmB,EAAqB,CAAW,EACzD,GAAwB,EAAiB,OACzC,IAAM,EAAqB,EACxB,IAAK,IAAQ,CAAE,KAAI,SAAU,EAAgBF,EAAAA,mBAAoB,EAAI,CAAM,CAAE,EAAE,CAAC,CAChF,QAAQ,CAAE,cAAe,EAAS,OAAS,CAAC,EAC/C,IAAK,GAAM,CAAE,KAAI,cAAc,EAAoB,CACjD,IAAM,EAAc,EAAS,OAAQ,GAAU,CAAC,EAA8B,EAAO,EAAI,CAAkB,CAAC,EACxG,EAAY,OAAS,GACvB,EAAW,KAAK,EAAoB,EAAM,EAAI,CAAW,CAAC,CAE9D,CACA,EAAW,KACT,GAAG,EAA6B,EAAa,EAAsB,IAAI,CAAI,GAAK,CAAC,EAAG,EAAQ,CAAW,CACzG,CACF,CACA,MAAO,CAAE,WAAY,EAAW,SAAS,CAAiB,EAAG,sBAAqB,CACpF,CAIA,SAAS,EAAkB,EAAiB,EAA0B,CACpE,OACEG,EAAAA,OAAO,QAAQ,EAAM,KAAK,EAAIA,EAAAA,OAAO,QAAQ,EAAK,KAAK,GACvD,EAAe,EAAK,KAAM,EAAM,IAAI,GACpC,EAAK,UAAY,EAAM,WACvB,EAAe,QAAQ,EAAK,IAAI,EAAI,EAAe,QAAQ,EAAM,IAAI,GACrE,EAAK,QAAU,EAAM,OAEzB,CAGA,SAAS,EAAqB,CAAE,UAAS,SAAyC,CAChF,IAAM,EAAoB,EAAQ,WAAa,SAC/C,OAAO,EACH,EAAQ,UAAU,OAAQ,GACxB,EAAM,KAAM,GAAS,EAAY,EAAM,EAAG,UAAW,EAAG,QAAS,CAAiB,CAAC,CACrF,EACA,EAAQ,SACd,CAOA,SAAS,EACP,EACA,EACA,EACS,CAIT,OAHK,EAAuB,IAAI,EAAM,MAAM,EAGrC,EAAmB,KACvB,GAAU,EAAM,KAAO,GAAM,EAAS,EAAM,GAAI,CAAE,GAAK,EAAM,SAAS,MAAM,CAAE,YAAa,IAAW,EAAM,MAAM,CACrH,EAJS,EAKX,CAEA,SAAS,EAAS,EAAwB,EAAiC,CACzE,IAAM,EACJ,EAAM,UAAY,EAAM,WACvB,EAAM,YAAc,EAAM,WAAa,EAAM,aAAe,EAAM,YAC/D,EACJ,EAAM,QAAU,EAAM,SAAY,EAAM,UAAY,EAAM,SAAW,EAAM,WAAa,EAAM,UAChG,OAAO,GAAgB,CACzB,CAEA,SAAS,EAAoB,EAAc,EAAqB,EAAsC,CACpG,GAAM,CAAE,YAAW,WAAY,EACzB,EAA6B,EAAS,MAAM,CAAE,YAAa,IAAW,6BAAyB,EACrG,MAAO,CACL,GAAG,EAAY,CAAE,OAAM,YAAW,SAAQ,EAAG,WAAY,CAAQ,EACjE,KAAM,EAAoB,CAAE,EAC5B,GAAI,GAA8B,EAAG,gBAAgB,OAAS,GAAK,CAAE,cAAe,EAAG,eAAgB,CACzG,CACF,CAGA,SAAgB,EAAoB,CAAE,OAAO,cAAe,iBAA0C,CACpG,OAAO,EAAgB,GAAG,EAAc,GAAG,IAAS,CACtD,CAEA,SAAS,EAAY,EAAyB,EAAyB,EAAsC,CAE3G,MAAO,CAAE,OAAM,MADD,EAAS,KAAM,GAAU,EAAM,QAAU,OAAO,EAAI,QAAU,UACtD,GAAG,EAAU,UAAS,CAC9C,CAEA,SAAS,EACP,EACA,EACA,EACiB,CACjB,OAAO,EAAW,QAAS,GAAc,EAAa,EAAW,EAAS,GAAS,EAAO,IAAU,EAAQ,CAAK,CAAC,CACpH,CAGA,SAAS,EACP,EACA,EACA,EACA,EACiB,CACjB,IAAM,EAAQ,EAAU,QAAQ,CAAO,EACjC,EAAQA,EAAAA,OAAO,SAAU,GAAc,EAAS,EAAO,EAAO,EAAU,CAAC,EAAU,IAAc,CAAC,EACxG,OAAO,EACH,CAAC,CAAE,OAAQF,EAAAA,aAAa,CAAS,EAAG,QAAO,QAAO,MAAO,EAAO,EAAM,CAAC,EAAU,IAAe,CAAC,EACjG,CAAC,CACP,CASA,SAAS,EAAY,EAAgB,EAAmB,EAAiB,EAAoB,GAAgB,CAC3G,OAAO,EAAK,YAAc,EACtB,EAAK,WAAa,IAAc,EAAK,UAAY,GAAY,GAAqB,EAAK,YAAc,GACrG,EAAK,WAAa,GAAW,EAAK,UAAY,EAAK,UAAY,CACrE,CAGA,SAAS,EACP,EACA,EAC6C,CAC7C,IAAM,kBAAe,IAAI,IACzB,IAAK,IAAM,KAAS,GAAsB,QAAU,CAAC,EACnD,IAAK,IAAM,KAAQ,EAAM,MACvB,GAAI,EAAM,IAAI,CAAI,EAAG,CACnB,IAAM,EAAS,EAAa,IAAI,CAAI,GAAK,CAAC,EAC1C,EAAO,KAAK,CAAK,EACjB,EAAa,IAAI,EAAM,CAAM,CAC/B,CAGJ,OAAO,CACT,CAOA,SAAS,EACP,CAAE,OAAM,UAAS,SACjB,EACA,EACA,CAAE,wBACW,CACb,IAAM,EAAqB,GACzB,EAAaG,EAAAA,qBAAsB,CAAE,qBAAoB,EAAG,GAAS,EAAO,IAAU,GAAS,CAAK,EAChG,EAA8B,CAClC,GAAG,EAAQ,YAAY,qBAAqB,IAAK,GAAU,EAAM,IAAK,IAAgB,CAAE,GAAG,EAAY,MAAK,EAAE,CAAC,EAC/G,GAAG,EAAgB,IAAK,GAAU,EAAM,YAAY,IAAK,GAAe,EAAkB,CAAU,CAAC,CAAC,CACxG,EACM,EAAc,GAClB,EAAM,OAAS,GACf,EAAkB,EAAM,YAAY,MAAM,CAAC,CAAC,OAAS,IACpD,CAAC,GAAS,EAAM,KAAM,GAAS,EAAK,UAAY,GAAK,EAAY,EAAM,EAAM,UAAW,EAAM,OAAO,CAAC,GACnG,kBAAe,IAAI,IACzB,IAAK,IAAM,KAAS,EAClB,IAAK,IAAM,KAAS,EAAM,OAAQ,GAAe,EAAW,CAAU,CAAC,EACrE,EAAa,IAAI,EAAO,CAAK,EAGjC,OAAO,EAAiB,CAAC,GAAG,EAAa,KAAK,CAAC,CAAC,CAAC,CAAC,KAAK,CAAE,SAAQ,aAAc,CAC7E,IAAM,EAAc,GAClB,EACE,EAAQ,QAAS,GAAW,CAG1B,IAAM,EAAa,KAAK,IACtB,GAAG,EAAQ,OAAQ,GAAU,EAAS,EAAO,CAAM,CAAC,CAAC,CAAC,IAAK,GAAU,EAAM,UAAU,CACvF,EACA,OAAQ,EAAa,IAAI,CAAM,GAAK,CAAC,EAAA,CAAG,OACrC,GAAS,EAAK,WAAa,KAAO,EAAa,CAClD,CACF,CAAC,CACH,CAAC,CAEE,OAAQ,GAAY,CAAC,EAAS,EAAQ,OAAQ,CAAM,CAAC,CAAC,CAEtD,UAAU,EAAM,IAAU,EAAU,EAAM,OAAO,EAAI,EAAU,EAAK,OAAO,CAAC,CAAC,CAC7E,IAAK,GAAY,EAAQ,MAAM,EAM9B,EAAS,EAAW,CAAoB,EAC9C,MAAO,CACL,GAAG,EACD,EACA,cAEA,EAAkB,IAAI,IAAI,EAAQ,QAAS,GAAW,EAAO,WAAW,CAAC,CAAC,CAAC,IAAI,CACjF,EAIA,SAAU,EAAO,OAAS,EAAI,EAAS,EAAW,CAAC,CACrD,CACF,CAAC,CACH,CAEA,SAAS,EAAkB,EAA2D,CACpF,GAAM,CAAE,eAAgB,EAExB,OADA,EAAA,QAAO,GAAG,EAAa,wBAAwB,EAAW,KAAK,kBAAkB,EAC1E,CAAE,GAAG,EAAY,aAAY,CACtC,CAEA,SAAS,EAAU,EAAiD,CAClE,OAAO,EAAY,QAAQ,EAAK,IAAe,EAAM,EAAW,WAAY,CAAC,CAC/E,CAGA,SAAS,EACP,EACyD,CACzD,IAAM,EAAmE,CAAC,EACpE,EAAU,EAAU,UACvB,EAAM,IAAU,EAAe,EAAK,KAAM,EAAM,IAAI,GAAK,EAAK,UAAY,EAAM,SACnF,EACA,IAAK,IAAM,KAAY,EAAS,CAC9B,IAAM,EAAO,EAAQ,GAAG,EAAE,EAC1B,GAAI,GAAQ,EAAS,EAAK,OAAQ,CAAQ,EACxC,EAAK,OAAO,QAAU,KAAK,IAAI,EAAK,OAAO,QAAS,EAAS,OAAO,EACpE,EAAK,QAAQ,KAAK,CAAQ,MACrB,CACL,GAAM,CAAE,OAAM,YAAW,WAAY,EACrC,EAAQ,KAAK,CAAE,OAAQ,CAAE,OAAM,YAAW,SAAQ,EAAG,QAAS,CAAC,CAAQ,CAAE,CAAC,CAC5E,CACF,CACA,OAAO,CACT,CAGA,SAAS,EAAS,EAAsB,EAA+B,CACrE,OAAO,EAAM,OAAS,EAAM,MAAQ,EAAM,WAAa,EAAM,WAAa,EAAM,SAAW,EAAM,OACnG,CAEA,SAAS,EAAS,EAAqB,EAA+B,CACpE,OAAO,EAAK,OAAS,EAAM,MAAQ,EAAK,WAAa,EAAM,SAAW,EAAM,WAAa,EAAK,OAChG,CAGA,SAAS,EAAe,EAAc,EAAuB,CAC3D,OAAO,EAAO,EAAQ,GAAK,IAAO,EACpC"}

package/dist/thresholdCheck.d.ts

@@ -9,12 +9,12 @@ export interface CheckedFile {
     /** The change to the file; when set, only what the change touches is checked. */
     hunks?: readonly LineHunk[];
 }
-/**
- * By file path, the 1-based lines that are neither blank nor comment-only, by which duplicated
- * blocks are measured and their copies ranked: of the checked files and of every other measured
- * file a copy may lie in.
- */
-export type CodeLineNumbersByFile = ReadonlyMap<string, ReadonlySet<number>>;
+/** The duplication of the measured files, in which duplicated blocks and their copies are found. */
+export interface CheckedDuplication {
+    crossFile?: CrossFileDuplicationMetrics;
+    /** The detection's `minSimilarityPercent`, the share of a block a copy of it holds at least. */
+    minSimilarityPercent: number;
+}
 export interface ExceededLimit {
     metric: string;
     value: number;
@@ -41,7 +41,7 @@ export interface Violation extends BlockLocation {
      * construct add to the complexity without being one.
      */
     largestBlocks?: CognitiveBlock[];
-    /** The other copies of the block, most code lines first, for kind `duplication`. */
+    /** The other copies of the block, most matched tokens first, for kind `duplication`. */
     partners?: BlockLocation[];
 }
 export interface CheckResult {
@@ -49,6 +49,6 @@ export interface CheckResult {
     checkedFunctionCount: number;
 }
 /** Every limit the files exceed, errors first, then ordered by path, then line. */
-export declare function checkThresholds(files: readonly CheckedFile[], crossFileDuplication: CrossFileDuplicationMetrics | undefined, limitsOf: (language: string) => LimitsByLevel, codeLineNumbersByFile: CodeLineNumbersByFile): CheckResult;
+export declare function checkThresholds(files: readonly CheckedFile[], duplication: CheckedDuplication, limitsOf: (language: string) => LimitsByLevel): CheckResult;
 /** How reports name a function: with the type or module it is a member of, so that it can be found. */
 export declare function qualifyFunctionName({ name, containerName }: FunctionMetrics): string;

package/dist/thresholdCheck.js

@@ -1,2 +1,2 @@
-import{duplicationThreshold as e,fileThresholds as t,functionThresholds as n,levels as r,metricNameOf as i}from"./thresholds.js";import a from"node:assert";const o=[`file`,`function`,`duplication`],s=new Set(n.filter(({includesNestedFunctions:e})=>e).map(e=>i(e)));function c(e,r,i,a){let o=y(r,new Set(e.map(({file:e})=>e))),s=[],c=0;for(let r of e){let{file:e,metrics:l,hunks:f}=r,m=i(l.language),_=f?.length===0?[]:g(t,l,m);_.length>0&&s.push(h({file:e,startLine:1,endLine:l.lines.total},`file`,_));let v=u(r);c+=v.length;let y=v.map(e=>({fn:e,exceeded:g(n,e,m)})).filter(({exceeded:e})=>e.length>0);for(let{fn:t,exceeded:n}of y){let r=n.filter(e=>!d(e,t,y));r.length>0&&s.push(p(e,t,r))}s.push(...b(r,o.get(e)??[],m,a))}return{violations:s.toSorted(l),checkedFunctionCount:c}}function l(e,t){return r.indexOf(t.level)-r.indexOf(e.level)||w(e.file,t.file)||e.startLine-t.startLine||o.indexOf(e.kind)-o.indexOf(t.kind)||e.endLine-t.endLine}function u({metrics:e,hunks:t}){let n=e.language===`python`;return t?e.functions.filter(e=>t.some(t=>v(t,e.startLine,e.endLine,n))):e.functions}function d(e,t,n){return s.has(e.metric)?n.some(n=>n.fn!==t&&f(n.fn,t)&&n.exceeded.some(({metric:t})=>t===e.metric)):!1}function f(e,t){let n=e.startLine<t.startLine||e.startLine===t.startLine&&e.startColumn<=t.startColumn,r=e.endLine>t.endLine||e.endLine===t.endLine&&e.endColumn>=t.endColumn;return n&&r}function p(e,t,n){let{startLine:r,endLine:i}=t,a=n.some(({metric:e})=>e===`functionCognitiveComplexity`);return{...h({file:e,startLine:r,endLine:i},`function`,n),name:m(t),...a&&t.cognitiveBlocks.length>0&&{largestBlocks:t.cognitiveBlocks}}}function m({name:e=`<anonymous>`,containerName:t}){return t?`${t}.${e}`:e}function h(e,t,n){return{kind:t,level:n.some(e=>e.level===`error`)?`error`:`warning`,...e,exceeded:n}}function g(e,t,n){return e.flatMap(e=>_(e,t,n,(e,t)=>e>t))}function _(e,t,n,a){let o=e.measure(t),s=r.findLast(t=>a(o,n[t][e.key]));return s?[{metric:i(e),value:o,level:s,limit:n[s][e.key]}]:[]}function v(e,t,n,r=!1){return e.headCount===0?e.headStart>=t&&(e.headStart<n||r&&e.headStart===n):e.headStart<=n&&e.headStart+e.headCount>t}function y(e,t){let n=/* @__PURE__ */ new Map;for(let r of e?.groups??[])for(let e of r.files)if(t.has(e)){let t=n.get(e)??[];t.push(r),n.set(e,t)}return n}function b({file:t,metrics:n,hunks:r},i,a,o){let s=t=>_(e,{codeLineCount:S(t,o)},a,(e,t)=>e>=t),c=[...n.duplication.duplicateBlockGroups.map(e=>e.map(e=>({...e,file:t}))),...i.map(e=>e.occurrences)],l=e=>e.file===t&&s(e).length>0&&(!r||r.some(t=>t.headCount>0&&v(t,e.startLine,e.endLine))),u=/* @__PURE__ */ new Map;for(let e of c)for(let t of e.filter(e=>l(e)))u.set(t,e);return x([...u.keys()]).map(({merged:e,sources:t})=>({...h(e,`duplication`,s(e)),partners:x(t.flatMap(e=>u.get(e)??[])).map(e=>e.merged).filter(t=>!C(t,e)).toSorted((e,t)=>S(t,o)-S(e,o))}))}function x(e){let t=[],n=e.toSorted((e,t)=>w(e.file,t.file)||e.startLine-t.startLine);for(let e of n){let n=t.at(-1);if(n&&C(n.merged,e))n.merged.endLine=Math.max(n.merged.endLine,e.endLine),n.sources.push(e);else{let{file:n,startLine:r,endLine:i}=e;t.push({merged:{file:n,startLine:r,endLine:i},sources:[e]})}}return t}function S(e,t){let n=t.get(e.file);a.ok(n,`No code lines for ${e.file}, which holds a duplicated block.`);let r=0;for(let t=e.startLine;t<=e.endLine;t++)n.has(t)&&(r+=1);return r}function C(e,t){return e.file===t.file&&e.startLine<=t.endLine&&t.startLine<=e.endLine}function w(e,t){return e<t?-1:+(e>t)}export{c as checkThresholds,m as qualifyFunctionName};
+import{duplicationThreshold as e,fileThresholds as t,functionThresholds as n,levels as r,metricNameOf as i}from"./thresholds.js";import a from"node:assert";const o=[`file`,`function`,`duplication`],s=new Set(n.filter(({includesNestedFunctions:e})=>e).map(e=>i(e)));function c(e,r,i){let a=y(r.crossFile,new Set(e.map(({file:e})=>e))),o=[],s=0;for(let c of e){let{file:e,metrics:l,hunks:f}=c,m=i(l.language),_=f?.length===0?[]:g(t,l,m);_.length>0&&o.push(h({file:e,startLine:1,endLine:l.lines.total},`file`,_));let v=u(c);s+=v.length;let y=v.map(e=>({fn:e,exceeded:g(n,e,m)})).filter(({exceeded:e})=>e.length>0);for(let{fn:t,exceeded:n}of y){let r=n.filter(e=>!d(e,t,y));r.length>0&&o.push(p(e,t,r))}o.push(...b(c,a.get(e)??[],m,r))}return{violations:o.toSorted(l),checkedFunctionCount:s}}function l(e,t){return r.indexOf(t.level)-r.indexOf(e.level)||E(e.file,t.file)||e.startLine-t.startLine||o.indexOf(e.kind)-o.indexOf(t.kind)||e.endLine-t.endLine}function u({metrics:e,hunks:t}){let n=e.language===`python`;return t?e.functions.filter(e=>t.some(t=>v(t,e.startLine,e.endLine,n))):e.functions}function d(e,t,n){return s.has(e.metric)?n.some(n=>n.fn!==t&&f(n.fn,t)&&n.exceeded.some(({metric:t})=>t===e.metric)):!1}function f(e,t){let n=e.startLine<t.startLine||e.startLine===t.startLine&&e.startColumn<=t.startColumn,r=e.endLine>t.endLine||e.endLine===t.endLine&&e.endColumn>=t.endColumn;return n&&r}function p(e,t,n){let{startLine:r,endLine:i}=t,a=n.some(({metric:e})=>e===`functionCognitiveComplexity`);return{...h({file:e,startLine:r,endLine:i},`function`,n),name:m(t),...a&&t.cognitiveBlocks.length>0&&{largestBlocks:t.cognitiveBlocks}}}function m({name:e=`<anonymous>`,containerName:t}){return t?`${t}.${e}`:e}function h(e,t,n){return{kind:t,level:n.some(e=>e.level===`error`)?`error`:`warning`,...e,exceeded:n}}function g(e,t,n){return e.flatMap(e=>_(e,t,n,(e,t)=>e>t))}function _(e,t,n,a){let o=e.measure(t),s=r.findLast(t=>a(o,n[t][e.key]));return s?[{metric:i(e),value:o,level:s,limit:n[s][e.key]}]:[]}function v(e,t,n,r=!1){return e.headCount===0?e.headStart>=t&&(e.headStart<n||r&&e.headStart===n):e.headStart<=n&&e.headStart+e.headCount>t}function y(e,t){let n=/* @__PURE__ */ new Map;for(let r of e?.groups??[])for(let e of r.files)if(t.has(e)){let t=n.get(e)??[];t.push(r),n.set(e,t)}return n}function b({file:t,metrics:n,hunks:r},i,a,{minSimilarityPercent:o}){let s=t=>_(e,{duplicatedLineCount:t},a,(e,t)=>e>=t),c=[...n.duplication.duplicateBlockGroups.map(e=>e.map(e=>({...e,file:t}))),...i.map(e=>e.occurrences.map(e=>x(e)))],l=e=>e.file===t&&s(e.lineNumbers.length).length>0&&(!r||r.some(t=>t.headCount>0&&v(t,e.startLine,e.endLine))),u=/* @__PURE__ */ new Map;for(let e of c)for(let t of e.filter(e=>l(e)))u.set(t,e);return C([...u.keys()]).map(({merged:e,sources:t})=>{let n=n=>C(t.flatMap(e=>{let r=Math.max(...t.filter(t=>w(t,e)).map(e=>e.tokenCount));return(u.get(e)??[]).filter(e=>e.tokenCount*100>=r*n)})).filter(t=>!T(t.merged,e)).toSorted((e,t)=>S(t.sources)-S(e.sources)).map(e=>e.merged),r=n(o);return{...h(e,`duplication`,s(new Set(t.flatMap(e=>e.lineNumbers)).size)),partners:r.length>0?r:n(0)}})}function x(e){let{lineNumbers:t}=e;return a.ok(t,`No matched lines for ${e.file}, a scanned file.`),{...e,lineNumbers:t}}function S(e){return e.reduce((e,t)=>e+t.tokenCount,0)}function C(e){let t=[],n=e.toSorted((e,t)=>E(e.file,t.file)||e.startLine-t.startLine);for(let e of n){let n=t.at(-1);if(n&&T(n.merged,e))n.merged.endLine=Math.max(n.merged.endLine,e.endLine),n.sources.push(e);else{let{file:n,startLine:r,endLine:i}=e;t.push({merged:{file:n,startLine:r,endLine:i},sources:[e]})}}return t}function w(e,t){return e.file===t.file&&e.startLine<=t.startLine&&t.endLine<=e.endLine}function T(e,t){return e.file===t.file&&e.startLine<=t.endLine&&t.startLine<=e.endLine}function E(e,t){return e<t?-1:+(e>t)}export{c as checkThresholds,m as qualifyFunctionName};
 //# sourceMappingURL=thresholdCheck.js.map

package/dist/thresholdCheck.js.map

@@ -1 +1 @@
-{"version":3,"file":"thresholdCheck.js","names":[],"sources":["../src/thresholdCheck.ts"],"sourcesContent":["import assert from 'node:assert';\nimport type { CrossFileDuplicateBlockGroup, CrossFileDuplicationMetrics } from './crossFileDuplication.js';\nimport type { LineHunk } from './git.js';\nimport {\n  duplicationThreshold,\n  fileThresholds,\n  functionThresholds,\n  levels,\n  metricNameOf,\n  type Level,\n  type LimitsByLevel,\n  type Threshold,\n} from './thresholds.js';\nimport type { CodeMetrics, CognitiveBlock, FunctionMetrics } from './types.js';\n\n/** A measured file to check, under the path cross-file duplication knows it by. */\nexport interface CheckedFile {\n  file: string;\n  metrics: CodeMetrics;\n  /** The change to the file; when set, only what the change touches is checked. */\n  hunks?: readonly LineHunk[];\n}\n\n/**\n * By file path, the 1-based lines that are neither blank nor comment-only, by which duplicated\n * blocks are measured and their copies ranked: of the checked files and of every other measured\n * file a copy may lie in.\n */\nexport type CodeLineNumbersByFile = ReadonlyMap<string, ReadonlySet<number>>;\n\nexport interface ExceededLimit {\n  metric: string;\n  value: number;\n  /** The most severe level whose limit the value violates. */\n  level: Level;\n  /** That level's limit: the value is allowed up to it; a duplicated block is allowed below it. */\n  limit: number;\n}\n\nexport interface BlockLocation {\n  file: string;\n  startLine: number;\n  endLine: number;\n}\n\nexport interface Violation extends BlockLocation {\n  kind: 'function' | 'file' | 'duplication';\n  /** The most severe level among `exceeded`. */\n  level: Level;\n  /** The function's name, qualified with the class it is a member of, for kind `function`. */\n  name?: string;\n  exceeded: ExceededLimit[];\n  /**\n   * The parts adding the most to the cognitive complexity, largest first, for a function that\n   * exceeds a limit of it and has such parts: operator sequences and jumps outside any branching\n   * construct add to the complexity without being one.\n   */\n  largestBlocks?: CognitiveBlock[];\n  /** The other copies of the block, most code lines first, for kind `duplication`. */\n  partners?: BlockLocation[];\n}\n\nexport interface CheckResult {\n  violations: Violation[];\n  checkedFunctionCount: number;\n}\n\nconst violationKinds: readonly Violation['kind'][] = ['file', 'function', 'duplication'];\n\nconst cognitiveComplexityMetric = 'functionCognitiveComplexity';\n\nconst nestedInclusiveMetrics = new Set(\n  functionThresholds\n    .filter(({ includesNestedFunctions }) => includesNestedFunctions)\n    .map((threshold) => metricNameOf(threshold))\n);\n\n/** Every limit the files exceed, errors first, then ordered by path, then line. */\nexport function checkThresholds(\n  files: readonly CheckedFile[],\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined,\n  limitsOf: (language: string) => LimitsByLevel,\n  codeLineNumbersByFile: CodeLineNumbersByFile\n): CheckResult {\n  const crossFileGroupsByFile = indexGroupsByFile(crossFileDuplication, new Set(files.map(({ file }) => file)));\n  const violations: Violation[] = [];\n  let checkedFunctionCount = 0;\n  for (const checkedFile of files) {\n    const { file, metrics, hunks } = checkedFile;\n    const limits = limitsOf(metrics.language);\n    // A rename or a mode change lists the file as changed without touching any of its lines.\n    const fileExceeded = hunks?.length === 0 ? [] : collectExceeded(fileThresholds, metrics, limits);\n    if (fileExceeded.length > 0) {\n      violations.push(toViolation({ file, startLine: 1, endLine: metrics.lines.total }, 'file', fileExceeded));\n    }\n    const checkedFunctions = listCheckedFunctions(checkedFile);\n    checkedFunctionCount += checkedFunctions.length;\n    const exceedingFunctions = checkedFunctions\n      .map((fn) => ({ fn, exceeded: collectExceeded(functionThresholds, fn, limits) }))\n      .filter(({ exceeded }) => exceeded.length > 0);\n    for (const { fn, exceeded } of exceedingFunctions) {\n      const ownExceeded = exceeded.filter((limit) => !isReportedByEnclosingFunction(limit, fn, exceedingFunctions));\n      if (ownExceeded.length > 0) {\n        violations.push(toFunctionViolation(file, fn, ownExceeded));\n      }\n    }\n    violations.push(\n      ...collectDuplicationViolations(checkedFile, crossFileGroupsByFile.get(file) ?? [], limits, codeLineNumbersByFile)\n    );\n  }\n  return { violations: violations.toSorted(compareViolations), checkedFunctionCount };\n}\n\n// The kind is compared before the end line because a file violation's end line is corrected after\n// this ordering (to the last line that exists), and the order must not depend on that correction.\nfunction compareViolations(left: Violation, right: Violation): number {\n  return (\n    levels.indexOf(right.level) - levels.indexOf(left.level) ||\n    compareStrings(left.file, right.file) ||\n    left.startLine - right.startLine ||\n    violationKinds.indexOf(left.kind) - violationKinds.indexOf(right.kind) ||\n    left.endLine - right.endLine\n  );\n}\n\n/** The file's functions or, for a changed file, those the change touches. */\nfunction listCheckedFunctions({ metrics, hunks }: CheckedFile): FunctionMetrics[] {\n  const endsByIndentation = metrics.language === 'python';\n  return hunks\n    ? metrics.functions.filter((fn) =>\n        hunks.some((hunk) => touchesSpan(hunk, fn.startLine, fn.endLine, endsByIndentation))\n      )\n    : metrics.functions;\n}\n\n/**\n * Whether a function enclosing `fn` exceeds the same limit with a value that covers `fn`, so\n * reporting it for `fn` too would count the same code twice. The enclosing value is at least that\n * of `fn` and both share the file's limits, so its level is never milder.\n */\nfunction isReportedByEnclosingFunction(\n  limit: ExceededLimit,\n  fn: FunctionMetrics,\n  exceedingFunctions: readonly { fn: FunctionMetrics; exceeded: ExceededLimit[] }[]\n): boolean {\n  if (!nestedInclusiveMetrics.has(limit.metric)) {\n    return false;\n  }\n  return exceedingFunctions.some(\n    (outer) => outer.fn !== fn && encloses(outer.fn, fn) && outer.exceeded.some(({ metric }) => metric === limit.metric)\n  );\n}\n\nfunction encloses(outer: FunctionMetrics, inner: FunctionMetrics): boolean {\n  const startsBefore =\n    outer.startLine < inner.startLine ||\n    (outer.startLine === inner.startLine && outer.startColumn <= inner.startColumn);\n  const endsAfter =\n    outer.endLine > inner.endLine || (outer.endLine === inner.endLine && outer.endColumn >= inner.endColumn);\n  return startsBefore && endsAfter;\n}\n\nfunction toFunctionViolation(file: string, fn: FunctionMetrics, exceeded: ExceededLimit[]): Violation {\n  const { startLine, endLine } = fn;\n  const exceedsCognitiveComplexity = exceeded.some(({ metric }) => metric === cognitiveComplexityMetric);\n  return {\n    ...toViolation({ file, startLine, endLine }, 'function', exceeded),\n    name: qualifyFunctionName(fn),\n    ...(exceedsCognitiveComplexity && fn.cognitiveBlocks.length > 0 && { largestBlocks: fn.cognitiveBlocks }),\n  };\n}\n\n/** How reports name a function: with the type or module it is a member of, so that it can be found. */\nexport function qualifyFunctionName({ name = '<anonymous>', containerName }: FunctionMetrics): string {\n  return containerName ? `${containerName}.${name}` : name;\n}\n\nfunction toViolation(location: BlockLocation, kind: Violation['kind'], exceeded: ExceededLimit[]): Violation {\n  const level = exceeded.some((limit) => limit.level === 'error') ? 'error' : 'warning';\n  return { kind, level, ...location, exceeded };\n}\n\nfunction collectExceeded<Subject>(\n  thresholds: readonly Threshold<Subject>[],\n  subject: Subject,\n  limits: LimitsByLevel\n): ExceededLimit[] {\n  return thresholds.flatMap((threshold) => findExceeded(threshold, subject, limits, (value, limit) => value > limit));\n}\n\n/** The most severe violated limit of the threshold, if any. */\nfunction findExceeded<Subject>(\n  threshold: Threshold<Subject>,\n  subject: Subject,\n  limits: LimitsByLevel,\n  violates: (value: number, limit: number) => boolean\n): ExceededLimit[] {\n  const value = threshold.measure(subject);\n  const level = levels.findLast((candidate) => violates(value, limits[candidate][threshold.key] as number));\n  return level\n    ? [{ metric: metricNameOf(threshold), value, level, limit: limits[level][threshold.key] as number }]\n    : [];\n}\n\n/**\n * Whether the hunk changes the head lines [startLine, endLine]. A hunk without head lines (a pure\n * deletion) sits between its headStart line and the next one. Where indentation ends a span, its\n * last line is a statement rather than a closing delimiter, so a deletion right after that line may\n * have removed the span's tail; `endsByIndentation` counts it as touching, since the head alone\n * cannot tell it from a deletion after the span.\n */\nfunction touchesSpan(hunk: LineHunk, startLine: number, endLine: number, endsByIndentation = false): boolean {\n  return hunk.headCount === 0\n    ? hunk.headStart >= startLine && (hunk.headStart < endLine || (endsByIndentation && hunk.headStart === endLine))\n    : hunk.headStart <= endLine && hunk.headStart + hunk.headCount > startLine;\n}\n\n/** The cross-file groups with an occurrence in each of the given files. */\nfunction indexGroupsByFile(\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined,\n  files: ReadonlySet<string>\n): Map<string, CrossFileDuplicateBlockGroup[]> {\n  const groupsByFile = new Map<string, CrossFileDuplicateBlockGroup[]>();\n  for (const group of crossFileDuplication?.groups ?? []) {\n    for (const file of group.files) {\n      if (files.has(file)) {\n        const groups = groupsByFile.get(file) ?? [];\n        groups.push(group);\n        groupsByFile.set(file, groups);\n      }\n    }\n  }\n  return groupsByFile;\n}\n\n/**\n * One violation per duplicated region of the file: its clone occurrences, within-file and\n * cross-file, that hold at least the code lines of a level (with `hunks`, only those overlapping\n * added lines), with overlapping occurrences merged into one region.\n */\nfunction collectDuplicationViolations(\n  { file, metrics, hunks }: CheckedFile,\n  crossFileGroups: readonly CrossFileDuplicateBlockGroup[],\n  limits: LimitsByLevel,\n  codeLineNumbersByFile: CodeLineNumbersByFile\n): Violation[] {\n  const findExceededLines = (block: BlockLocation): ExceededLimit[] =>\n    findExceeded(\n      duplicationThreshold,\n      { codeLineCount: countCodeLines(block, codeLineNumbersByFile) },\n      limits,\n      (value, limit) => value >= limit\n    );\n  const groups: BlockLocation[][] = [\n    ...metrics.duplication.duplicateBlockGroups.map((group) => group.map((occurrence) => ({ ...occurrence, file }))),\n    ...crossFileGroups.map((group) => group.occurrences),\n  ];\n  const isReported = (block: BlockLocation): boolean =>\n    block.file === file &&\n    findExceededLines(block).length > 0 &&\n    (!hunks || hunks.some((hunk) => hunk.headCount > 0 && touchesSpan(hunk, block.startLine, block.endLine)));\n  const copiesByBlock = new Map<BlockLocation, BlockLocation[]>();\n  for (const group of groups) {\n    for (const block of group.filter((occurrence) => isReported(occurrence))) {\n      copiesByBlock.set(block, group);\n    }\n  }\n  return mergeOverlapping([...copiesByBlock.keys()]).map(({ merged, sources }) => ({\n    ...toViolation(merged, 'duplication', findExceededLines(merged)),\n    // The region's own occurrences are among the copies of the groups it belongs to.\n    // The region merges every clone that overlaps it, so its copies range from whole copies of it\n    // to a few lines matching one fragment; the largest say the most about what to share.\n    partners: mergeOverlapping(sources.flatMap((source) => copiesByBlock.get(source) ?? []))\n      .map((partner) => partner.merged)\n      .filter((partner) => !overlaps(partner, merged))\n      .toSorted(\n        (left, right) => countCodeLines(right, codeLineNumbersByFile) - countCodeLines(left, codeLineNumbersByFile)\n      ),\n  }));\n}\n\n/** The locations with those overlapping in the same file merged, ordered by file, then line. */\nfunction mergeOverlapping(locations: readonly BlockLocation[]): { merged: BlockLocation; sources: BlockLocation[] }[] {\n  const regions: { merged: BlockLocation; sources: BlockLocation[] }[] = [];\n  const ordered = locations.toSorted(\n    (left, right) => compareStrings(left.file, right.file) || left.startLine - right.startLine\n  );\n  for (const location of ordered) {\n    const last = regions.at(-1);\n    if (last && overlaps(last.merged, location)) {\n      last.merged.endLine = Math.max(last.merged.endLine, location.endLine);\n      last.sources.push(location);\n    } else {\n      const { file, startLine, endLine } = location;\n      regions.push({ merged: { file, startLine, endLine }, sources: [location] });\n    }\n  }\n  return regions;\n}\n\n/** The length of a duplicated block. */\nfunction countCodeLines(block: BlockLocation, codeLineNumbersByFile: CodeLineNumbersByFile): number {\n  const codeLineNumbers = codeLineNumbersByFile.get(block.file);\n  assert.ok(codeLineNumbers, `No code lines for ${block.file}, which holds a duplicated block.`);\n  let count = 0;\n  for (let line = block.startLine; line <= block.endLine; line++) {\n    if (codeLineNumbers.has(line)) count += 1;\n  }\n  return count;\n}\n\nfunction overlaps(left: BlockLocation, right: BlockLocation): boolean {\n  return left.file === right.file && left.startLine <= right.endLine && right.startLine <= left.endLine;\n}\n\n/** Code-unit order, so reports do not depend on the locale. */\nfunction compareStrings(left: string, right: string): number {\n  return left < right ? -1 : left > right ? 1 : 0;\n}\n"],"mappings":"4JAmEA,MAAM,EAA+C,CAAC,OAAQ,WAAY,aAAa,EAIjF,EAAyB,IAAI,IACjC,EACG,QAAQ,CAAE,6BAA8B,CAAuB,CAAC,CAChE,IAAK,GAAc,EAAa,CAAS,CAAC,CAC/C,EAGA,SAAgB,EACd,EACA,EACA,EACA,EACa,CACb,IAAM,EAAwB,EAAkB,EAAsB,IAAI,IAAI,EAAM,KAAK,CAAE,UAAW,CAAI,CAAC,CAAC,EACtG,EAA0B,CAAC,EAC7B,EAAuB,EAC3B,IAAK,IAAM,KAAe,EAAO,CAC/B,GAAM,CAAE,OAAM,UAAS,SAAU,EAC3B,EAAS,EAAS,EAAQ,QAAQ,EAElC,EAAe,GAAO,SAAW,EAAI,CAAC,EAAI,EAAgB,EAAgB,EAAS,CAAM,EAC3F,EAAa,OAAS,GACxB,EAAW,KAAK,EAAY,CAAE,OAAM,UAAW,EAAG,QAAS,EAAQ,MAAM,KAAM,EAAG,OAAQ,CAAY,CAAC,EAEzG,IAAM,EAAmB,EAAqB,CAAW,EACzD,GAAwB,EAAiB,OACzC,IAAM,EAAqB,EACxB,IAAK,IAAQ,CAAE,KAAI,SAAU,EAAgB,EAAoB,EAAI,CAAM,CAAE,EAAE,CAAC,CAChF,QAAQ,CAAE,cAAe,EAAS,OAAS,CAAC,EAC/C,IAAK,GAAM,CAAE,KAAI,cAAc,EAAoB,CACjD,IAAM,EAAc,EAAS,OAAQ,GAAU,CAAC,EAA8B,EAAO,EAAI,CAAkB,CAAC,EACxG,EAAY,OAAS,GACvB,EAAW,KAAK,EAAoB,EAAM,EAAI,CAAW,CAAC,CAE9D,CACA,EAAW,KACT,GAAG,EAA6B,EAAa,EAAsB,IAAI,CAAI,GAAK,CAAC,EAAG,EAAQ,CAAqB,CACnH,CACF,CACA,MAAO,CAAE,WAAY,EAAW,SAAS,CAAiB,EAAG,sBAAqB,CACpF,CAIA,SAAS,EAAkB,EAAiB,EAA0B,CACpE,OACE,EAAO,QAAQ,EAAM,KAAK,EAAI,EAAO,QAAQ,EAAK,KAAK,GACvD,EAAe,EAAK,KAAM,EAAM,IAAI,GACpC,EAAK,UAAY,EAAM,WACvB,EAAe,QAAQ,EAAK,IAAI,EAAI,EAAe,QAAQ,EAAM,IAAI,GACrE,EAAK,QAAU,EAAM,OAEzB,CAGA,SAAS,EAAqB,CAAE,UAAS,SAAyC,CAChF,IAAM,EAAoB,EAAQ,WAAa,SAC/C,OAAO,EACH,EAAQ,UAAU,OAAQ,GACxB,EAAM,KAAM,GAAS,EAAY,EAAM,EAAG,UAAW,EAAG,QAAS,CAAiB,CAAC,CACrF,EACA,EAAQ,SACd,CAOA,SAAS,EACP,EACA,EACA,EACS,CAIT,OAHK,EAAuB,IAAI,EAAM,MAAM,EAGrC,EAAmB,KACvB,GAAU,EAAM,KAAO,GAAM,EAAS,EAAM,GAAI,CAAE,GAAK,EAAM,SAAS,MAAM,CAAE,YAAa,IAAW,EAAM,MAAM,CACrH,EAJS,EAKX,CAEA,SAAS,EAAS,EAAwB,EAAiC,CACzE,IAAM,EACJ,EAAM,UAAY,EAAM,WACvB,EAAM,YAAc,EAAM,WAAa,EAAM,aAAe,EAAM,YAC/D,EACJ,EAAM,QAAU,EAAM,SAAY,EAAM,UAAY,EAAM,SAAW,EAAM,WAAa,EAAM,UAChG,OAAO,GAAgB,CACzB,CAEA,SAAS,EAAoB,EAAc,EAAqB,EAAsC,CACpG,GAAM,CAAE,YAAW,WAAY,EACzB,EAA6B,EAAS,MAAM,CAAE,YAAa,IAAW,6BAAyB,EACrG,MAAO,CACL,GAAG,EAAY,CAAE,OAAM,YAAW,SAAQ,EAAG,WAAY,CAAQ,EACjE,KAAM,EAAoB,CAAE,EAC5B,GAAI,GAA8B,EAAG,gBAAgB,OAAS,GAAK,CAAE,cAAe,EAAG,eAAgB,CACzG,CACF,CAGA,SAAgB,EAAoB,CAAE,OAAO,cAAe,iBAA0C,CACpG,OAAO,EAAgB,GAAG,EAAc,GAAG,IAAS,CACtD,CAEA,SAAS,EAAY,EAAyB,EAAyB,EAAsC,CAE3G,MAAO,CAAE,OAAM,MADD,EAAS,KAAM,GAAU,EAAM,QAAU,OAAO,EAAI,QAAU,UACtD,GAAG,EAAU,UAAS,CAC9C,CAEA,SAAS,EACP,EACA,EACA,EACiB,CACjB,OAAO,EAAW,QAAS,GAAc,EAAa,EAAW,EAAS,GAAS,EAAO,IAAU,EAAQ,CAAK,CAAC,CACpH,CAGA,SAAS,EACP,EACA,EACA,EACA,EACiB,CACjB,IAAM,EAAQ,EAAU,QAAQ,CAAO,EACjC,EAAQ,EAAO,SAAU,GAAc,EAAS,EAAO,EAAO,EAAU,CAAC,EAAU,IAAc,CAAC,EACxG,OAAO,EACH,CAAC,CAAE,OAAQ,EAAa,CAAS,EAAG,QAAO,QAAO,MAAO,EAAO,EAAM,CAAC,EAAU,IAAe,CAAC,EACjG,CAAC,CACP,CASA,SAAS,EAAY,EAAgB,EAAmB,EAAiB,EAAoB,GAAgB,CAC3G,OAAO,EAAK,YAAc,EACtB,EAAK,WAAa,IAAc,EAAK,UAAY,GAAY,GAAqB,EAAK,YAAc,GACrG,EAAK,WAAa,GAAW,EAAK,UAAY,EAAK,UAAY,CACrE,CAGA,SAAS,EACP,EACA,EAC6C,CAC7C,IAAM,kBAAe,IAAI,IACzB,IAAK,IAAM,KAAS,GAAsB,QAAU,CAAC,EACnD,IAAK,IAAM,KAAQ,EAAM,MACvB,GAAI,EAAM,IAAI,CAAI,EAAG,CACnB,IAAM,EAAS,EAAa,IAAI,CAAI,GAAK,CAAC,EAC1C,EAAO,KAAK,CAAK,EACjB,EAAa,IAAI,EAAM,CAAM,CAC/B,CAGJ,OAAO,CACT,CAOA,SAAS,EACP,CAAE,OAAM,UAAS,SACjB,EACA,EACA,EACa,CACb,IAAM,EAAqB,GACzB,EACE,EACA,CAAE,cAAe,EAAe,EAAO,CAAqB,CAAE,EAC9D,GACC,EAAO,IAAU,GAAS,CAC7B,EACI,EAA4B,CAChC,GAAG,EAAQ,YAAY,qBAAqB,IAAK,GAAU,EAAM,IAAK,IAAgB,CAAE,GAAG,EAAY,MAAK,EAAE,CAAC,EAC/G,GAAG,EAAgB,IAAK,GAAU,EAAM,WAAW,CACrD,EACM,EAAc,GAClB,EAAM,OAAS,GACf,EAAkB,CAAK,CAAC,CAAC,OAAS,IACjC,CAAC,GAAS,EAAM,KAAM,GAAS,EAAK,UAAY,GAAK,EAAY,EAAM,EAAM,UAAW,EAAM,OAAO,CAAC,GACnG,kBAAgB,IAAI,IAC1B,IAAK,IAAM,KAAS,EAClB,IAAK,IAAM,KAAS,EAAM,OAAQ,GAAe,EAAW,CAAU,CAAC,EACrE,EAAc,IAAI,EAAO,CAAK,EAGlC,OAAO,EAAiB,CAAC,GAAG,EAAc,KAAK,CAAC,CAAC,CAAC,CAAC,KAAK,CAAE,SAAQ,cAAe,CAC/E,GAAG,EAAY,EAAQ,cAAe,EAAkB,CAAM,CAAC,EAI/D,SAAU,EAAiB,EAAQ,QAAS,GAAW,EAAc,IAAI,CAAM,GAAK,CAAC,CAAC,CAAC,CAAC,CACrF,IAAK,GAAY,EAAQ,MAAM,CAAC,CAChC,OAAQ,GAAY,CAAC,EAAS,EAAS,CAAM,CAAC,CAAC,CAC/C,UACE,EAAM,IAAU,EAAe,EAAO,CAAqB,EAAI,EAAe,EAAM,CAAqB,CAC5G,CACJ,EAAE,CACJ,CAGA,SAAS,EAAiB,EAA4F,CACpH,IAAM,EAAiE,CAAC,EAClE,EAAU,EAAU,UACvB,EAAM,IAAU,EAAe,EAAK,KAAM,EAAM,IAAI,GAAK,EAAK,UAAY,EAAM,SACnF,EACA,IAAK,IAAM,KAAY,EAAS,CAC9B,IAAM,EAAO,EAAQ,GAAG,EAAE,EAC1B,GAAI,GAAQ,EAAS,EAAK,OAAQ,CAAQ,EACxC,EAAK,OAAO,QAAU,KAAK,IAAI,EAAK,OAAO,QAAS,EAAS,OAAO,EACpE,EAAK,QAAQ,KAAK,CAAQ,MACrB,CACL,GAAM,CAAE,OAAM,YAAW,WAAY,EACrC,EAAQ,KAAK,CAAE,OAAQ,CAAE,OAAM,YAAW,SAAQ,EAAG,QAAS,CAAC,CAAQ,CAAE,CAAC,CAC5E,CACF,CACA,OAAO,CACT,CAGA,SAAS,EAAe,EAAsB,EAAsD,CAClG,IAAM,EAAkB,EAAsB,IAAI,EAAM,IAAI,EAC5D,EAAO,GAAG,EAAiB,qBAAqB,EAAM,KAAK,kCAAkC,EAC7F,IAAI,EAAQ,EACZ,IAAK,IAAI,EAAO,EAAM,UAAW,GAAQ,EAAM,QAAS,IAClD,EAAgB,IAAI,CAAI,IAAG,GAAS,GAE1C,OAAO,CACT,CAEA,SAAS,EAAS,EAAqB,EAA+B,CACpE,OAAO,EAAK,OAAS,EAAM,MAAQ,EAAK,WAAa,EAAM,SAAW,EAAM,WAAa,EAAK,OAChG,CAGA,SAAS,EAAe,EAAc,EAAuB,CAC3D,OAAO,EAAO,EAAQ,GAAK,IAAO,EACpC"}
+{"version":3,"file":"thresholdCheck.js","names":[],"sources":["../src/thresholdCheck.ts"],"sourcesContent":["import assert from 'node:assert';\nimport type {\n  CrossFileDuplicateBlockGroup,\n  CrossFileDuplicateOccurrence,\n  CrossFileDuplicationMetrics,\n} from './crossFileDuplication.js';\nimport type { LineHunk } from './git.js';\nimport {\n  duplicationThreshold,\n  fileThresholds,\n  functionThresholds,\n  levels,\n  metricNameOf,\n  type Level,\n  type LimitsByLevel,\n  type Threshold,\n} from './thresholds.js';\nimport type { CodeMetrics, CognitiveBlock, FunctionMetrics } from './types.js';\n\n/** A measured file to check, under the path cross-file duplication knows it by. */\nexport interface CheckedFile {\n  file: string;\n  metrics: CodeMetrics;\n  /** The change to the file; when set, only what the change touches is checked. */\n  hunks?: readonly LineHunk[];\n}\n\n/** The duplication of the measured files, in which duplicated blocks and their copies are found. */\nexport interface CheckedDuplication {\n  crossFile?: CrossFileDuplicationMetrics;\n  /** The detection's `minSimilarityPercent`, the share of a block a copy of it holds at least. */\n  minSimilarityPercent: number;\n}\n\nexport interface ExceededLimit {\n  metric: string;\n  value: number;\n  /** The most severe level whose limit the value violates. */\n  level: Level;\n  /** That level's limit: the value is allowed up to it; a duplicated block is allowed below it. */\n  limit: number;\n}\n\nexport interface BlockLocation {\n  file: string;\n  startLine: number;\n  endLine: number;\n}\n\n/** One copy of a clone group. */\ninterface CloneOccurrence extends BlockLocation {\n  /**\n   * Its code lines matched in another copy, whose number is the length of a duplicated block. Its\n   * line span would also count the lines between the matched parts of one copy, which the copies\n   * do not share; a block matched as a whole by a near-miss copy holds all its code lines.\n   */\n  lineNumbers: readonly number[];\n  tokenCount: number;\n}\n\nexport interface Violation extends BlockLocation {\n  kind: 'function' | 'file' | 'duplication';\n  /** The most severe level among `exceeded`. */\n  level: Level;\n  /** The function's name, qualified with the class it is a member of, for kind `function`. */\n  name?: string;\n  exceeded: ExceededLimit[];\n  /**\n   * The parts adding the most to the cognitive complexity, largest first, for a function that\n   * exceeds a limit of it and has such parts: operator sequences and jumps outside any branching\n   * construct add to the complexity without being one.\n   */\n  largestBlocks?: CognitiveBlock[];\n  /** The other copies of the block, most matched tokens first, for kind `duplication`. */\n  partners?: BlockLocation[];\n}\n\nexport interface CheckResult {\n  violations: Violation[];\n  checkedFunctionCount: number;\n}\n\nconst violationKinds: readonly Violation['kind'][] = ['file', 'function', 'duplication'];\n\nconst cognitiveComplexityMetric = 'functionCognitiveComplexity';\n\nconst nestedInclusiveMetrics = new Set(\n  functionThresholds\n    .filter(({ includesNestedFunctions }) => includesNestedFunctions)\n    .map((threshold) => metricNameOf(threshold))\n);\n\n/** Every limit the files exceed, errors first, then ordered by path, then line. */\nexport function checkThresholds(\n  files: readonly CheckedFile[],\n  duplication: CheckedDuplication,\n  limitsOf: (language: string) => LimitsByLevel\n): CheckResult {\n  const crossFileGroupsByFile = indexGroupsByFile(duplication.crossFile, new Set(files.map(({ file }) => file)));\n  const violations: Violation[] = [];\n  let checkedFunctionCount = 0;\n  for (const checkedFile of files) {\n    const { file, metrics, hunks } = checkedFile;\n    const limits = limitsOf(metrics.language);\n    // A rename or a mode change lists the file as changed without touching any of its lines.\n    const fileExceeded = hunks?.length === 0 ? [] : collectExceeded(fileThresholds, metrics, limits);\n    if (fileExceeded.length > 0) {\n      violations.push(toViolation({ file, startLine: 1, endLine: metrics.lines.total }, 'file', fileExceeded));\n    }\n    const checkedFunctions = listCheckedFunctions(checkedFile);\n    checkedFunctionCount += checkedFunctions.length;\n    const exceedingFunctions = checkedFunctions\n      .map((fn) => ({ fn, exceeded: collectExceeded(functionThresholds, fn, limits) }))\n      .filter(({ exceeded }) => exceeded.length > 0);\n    for (const { fn, exceeded } of exceedingFunctions) {\n      const ownExceeded = exceeded.filter((limit) => !isReportedByEnclosingFunction(limit, fn, exceedingFunctions));\n      if (ownExceeded.length > 0) {\n        violations.push(toFunctionViolation(file, fn, ownExceeded));\n      }\n    }\n    violations.push(\n      ...collectDuplicationViolations(checkedFile, crossFileGroupsByFile.get(file) ?? [], limits, duplication)\n    );\n  }\n  return { violations: violations.toSorted(compareViolations), checkedFunctionCount };\n}\n\n// The kind is compared before the end line because a file violation's end line is corrected after\n// this ordering (to the last line that exists), and the order must not depend on that correction.\nfunction compareViolations(left: Violation, right: Violation): number {\n  return (\n    levels.indexOf(right.level) - levels.indexOf(left.level) ||\n    compareStrings(left.file, right.file) ||\n    left.startLine - right.startLine ||\n    violationKinds.indexOf(left.kind) - violationKinds.indexOf(right.kind) ||\n    left.endLine - right.endLine\n  );\n}\n\n/** The file's functions or, for a changed file, those the change touches. */\nfunction listCheckedFunctions({ metrics, hunks }: CheckedFile): FunctionMetrics[] {\n  const endsByIndentation = metrics.language === 'python';\n  return hunks\n    ? metrics.functions.filter((fn) =>\n        hunks.some((hunk) => touchesSpan(hunk, fn.startLine, fn.endLine, endsByIndentation))\n      )\n    : metrics.functions;\n}\n\n/**\n * Whether a function enclosing `fn` exceeds the same limit with a value that covers `fn`, so\n * reporting it for `fn` too would count the same code twice. The enclosing value is at least that\n * of `fn` and both share the file's limits, so its level is never milder.\n */\nfunction isReportedByEnclosingFunction(\n  limit: ExceededLimit,\n  fn: FunctionMetrics,\n  exceedingFunctions: readonly { fn: FunctionMetrics; exceeded: ExceededLimit[] }[]\n): boolean {\n  if (!nestedInclusiveMetrics.has(limit.metric)) {\n    return false;\n  }\n  return exceedingFunctions.some(\n    (outer) => outer.fn !== fn && encloses(outer.fn, fn) && outer.exceeded.some(({ metric }) => metric === limit.metric)\n  );\n}\n\nfunction encloses(outer: FunctionMetrics, inner: FunctionMetrics): boolean {\n  const startsBefore =\n    outer.startLine < inner.startLine ||\n    (outer.startLine === inner.startLine && outer.startColumn <= inner.startColumn);\n  const endsAfter =\n    outer.endLine > inner.endLine || (outer.endLine === inner.endLine && outer.endColumn >= inner.endColumn);\n  return startsBefore && endsAfter;\n}\n\nfunction toFunctionViolation(file: string, fn: FunctionMetrics, exceeded: ExceededLimit[]): Violation {\n  const { startLine, endLine } = fn;\n  const exceedsCognitiveComplexity = exceeded.some(({ metric }) => metric === cognitiveComplexityMetric);\n  return {\n    ...toViolation({ file, startLine, endLine }, 'function', exceeded),\n    name: qualifyFunctionName(fn),\n    ...(exceedsCognitiveComplexity && fn.cognitiveBlocks.length > 0 && { largestBlocks: fn.cognitiveBlocks }),\n  };\n}\n\n/** How reports name a function: with the type or module it is a member of, so that it can be found. */\nexport function qualifyFunctionName({ name = '<anonymous>', containerName }: FunctionMetrics): string {\n  return containerName ? `${containerName}.${name}` : name;\n}\n\nfunction toViolation(location: BlockLocation, kind: Violation['kind'], exceeded: ExceededLimit[]): Violation {\n  const level = exceeded.some((limit) => limit.level === 'error') ? 'error' : 'warning';\n  return { kind, level, ...location, exceeded };\n}\n\nfunction collectExceeded<Subject>(\n  thresholds: readonly Threshold<Subject>[],\n  subject: Subject,\n  limits: LimitsByLevel\n): ExceededLimit[] {\n  return thresholds.flatMap((threshold) => findExceeded(threshold, subject, limits, (value, limit) => value > limit));\n}\n\n/** The most severe violated limit of the threshold, if any. */\nfunction findExceeded<Subject>(\n  threshold: Threshold<Subject>,\n  subject: Subject,\n  limits: LimitsByLevel,\n  violates: (value: number, limit: number) => boolean\n): ExceededLimit[] {\n  const value = threshold.measure(subject);\n  const level = levels.findLast((candidate) => violates(value, limits[candidate][threshold.key] as number));\n  return level\n    ? [{ metric: metricNameOf(threshold), value, level, limit: limits[level][threshold.key] as number }]\n    : [];\n}\n\n/**\n * Whether the hunk changes the head lines [startLine, endLine]. A hunk without head lines (a pure\n * deletion) sits between its headStart line and the next one. Where indentation ends a span, its\n * last line is a statement rather than a closing delimiter, so a deletion right after that line may\n * have removed the span's tail; `endsByIndentation` counts it as touching, since the head alone\n * cannot tell it from a deletion after the span.\n */\nfunction touchesSpan(hunk: LineHunk, startLine: number, endLine: number, endsByIndentation = false): boolean {\n  return hunk.headCount === 0\n    ? hunk.headStart >= startLine && (hunk.headStart < endLine || (endsByIndentation && hunk.headStart === endLine))\n    : hunk.headStart <= endLine && hunk.headStart + hunk.headCount > startLine;\n}\n\n/** The cross-file groups with an occurrence in each of the given files. */\nfunction indexGroupsByFile(\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined,\n  files: ReadonlySet<string>\n): Map<string, CrossFileDuplicateBlockGroup[]> {\n  const groupsByFile = new Map<string, CrossFileDuplicateBlockGroup[]>();\n  for (const group of crossFileDuplication?.groups ?? []) {\n    for (const file of group.files) {\n      if (files.has(file)) {\n        const groups = groupsByFile.get(file) ?? [];\n        groups.push(group);\n        groupsByFile.set(file, groups);\n      }\n    }\n  }\n  return groupsByFile;\n}\n\n/**\n * One violation per duplicated region of the file: its clone occurrences, within-file and\n * cross-file, that hold at least the duplicated lines of a level (with `hunks`, only those\n * overlapping added lines), with overlapping occurrences merged into one region.\n */\nfunction collectDuplicationViolations(\n  { file, metrics, hunks }: CheckedFile,\n  crossFileGroups: readonly CrossFileDuplicateBlockGroup[],\n  limits: LimitsByLevel,\n  { minSimilarityPercent }: CheckedDuplication\n): Violation[] {\n  const findExceededLines = (duplicatedLineCount: number): ExceededLimit[] =>\n    findExceeded(duplicationThreshold, { duplicatedLineCount }, limits, (value, limit) => value >= limit);\n  const groups: CloneOccurrence[][] = [\n    ...metrics.duplication.duplicateBlockGroups.map((group) => group.map((occurrence) => ({ ...occurrence, file }))),\n    ...crossFileGroups.map((group) => group.occurrences.map((occurrence) => toCloneOccurrence(occurrence))),\n  ];\n  const isReported = (block: CloneOccurrence): boolean =>\n    block.file === file &&\n    findExceededLines(block.lineNumbers.length).length > 0 &&\n    (!hunks || hunks.some((hunk) => hunk.headCount > 0 && touchesSpan(hunk, block.startLine, block.endLine)));\n  const groupByBlock = new Map<CloneOccurrence, CloneOccurrence[]>();\n  for (const group of groups) {\n    for (const block of group.filter((occurrence) => isReported(occurrence))) {\n      groupByBlock.set(block, group);\n    }\n  }\n  return mergeOverlapping([...groupByBlock.keys()]).map(({ merged, sources }) => {\n    const listCopies = (minTokenPercent: number): BlockLocation[] =>\n      mergeOverlapping(\n        sources.flatMap((source) => {\n          // The region merges every clone that overlaps it, so a copy of an occurrence lying\n          // within a larger one is measured against the larger one.\n          const tokenCount = Math.max(\n            ...sources.filter((outer) => contains(outer, source)).map((outer) => outer.tokenCount)\n          );\n          return (groupByBlock.get(source) ?? []).filter(\n            (copy) => copy.tokenCount * 100 >= tokenCount * minTokenPercent\n          );\n        })\n      )\n        // The region's own occurrences are among the copies of the groups it belongs to.\n        .filter((partner) => !overlaps(partner.merged, merged))\n        // The largest say the most about what to share.\n        .toSorted((left, right) => sumTokens(right.sources) - sumTokens(left.sources))\n        .map((partner) => partner.merged);\n    // A group joins a block with every place repeating a part of it: a function another file\n    // copies whole is grouped with each file holding a few of its lines. Such a fragment is not a\n    // copy of the block: sharing it would not remove the block, and a report listing it reads as\n    // one more copy. Sizes are compared in tokens, since the same code may be laid out over any\n    // number of lines.\n    const copies = listCopies(minSimilarityPercent);\n    return {\n      ...toViolation(\n        merged,\n        'duplication',\n        // Occurrences that overlap share lines, which count once.\n        findExceededLines(new Set(sources.flatMap((source) => source.lineNumbers)).size)\n      ),\n      // A group does not always hold the copy a block was matched with as a whole: within a file,\n      // a copy that already belongs to an exact group is listed as its part in that group. The\n      // fragments then are all that locates the block's copies.\n      partners: copies.length > 0 ? copies : listCopies(0),\n    };\n  });\n}\n\nfunction toCloneOccurrence(occurrence: CrossFileDuplicateOccurrence): CloneOccurrence {\n  const { lineNumbers } = occurrence;\n  assert.ok(lineNumbers, `No matched lines for ${occurrence.file}, a scanned file.`);\n  return { ...occurrence, lineNumbers };\n}\n\nfunction sumTokens(occurrences: readonly CloneOccurrence[]): number {\n  return occurrences.reduce((sum, occurrence) => sum + occurrence.tokenCount, 0);\n}\n\n/** The locations with those overlapping in the same file merged, ordered by file, then line. */\nfunction mergeOverlapping(\n  locations: readonly CloneOccurrence[]\n): { merged: BlockLocation; sources: CloneOccurrence[] }[] {\n  const regions: { merged: BlockLocation; sources: CloneOccurrence[] }[] = [];\n  const ordered = locations.toSorted(\n    (left, right) => compareStrings(left.file, right.file) || left.startLine - right.startLine\n  );\n  for (const location of ordered) {\n    const last = regions.at(-1);\n    if (last && overlaps(last.merged, location)) {\n      last.merged.endLine = Math.max(last.merged.endLine, location.endLine);\n      last.sources.push(location);\n    } else {\n      const { file, startLine, endLine } = location;\n      regions.push({ merged: { file, startLine, endLine }, sources: [location] });\n    }\n  }\n  return regions;\n}\n\n/** Whether `outer` spans every line of `inner`; a block contains itself. */\nfunction contains(outer: BlockLocation, inner: BlockLocation): boolean {\n  return outer.file === inner.file && outer.startLine <= inner.startLine && inner.endLine <= outer.endLine;\n}\n\nfunction overlaps(left: BlockLocation, right: BlockLocation): boolean {\n  return left.file === right.file && left.startLine <= right.endLine && right.startLine <= left.endLine;\n}\n\n/** Code-unit order, so reports do not depend on the locale. */\nfunction compareStrings(left: string, right: string): number {\n  return left < right ? -1 : left > right ? 1 : 0;\n}\n"],"mappings":"4JAkFA,MAAM,EAA+C,CAAC,OAAQ,WAAY,aAAa,EAIjF,EAAyB,IAAI,IACjC,EACG,QAAQ,CAAE,6BAA8B,CAAuB,CAAC,CAChE,IAAK,GAAc,EAAa,CAAS,CAAC,CAC/C,EAGA,SAAgB,EACd,EACA,EACA,EACa,CACb,IAAM,EAAwB,EAAkB,EAAY,UAAW,IAAI,IAAI,EAAM,KAAK,CAAE,UAAW,CAAI,CAAC,CAAC,EACvG,EAA0B,CAAC,EAC7B,EAAuB,EAC3B,IAAK,IAAM,KAAe,EAAO,CAC/B,GAAM,CAAE,OAAM,UAAS,SAAU,EAC3B,EAAS,EAAS,EAAQ,QAAQ,EAElC,EAAe,GAAO,SAAW,EAAI,CAAC,EAAI,EAAgB,EAAgB,EAAS,CAAM,EAC3F,EAAa,OAAS,GACxB,EAAW,KAAK,EAAY,CAAE,OAAM,UAAW,EAAG,QAAS,EAAQ,MAAM,KAAM,EAAG,OAAQ,CAAY,CAAC,EAEzG,IAAM,EAAmB,EAAqB,CAAW,EACzD,GAAwB,EAAiB,OACzC,IAAM,EAAqB,EACxB,IAAK,IAAQ,CAAE,KAAI,SAAU,EAAgB,EAAoB,EAAI,CAAM,CAAE,EAAE,CAAC,CAChF,QAAQ,CAAE,cAAe,EAAS,OAAS,CAAC,EAC/C,IAAK,GAAM,CAAE,KAAI,cAAc,EAAoB,CACjD,IAAM,EAAc,EAAS,OAAQ,GAAU,CAAC,EAA8B,EAAO,EAAI,CAAkB,CAAC,EACxG,EAAY,OAAS,GACvB,EAAW,KAAK,EAAoB,EAAM,EAAI,CAAW,CAAC,CAE9D,CACA,EAAW,KACT,GAAG,EAA6B,EAAa,EAAsB,IAAI,CAAI,GAAK,CAAC,EAAG,EAAQ,CAAW,CACzG,CACF,CACA,MAAO,CAAE,WAAY,EAAW,SAAS,CAAiB,EAAG,sBAAqB,CACpF,CAIA,SAAS,EAAkB,EAAiB,EAA0B,CACpE,OACE,EAAO,QAAQ,EAAM,KAAK,EAAI,EAAO,QAAQ,EAAK,KAAK,GACvD,EAAe,EAAK,KAAM,EAAM,IAAI,GACpC,EAAK,UAAY,EAAM,WACvB,EAAe,QAAQ,EAAK,IAAI,EAAI,EAAe,QAAQ,EAAM,IAAI,GACrE,EAAK,QAAU,EAAM,OAEzB,CAGA,SAAS,EAAqB,CAAE,UAAS,SAAyC,CAChF,IAAM,EAAoB,EAAQ,WAAa,SAC/C,OAAO,EACH,EAAQ,UAAU,OAAQ,GACxB,EAAM,KAAM,GAAS,EAAY,EAAM,EAAG,UAAW,EAAG,QAAS,CAAiB,CAAC,CACrF,EACA,EAAQ,SACd,CAOA,SAAS,EACP,EACA,EACA,EACS,CAIT,OAHK,EAAuB,IAAI,EAAM,MAAM,EAGrC,EAAmB,KACvB,GAAU,EAAM,KAAO,GAAM,EAAS,EAAM,GAAI,CAAE,GAAK,EAAM,SAAS,MAAM,CAAE,YAAa,IAAW,EAAM,MAAM,CACrH,EAJS,EAKX,CAEA,SAAS,EAAS,EAAwB,EAAiC,CACzE,IAAM,EACJ,EAAM,UAAY,EAAM,WACvB,EAAM,YAAc,EAAM,WAAa,EAAM,aAAe,EAAM,YAC/D,EACJ,EAAM,QAAU,EAAM,SAAY,EAAM,UAAY,EAAM,SAAW,EAAM,WAAa,EAAM,UAChG,OAAO,GAAgB,CACzB,CAEA,SAAS,EAAoB,EAAc,EAAqB,EAAsC,CACpG,GAAM,CAAE,YAAW,WAAY,EACzB,EAA6B,EAAS,MAAM,CAAE,YAAa,IAAW,6BAAyB,EACrG,MAAO,CACL,GAAG,EAAY,CAAE,OAAM,YAAW,SAAQ,EAAG,WAAY,CAAQ,EACjE,KAAM,EAAoB,CAAE,EAC5B,GAAI,GAA8B,EAAG,gBAAgB,OAAS,GAAK,CAAE,cAAe,EAAG,eAAgB,CACzG,CACF,CAGA,SAAgB,EAAoB,CAAE,OAAO,cAAe,iBAA0C,CACpG,OAAO,EAAgB,GAAG,EAAc,GAAG,IAAS,CACtD,CAEA,SAAS,EAAY,EAAyB,EAAyB,EAAsC,CAE3G,MAAO,CAAE,OAAM,MADD,EAAS,KAAM,GAAU,EAAM,QAAU,OAAO,EAAI,QAAU,UACtD,GAAG,EAAU,UAAS,CAC9C,CAEA,SAAS,EACP,EACA,EACA,EACiB,CACjB,OAAO,EAAW,QAAS,GAAc,EAAa,EAAW,EAAS,GAAS,EAAO,IAAU,EAAQ,CAAK,CAAC,CACpH,CAGA,SAAS,EACP,EACA,EACA,EACA,EACiB,CACjB,IAAM,EAAQ,EAAU,QAAQ,CAAO,EACjC,EAAQ,EAAO,SAAU,GAAc,EAAS,EAAO,EAAO,EAAU,CAAC,EAAU,IAAc,CAAC,EACxG,OAAO,EACH,CAAC,CAAE,OAAQ,EAAa,CAAS,EAAG,QAAO,QAAO,MAAO,EAAO,EAAM,CAAC,EAAU,IAAe,CAAC,EACjG,CAAC,CACP,CASA,SAAS,EAAY,EAAgB,EAAmB,EAAiB,EAAoB,GAAgB,CAC3G,OAAO,EAAK,YAAc,EACtB,EAAK,WAAa,IAAc,EAAK,UAAY,GAAY,GAAqB,EAAK,YAAc,GACrG,EAAK,WAAa,GAAW,EAAK,UAAY,EAAK,UAAY,CACrE,CAGA,SAAS,EACP,EACA,EAC6C,CAC7C,IAAM,kBAAe,IAAI,IACzB,IAAK,IAAM,KAAS,GAAsB,QAAU,CAAC,EACnD,IAAK,IAAM,KAAQ,EAAM,MACvB,GAAI,EAAM,IAAI,CAAI,EAAG,CACnB,IAAM,EAAS,EAAa,IAAI,CAAI,GAAK,CAAC,EAC1C,EAAO,KAAK,CAAK,EACjB,EAAa,IAAI,EAAM,CAAM,CAC/B,CAGJ,OAAO,CACT,CAOA,SAAS,EACP,CAAE,OAAM,UAAS,SACjB,EACA,EACA,CAAE,wBACW,CACb,IAAM,EAAqB,GACzB,EAAa,EAAsB,CAAE,qBAAoB,EAAG,GAAS,EAAO,IAAU,GAAS,CAAK,EAChG,EAA8B,CAClC,GAAG,EAAQ,YAAY,qBAAqB,IAAK,GAAU,EAAM,IAAK,IAAgB,CAAE,GAAG,EAAY,MAAK,EAAE,CAAC,EAC/G,GAAG,EAAgB,IAAK,GAAU,EAAM,YAAY,IAAK,GAAe,EAAkB,CAAU,CAAC,CAAC,CACxG,EACM,EAAc,GAClB,EAAM,OAAS,GACf,EAAkB,EAAM,YAAY,MAAM,CAAC,CAAC,OAAS,IACpD,CAAC,GAAS,EAAM,KAAM,GAAS,EAAK,UAAY,GAAK,EAAY,EAAM,EAAM,UAAW,EAAM,OAAO,CAAC,GACnG,kBAAe,IAAI,IACzB,IAAK,IAAM,KAAS,EAClB,IAAK,IAAM,KAAS,EAAM,OAAQ,GAAe,EAAW,CAAU,CAAC,EACrE,EAAa,IAAI,EAAO,CAAK,EAGjC,OAAO,EAAiB,CAAC,GAAG,EAAa,KAAK,CAAC,CAAC,CAAC,CAAC,KAAK,CAAE,SAAQ,aAAc,CAC7E,IAAM,EAAc,GAClB,EACE,EAAQ,QAAS,GAAW,CAG1B,IAAM,EAAa,KAAK,IACtB,GAAG,EAAQ,OAAQ,GAAU,EAAS,EAAO,CAAM,CAAC,CAAC,CAAC,IAAK,GAAU,EAAM,UAAU,CACvF,EACA,OAAQ,EAAa,IAAI,CAAM,GAAK,CAAC,EAAA,CAAG,OACrC,GAAS,EAAK,WAAa,KAAO,EAAa,CAClD,CACF,CAAC,CACH,CAAC,CAEE,OAAQ,GAAY,CAAC,EAAS,EAAQ,OAAQ,CAAM,CAAC,CAAC,CAEtD,UAAU,EAAM,IAAU,EAAU,EAAM,OAAO,EAAI,EAAU,EAAK,OAAO,CAAC,CAAC,CAC7E,IAAK,GAAY,EAAQ,MAAM,EAM9B,EAAS,EAAW,CAAoB,EAC9C,MAAO,CACL,GAAG,EACD,EACA,cAEA,EAAkB,IAAI,IAAI,EAAQ,QAAS,GAAW,EAAO,WAAW,CAAC,CAAC,CAAC,IAAI,CACjF,EAIA,SAAU,EAAO,OAAS,EAAI,EAAS,EAAW,CAAC,CACrD,CACF,CAAC,CACH,CAEA,SAAS,EAAkB,EAA2D,CACpF,GAAM,CAAE,eAAgB,EAExB,OADA,EAAO,GAAG,EAAa,wBAAwB,EAAW,KAAK,kBAAkB,EAC1E,CAAE,GAAG,EAAY,aAAY,CACtC,CAEA,SAAS,EAAU,EAAiD,CAClE,OAAO,EAAY,QAAQ,EAAK,IAAe,EAAM,EAAW,WAAY,CAAC,CAC/E,CAGA,SAAS,EACP,EACyD,CACzD,IAAM,EAAmE,CAAC,EACpE,EAAU,EAAU,UACvB,EAAM,IAAU,EAAe,EAAK,KAAM,EAAM,IAAI,GAAK,EAAK,UAAY,EAAM,SACnF,EACA,IAAK,IAAM,KAAY,EAAS,CAC9B,IAAM,EAAO,EAAQ,GAAG,EAAE,EAC1B,GAAI,GAAQ,EAAS,EAAK,OAAQ,CAAQ,EACxC,EAAK,OAAO,QAAU,KAAK,IAAI,EAAK,OAAO,QAAS,EAAS,OAAO,EACpE,EAAK,QAAQ,KAAK,CAAQ,MACrB,CACL,GAAM,CAAE,OAAM,YAAW,WAAY,EACrC,EAAQ,KAAK,CAAE,OAAQ,CAAE,OAAM,YAAW,SAAQ,EAAG,QAAS,CAAC,CAAQ,CAAE,CAAC,CAC5E,CACF,CACA,OAAO,CACT,CAGA,SAAS,EAAS,EAAsB,EAA+B,CACrE,OAAO,EAAM,OAAS,EAAM,MAAQ,EAAM,WAAa,EAAM,WAAa,EAAM,SAAW,EAAM,OACnG,CAEA,SAAS,EAAS,EAAqB,EAA+B,CACpE,OAAO,EAAK,OAAS,EAAM,MAAQ,EAAK,WAAa,EAAM,SAAW,EAAM,WAAa,EAAK,OAChG,CAGA,SAAS,EAAe,EAAc,EAAuB,CAC3D,OAAO,EAAO,EAAQ,GAAK,IAAO,EACpC"}

package/dist/thresholds.cjs

@@ -1,2 +1,2 @@
-"use strict";const e=[`warning`,`error`],t=[{key:`maxFunctionCognitiveComplexity`,label:`cognitive complexity`,defaultLimits:{warning:15,error:30},hint:`flatten nested branching with early returns and extract nested blocks into named functions.`,measure:e=>e.cognitiveComplexity,includesNestedFunctions:!0},{key:`maxFunctionCyclomaticComplexity`,label:`cyclomatic complexity`,defaultLimits:{warning:1/0,error:1/0},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`,defaultLimits:{warning:30,error:60},hint:`move each group of statements that forms one step into a named function.`,measure:e=>e.ncss,includesNestedFunctions:!0},{key:`maxFunctionNestingDepth`,label:`nesting depth`,defaultLimits:{warning:1/0,error:1/0},hint:`replace nested conditions with guard clauses or move inner blocks into functions.`,measure:e=>e.nestingDepth},{key:`maxFunctionParameterCount`,label:`parameters`,defaultLimits:{warning:6,error:1/0},hint:`group related parameters into one object or split the function.`,measure:e=>e.parameterCount},{key:`maxFunctionHalsteadVolume`,label:`Halstead volume`,defaultLimits:{warning:1/0,error:1/0},hint:`the function holds too many operators and operands; move cohesive parts into helper functions.`,measure:e=>e.halstead.volume,includesNestedFunctions:!0},{key:`maxFunctionHalsteadDifficulty`,label:`Halstead difficulty`,defaultLimits:{warning:1/0,error:1/0},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`,defaultLimits:{warning:1/0,error:1/0},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`,defaultLimits:{warning:1/0,error:1/0},hint:`too many values flow between the variables of the function; split it so each part works on fewer variables.`,measure:e=>e.depDegree,includesNestedFunctions:!0}],n=[{key:`maxFileNcss`,label:`file NCSS`,defaultLimits:{warning:400,error:1e3},hint:`split the file into modules with one responsibility each, for example by moving a class or the helpers that serve only one function into a file of their own.`,measure:e=>e.ncssCount}],r={key:`minDuplicateLines`,label:`duplicated lines`,defaultLimits:{warning:15,error:1/0},hint:`extract the repeated code into one shared function or module and call it from every location.`,measure:e=>e.codeLineCount},i=[...t,...n,r];function a(e){let t=e.key.slice(3);return t.charAt(0).toLowerCase()+t.slice(1)}function o(e,t){let n=(n,r)=>{let a=t?.[n];return{...Object.fromEntries(i.map(({key:e,defaultLimits:t})=>[e,t[n]])),...a?.limits,...a&&Object.hasOwn(a.languages,r)?a.languages[r]:void 0,...e[n]}};return e=>({warning:n(`warning`,e),error:n(`error`,e)})}exports.duplicationThreshold=r,exports.fileThresholds=n,exports.functionThresholds=t,exports.levels=e,exports.metricNameOf=a,exports.resolveLimits=o,exports.thresholds=i;
+"use strict";const e=[`warning`,`error`],t=[{key:`maxFunctionCognitiveComplexity`,label:`cognitive complexity`,defaultLimits:{warning:15,error:30},hint:`flatten nested branching with early returns and extract nested blocks into named functions.`,measure:e=>e.cognitiveComplexity,includesNestedFunctions:!0},{key:`maxFunctionCyclomaticComplexity`,label:`cyclomatic complexity`,defaultLimits:{warning:1/0,error:1/0},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`,defaultLimits:{warning:30,error:60},hint:`move each group of statements that forms one step into a named function.`,measure:e=>e.ncss,includesNestedFunctions:!0},{key:`maxFunctionNestingDepth`,label:`nesting depth`,defaultLimits:{warning:1/0,error:1/0},hint:`replace nested conditions with guard clauses or move inner blocks into functions.`,measure:e=>e.nestingDepth},{key:`maxFunctionParameterCount`,label:`parameters`,defaultLimits:{warning:6,error:1/0},hint:`group related parameters into one object or split the function.`,measure:e=>e.parameterCount},{key:`maxFunctionHalsteadVolume`,label:`Halstead volume`,defaultLimits:{warning:1/0,error:1/0},hint:`the function holds too many operators and operands; move cohesive parts into helper functions.`,measure:e=>e.halstead.volume,includesNestedFunctions:!0},{key:`maxFunctionHalsteadDifficulty`,label:`Halstead difficulty`,defaultLimits:{warning:1/0,error:1/0},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`,defaultLimits:{warning:1/0,error:1/0},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`,defaultLimits:{warning:1/0,error:1/0},hint:`too many values flow between the variables of the function; split it so each part works on fewer variables.`,measure:e=>e.depDegree,includesNestedFunctions:!0}],n=[{key:`maxFileNcss`,label:`file NCSS`,defaultLimits:{warning:400,error:1e3},hint:`split the file into modules with one responsibility each, for example by moving a class or the helpers that serve only one function into a file of their own.`,measure:e=>e.ncssCount}],r={key:`minDuplicateLines`,label:`duplicated lines`,defaultLimits:{warning:15,error:1/0},hint:`extract the repeated code into one shared function or module and call it from every location.`,measure:e=>e.duplicatedLineCount},i=[...t,...n,r];function a(e){let t=e.key.slice(3);return t.charAt(0).toLowerCase()+t.slice(1)}function o(e,t){let n=(n,r)=>{let a=t?.[n];return{...Object.fromEntries(i.map(({key:e,defaultLimits:t})=>[e,t[n]])),...a?.limits,...a&&Object.hasOwn(a.languages,r)?a.languages[r]:void 0,...e[n]}};return e=>({warning:n(`warning`,e),error:n(`error`,e)})}exports.duplicationThreshold=r,exports.fileThresholds=n,exports.functionThresholds=t,exports.levels=e,exports.metricNameOf=a,exports.resolveLimits=o,exports.thresholds=i;
 //# sourceMappingURL=thresholds.cjs.map