code-gauge 9.0.0 → 10.1.0

package/dist/cli.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"cli.cjs","names":["getErrorMessage","Command","levels","thresholds","runCheckCommand","configFileName","duplicationThreshold","fileThresholds","functionThresholds","resolveTarget","configSearchDirectory","loadConfig","resolveOptions","scanTarget","loadExclusion","formatPath","collectDuplicatedLineNumbers","InvalidArgumentError","parsePositiveInteger"],"sources":["../src/cli.ts"],"sourcesContent":["#!/usr/bin/env node\n\nimport { Command, InvalidArgumentError } from 'commander';\nimport { getErrorMessage, parsePositiveInteger } from '@willbooster/shared-lib';\nimport { type CliOptions, configFileName, loadConfig, type ResolvedOptions, resolveOptions } from './cliConfig.js';\nimport type { CrossFileDuplicationMetrics } from './crossFileDuplication.js';\nimport { runCheckCommand, type CheckCliOptions } from './checkCommand.js';\nimport { loadExclusion } from './exclusion.js';\nimport {\n  addCrossFileDuplication,\n  collectDuplicatedLineNumbers,\n  configSearchDirectory,\n  formatPath,\n  resolveTarget,\n  scanTarget,\n  writeStderr,\n  writeStdout,\n  type FileMetrics,\n  type ScanResult,\n} from './scan.js';\nimport {\n  duplicationThreshold,\n  fileThresholds,\n  functionThresholds,\n  levels,\n  thresholds,\n  type Level,\n  type Limits,\n} from './thresholds.js';\nimport type { FunctionMetrics } from './types.js';\n\n/** The worst (highest-cognitive-complexity) function of a file, reported as the ranking evidence. */\ninterface WorstFunction {\n  name: string;\n  startLine: number;\n  endLine: number;\n  cognitiveComplexity: number;\n  ncss: number;\n  nestingDepth: number;\n}\n\n/**\n * One ranked refactoring candidate. The score is the sum of the file's repo-relative percentile\n * ranks (each in [0, 1)) over three dimensions: worst-function cognitive complexity, duplicated\n * lines (within-file and cross-file combined), and file NCSS. Ranking is relative to the scanned\n * project, so no absolute threshold is involved.\n */\ninterface RankedFile {\n  file: string;\n  score: number;\n  worstFunction?: WorstFunction;\n  /** Distinct lines covered by within-file duplicate blocks or cross-file duplicate occurrences. */\n  duplicatedLineCount: number;\n  /** duplicatedLineCount / code lines (0 when the file has no code). */\n  duplicatedLineRatio: number;\n  /** Other files sharing cross-file duplicate blocks; filled only for the reported top files. */\n  crossFilePartners: string[];\n  ncss: number;\n  codeLines: number;\n}\n\n/** Caps how many cross-file partners a ranked file lists so the report stays scannable. */\nconst maxCrossFilePartners = 3;\n\n// oxlint-disable-next-line unicorn/prefer-top-level-await -- CommonJS build output cannot preserve top-level await.\nvoid main().catch((error: unknown) => {\n  writeStderr(`Error: ${getErrorMessage(error)}\\n`);\n  process.exitCode = 1;\n});\n\nasync function main(): Promise<void> {\n  const program = new Command()\n    .name('code-gauge')\n    .description('Measure code metrics: rank refactoring candidates (the default command) or check thresholds.');\n\n  addSharedOptions(\n    program\n      .command('rank', { isDefault: true })\n      .description('Rank the files of a project by refactoring priority')\n      .argument('[target]', 'file or directory to measure', '.')\n  )\n    .option('--top <number>', 'number of top-ranked files to report (default: 10)', parsePositiveIntegerOption)\n    .option('--fail-on-error', 'exit with code 1 when files or directories cannot be scanned')\n    .action(runRankCommand);\n\n  const check = addSharedOptions(\n    program\n      .command('check')\n      .description(\n        'Report every exceeded threshold as a warning or an error (exit 1 on errors, 2 when files cannot be measured); with --base, only in what the change touches'\n      )\n      .argument('[target]', 'file or directory to check', '.')\n  ).option('--base <ref>', 'check only what the working tree changed since the merge-base of this git ref and HEAD');\n  for (const level of levels) {\n    for (const { key, defaultLimits } of thresholds) {\n      const defaultLimit = defaultLimits[level];\n      check.option(\n        `--${level}-${toKebabCase(key)} <number>`,\n        `${describeLimit(key, level)} (default ${Number.isFinite(defaultLimit) ? defaultLimit : 'off'}); \"off\" disables it`,\n        parseLimit\n      );\n    }\n  }\n  // Exit code 1 means error-level violations, so a rejected command line must not use it.\n  check.exitOverride((error) => process.exit(error.exitCode === 0 ? 0 : 2));\n  check.action(async (target: string, cliOptions: CheckCliOptions & Record<string, unknown>) => {\n    const readLimits = (level: Level): Limits =>\n      Object.fromEntries(\n        thresholds.flatMap(({ key }) => {\n          const limit = cliOptions[`${level}${key.charAt(0).toUpperCase()}${key.slice(1)}`];\n          return limit === undefined ? [] : [[key, limit as number]];\n        })\n      );\n    await runCheckCommand(target, cliOptions, { warning: readLimits('warning'), error: readLimits('error') });\n  });\n\n  await program.parseAsync();\n}\n\n/** Registers the options shared by the ranking and check commands. */\nfunction addSharedOptions(command: Command): Command {\n  return command\n    .option('--config <path>', `config file to use instead of the auto-detected ${configFileName}`)\n    .option(\n      '--duplication-min-tokens <number>',\n      'minimum normalized token count for a duplicate region (default 40)',\n      parsePositiveIntegerOption\n    )\n    .option(\n      '--duplication-max-gap-tokens <number>',\n      'maximum token gap merged into one gapped clone group; 0 disables merging (default 30)',\n      parseNonNegativeInteger\n    )\n    .option(\n      '--duplication-min-similarity-percent <number>',\n      'minimum similarity percent (1-100) for near-miss (Type-3) clone blocks; 100 reports exact matches only (default 70)',\n      parsePercentInteger\n    )\n    .option('--include-tests', 'include test files and test directories')\n    .option('--json', 'print JSON output');\n}\n\nfunction describeLimit(key: string, level: Level): string {\n  if (key === duplicationThreshold.key) {\n    return `line count from which a duplicated block is ${level === 'error' ? 'an error' : 'a warning'}`;\n  }\n  const fileThreshold = fileThresholds.find((threshold) => threshold.key === key);\n  const subject = fileThreshold\n    ? fileThreshold.label\n    : `${functionThresholds.find((threshold) => threshold.key === key)?.label} of a function`;\n  return `maximum ${subject} without ${level === 'error' ? 'an error' : 'a warning'}`;\n}\n\nfunction toKebabCase(key: string): string {\n  return key.replaceAll(/[A-Z]/gu, (letter) => `-${letter.toLowerCase()}`);\n}\n\nasync function runRankCommand(target: string, cliOptions: CliOptions): Promise<void> {\n  const resolvedTarget = resolveTarget(target);\n  const searchDirectory = await configSearchDirectory(resolvedTarget);\n  const config = await loadConfig(cliOptions.config, searchDirectory);\n  const options = resolveOptions(cliOptions, config);\n  const result = await scanTarget(resolvedTarget, {\n    ...options,\n    loadExclusion: (absolutePaths) => loadExclusion(searchDirectory, options.exclude, absolutePaths),\n  });\n  addCrossFileDuplication(result, options);\n  const rankedFiles = rankFiles(result, options.top);\n\n  if (options.json) {\n    printJson(result, rankedFiles, options);\n  } else {\n    printTextReport(resolvedTarget, result, rankedFiles, options);\n  }\n\n  if (result.fatalError || (options.failOnError && result.errors.length > 0)) {\n    process.exitCode = 1;\n  }\n}\n\nfunction rankFiles(result: ScanResult, top: number): RankedFile[] {\n  const candidates = result.files.map(({ file, metrics }) => {\n    const formattedFile = formatPath(file, result.displayRoot);\n    const duplicatedLines = collectDuplicatedLineNumbers(metrics, result.crossFileDuplication, formattedFile);\n    return {\n      file: formattedFile,\n      worstFunction: findWorstFunction(metrics.functions),\n      duplicatedLineCount: duplicatedLines.size,\n      duplicatedLineRatio: metrics.lines.code === 0 ? 0 : duplicatedLines.size / metrics.lines.code,\n      crossFilePartners: [] as string[],\n      ncss: metrics.ncssCount,\n      codeLines: metrics.lines.code,\n    };\n  });\n\n  const cognitivePercentile = makePercentile(candidates.map((c) => c.worstFunction?.cognitiveComplexity ?? 0));\n  const duplicationPercentile = makePercentile(candidates.map((c) => c.duplicatedLineCount));\n  const ncssPercentile = makePercentile(candidates.map((c) => c.ncss));\n\n  const ranked = candidates\n    .map((candidate) => ({\n      ...candidate,\n      score:\n        cognitivePercentile(candidate.worstFunction?.cognitiveComplexity ?? 0) +\n        duplicationPercentile(candidate.duplicatedLineCount) +\n        ncssPercentile(candidate.ncss),\n    }))\n    .toSorted(\n      (left, right) => right.score - left.score || right.ncss - left.ncss || left.file.localeCompare(right.file)\n    );\n  // Partner evidence is attached only to the files that will be reported: expanding partners for\n  // every scanned file first would retain TH(F^2) strings when one clone group spans F files, and\n  // rescanning all groups per file would be O(files x groups) on the post-scan ranking step.\n  attachCrossFilePartners(ranked.slice(0, top), result.crossFileDuplication);\n  return ranked;\n}\n\n/** One pass over the groups fills the reported files' partner lists (other files sharing a group). */\nfunction attachCrossFilePartners(\n  reportedFiles: RankedFile[],\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined\n): void {\n  if (!crossFileDuplication) {\n    return;\n  }\n  const partnersByFile = new Map(reportedFiles.map((ranked) => [ranked.file, new Set<string>()]));\n  for (const group of crossFileDuplication.groups) {\n    for (const file of group.files) {\n      const partners = partnersByFile.get(file);\n      if (!partners) {\n        continue;\n      }\n      for (const partner of group.files) {\n        if (partner !== file) {\n          partners.add(partner);\n        }\n      }\n    }\n  }\n  for (const ranked of reportedFiles) {\n    ranked.crossFilePartners = [...(partnersByFile.get(ranked.file) ?? [])].toSorted();\n  }\n}\n\n/**\n * Percentile rank within the scanned project: the fraction of files with a strictly smaller\n * value, in [0, 1). Relative ranking needs no absolute threshold, which sidesteps the metric\n * calibration problem entirely — the top of the list is worth refactoring first regardless of\n * where any cutoff would sit.\n */\nfunction makePercentile(values: number[]): (value: number) => number {\n  const sorted = values.toSorted((left, right) => left - right);\n  return (value: number) => {\n    let low = 0;\n    let high = sorted.length;\n    while (low < high) {\n      const middle = (low + high) >> 1;\n      if ((sorted[middle] as number) < value) {\n        low = middle + 1;\n      } else {\n        high = middle;\n      }\n    }\n    return sorted.length === 0 ? 0 : low / sorted.length;\n  };\n}\n\n/** The highest-cognitive-complexity function; NCSS breaks ties so the larger body is reported. */\nfunction findWorstFunction(functions: FunctionMetrics[]): WorstFunction | undefined {\n  let worst: FunctionMetrics | undefined;\n  for (const fn of functions) {\n    if (\n      !worst ||\n      fn.cognitiveComplexity > worst.cognitiveComplexity ||\n      (fn.cognitiveComplexity === worst.cognitiveComplexity && fn.ncss > worst.ncss)\n    ) {\n      worst = fn;\n    }\n  }\n  if (!worst) {\n    return undefined;\n  }\n  return {\n    name: worst.name ?? '<anonymous>',\n    startLine: worst.startLine,\n    endLine: worst.endLine,\n    cognitiveComplexity: worst.cognitiveComplexity,\n    ncss: worst.ncss,\n    nestingDepth: worst.nestingDepth,\n  };\n}\n\nfunction printJson(result: ScanResult, rankedFiles: RankedFile[], options: ResolvedOptions): void {\n  const reportedFiles = rankedFiles.slice(0, options.top);\n  writeStdout(\n    JSON.stringify(\n      {\n        summary: summarize(result.files),\n        totalRankedFiles: rankedFiles.length,\n        truncated: reportedFiles.length < rankedFiles.length,\n        files: reportedFiles,\n        errors: result.errors,\n        warnings: result.warnings,\n      },\n      undefined,\n      2\n    ) + '\\n'\n  );\n}\n\nfunction printTextReport(\n  target: string,\n  result: ScanResult,\n  rankedFiles: RankedFile[],\n  options: ResolvedOptions\n): void {\n  if (result.fatalError) {\n    writeStderr(`Error: ${result.fatalError}\\n`);\n    return;\n  }\n\n  const summary = summarize(result.files);\n  writeStdout(\n    `Measured ${summary.fileCount} files under ${target} (code LOC ${summary.linesOfCode}, NCSS ${summary.ncssCount}, functions ${summary.functionCount})\\n`\n  );\n\n  if (rankedFiles.length === 0) {\n    writeStdout('No measurable files found.\\n');\n  } else {\n    const reportedFiles = rankedFiles.slice(0, options.top);\n    const totalSuffix = rankedFiles.length > reportedFiles.length ? ` of ${rankedFiles.length}` : '';\n    writeStdout(`\\nRefactoring candidates (top ${reportedFiles.length}${totalSuffix}):\\n`);\n    for (const [index, ranked] of reportedFiles.entries()) {\n      writeStdout(`${index + 1}. ${formatRankedFile(ranked)}\\n`);\n    }\n  }\n\n  if (result.warnings.length > 0) {\n    writeStderr(`\\nDegraded ${result.warnings.length} files (measured, but excluded from cross-file matching):\\n`);\n    for (const warning of result.warnings.slice(0, 10)) {\n      writeStderr(`- ${warning}\\n`);\n    }\n    if (result.warnings.length > 10) {\n      writeStderr(`- ... ${result.warnings.length - 10} more\\n`);\n    }\n  }\n\n  if (result.errors.length > 0) {\n    writeStderr(`\\nSkipped ${result.errors.length} files or directories:\\n`);\n    for (const error of result.errors.slice(0, 10)) {\n      writeStderr(`- ${error}\\n`);\n    }\n    if (result.errors.length > 10) {\n      writeStderr(`- ... ${result.errors.length - 10} more\\n`);\n    }\n  }\n}\n\n/** One ranked file as a single line: the location, the evidence, and where the duplication points. */\nfunction formatRankedFile(ranked: RankedFile): string {\n  const reasons: string[] = [];\n  if (ranked.worstFunction) {\n    const fn = ranked.worstFunction;\n    reasons.push(\n      `worst function ${fn.name} (L${fn.startLine}-${fn.endLine}) cognitive ${fn.cognitiveComplexity}, NCSS ${fn.ncss}, nesting ${fn.nestingDepth}`\n    );\n  }\n  if (ranked.duplicatedLineCount > 0) {\n    const partnersSuffix =\n      ranked.crossFilePartners.length > 0\n        ? `, shared with ${ranked.crossFilePartners.slice(0, maxCrossFilePartners).join(', ')}${ranked.crossFilePartners.length > maxCrossFilePartners ? ', ...' : ''}`\n        : '';\n    reasons.push(\n      `duplicated lines ${ranked.duplicatedLineCount} (${Math.round(ranked.duplicatedLineRatio * 100)}%${partnersSuffix})`\n    );\n  }\n  reasons.push(`file NCSS ${ranked.ncss}`);\n  return `${ranked.file} (score ${ranked.score.toFixed(2)}): ${reasons.join('; ')}`;\n}\n\nfunction summarize(files: FileMetrics[]): {\n  fileCount: number;\n  functionCount: number;\n  linesOfCode: number;\n  maxCognitiveComplexity: number;\n  ncssCount: number;\n} {\n  let functionCount = 0;\n  let linesOfCode = 0;\n  let maxCognitiveComplexity = 0;\n  let ncssCount = 0;\n\n  for (const file of files) {\n    functionCount += file.metrics.functions.length;\n    linesOfCode += file.metrics.lines.code;\n    maxCognitiveComplexity = Math.max(maxCognitiveComplexity, file.metrics.maxCognitiveComplexity);\n    ncssCount += file.metrics.ncssCount;\n  }\n\n  return { fileCount: files.length, functionCount, linesOfCode, maxCognitiveComplexity, ncssCount };\n}\n\nfunction parseLimit(value: string): number {\n  if (value === 'off') {\n    return Infinity;\n  }\n  const parsed = value.trim() === '' ? Number.NaN : Number(value);\n  if (!Number.isFinite(parsed) || parsed < 0) {\n    throw new InvalidArgumentError('Expected a non-negative number or \"off\".');\n  }\n  return parsed;\n}\n\nfunction parsePercentInteger(value: string): number {\n  const parsed = parsePositiveIntegerOption(value);\n  if (parsed > 100) {\n    throw new InvalidArgumentError('Expected an integer between 1 and 100.');\n  }\n  return parsed;\n}\n\nfunction parseNonNegativeInteger(value: string): number {\n  if (!/^\\d+$/u.test(value)) {\n    throw new InvalidArgumentError('Expected a non-negative integer.');\n  }\n\n  const parsed = Number(value);\n  if (!Number.isSafeInteger(parsed) || parsed < 0) {\n    throw new InvalidArgumentError('Expected a non-negative integer.');\n  }\n  return parsed;\n}\n\nfunction parsePositiveIntegerOption(value: string): number {\n  const parsed = parsePositiveInteger(value);\n  if (parsed === undefined) {\n    throw new InvalidArgumentError('Expected a positive integer.');\n  }\n  return parsed;\n}\n"],"mappings":";mOAiEK,EAAK,CAAC,CAAC,MAAO,GAAmB,CACpC,EAAA,YAAY,WAAA,EAAUA,EAAAA,gBAAAA,CAAgB,CAAK,EAAE,GAAG,EAChD,QAAQ,SAAW,CACrB,CAAC,EAED,eAAe,GAAsB,CACnC,IAAM,EAAU,IAAIC,EAAAA,QAAQ,CAAC,CAC1B,KAAK,YAAY,CAAC,CAClB,YAAY,8FAA8F,EAE7G,EACE,EACG,QAAQ,OAAQ,CAAE,UAAW,EAAK,CAAC,CAAC,CACpC,YAAY,qDAAqD,CAAC,CAClE,SAAS,WAAY,+BAAgC,GAAG,CAC7D,CAAC,CACE,OAAO,iBAAkB,qDAAsD,CAA0B,CAAC,CAC1G,OAAO,kBAAmB,8DAA8D,CAAC,CACzF,OAAO,CAAc,EAExB,IAAM,EAAQ,EACZ,EACG,QAAQ,OAAO,CAAC,CAChB,YACC,4JACF,CAAC,CACA,SAAS,WAAY,6BAA8B,GAAG,CAC3D,CAAC,CAAC,OAAO,eAAgB,wFAAwF,EACjH,IAAK,IAAM,KAASC,EAAAA,OAClB,IAAK,GAAM,CAAE,MAAK,mBAAmBC,EAAAA,WAAY,CAC/C,IAAM,EAAe,EAAc,GACnC,EAAM,OACJ,KAAK,EAAM,GAAG,EAAY,CAAG,EAAE,WAC/B,GAAG,EAAc,EAAK,CAAK,EAAE,YAAY,OAAO,SAAS,CAAY,EAAI,EAAe,MAAM,sBAC9F,CACF,CACF,CAGF,EAAM,aAAc,GAAU,QAAQ,KAAK,EAAM,WAAa,EAAI,EAAI,CAAC,CAAC,EACxE,EAAM,OAAO,MAAO,EAAgB,IAA0D,CAC5F,IAAM,EAAc,GAClB,OAAO,YACLA,EAAAA,WAAW,SAAS,CAAE,SAAU,CAC9B,IAAM,EAAQ,EAAW,GAAG,IAAQ,EAAI,OAAO,CAAC,CAAC,CAAC,YAAY,IAAI,EAAI,MAAM,CAAC,KAC7E,OAAO,IAAU,IAAA,GAAY,CAAC,EAAI,CAAC,CAAC,EAAK,CAAe,CAAC,CAC3D,CAAC,CACH,EACF,MAAMC,EAAAA,gBAAgB,EAAQ,EAAY,CAAE,QAAS,EAAW,SAAS,EAAG,MAAO,EAAW,OAAO,CAAE,CAAC,CAC1G,CAAC,EAED,MAAM,EAAQ,WAAW,CAC3B,CAGA,SAAS,EAAiB,EAA2B,CACnD,OAAO,EACJ,OAAO,kBAAmB,mDAAmDC,EAAAA,gBAAgB,CAAC,CAC9F,OACC,oCACA,qEACA,CACF,CAAC,CACA,OACC,wCACA,wFACA,CACF,CAAC,CACA,OACC,gDACA,sHACA,CACF,CAAC,CACA,OAAO,kBAAmB,yCAAyC,CAAC,CACpE,OAAO,SAAU,mBAAmB,CACzC,CAEA,SAAS,EAAc,EAAa,EAAsB,CACxD,GAAI,IAAQC,EAAAA,qBAAqB,IAC/B,MAAO,+CAA+C,IAAU,QAAU,WAAa,cAEzF,IAAM,EAAgBC,EAAAA,eAAe,KAAM,GAAc,EAAU,MAAQ,CAAG,EAI9E,MAAO,WAHS,EACZ,EAAc,MACd,GAAGC,EAAAA,mBAAmB,KAAM,GAAc,EAAU,MAAQ,CAAG,CAAC,EAAE,MAAM,gBAClD,WAAW,IAAU,QAAU,WAAa,aACxE,CAEA,SAAS,EAAY,EAAqB,CACxC,OAAO,EAAI,WAAW,UAAY,GAAW,IAAI,EAAO,YAAY,GAAG,CACzE,CAEA,eAAe,EAAe,EAAgB,EAAuC,CACnF,IAAM,EAAiBC,EAAAA,cAAc,CAAM,EACrC,EAAkB,MAAMC,EAAAA,sBAAsB,CAAc,EAC5D,EAAS,MAAMC,EAAAA,WAAW,EAAW,OAAQ,CAAe,EAC5D,EAAUC,EAAAA,eAAe,EAAY,CAAM,EAC3C,EAAS,MAAMC,EAAAA,WAAW,EAAgB,CAC9C,GAAG,EACH,cAAgB,GAAkBC,EAAAA,cAAc,EAAiB,EAAQ,QAAS,CAAa,CACjG,CAAC,EACD,EAAA,wBAAwB,EAAQ,CAAO,EACvC,IAAM,EAAc,EAAU,EAAQ,EAAQ,GAAG,EAE7C,EAAQ,KACV,EAAU,EAAQ,EAAa,CAAO,EAEtC,EAAgB,EAAgB,EAAQ,EAAa,CAAO,GAG1D,EAAO,YAAe,EAAQ,aAAe,EAAO,OAAO,OAAS,KACtE,QAAQ,SAAW,EAEvB,CAEA,SAAS,EAAU,EAAoB,EAA2B,CAChE,IAAM,EAAa,EAAO,MAAM,KAAK,CAAE,OAAM,aAAc,CACzD,IAAM,EAAgBC,EAAAA,WAAW,EAAM,EAAO,WAAW,EACnD,EAAkBC,EAAAA,6BAA6B,EAAS,EAAO,qBAAsB,CAAa,EACxG,MAAO,CACL,KAAM,EACN,cAAe,EAAkB,EAAQ,SAAS,EAClD,oBAAqB,EAAgB,KACrC,oBAAqB,EAAQ,MAAM,OAAS,EAAI,EAAI,EAAgB,KAAO,EAAQ,MAAM,KACzF,kBAAmB,CAAC,EACpB,KAAM,EAAQ,UACd,UAAW,EAAQ,MAAM,IAC3B,CACF,CAAC,EAEK,EAAsB,EAAe,EAAW,IAAK,GAAM,EAAE,eAAe,qBAAuB,CAAC,CAAC,EACrG,EAAwB,EAAe,EAAW,IAAK,GAAM,EAAE,mBAAmB,CAAC,EACnF,EAAiB,EAAe,EAAW,IAAK,GAAM,EAAE,IAAI,CAAC,EAE7D,EAAS,EACZ,IAAK,IAAe,CACnB,GAAG,EACH,MACE,EAAoB,EAAU,eAAe,qBAAuB,CAAC,EACrE,EAAsB,EAAU,mBAAmB,EACnD,EAAe,EAAU,IAAI,CACjC,EAAE,CAAC,CACF,UACE,EAAM,IAAU,EAAM,MAAQ,EAAK,OAAS,EAAM,KAAO,EAAK,MAAQ,EAAK,KAAK,cAAc,EAAM,IAAI,CAC3G,EAKF,OADA,EAAwB,EAAO,MAAM,EAAG,CAAG,EAAG,EAAO,oBAAoB,EAClE,CACT,CAGA,SAAS,EACP,EACA,EACM,CACN,GAAI,CAAC,EACH,OAEF,IAAM,EAAiB,IAAI,IAAI,EAAc,IAAK,GAAW,CAAC,EAAO,qBAAM,IAAI,GAAa,CAAC,CAAC,EAC9F,IAAK,IAAM,KAAS,EAAqB,OACvC,IAAK,IAAM,KAAQ,EAAM,MAAO,CAC9B,IAAM,EAAW,EAAe,IAAI,CAAI,EACnC,KAGL,IAAK,IAAM,KAAW,EAAM,MACtB,IAAY,GACd,EAAS,IAAI,CAAO,CAG1B,CAEF,IAAK,IAAM,KAAU,EACnB,EAAO,kBAAoB,CAAC,GAAI,EAAe,IAAI,EAAO,IAAI,GAAK,CAAC,CAAE,CAAC,CAAC,SAAS,CAErF,CAQA,SAAS,EAAe,EAA6C,CACnE,IAAM,EAAS,EAAO,UAAU,EAAM,IAAU,EAAO,CAAK,EAC5D,MAAQ,IAAkB,CACxB,IAAI,EAAM,EACN,EAAO,EAAO,OAClB,KAAO,EAAM,GAAM,CACjB,IAAM,EAAU,EAAM,GAAS,EAC1B,EAAO,GAAqB,EAC/B,EAAM,EAAS,EAEf,EAAO,CAEX,CACA,OAAO,EAAO,SAAW,EAAI,EAAI,EAAM,EAAO,MAChD,CACF,CAGA,SAAS,EAAkB,EAAyD,CAClF,IAAI,EACJ,IAAK,IAAM,KAAM,GAEb,CAAC,GACD,EAAG,oBAAsB,EAAM,qBAC9B,EAAG,sBAAwB,EAAM,qBAAuB,EAAG,KAAO,EAAM,QAEzE,EAAQ,GAGP,KAGL,MAAO,CACL,KAAM,EAAM,MAAQ,cACpB,UAAW,EAAM,UACjB,QAAS,EAAM,QACf,oBAAqB,EAAM,oBAC3B,KAAM,EAAM,KACZ,aAAc,EAAM,YACtB,CACF,CAEA,SAAS,EAAU,EAAoB,EAA2B,EAAgC,CAChG,IAAM,EAAgB,EAAY,MAAM,EAAG,EAAQ,GAAG,EACtD,EAAA,YACE,KAAK,UACH,CACE,QAAS,EAAU,EAAO,KAAK,EAC/B,iBAAkB,EAAY,OAC9B,UAAW,EAAc,OAAS,EAAY,OAC9C,MAAO,EACP,OAAQ,EAAO,OACf,SAAU,EAAO,QACnB,EACA,IAAA,GACA,CACF,EAAI;CACN,CACF,CAEA,SAAS,EACP,EACA,EACA,EACA,EACM,CACN,GAAI,EAAO,WAAY,CACrB,EAAA,YAAY,UAAU,EAAO,WAAW,GAAG,EAC3C,MACF,CAEA,IAAM,EAAU,EAAU,EAAO,KAAK,EAKtC,GAJA,EAAA,YACE,YAAY,EAAQ,UAAU,eAAe,EAAO,aAAa,EAAQ,YAAY,SAAS,EAAQ,UAAU,cAAc,EAAQ,cAAc,IACtJ,EAEI,EAAY,SAAW,EACzB,EAAA,YAAY;CAA8B,MACrC,CACL,IAAM,EAAgB,EAAY,MAAM,EAAG,EAAQ,GAAG,EAChD,EAAc,EAAY,OAAS,EAAc,OAAS,OAAO,EAAY,SAAW,GAC9F,EAAA,YAAY,iCAAiC,EAAc,SAAS,EAAY,KAAK,EACrF,IAAK,GAAM,CAAC,EAAO,KAAW,EAAc,QAAQ,EAClD,EAAA,YAAY,GAAG,EAAQ,EAAE,IAAI,EAAiB,CAAM,EAAE,GAAG,CAE7D,CAEA,GAAI,EAAO,SAAS,OAAS,EAAG,CAC9B,EAAA,YAAY,cAAc,EAAO,SAAS,OAAO,4DAA4D,EAC7G,IAAK,IAAM,KAAW,EAAO,SAAS,MAAM,EAAG,EAAE,EAC/C,EAAA,YAAY,KAAK,EAAQ,GAAG,EAE1B,EAAO,SAAS,OAAS,IAC3B,EAAA,YAAY,SAAS,EAAO,SAAS,OAAS,GAAG,QAAQ,CAE7D,CAEA,GAAI,EAAO,OAAO,OAAS,EAAG,CAC5B,EAAA,YAAY,aAAa,EAAO,OAAO,OAAO,yBAAyB,EACvE,IAAK,IAAM,KAAS,EAAO,OAAO,MAAM,EAAG,EAAE,EAC3C,EAAA,YAAY,KAAK,EAAM,GAAG,EAExB,EAAO,OAAO,OAAS,IACzB,EAAA,YAAY,SAAS,EAAO,OAAO,OAAS,GAAG,QAAQ,CAE3D,CACF,CAGA,SAAS,EAAiB,EAA4B,CACpD,IAAM,EAAoB,CAAC,EAC3B,GAAI,EAAO,cAAe,CACxB,IAAM,EAAK,EAAO,cAClB,EAAQ,KACN,kBAAkB,EAAG,KAAK,KAAK,EAAG,UAAU,GAAG,EAAG,QAAQ,cAAc,EAAG,oBAAoB,SAAS,EAAG,KAAK,YAAY,EAAG,cACjI,CACF,CACA,GAAI,EAAO,oBAAsB,EAAG,CAClC,IAAM,EACJ,EAAO,kBAAkB,OAAS,EAC9B,iBAAiB,EAAO,kBAAkB,MAAM,EAAG,CAAoB,CAAC,CAAC,KAAK,IAAI,IAAI,EAAO,kBAAkB,OAAS,EAAuB,QAAU,KACzJ,GACN,EAAQ,KACN,oBAAoB,EAAO,oBAAoB,IAAI,KAAK,MAAM,EAAO,oBAAsB,GAAG,EAAE,GAAG,EAAe,EACpH,CACF,CAEA,OADA,EAAQ,KAAK,aAAa,EAAO,MAAM,EAChC,GAAG,EAAO,KAAK,UAAU,EAAO,MAAM,QAAQ,CAAC,EAAE,KAAK,EAAQ,KAAK,IAAI,GAChF,CAEA,SAAS,EAAU,EAMjB,CACA,IAAI,EAAgB,EAChB,EAAc,EACd,EAAyB,EACzB,EAAY,EAEhB,IAAK,IAAM,KAAQ,EACjB,GAAiB,EAAK,QAAQ,UAAU,OACxC,GAAe,EAAK,QAAQ,MAAM,KAClC,EAAyB,KAAK,IAAI,EAAwB,EAAK,QAAQ,sBAAsB,EAC7F,GAAa,EAAK,QAAQ,UAG5B,MAAO,CAAE,UAAW,EAAM,OAAQ,gBAAe,cAAa,yBAAwB,WAAU,CAClG,CAEA,SAAS,EAAW,EAAuB,CACzC,GAAI,IAAU,MACZ,MAAO,KAET,IAAM,EAAS,EAAM,KAAK,IAAM,GAAK,IAAa,OAAO,CAAK,EAC9D,GAAI,CAAC,OAAO,SAAS,CAAM,GAAK,EAAS,EACvC,MAAM,IAAIC,EAAAA,qBAAqB,0CAA0C,EAE3E,OAAO,CACT,CAEA,SAAS,EAAoB,EAAuB,CAClD,IAAM,EAAS,EAA2B,CAAK,EAC/C,GAAI,EAAS,IACX,MAAM,IAAIA,EAAAA,qBAAqB,wCAAwC,EAEzE,OAAO,CACT,CAEA,SAAS,EAAwB,EAAuB,CACtD,GAAI,CAAC,SAAS,KAAK,CAAK,EACtB,MAAM,IAAIA,EAAAA,qBAAqB,kCAAkC,EAGnE,IAAM,EAAS,OAAO,CAAK,EAC3B,GAAI,CAAC,OAAO,cAAc,CAAM,GAAK,EAAS,EAC5C,MAAM,IAAIA,EAAAA,qBAAqB,kCAAkC,EAEnE,OAAO,CACT,CAEA,SAAS,EAA2B,EAAuB,CACzD,IAAM,GAAA,EAASC,EAAAA,qBAAAA,CAAqB,CAAK,EACzC,GAAI,IAAW,IAAA,GACb,MAAM,IAAID,EAAAA,qBAAqB,8BAA8B,EAE/D,OAAO,CACT"}
+{"version":3,"file":"cli.cjs","names":["getErrorMessage","Command","levels","thresholds","runCheckCommand","configFileName","duplicationThreshold","fileThresholds","functionThresholds","resolveTarget","configSearchDirectory","loadConfig","resolveOptions","scanTarget","loadExclusion","formatPath","collectDuplicatedLineNumbers","qualifyFunctionName","InvalidArgumentError","parsePositiveInteger"],"sources":["../src/cli.ts"],"sourcesContent":["#!/usr/bin/env node\n\nimport { Command, InvalidArgumentError } from 'commander';\nimport { getErrorMessage, parsePositiveInteger } from '@willbooster/shared-lib';\nimport { type CliOptions, configFileName, loadConfig, type ResolvedOptions, resolveOptions } from './cliConfig.js';\nimport type { CrossFileDuplicationMetrics } from './crossFileDuplication.js';\nimport { runCheckCommand, type CheckCliOptions } from './checkCommand.js';\nimport { loadExclusion } from './exclusion.js';\nimport {\n  addCrossFileDuplication,\n  collectDuplicatedLineNumbers,\n  configSearchDirectory,\n  formatPath,\n  resolveTarget,\n  scanTarget,\n  writeStderr,\n  writeStdout,\n  type FileMetrics,\n  type ScanResult,\n} from './scan.js';\nimport {\n  duplicationThreshold,\n  fileThresholds,\n  functionThresholds,\n  levels,\n  thresholds,\n  type Level,\n  type Limits,\n} from './thresholds.js';\nimport { qualifyFunctionName } from './thresholdCheck.js';\nimport type { FunctionMetrics } from './types.js';\n\n/** The worst (highest-cognitive-complexity) function of a file, reported as the ranking evidence. */\ninterface WorstFunction {\n  name: string;\n  startLine: number;\n  endLine: number;\n  cognitiveComplexity: number;\n  ncss: number;\n  nestingDepth: number;\n}\n\n/**\n * One ranked refactoring candidate. The score is the sum of the file's repo-relative percentile\n * ranks (each in [0, 1)) over three dimensions: worst-function cognitive complexity, duplicated\n * lines (within-file and cross-file combined), and file NCSS. Ranking is relative to the scanned\n * project, so no absolute threshold is involved.\n */\ninterface RankedFile {\n  file: string;\n  score: number;\n  worstFunction?: WorstFunction;\n  /** Distinct lines covered by within-file duplicate blocks or cross-file duplicate occurrences. */\n  duplicatedLineCount: number;\n  /** duplicatedLineCount / code lines (0 when the file has no code). */\n  duplicatedLineRatio: number;\n  /** Other files sharing cross-file duplicate blocks; filled only for the reported top files. */\n  crossFilePartners: string[];\n  ncss: number;\n  codeLines: number;\n}\n\n/** Caps how many cross-file partners a ranked file lists so the report stays scannable. */\nconst maxCrossFilePartners = 3;\n\n// oxlint-disable-next-line unicorn/prefer-top-level-await -- CommonJS build output cannot preserve top-level await.\nvoid main().catch((error: unknown) => {\n  writeStderr(`Error: ${getErrorMessage(error)}\\n`);\n  process.exitCode = 1;\n});\n\nasync function main(): Promise<void> {\n  const program = new Command()\n    .name('code-gauge')\n    .description('Measure code metrics: rank refactoring candidates (the default command) or check thresholds.');\n\n  addSharedOptions(\n    program\n      .command('rank', { isDefault: true })\n      .description('Rank the files of a project by refactoring priority')\n      .argument('[target]', 'file or directory to measure', '.')\n  )\n    .option('--top <number>', 'number of top-ranked files to report (default: 10)', parsePositiveIntegerOption)\n    .option('--fail-on-error', 'exit with code 1 when files or directories cannot be scanned')\n    .action(runRankCommand);\n\n  const check = addSharedOptions(\n    program\n      .command('check')\n      .description(\n        'Report every exceeded threshold as a warning or an error (exit 1 on errors, 2 when files cannot be measured); with --base, only in what the change touches'\n      )\n      .argument('[target]', 'file or directory to check', '.')\n  ).option('--base <ref>', 'check only what the working tree changed since the merge-base of this git ref and HEAD');\n  for (const level of levels) {\n    for (const { key, defaultLimits } of thresholds) {\n      const defaultLimit = defaultLimits[level];\n      check.option(\n        `--${level}-${toKebabCase(key)} <number>`,\n        `${describeLimit(key, level)} (default ${Number.isFinite(defaultLimit) ? defaultLimit : 'off'}); \"off\" disables it`,\n        parseLimit\n      );\n    }\n  }\n  // Exit code 1 means error-level violations, so a rejected command line must not use it.\n  check.exitOverride((error) => process.exit(error.exitCode === 0 ? 0 : 2));\n  check.action(async (target: string, cliOptions: CheckCliOptions & Record<string, unknown>) => {\n    const readLimits = (level: Level): Limits =>\n      Object.fromEntries(\n        thresholds.flatMap(({ key }) => {\n          const limit = cliOptions[`${level}${key.charAt(0).toUpperCase()}${key.slice(1)}`];\n          return limit === undefined ? [] : [[key, limit as number]];\n        })\n      );\n    await runCheckCommand(target, cliOptions, { warning: readLimits('warning'), error: readLimits('error') });\n  });\n\n  await program.parseAsync();\n}\n\n/** Registers the options shared by the ranking and check commands. */\nfunction addSharedOptions(command: Command): Command {\n  return command\n    .option('--config <path>', `config file to use instead of the auto-detected ${configFileName}`)\n    .option(\n      '--duplication-min-tokens <number>',\n      'minimum normalized token count for a duplicate region (default 40)',\n      parsePositiveIntegerOption\n    )\n    .option(\n      '--duplication-max-gap-tokens <number>',\n      'maximum token gap merged into one gapped clone group; 0 disables merging (default 30)',\n      parseNonNegativeInteger\n    )\n    .option(\n      '--duplication-min-similarity-percent <number>',\n      'minimum similarity percent (1-100) for near-miss (Type-3) clone blocks; 100 reports exact matches only (default 70)',\n      parsePercentInteger\n    )\n    .option('--include-tests', 'include test files and test directories')\n    .option('--json', 'print JSON output');\n}\n\nfunction describeLimit(key: string, level: Level): string {\n  if (key === duplicationThreshold.key) {\n    return `code-line count from which a duplicated block is ${level === 'error' ? 'an error' : 'a warning'}`;\n  }\n  const fileThreshold = fileThresholds.find((threshold) => threshold.key === key);\n  const subject = fileThreshold\n    ? fileThreshold.label\n    : `${functionThresholds.find((threshold) => threshold.key === key)?.label} of a function`;\n  return `maximum ${subject} without ${level === 'error' ? 'an error' : 'a warning'}`;\n}\n\nfunction toKebabCase(key: string): string {\n  return key.replaceAll(/[A-Z]/gu, (letter) => `-${letter.toLowerCase()}`);\n}\n\nasync function runRankCommand(target: string, cliOptions: CliOptions): Promise<void> {\n  const resolvedTarget = resolveTarget(target);\n  const searchDirectory = await configSearchDirectory(resolvedTarget);\n  const config = await loadConfig(cliOptions.config, searchDirectory);\n  const options = resolveOptions(cliOptions, config);\n  const result = await scanTarget(resolvedTarget, {\n    ...options,\n    loadExclusion: (absolutePaths) => loadExclusion(searchDirectory, options.exclude, absolutePaths),\n  });\n  addCrossFileDuplication(result, options);\n  const rankedFiles = rankFiles(result, options.top);\n\n  if (options.json) {\n    printJson(result, rankedFiles, options);\n  } else {\n    printTextReport(resolvedTarget, result, rankedFiles, options);\n  }\n\n  if (result.fatalError || (options.failOnError && result.errors.length > 0)) {\n    process.exitCode = 1;\n  }\n}\n\nfunction rankFiles(result: ScanResult, top: number): RankedFile[] {\n  const candidates = result.files.map(({ file, metrics }) => {\n    const formattedFile = formatPath(file, result.displayRoot);\n    const duplicatedLines = collectDuplicatedLineNumbers(metrics, result.crossFileDuplication, formattedFile);\n    return {\n      file: formattedFile,\n      worstFunction: findWorstFunction(metrics.functions),\n      duplicatedLineCount: duplicatedLines.size,\n      duplicatedLineRatio: metrics.lines.code === 0 ? 0 : duplicatedLines.size / metrics.lines.code,\n      crossFilePartners: [] as string[],\n      ncss: metrics.ncssCount,\n      codeLines: metrics.lines.code,\n    };\n  });\n\n  const cognitivePercentile = makePercentile(candidates.map((c) => c.worstFunction?.cognitiveComplexity ?? 0));\n  const duplicationPercentile = makePercentile(candidates.map((c) => c.duplicatedLineCount));\n  const ncssPercentile = makePercentile(candidates.map((c) => c.ncss));\n\n  const ranked = candidates\n    .map((candidate) => ({\n      ...candidate,\n      score:\n        cognitivePercentile(candidate.worstFunction?.cognitiveComplexity ?? 0) +\n        duplicationPercentile(candidate.duplicatedLineCount) +\n        ncssPercentile(candidate.ncss),\n    }))\n    .toSorted(\n      (left, right) => right.score - left.score || right.ncss - left.ncss || left.file.localeCompare(right.file)\n    );\n  // Partner evidence is attached only to the files that will be reported: expanding partners for\n  // every scanned file first would retain TH(F^2) strings when one clone group spans F files, and\n  // rescanning all groups per file would be O(files x groups) on the post-scan ranking step.\n  attachCrossFilePartners(ranked.slice(0, top), result.crossFileDuplication);\n  return ranked;\n}\n\n/** One pass over the groups fills the reported files' partner lists (other files sharing a group). */\nfunction attachCrossFilePartners(\n  reportedFiles: RankedFile[],\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined\n): void {\n  if (!crossFileDuplication) {\n    return;\n  }\n  const partnersByFile = new Map(reportedFiles.map((ranked) => [ranked.file, new Set<string>()]));\n  for (const group of crossFileDuplication.groups) {\n    for (const file of group.files) {\n      const partners = partnersByFile.get(file);\n      if (!partners) {\n        continue;\n      }\n      for (const partner of group.files) {\n        if (partner !== file) {\n          partners.add(partner);\n        }\n      }\n    }\n  }\n  for (const ranked of reportedFiles) {\n    ranked.crossFilePartners = [...(partnersByFile.get(ranked.file) ?? [])].toSorted();\n  }\n}\n\n/**\n * Percentile rank within the scanned project: the fraction of files with a strictly smaller\n * value, in [0, 1). Relative ranking needs no absolute threshold, which sidesteps the metric\n * calibration problem entirely — the top of the list is worth refactoring first regardless of\n * where any cutoff would sit.\n */\nfunction makePercentile(values: number[]): (value: number) => number {\n  const sorted = values.toSorted((left, right) => left - right);\n  return (value: number) => {\n    let low = 0;\n    let high = sorted.length;\n    while (low < high) {\n      const middle = (low + high) >> 1;\n      if ((sorted[middle] as number) < value) {\n        low = middle + 1;\n      } else {\n        high = middle;\n      }\n    }\n    return sorted.length === 0 ? 0 : low / sorted.length;\n  };\n}\n\n/** The highest-cognitive-complexity function; NCSS breaks ties so the larger body is reported. */\nfunction findWorstFunction(functions: FunctionMetrics[]): WorstFunction | undefined {\n  let worst: FunctionMetrics | undefined;\n  for (const fn of functions) {\n    if (\n      !worst ||\n      fn.cognitiveComplexity > worst.cognitiveComplexity ||\n      (fn.cognitiveComplexity === worst.cognitiveComplexity && fn.ncss > worst.ncss)\n    ) {\n      worst = fn;\n    }\n  }\n  if (!worst) {\n    return undefined;\n  }\n  return {\n    name: qualifyFunctionName(worst),\n    startLine: worst.startLine,\n    endLine: worst.endLine,\n    cognitiveComplexity: worst.cognitiveComplexity,\n    ncss: worst.ncss,\n    nestingDepth: worst.nestingDepth,\n  };\n}\n\nfunction printJson(result: ScanResult, rankedFiles: RankedFile[], options: ResolvedOptions): void {\n  const reportedFiles = rankedFiles.slice(0, options.top);\n  writeStdout(\n    JSON.stringify(\n      {\n        summary: summarize(result.files),\n        totalRankedFiles: rankedFiles.length,\n        truncated: reportedFiles.length < rankedFiles.length,\n        files: reportedFiles,\n        errors: result.errors,\n        warnings: result.warnings,\n      },\n      undefined,\n      2\n    ) + '\\n'\n  );\n}\n\nfunction printTextReport(\n  target: string,\n  result: ScanResult,\n  rankedFiles: RankedFile[],\n  options: ResolvedOptions\n): void {\n  if (result.fatalError) {\n    writeStderr(`Error: ${result.fatalError}\\n`);\n    return;\n  }\n\n  const summary = summarize(result.files);\n  writeStdout(\n    `Measured ${summary.fileCount} files under ${target} (code LOC ${summary.linesOfCode}, NCSS ${summary.ncssCount}, functions ${summary.functionCount})\\n`\n  );\n\n  if (rankedFiles.length === 0) {\n    writeStdout('No measurable files found.\\n');\n  } else {\n    const reportedFiles = rankedFiles.slice(0, options.top);\n    const totalSuffix = rankedFiles.length > reportedFiles.length ? ` of ${rankedFiles.length}` : '';\n    writeStdout(`\\nRefactoring candidates (top ${reportedFiles.length}${totalSuffix}):\\n`);\n    for (const [index, ranked] of reportedFiles.entries()) {\n      writeStdout(`${index + 1}. ${formatRankedFile(ranked)}\\n`);\n    }\n  }\n\n  if (result.warnings.length > 0) {\n    writeStderr(`\\nDegraded ${result.warnings.length} files (measured, but excluded from cross-file matching):\\n`);\n    for (const warning of result.warnings.slice(0, 10)) {\n      writeStderr(`- ${warning}\\n`);\n    }\n    if (result.warnings.length > 10) {\n      writeStderr(`- ... ${result.warnings.length - 10} more\\n`);\n    }\n  }\n\n  if (result.errors.length > 0) {\n    writeStderr(`\\nSkipped ${result.errors.length} files or directories:\\n`);\n    for (const error of result.errors.slice(0, 10)) {\n      writeStderr(`- ${error}\\n`);\n    }\n    if (result.errors.length > 10) {\n      writeStderr(`- ... ${result.errors.length - 10} more\\n`);\n    }\n  }\n}\n\n/** One ranked file as a single line: the location, the evidence, and where the duplication points. */\nfunction formatRankedFile(ranked: RankedFile): string {\n  const reasons: string[] = [];\n  if (ranked.worstFunction) {\n    const fn = ranked.worstFunction;\n    reasons.push(\n      `worst function ${fn.name} (L${fn.startLine}-${fn.endLine}) cognitive ${fn.cognitiveComplexity}, NCSS ${fn.ncss}, nesting ${fn.nestingDepth}`\n    );\n  }\n  if (ranked.duplicatedLineCount > 0) {\n    const partnersSuffix =\n      ranked.crossFilePartners.length > 0\n        ? `, shared with ${ranked.crossFilePartners.slice(0, maxCrossFilePartners).join(', ')}${ranked.crossFilePartners.length > maxCrossFilePartners ? ', ...' : ''}`\n        : '';\n    reasons.push(\n      `duplicated lines ${ranked.duplicatedLineCount} (${Math.round(ranked.duplicatedLineRatio * 100)}%${partnersSuffix})`\n    );\n  }\n  reasons.push(`file NCSS ${ranked.ncss}`);\n  return `${ranked.file} (score ${ranked.score.toFixed(2)}): ${reasons.join('; ')}`;\n}\n\nfunction summarize(files: FileMetrics[]): {\n  fileCount: number;\n  functionCount: number;\n  linesOfCode: number;\n  maxCognitiveComplexity: number;\n  ncssCount: number;\n} {\n  let functionCount = 0;\n  let linesOfCode = 0;\n  let maxCognitiveComplexity = 0;\n  let ncssCount = 0;\n\n  for (const file of files) {\n    functionCount += file.metrics.functions.length;\n    linesOfCode += file.metrics.lines.code;\n    maxCognitiveComplexity = Math.max(maxCognitiveComplexity, file.metrics.maxCognitiveComplexity);\n    ncssCount += file.metrics.ncssCount;\n  }\n\n  return { fileCount: files.length, functionCount, linesOfCode, maxCognitiveComplexity, ncssCount };\n}\n\nfunction parseLimit(value: string): number {\n  if (value === 'off') {\n    return Infinity;\n  }\n  const parsed = value.trim() === '' ? Number.NaN : Number(value);\n  if (!Number.isFinite(parsed) || parsed < 0) {\n    throw new InvalidArgumentError('Expected a non-negative number or \"off\".');\n  }\n  return parsed;\n}\n\nfunction parsePercentInteger(value: string): number {\n  const parsed = parsePositiveIntegerOption(value);\n  if (parsed > 100) {\n    throw new InvalidArgumentError('Expected an integer between 1 and 100.');\n  }\n  return parsed;\n}\n\nfunction parseNonNegativeInteger(value: string): number {\n  if (!/^\\d+$/u.test(value)) {\n    throw new InvalidArgumentError('Expected a non-negative integer.');\n  }\n\n  const parsed = Number(value);\n  if (!Number.isSafeInteger(parsed) || parsed < 0) {\n    throw new InvalidArgumentError('Expected a non-negative integer.');\n  }\n  return parsed;\n}\n\nfunction parsePositiveIntegerOption(value: string): number {\n  const parsed = parsePositiveInteger(value);\n  if (parsed === undefined) {\n    throw new InvalidArgumentError('Expected a positive integer.');\n  }\n  return parsed;\n}\n"],"mappings":";qQAkEK,EAAK,CAAC,CAAC,MAAO,GAAmB,CACpC,EAAA,YAAY,WAAA,EAAUA,EAAAA,gBAAAA,CAAgB,CAAK,EAAE,GAAG,EAChD,QAAQ,SAAW,CACrB,CAAC,EAED,eAAe,GAAsB,CACnC,IAAM,EAAU,IAAIC,EAAAA,QAAQ,CAAC,CAC1B,KAAK,YAAY,CAAC,CAClB,YAAY,8FAA8F,EAE7G,EACE,EACG,QAAQ,OAAQ,CAAE,UAAW,EAAK,CAAC,CAAC,CACpC,YAAY,qDAAqD,CAAC,CAClE,SAAS,WAAY,+BAAgC,GAAG,CAC7D,CAAC,CACE,OAAO,iBAAkB,qDAAsD,CAA0B,CAAC,CAC1G,OAAO,kBAAmB,8DAA8D,CAAC,CACzF,OAAO,CAAc,EAExB,IAAM,EAAQ,EACZ,EACG,QAAQ,OAAO,CAAC,CAChB,YACC,4JACF,CAAC,CACA,SAAS,WAAY,6BAA8B,GAAG,CAC3D,CAAC,CAAC,OAAO,eAAgB,wFAAwF,EACjH,IAAK,IAAM,KAASC,EAAAA,OAClB,IAAK,GAAM,CAAE,MAAK,mBAAmBC,EAAAA,WAAY,CAC/C,IAAM,EAAe,EAAc,GACnC,EAAM,OACJ,KAAK,EAAM,GAAG,EAAY,CAAG,EAAE,WAC/B,GAAG,EAAc,EAAK,CAAK,EAAE,YAAY,OAAO,SAAS,CAAY,EAAI,EAAe,MAAM,sBAC9F,CACF,CACF,CAGF,EAAM,aAAc,GAAU,QAAQ,KAAK,EAAM,WAAa,EAAI,EAAI,CAAC,CAAC,EACxE,EAAM,OAAO,MAAO,EAAgB,IAA0D,CAC5F,IAAM,EAAc,GAClB,OAAO,YACLA,EAAAA,WAAW,SAAS,CAAE,SAAU,CAC9B,IAAM,EAAQ,EAAW,GAAG,IAAQ,EAAI,OAAO,CAAC,CAAC,CAAC,YAAY,IAAI,EAAI,MAAM,CAAC,KAC7E,OAAO,IAAU,IAAA,GAAY,CAAC,EAAI,CAAC,CAAC,EAAK,CAAe,CAAC,CAC3D,CAAC,CACH,EACF,MAAMC,EAAAA,gBAAgB,EAAQ,EAAY,CAAE,QAAS,EAAW,SAAS,EAAG,MAAO,EAAW,OAAO,CAAE,CAAC,CAC1G,CAAC,EAED,MAAM,EAAQ,WAAW,CAC3B,CAGA,SAAS,EAAiB,EAA2B,CACnD,OAAO,EACJ,OAAO,kBAAmB,mDAAmDC,EAAAA,gBAAgB,CAAC,CAC9F,OACC,oCACA,qEACA,CACF,CAAC,CACA,OACC,wCACA,wFACA,CACF,CAAC,CACA,OACC,gDACA,sHACA,CACF,CAAC,CACA,OAAO,kBAAmB,yCAAyC,CAAC,CACpE,OAAO,SAAU,mBAAmB,CACzC,CAEA,SAAS,EAAc,EAAa,EAAsB,CACxD,GAAI,IAAQC,EAAAA,qBAAqB,IAC/B,MAAO,oDAAoD,IAAU,QAAU,WAAa,cAE9F,IAAM,EAAgBC,EAAAA,eAAe,KAAM,GAAc,EAAU,MAAQ,CAAG,EAI9E,MAAO,WAHS,EACZ,EAAc,MACd,GAAGC,EAAAA,mBAAmB,KAAM,GAAc,EAAU,MAAQ,CAAG,CAAC,EAAE,MAAM,gBAClD,WAAW,IAAU,QAAU,WAAa,aACxE,CAEA,SAAS,EAAY,EAAqB,CACxC,OAAO,EAAI,WAAW,UAAY,GAAW,IAAI,EAAO,YAAY,GAAG,CACzE,CAEA,eAAe,EAAe,EAAgB,EAAuC,CACnF,IAAM,EAAiBC,EAAAA,cAAc,CAAM,EACrC,EAAkB,MAAMC,EAAAA,sBAAsB,CAAc,EAC5D,EAAS,MAAMC,EAAAA,WAAW,EAAW,OAAQ,CAAe,EAC5D,EAAUC,EAAAA,eAAe,EAAY,CAAM,EAC3C,EAAS,MAAMC,EAAAA,WAAW,EAAgB,CAC9C,GAAG,EACH,cAAgB,GAAkBC,EAAAA,cAAc,EAAiB,EAAQ,QAAS,CAAa,CACjG,CAAC,EACD,EAAA,wBAAwB,EAAQ,CAAO,EACvC,IAAM,EAAc,EAAU,EAAQ,EAAQ,GAAG,EAE7C,EAAQ,KACV,EAAU,EAAQ,EAAa,CAAO,EAEtC,EAAgB,EAAgB,EAAQ,EAAa,CAAO,GAG1D,EAAO,YAAe,EAAQ,aAAe,EAAO,OAAO,OAAS,KACtE,QAAQ,SAAW,EAEvB,CAEA,SAAS,EAAU,EAAoB,EAA2B,CAChE,IAAM,EAAa,EAAO,MAAM,KAAK,CAAE,OAAM,aAAc,CACzD,IAAM,EAAgBC,EAAAA,WAAW,EAAM,EAAO,WAAW,EACnD,EAAkBC,EAAAA,6BAA6B,EAAS,EAAO,qBAAsB,CAAa,EACxG,MAAO,CACL,KAAM,EACN,cAAe,EAAkB,EAAQ,SAAS,EAClD,oBAAqB,EAAgB,KACrC,oBAAqB,EAAQ,MAAM,OAAS,EAAI,EAAI,EAAgB,KAAO,EAAQ,MAAM,KACzF,kBAAmB,CAAC,EACpB,KAAM,EAAQ,UACd,UAAW,EAAQ,MAAM,IAC3B,CACF,CAAC,EAEK,EAAsB,EAAe,EAAW,IAAK,GAAM,EAAE,eAAe,qBAAuB,CAAC,CAAC,EACrG,EAAwB,EAAe,EAAW,IAAK,GAAM,EAAE,mBAAmB,CAAC,EACnF,EAAiB,EAAe,EAAW,IAAK,GAAM,EAAE,IAAI,CAAC,EAE7D,EAAS,EACZ,IAAK,IAAe,CACnB,GAAG,EACH,MACE,EAAoB,EAAU,eAAe,qBAAuB,CAAC,EACrE,EAAsB,EAAU,mBAAmB,EACnD,EAAe,EAAU,IAAI,CACjC,EAAE,CAAC,CACF,UACE,EAAM,IAAU,EAAM,MAAQ,EAAK,OAAS,EAAM,KAAO,EAAK,MAAQ,EAAK,KAAK,cAAc,EAAM,IAAI,CAC3G,EAKF,OADA,EAAwB,EAAO,MAAM,EAAG,CAAG,EAAG,EAAO,oBAAoB,EAClE,CACT,CAGA,SAAS,EACP,EACA,EACM,CACN,GAAI,CAAC,EACH,OAEF,IAAM,EAAiB,IAAI,IAAI,EAAc,IAAK,GAAW,CAAC,EAAO,qBAAM,IAAI,GAAa,CAAC,CAAC,EAC9F,IAAK,IAAM,KAAS,EAAqB,OACvC,IAAK,IAAM,KAAQ,EAAM,MAAO,CAC9B,IAAM,EAAW,EAAe,IAAI,CAAI,EACnC,KAGL,IAAK,IAAM,KAAW,EAAM,MACtB,IAAY,GACd,EAAS,IAAI,CAAO,CAG1B,CAEF,IAAK,IAAM,KAAU,EACnB,EAAO,kBAAoB,CAAC,GAAI,EAAe,IAAI,EAAO,IAAI,GAAK,CAAC,CAAE,CAAC,CAAC,SAAS,CAErF,CAQA,SAAS,EAAe,EAA6C,CACnE,IAAM,EAAS,EAAO,UAAU,EAAM,IAAU,EAAO,CAAK,EAC5D,MAAQ,IAAkB,CACxB,IAAI,EAAM,EACN,EAAO,EAAO,OAClB,KAAO,EAAM,GAAM,CACjB,IAAM,EAAU,EAAM,GAAS,EAC1B,EAAO,GAAqB,EAC/B,EAAM,EAAS,EAEf,EAAO,CAEX,CACA,OAAO,EAAO,SAAW,EAAI,EAAI,EAAM,EAAO,MAChD,CACF,CAGA,SAAS,EAAkB,EAAyD,CAClF,IAAI,EACJ,IAAK,IAAM,KAAM,GAEb,CAAC,GACD,EAAG,oBAAsB,EAAM,qBAC9B,EAAG,sBAAwB,EAAM,qBAAuB,EAAG,KAAO,EAAM,QAEzE,EAAQ,GAGP,KAGL,MAAO,CACL,KAAMC,EAAAA,oBAAoB,CAAK,EAC/B,UAAW,EAAM,UACjB,QAAS,EAAM,QACf,oBAAqB,EAAM,oBAC3B,KAAM,EAAM,KACZ,aAAc,EAAM,YACtB,CACF,CAEA,SAAS,EAAU,EAAoB,EAA2B,EAAgC,CAChG,IAAM,EAAgB,EAAY,MAAM,EAAG,EAAQ,GAAG,EACtD,EAAA,YACE,KAAK,UACH,CACE,QAAS,EAAU,EAAO,KAAK,EAC/B,iBAAkB,EAAY,OAC9B,UAAW,EAAc,OAAS,EAAY,OAC9C,MAAO,EACP,OAAQ,EAAO,OACf,SAAU,EAAO,QACnB,EACA,IAAA,GACA,CACF,EAAI;CACN,CACF,CAEA,SAAS,EACP,EACA,EACA,EACA,EACM,CACN,GAAI,EAAO,WAAY,CACrB,EAAA,YAAY,UAAU,EAAO,WAAW,GAAG,EAC3C,MACF,CAEA,IAAM,EAAU,EAAU,EAAO,KAAK,EAKtC,GAJA,EAAA,YACE,YAAY,EAAQ,UAAU,eAAe,EAAO,aAAa,EAAQ,YAAY,SAAS,EAAQ,UAAU,cAAc,EAAQ,cAAc,IACtJ,EAEI,EAAY,SAAW,EACzB,EAAA,YAAY;CAA8B,MACrC,CACL,IAAM,EAAgB,EAAY,MAAM,EAAG,EAAQ,GAAG,EAChD,EAAc,EAAY,OAAS,EAAc,OAAS,OAAO,EAAY,SAAW,GAC9F,EAAA,YAAY,iCAAiC,EAAc,SAAS,EAAY,KAAK,EACrF,IAAK,GAAM,CAAC,EAAO,KAAW,EAAc,QAAQ,EAClD,EAAA,YAAY,GAAG,EAAQ,EAAE,IAAI,EAAiB,CAAM,EAAE,GAAG,CAE7D,CAEA,GAAI,EAAO,SAAS,OAAS,EAAG,CAC9B,EAAA,YAAY,cAAc,EAAO,SAAS,OAAO,4DAA4D,EAC7G,IAAK,IAAM,KAAW,EAAO,SAAS,MAAM,EAAG,EAAE,EAC/C,EAAA,YAAY,KAAK,EAAQ,GAAG,EAE1B,EAAO,SAAS,OAAS,IAC3B,EAAA,YAAY,SAAS,EAAO,SAAS,OAAS,GAAG,QAAQ,CAE7D,CAEA,GAAI,EAAO,OAAO,OAAS,EAAG,CAC5B,EAAA,YAAY,aAAa,EAAO,OAAO,OAAO,yBAAyB,EACvE,IAAK,IAAM,KAAS,EAAO,OAAO,MAAM,EAAG,EAAE,EAC3C,EAAA,YAAY,KAAK,EAAM,GAAG,EAExB,EAAO,OAAO,OAAS,IACzB,EAAA,YAAY,SAAS,EAAO,OAAO,OAAS,GAAG,QAAQ,CAE3D,CACF,CAGA,SAAS,EAAiB,EAA4B,CACpD,IAAM,EAAoB,CAAC,EAC3B,GAAI,EAAO,cAAe,CACxB,IAAM,EAAK,EAAO,cAClB,EAAQ,KACN,kBAAkB,EAAG,KAAK,KAAK,EAAG,UAAU,GAAG,EAAG,QAAQ,cAAc,EAAG,oBAAoB,SAAS,EAAG,KAAK,YAAY,EAAG,cACjI,CACF,CACA,GAAI,EAAO,oBAAsB,EAAG,CAClC,IAAM,EACJ,EAAO,kBAAkB,OAAS,EAC9B,iBAAiB,EAAO,kBAAkB,MAAM,EAAG,CAAoB,CAAC,CAAC,KAAK,IAAI,IAAI,EAAO,kBAAkB,OAAS,EAAuB,QAAU,KACzJ,GACN,EAAQ,KACN,oBAAoB,EAAO,oBAAoB,IAAI,KAAK,MAAM,EAAO,oBAAsB,GAAG,EAAE,GAAG,EAAe,EACpH,CACF,CAEA,OADA,EAAQ,KAAK,aAAa,EAAO,MAAM,EAChC,GAAG,EAAO,KAAK,UAAU,EAAO,MAAM,QAAQ,CAAC,EAAE,KAAK,EAAQ,KAAK,IAAI,GAChF,CAEA,SAAS,EAAU,EAMjB,CACA,IAAI,EAAgB,EAChB,EAAc,EACd,EAAyB,EACzB,EAAY,EAEhB,IAAK,IAAM,KAAQ,EACjB,GAAiB,EAAK,QAAQ,UAAU,OACxC,GAAe,EAAK,QAAQ,MAAM,KAClC,EAAyB,KAAK,IAAI,EAAwB,EAAK,QAAQ,sBAAsB,EAC7F,GAAa,EAAK,QAAQ,UAG5B,MAAO,CAAE,UAAW,EAAM,OAAQ,gBAAe,cAAa,yBAAwB,WAAU,CAClG,CAEA,SAAS,EAAW,EAAuB,CACzC,GAAI,IAAU,MACZ,MAAO,KAET,IAAM,EAAS,EAAM,KAAK,IAAM,GAAK,IAAa,OAAO,CAAK,EAC9D,GAAI,CAAC,OAAO,SAAS,CAAM,GAAK,EAAS,EACvC,MAAM,IAAIC,EAAAA,qBAAqB,0CAA0C,EAE3E,OAAO,CACT,CAEA,SAAS,EAAoB,EAAuB,CAClD,IAAM,EAAS,EAA2B,CAAK,EAC/C,GAAI,EAAS,IACX,MAAM,IAAIA,EAAAA,qBAAqB,wCAAwC,EAEzE,OAAO,CACT,CAEA,SAAS,EAAwB,EAAuB,CACtD,GAAI,CAAC,SAAS,KAAK,CAAK,EACtB,MAAM,IAAIA,EAAAA,qBAAqB,kCAAkC,EAGnE,IAAM,EAAS,OAAO,CAAK,EAC3B,GAAI,CAAC,OAAO,cAAc,CAAM,GAAK,EAAS,EAC5C,MAAM,IAAIA,EAAAA,qBAAqB,kCAAkC,EAEnE,OAAO,CACT,CAEA,SAAS,EAA2B,EAAuB,CACzD,IAAM,GAAA,EAASC,EAAAA,qBAAAA,CAAqB,CAAK,EACzC,GAAI,IAAW,IAAA,GACb,MAAM,IAAID,EAAAA,qBAAqB,8BAA8B,EAE/D,OAAO,CACT"}

package/dist/cli.js

@@ -1,5 +1,5 @@
 #!/usr/bin/env node
-import{duplicationThreshold as e,fileThresholds as t,functionThresholds as n,levels as r,thresholds as i}from"./thresholds.js";import{configFileName as a,loadConfig as o,resolveOptions as s}from"./cliConfig.js";import{loadExclusion as c}from"./exclusion.js";import{addCrossFileDuplication as l,collectDuplicatedLineNumbers as u,configSearchDirectory as d,formatPath as f,resolveTarget as p,scanTarget as m,writeStderr as h,writeStdout as g}from"./scan.js";import{runCheckCommand as _}from"./checkCommand.js";import{getErrorMessage as v,parsePositiveInteger as y}from"@willbooster/shared-lib";import{Command as b,InvalidArgumentError as x}from"commander";S().catch(e=>{h(`Error: ${v(e)}\n`),process.exitCode=1});async function S(){let e=new b().name(`code-gauge`).description(`Measure code metrics: rank refactoring candidates (the default command) or check thresholds.`);C(e.command(`rank`,{isDefault:!0}).description(`Rank the files of a project by refactoring priority`).argument(`[target]`,`file or directory to measure`,`.`)).option(`--top <number>`,`number of top-ranked files to report (default: 10)`,R).option(`--fail-on-error`,`exit with code 1 when files or directories cannot be scanned`).action(E);let t=C(e.command(`check`).description(`Report every exceeded threshold as a warning or an error (exit 1 on errors, 2 when files cannot be measured); with --base, only in what the change touches`).argument(`[target]`,`file or directory to check`,`.`)).option(`--base <ref>`,`check only what the working tree changed since the merge-base of this git ref and HEAD`);for(let e of r)for(let{key:n,defaultLimits:r}of i){let i=r[e];t.option(`--${e}-${T(n)} <number>`,`${w(n,e)} (default ${Number.isFinite(i)?i:`off`}); "off" disables it`,F)}t.exitOverride(e=>process.exit(e.exitCode===0?0:2)),t.action(async(e,t)=>{let n=e=>Object.fromEntries(i.flatMap(({key:n})=>{let r=t[`${e}${n.charAt(0).toUpperCase()}${n.slice(1)}`];return r===void 0?[]:[[n,r]]}));await _(e,t,{warning:n(`warning`),error:n(`error`)})}),await e.parseAsync()}function C(e){return e.option(`--config <path>`,`config file to use instead of the auto-detected ${a}`).option(`--duplication-min-tokens <number>`,`minimum normalized token count for a duplicate region (default 40)`,R).option(`--duplication-max-gap-tokens <number>`,`maximum token gap merged into one gapped clone group; 0 disables merging (default 30)`,L).option(`--duplication-min-similarity-percent <number>`,`minimum similarity percent (1-100) for near-miss (Type-3) clone blocks; 100 reports exact matches only (default 70)`,I).option(`--include-tests`,`include test files and test directories`).option(`--json`,`print JSON output`)}function w(r,i){if(r===e.key)return`line count from which a duplicated block is ${i===`error`?`an error`:`a warning`}`;let a=t.find(e=>e.key===r);return`maximum ${a?a.label:`${n.find(e=>e.key===r)?.label} of a function`} without ${i===`error`?`an error`:`a warning`}`}function T(e){return e.replaceAll(/[A-Z]/gu,e=>`-${e.toLowerCase()}`)}async function E(e,t){let n=p(e),r=await d(n),i=await o(t.config,r),a=s(t,i),u=await m(n,{...a,loadExclusion:e=>c(r,a.exclude,e)});l(u,a);let f=D(u,a.top);a.json?j(u,f,a):M(n,u,f,a),(u.fatalError||a.failOnError&&u.errors.length>0)&&(process.exitCode=1)}function D(e,t){let n=e.files.map(({file:t,metrics:n})=>{let r=f(t,e.displayRoot),i=u(n,e.crossFileDuplication,r);return{file:r,worstFunction:A(n.functions),duplicatedLineCount:i.size,duplicatedLineRatio:n.lines.code===0?0:i.size/n.lines.code,crossFilePartners:[],ncss:n.ncssCount,codeLines:n.lines.code}}),r=k(n.map(e=>e.worstFunction?.cognitiveComplexity??0)),i=k(n.map(e=>e.duplicatedLineCount)),a=k(n.map(e=>e.ncss)),o=n.map(e=>({...e,score:r(e.worstFunction?.cognitiveComplexity??0)+i(e.duplicatedLineCount)+a(e.ncss)})).toSorted((e,t)=>t.score-e.score||t.ncss-e.ncss||e.file.localeCompare(t.file));return O(o.slice(0,t),e.crossFileDuplication),o}function O(e,t){if(!t)return;let n=new Map(e.map(e=>[e.file,/* @__PURE__ */ new Set]));for(let e of t.groups)for(let t of e.files){let r=n.get(t);if(r)for(let n of e.files)n!==t&&r.add(n)}for(let t of e)t.crossFilePartners=[...n.get(t.file)??[]].toSorted()}function k(e){let t=e.toSorted((e,t)=>e-t);return e=>{let n=0,r=t.length;for(;n<r;){let i=n+r>>1;t[i]<e?n=i+1:r=i}return t.length===0?0:n/t.length}}function A(e){let t;for(let n of e)(!t||n.cognitiveComplexity>t.cognitiveComplexity||n.cognitiveComplexity===t.cognitiveComplexity&&n.ncss>t.ncss)&&(t=n);if(t)return{name:t.name??`<anonymous>`,startLine:t.startLine,endLine:t.endLine,cognitiveComplexity:t.cognitiveComplexity,ncss:t.ncss,nestingDepth:t.nestingDepth}}function j(e,t,n){let r=t.slice(0,n.top);g(JSON.stringify({summary:P(e.files),totalRankedFiles:t.length,truncated:r.length<t.length,files:r,errors:e.errors,warnings:e.warnings},void 0,2)+`
-`)}function M(e,t,n,r){if(t.fatalError){h(`Error: ${t.fatalError}\n`);return}let i=P(t.files);if(g(`Measured ${i.fileCount} files under ${e} (code LOC ${i.linesOfCode}, NCSS ${i.ncssCount}, functions ${i.functionCount})\n`),n.length===0)g(`No measurable files found.
-`);else{let e=n.slice(0,r.top),t=n.length>e.length?` of ${n.length}`:``;g(`\nRefactoring candidates (top ${e.length}${t}):\n`);for(let[t,n]of e.entries())g(`${t+1}. ${N(n)}\n`)}if(t.warnings.length>0){h(`\nDegraded ${t.warnings.length} files (measured, but excluded from cross-file matching):\n`);for(let e of t.warnings.slice(0,10))h(`- ${e}\n`);t.warnings.length>10&&h(`- ... ${t.warnings.length-10} more\n`)}if(t.errors.length>0){h(`\nSkipped ${t.errors.length} files or directories:\n`);for(let e of t.errors.slice(0,10))h(`- ${e}\n`);t.errors.length>10&&h(`- ... ${t.errors.length-10} more\n`)}}function N(e){let t=[];if(e.worstFunction){let n=e.worstFunction;t.push(`worst function ${n.name} (L${n.startLine}-${n.endLine}) cognitive ${n.cognitiveComplexity}, NCSS ${n.ncss}, nesting ${n.nestingDepth}`)}if(e.duplicatedLineCount>0){let n=e.crossFilePartners.length>0?`, shared with ${e.crossFilePartners.slice(0,3).join(`, `)}${e.crossFilePartners.length>3?`, ...`:``}`:``;t.push(`duplicated lines ${e.duplicatedLineCount} (${Math.round(e.duplicatedLineRatio*100)}%${n})`)}return t.push(`file NCSS ${e.ncss}`),`${e.file} (score ${e.score.toFixed(2)}): ${t.join(`; `)}`}function P(e){let t=0,n=0,r=0,i=0;for(let a of e)t+=a.metrics.functions.length,n+=a.metrics.lines.code,r=Math.max(r,a.metrics.maxCognitiveComplexity),i+=a.metrics.ncssCount;return{fileCount:e.length,functionCount:t,linesOfCode:n,maxCognitiveComplexity:r,ncssCount:i}}function F(e){if(e===`off`)return 1/0;let t=e.trim()===``?NaN:Number(e);if(!Number.isFinite(t)||t<0)throw new x(`Expected a non-negative number or "off".`);return t}function I(e){let t=R(e);if(t>100)throw new x(`Expected an integer between 1 and 100.`);return t}function L(e){if(!/^\d+$/u.test(e))throw new x(`Expected a non-negative integer.`);let t=Number(e);if(!Number.isSafeInteger(t)||t<0)throw new x(`Expected a non-negative integer.`);return t}function R(e){let t=y(e);if(t===void 0)throw new x(`Expected a positive integer.`);return t}
+import{duplicationThreshold as e,fileThresholds as t,functionThresholds as n,levels as r,thresholds as i}from"./thresholds.js";import{configFileName as a,loadConfig as o,resolveOptions as s}from"./cliConfig.js";import{loadExclusion as c}from"./exclusion.js";import{addCrossFileDuplication as l,collectDuplicatedLineNumbers as u,configSearchDirectory as d,formatPath as f,resolveTarget as p,scanTarget as m,writeStderr as h,writeStdout as g}from"./scan.js";import{qualifyFunctionName as _}from"./thresholdCheck.js";import{runCheckCommand as v}from"./checkCommand.js";import{getErrorMessage as y,parsePositiveInteger as b}from"@willbooster/shared-lib";import{Command as x,InvalidArgumentError as S}from"commander";C().catch(e=>{h(`Error: ${y(e)}\n`),process.exitCode=1});async function C(){let e=new x().name(`code-gauge`).description(`Measure code metrics: rank refactoring candidates (the default command) or check thresholds.`);w(e.command(`rank`,{isDefault:!0}).description(`Rank the files of a project by refactoring priority`).argument(`[target]`,`file or directory to measure`,`.`)).option(`--top <number>`,`number of top-ranked files to report (default: 10)`,z).option(`--fail-on-error`,`exit with code 1 when files or directories cannot be scanned`).action(D);let t=w(e.command(`check`).description(`Report every exceeded threshold as a warning or an error (exit 1 on errors, 2 when files cannot be measured); with --base, only in what the change touches`).argument(`[target]`,`file or directory to check`,`.`)).option(`--base <ref>`,`check only what the working tree changed since the merge-base of this git ref and HEAD`);for(let e of r)for(let{key:n,defaultLimits:r}of i){let i=r[e];t.option(`--${e}-${E(n)} <number>`,`${T(n,e)} (default ${Number.isFinite(i)?i:`off`}); "off" disables it`,I)}t.exitOverride(e=>process.exit(e.exitCode===0?0:2)),t.action(async(e,t)=>{let n=e=>Object.fromEntries(i.flatMap(({key:n})=>{let r=t[`${e}${n.charAt(0).toUpperCase()}${n.slice(1)}`];return r===void 0?[]:[[n,r]]}));await v(e,t,{warning:n(`warning`),error:n(`error`)})}),await e.parseAsync()}function w(e){return e.option(`--config <path>`,`config file to use instead of the auto-detected ${a}`).option(`--duplication-min-tokens <number>`,`minimum normalized token count for a duplicate region (default 40)`,z).option(`--duplication-max-gap-tokens <number>`,`maximum token gap merged into one gapped clone group; 0 disables merging (default 30)`,R).option(`--duplication-min-similarity-percent <number>`,`minimum similarity percent (1-100) for near-miss (Type-3) clone blocks; 100 reports exact matches only (default 70)`,L).option(`--include-tests`,`include test files and test directories`).option(`--json`,`print JSON output`)}function T(r,i){if(r===e.key)return`code-line count from which a duplicated block is ${i===`error`?`an error`:`a warning`}`;let a=t.find(e=>e.key===r);return`maximum ${a?a.label:`${n.find(e=>e.key===r)?.label} of a function`} without ${i===`error`?`an error`:`a warning`}`}function E(e){return e.replaceAll(/[A-Z]/gu,e=>`-${e.toLowerCase()}`)}async function D(e,t){let n=p(e),r=await d(n),i=await o(t.config,r),a=s(t,i),u=await m(n,{...a,loadExclusion:e=>c(r,a.exclude,e)});l(u,a);let f=O(u,a.top);a.json?M(u,f,a):N(n,u,f,a),(u.fatalError||a.failOnError&&u.errors.length>0)&&(process.exitCode=1)}function O(e,t){let n=e.files.map(({file:t,metrics:n})=>{let r=f(t,e.displayRoot),i=u(n,e.crossFileDuplication,r);return{file:r,worstFunction:j(n.functions),duplicatedLineCount:i.size,duplicatedLineRatio:n.lines.code===0?0:i.size/n.lines.code,crossFilePartners:[],ncss:n.ncssCount,codeLines:n.lines.code}}),r=A(n.map(e=>e.worstFunction?.cognitiveComplexity??0)),i=A(n.map(e=>e.duplicatedLineCount)),a=A(n.map(e=>e.ncss)),o=n.map(e=>({...e,score:r(e.worstFunction?.cognitiveComplexity??0)+i(e.duplicatedLineCount)+a(e.ncss)})).toSorted((e,t)=>t.score-e.score||t.ncss-e.ncss||e.file.localeCompare(t.file));return k(o.slice(0,t),e.crossFileDuplication),o}function k(e,t){if(!t)return;let n=new Map(e.map(e=>[e.file,/* @__PURE__ */ new Set]));for(let e of t.groups)for(let t of e.files){let r=n.get(t);if(r)for(let n of e.files)n!==t&&r.add(n)}for(let t of e)t.crossFilePartners=[...n.get(t.file)??[]].toSorted()}function A(e){let t=e.toSorted((e,t)=>e-t);return e=>{let n=0,r=t.length;for(;n<r;){let i=n+r>>1;t[i]<e?n=i+1:r=i}return t.length===0?0:n/t.length}}function j(e){let t;for(let n of e)(!t||n.cognitiveComplexity>t.cognitiveComplexity||n.cognitiveComplexity===t.cognitiveComplexity&&n.ncss>t.ncss)&&(t=n);if(t)return{name:_(t),startLine:t.startLine,endLine:t.endLine,cognitiveComplexity:t.cognitiveComplexity,ncss:t.ncss,nestingDepth:t.nestingDepth}}function M(e,t,n){let r=t.slice(0,n.top);g(JSON.stringify({summary:F(e.files),totalRankedFiles:t.length,truncated:r.length<t.length,files:r,errors:e.errors,warnings:e.warnings},void 0,2)+`
+`)}function N(e,t,n,r){if(t.fatalError){h(`Error: ${t.fatalError}\n`);return}let i=F(t.files);if(g(`Measured ${i.fileCount} files under ${e} (code LOC ${i.linesOfCode}, NCSS ${i.ncssCount}, functions ${i.functionCount})\n`),n.length===0)g(`No measurable files found.
+`);else{let e=n.slice(0,r.top),t=n.length>e.length?` of ${n.length}`:``;g(`\nRefactoring candidates (top ${e.length}${t}):\n`);for(let[t,n]of e.entries())g(`${t+1}. ${P(n)}\n`)}if(t.warnings.length>0){h(`\nDegraded ${t.warnings.length} files (measured, but excluded from cross-file matching):\n`);for(let e of t.warnings.slice(0,10))h(`- ${e}\n`);t.warnings.length>10&&h(`- ... ${t.warnings.length-10} more\n`)}if(t.errors.length>0){h(`\nSkipped ${t.errors.length} files or directories:\n`);for(let e of t.errors.slice(0,10))h(`- ${e}\n`);t.errors.length>10&&h(`- ... ${t.errors.length-10} more\n`)}}function P(e){let t=[];if(e.worstFunction){let n=e.worstFunction;t.push(`worst function ${n.name} (L${n.startLine}-${n.endLine}) cognitive ${n.cognitiveComplexity}, NCSS ${n.ncss}, nesting ${n.nestingDepth}`)}if(e.duplicatedLineCount>0){let n=e.crossFilePartners.length>0?`, shared with ${e.crossFilePartners.slice(0,3).join(`, `)}${e.crossFilePartners.length>3?`, ...`:``}`:``;t.push(`duplicated lines ${e.duplicatedLineCount} (${Math.round(e.duplicatedLineRatio*100)}%${n})`)}return t.push(`file NCSS ${e.ncss}`),`${e.file} (score ${e.score.toFixed(2)}): ${t.join(`; `)}`}function F(e){let t=0,n=0,r=0,i=0;for(let a of e)t+=a.metrics.functions.length,n+=a.metrics.lines.code,r=Math.max(r,a.metrics.maxCognitiveComplexity),i+=a.metrics.ncssCount;return{fileCount:e.length,functionCount:t,linesOfCode:n,maxCognitiveComplexity:r,ncssCount:i}}function I(e){if(e===`off`)return 1/0;let t=e.trim()===``?NaN:Number(e);if(!Number.isFinite(t)||t<0)throw new S(`Expected a non-negative number or "off".`);return t}function L(e){let t=z(e);if(t>100)throw new S(`Expected an integer between 1 and 100.`);return t}function R(e){if(!/^\d+$/u.test(e))throw new S(`Expected a non-negative integer.`);let t=Number(e);if(!Number.isSafeInteger(t)||t<0)throw new S(`Expected a non-negative integer.`);return t}function z(e){let t=b(e);if(t===void 0)throw new S(`Expected a positive integer.`);return t}
 //# sourceMappingURL=cli.js.map

package/dist/cli.js.map

@@ -1 +1 @@
-{"version":3,"file":"cli.js","names":[],"sources":["../src/cli.ts"],"sourcesContent":["#!/usr/bin/env node\n\nimport { Command, InvalidArgumentError } from 'commander';\nimport { getErrorMessage, parsePositiveInteger } from '@willbooster/shared-lib';\nimport { type CliOptions, configFileName, loadConfig, type ResolvedOptions, resolveOptions } from './cliConfig.js';\nimport type { CrossFileDuplicationMetrics } from './crossFileDuplication.js';\nimport { runCheckCommand, type CheckCliOptions } from './checkCommand.js';\nimport { loadExclusion } from './exclusion.js';\nimport {\n  addCrossFileDuplication,\n  collectDuplicatedLineNumbers,\n  configSearchDirectory,\n  formatPath,\n  resolveTarget,\n  scanTarget,\n  writeStderr,\n  writeStdout,\n  type FileMetrics,\n  type ScanResult,\n} from './scan.js';\nimport {\n  duplicationThreshold,\n  fileThresholds,\n  functionThresholds,\n  levels,\n  thresholds,\n  type Level,\n  type Limits,\n} from './thresholds.js';\nimport type { FunctionMetrics } from './types.js';\n\n/** The worst (highest-cognitive-complexity) function of a file, reported as the ranking evidence. */\ninterface WorstFunction {\n  name: string;\n  startLine: number;\n  endLine: number;\n  cognitiveComplexity: number;\n  ncss: number;\n  nestingDepth: number;\n}\n\n/**\n * One ranked refactoring candidate. The score is the sum of the file's repo-relative percentile\n * ranks (each in [0, 1)) over three dimensions: worst-function cognitive complexity, duplicated\n * lines (within-file and cross-file combined), and file NCSS. Ranking is relative to the scanned\n * project, so no absolute threshold is involved.\n */\ninterface RankedFile {\n  file: string;\n  score: number;\n  worstFunction?: WorstFunction;\n  /** Distinct lines covered by within-file duplicate blocks or cross-file duplicate occurrences. */\n  duplicatedLineCount: number;\n  /** duplicatedLineCount / code lines (0 when the file has no code). */\n  duplicatedLineRatio: number;\n  /** Other files sharing cross-file duplicate blocks; filled only for the reported top files. */\n  crossFilePartners: string[];\n  ncss: number;\n  codeLines: number;\n}\n\n/** Caps how many cross-file partners a ranked file lists so the report stays scannable. */\nconst maxCrossFilePartners = 3;\n\n// oxlint-disable-next-line unicorn/prefer-top-level-await -- CommonJS build output cannot preserve top-level await.\nvoid main().catch((error: unknown) => {\n  writeStderr(`Error: ${getErrorMessage(error)}\\n`);\n  process.exitCode = 1;\n});\n\nasync function main(): Promise<void> {\n  const program = new Command()\n    .name('code-gauge')\n    .description('Measure code metrics: rank refactoring candidates (the default command) or check thresholds.');\n\n  addSharedOptions(\n    program\n      .command('rank', { isDefault: true })\n      .description('Rank the files of a project by refactoring priority')\n      .argument('[target]', 'file or directory to measure', '.')\n  )\n    .option('--top <number>', 'number of top-ranked files to report (default: 10)', parsePositiveIntegerOption)\n    .option('--fail-on-error', 'exit with code 1 when files or directories cannot be scanned')\n    .action(runRankCommand);\n\n  const check = addSharedOptions(\n    program\n      .command('check')\n      .description(\n        'Report every exceeded threshold as a warning or an error (exit 1 on errors, 2 when files cannot be measured); with --base, only in what the change touches'\n      )\n      .argument('[target]', 'file or directory to check', '.')\n  ).option('--base <ref>', 'check only what the working tree changed since the merge-base of this git ref and HEAD');\n  for (const level of levels) {\n    for (const { key, defaultLimits } of thresholds) {\n      const defaultLimit = defaultLimits[level];\n      check.option(\n        `--${level}-${toKebabCase(key)} <number>`,\n        `${describeLimit(key, level)} (default ${Number.isFinite(defaultLimit) ? defaultLimit : 'off'}); \"off\" disables it`,\n        parseLimit\n      );\n    }\n  }\n  // Exit code 1 means error-level violations, so a rejected command line must not use it.\n  check.exitOverride((error) => process.exit(error.exitCode === 0 ? 0 : 2));\n  check.action(async (target: string, cliOptions: CheckCliOptions & Record<string, unknown>) => {\n    const readLimits = (level: Level): Limits =>\n      Object.fromEntries(\n        thresholds.flatMap(({ key }) => {\n          const limit = cliOptions[`${level}${key.charAt(0).toUpperCase()}${key.slice(1)}`];\n          return limit === undefined ? [] : [[key, limit as number]];\n        })\n      );\n    await runCheckCommand(target, cliOptions, { warning: readLimits('warning'), error: readLimits('error') });\n  });\n\n  await program.parseAsync();\n}\n\n/** Registers the options shared by the ranking and check commands. */\nfunction addSharedOptions(command: Command): Command {\n  return command\n    .option('--config <path>', `config file to use instead of the auto-detected ${configFileName}`)\n    .option(\n      '--duplication-min-tokens <number>',\n      'minimum normalized token count for a duplicate region (default 40)',\n      parsePositiveIntegerOption\n    )\n    .option(\n      '--duplication-max-gap-tokens <number>',\n      'maximum token gap merged into one gapped clone group; 0 disables merging (default 30)',\n      parseNonNegativeInteger\n    )\n    .option(\n      '--duplication-min-similarity-percent <number>',\n      'minimum similarity percent (1-100) for near-miss (Type-3) clone blocks; 100 reports exact matches only (default 70)',\n      parsePercentInteger\n    )\n    .option('--include-tests', 'include test files and test directories')\n    .option('--json', 'print JSON output');\n}\n\nfunction describeLimit(key: string, level: Level): string {\n  if (key === duplicationThreshold.key) {\n    return `line count from which a duplicated block is ${level === 'error' ? 'an error' : 'a warning'}`;\n  }\n  const fileThreshold = fileThresholds.find((threshold) => threshold.key === key);\n  const subject = fileThreshold\n    ? fileThreshold.label\n    : `${functionThresholds.find((threshold) => threshold.key === key)?.label} of a function`;\n  return `maximum ${subject} without ${level === 'error' ? 'an error' : 'a warning'}`;\n}\n\nfunction toKebabCase(key: string): string {\n  return key.replaceAll(/[A-Z]/gu, (letter) => `-${letter.toLowerCase()}`);\n}\n\nasync function runRankCommand(target: string, cliOptions: CliOptions): Promise<void> {\n  const resolvedTarget = resolveTarget(target);\n  const searchDirectory = await configSearchDirectory(resolvedTarget);\n  const config = await loadConfig(cliOptions.config, searchDirectory);\n  const options = resolveOptions(cliOptions, config);\n  const result = await scanTarget(resolvedTarget, {\n    ...options,\n    loadExclusion: (absolutePaths) => loadExclusion(searchDirectory, options.exclude, absolutePaths),\n  });\n  addCrossFileDuplication(result, options);\n  const rankedFiles = rankFiles(result, options.top);\n\n  if (options.json) {\n    printJson(result, rankedFiles, options);\n  } else {\n    printTextReport(resolvedTarget, result, rankedFiles, options);\n  }\n\n  if (result.fatalError || (options.failOnError && result.errors.length > 0)) {\n    process.exitCode = 1;\n  }\n}\n\nfunction rankFiles(result: ScanResult, top: number): RankedFile[] {\n  const candidates = result.files.map(({ file, metrics }) => {\n    const formattedFile = formatPath(file, result.displayRoot);\n    const duplicatedLines = collectDuplicatedLineNumbers(metrics, result.crossFileDuplication, formattedFile);\n    return {\n      file: formattedFile,\n      worstFunction: findWorstFunction(metrics.functions),\n      duplicatedLineCount: duplicatedLines.size,\n      duplicatedLineRatio: metrics.lines.code === 0 ? 0 : duplicatedLines.size / metrics.lines.code,\n      crossFilePartners: [] as string[],\n      ncss: metrics.ncssCount,\n      codeLines: metrics.lines.code,\n    };\n  });\n\n  const cognitivePercentile = makePercentile(candidates.map((c) => c.worstFunction?.cognitiveComplexity ?? 0));\n  const duplicationPercentile = makePercentile(candidates.map((c) => c.duplicatedLineCount));\n  const ncssPercentile = makePercentile(candidates.map((c) => c.ncss));\n\n  const ranked = candidates\n    .map((candidate) => ({\n      ...candidate,\n      score:\n        cognitivePercentile(candidate.worstFunction?.cognitiveComplexity ?? 0) +\n        duplicationPercentile(candidate.duplicatedLineCount) +\n        ncssPercentile(candidate.ncss),\n    }))\n    .toSorted(\n      (left, right) => right.score - left.score || right.ncss - left.ncss || left.file.localeCompare(right.file)\n    );\n  // Partner evidence is attached only to the files that will be reported: expanding partners for\n  // every scanned file first would retain TH(F^2) strings when one clone group spans F files, and\n  // rescanning all groups per file would be O(files x groups) on the post-scan ranking step.\n  attachCrossFilePartners(ranked.slice(0, top), result.crossFileDuplication);\n  return ranked;\n}\n\n/** One pass over the groups fills the reported files' partner lists (other files sharing a group). */\nfunction attachCrossFilePartners(\n  reportedFiles: RankedFile[],\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined\n): void {\n  if (!crossFileDuplication) {\n    return;\n  }\n  const partnersByFile = new Map(reportedFiles.map((ranked) => [ranked.file, new Set<string>()]));\n  for (const group of crossFileDuplication.groups) {\n    for (const file of group.files) {\n      const partners = partnersByFile.get(file);\n      if (!partners) {\n        continue;\n      }\n      for (const partner of group.files) {\n        if (partner !== file) {\n          partners.add(partner);\n        }\n      }\n    }\n  }\n  for (const ranked of reportedFiles) {\n    ranked.crossFilePartners = [...(partnersByFile.get(ranked.file) ?? [])].toSorted();\n  }\n}\n\n/**\n * Percentile rank within the scanned project: the fraction of files with a strictly smaller\n * value, in [0, 1). Relative ranking needs no absolute threshold, which sidesteps the metric\n * calibration problem entirely — the top of the list is worth refactoring first regardless of\n * where any cutoff would sit.\n */\nfunction makePercentile(values: number[]): (value: number) => number {\n  const sorted = values.toSorted((left, right) => left - right);\n  return (value: number) => {\n    let low = 0;\n    let high = sorted.length;\n    while (low < high) {\n      const middle = (low + high) >> 1;\n      if ((sorted[middle] as number) < value) {\n        low = middle + 1;\n      } else {\n        high = middle;\n      }\n    }\n    return sorted.length === 0 ? 0 : low / sorted.length;\n  };\n}\n\n/** The highest-cognitive-complexity function; NCSS breaks ties so the larger body is reported. */\nfunction findWorstFunction(functions: FunctionMetrics[]): WorstFunction | undefined {\n  let worst: FunctionMetrics | undefined;\n  for (const fn of functions) {\n    if (\n      !worst ||\n      fn.cognitiveComplexity > worst.cognitiveComplexity ||\n      (fn.cognitiveComplexity === worst.cognitiveComplexity && fn.ncss > worst.ncss)\n    ) {\n      worst = fn;\n    }\n  }\n  if (!worst) {\n    return undefined;\n  }\n  return {\n    name: worst.name ?? '<anonymous>',\n    startLine: worst.startLine,\n    endLine: worst.endLine,\n    cognitiveComplexity: worst.cognitiveComplexity,\n    ncss: worst.ncss,\n    nestingDepth: worst.nestingDepth,\n  };\n}\n\nfunction printJson(result: ScanResult, rankedFiles: RankedFile[], options: ResolvedOptions): void {\n  const reportedFiles = rankedFiles.slice(0, options.top);\n  writeStdout(\n    JSON.stringify(\n      {\n        summary: summarize(result.files),\n        totalRankedFiles: rankedFiles.length,\n        truncated: reportedFiles.length < rankedFiles.length,\n        files: reportedFiles,\n        errors: result.errors,\n        warnings: result.warnings,\n      },\n      undefined,\n      2\n    ) + '\\n'\n  );\n}\n\nfunction printTextReport(\n  target: string,\n  result: ScanResult,\n  rankedFiles: RankedFile[],\n  options: ResolvedOptions\n): void {\n  if (result.fatalError) {\n    writeStderr(`Error: ${result.fatalError}\\n`);\n    return;\n  }\n\n  const summary = summarize(result.files);\n  writeStdout(\n    `Measured ${summary.fileCount} files under ${target} (code LOC ${summary.linesOfCode}, NCSS ${summary.ncssCount}, functions ${summary.functionCount})\\n`\n  );\n\n  if (rankedFiles.length === 0) {\n    writeStdout('No measurable files found.\\n');\n  } else {\n    const reportedFiles = rankedFiles.slice(0, options.top);\n    const totalSuffix = rankedFiles.length > reportedFiles.length ? ` of ${rankedFiles.length}` : '';\n    writeStdout(`\\nRefactoring candidates (top ${reportedFiles.length}${totalSuffix}):\\n`);\n    for (const [index, ranked] of reportedFiles.entries()) {\n      writeStdout(`${index + 1}. ${formatRankedFile(ranked)}\\n`);\n    }\n  }\n\n  if (result.warnings.length > 0) {\n    writeStderr(`\\nDegraded ${result.warnings.length} files (measured, but excluded from cross-file matching):\\n`);\n    for (const warning of result.warnings.slice(0, 10)) {\n      writeStderr(`- ${warning}\\n`);\n    }\n    if (result.warnings.length > 10) {\n      writeStderr(`- ... ${result.warnings.length - 10} more\\n`);\n    }\n  }\n\n  if (result.errors.length > 0) {\n    writeStderr(`\\nSkipped ${result.errors.length} files or directories:\\n`);\n    for (const error of result.errors.slice(0, 10)) {\n      writeStderr(`- ${error}\\n`);\n    }\n    if (result.errors.length > 10) {\n      writeStderr(`- ... ${result.errors.length - 10} more\\n`);\n    }\n  }\n}\n\n/** One ranked file as a single line: the location, the evidence, and where the duplication points. */\nfunction formatRankedFile(ranked: RankedFile): string {\n  const reasons: string[] = [];\n  if (ranked.worstFunction) {\n    const fn = ranked.worstFunction;\n    reasons.push(\n      `worst function ${fn.name} (L${fn.startLine}-${fn.endLine}) cognitive ${fn.cognitiveComplexity}, NCSS ${fn.ncss}, nesting ${fn.nestingDepth}`\n    );\n  }\n  if (ranked.duplicatedLineCount > 0) {\n    const partnersSuffix =\n      ranked.crossFilePartners.length > 0\n        ? `, shared with ${ranked.crossFilePartners.slice(0, maxCrossFilePartners).join(', ')}${ranked.crossFilePartners.length > maxCrossFilePartners ? ', ...' : ''}`\n        : '';\n    reasons.push(\n      `duplicated lines ${ranked.duplicatedLineCount} (${Math.round(ranked.duplicatedLineRatio * 100)}%${partnersSuffix})`\n    );\n  }\n  reasons.push(`file NCSS ${ranked.ncss}`);\n  return `${ranked.file} (score ${ranked.score.toFixed(2)}): ${reasons.join('; ')}`;\n}\n\nfunction summarize(files: FileMetrics[]): {\n  fileCount: number;\n  functionCount: number;\n  linesOfCode: number;\n  maxCognitiveComplexity: number;\n  ncssCount: number;\n} {\n  let functionCount = 0;\n  let linesOfCode = 0;\n  let maxCognitiveComplexity = 0;\n  let ncssCount = 0;\n\n  for (const file of files) {\n    functionCount += file.metrics.functions.length;\n    linesOfCode += file.metrics.lines.code;\n    maxCognitiveComplexity = Math.max(maxCognitiveComplexity, file.metrics.maxCognitiveComplexity);\n    ncssCount += file.metrics.ncssCount;\n  }\n\n  return { fileCount: files.length, functionCount, linesOfCode, maxCognitiveComplexity, ncssCount };\n}\n\nfunction parseLimit(value: string): number {\n  if (value === 'off') {\n    return Infinity;\n  }\n  const parsed = value.trim() === '' ? Number.NaN : Number(value);\n  if (!Number.isFinite(parsed) || parsed < 0) {\n    throw new InvalidArgumentError('Expected a non-negative number or \"off\".');\n  }\n  return parsed;\n}\n\nfunction parsePercentInteger(value: string): number {\n  const parsed = parsePositiveIntegerOption(value);\n  if (parsed > 100) {\n    throw new InvalidArgumentError('Expected an integer between 1 and 100.');\n  }\n  return parsed;\n}\n\nfunction parseNonNegativeInteger(value: string): number {\n  if (!/^\\d+$/u.test(value)) {\n    throw new InvalidArgumentError('Expected a non-negative integer.');\n  }\n\n  const parsed = Number(value);\n  if (!Number.isSafeInteger(parsed) || parsed < 0) {\n    throw new InvalidArgumentError('Expected a non-negative integer.');\n  }\n  return parsed;\n}\n\nfunction parsePositiveIntegerOption(value: string): number {\n  const parsed = parsePositiveInteger(value);\n  if (parsed === undefined) {\n    throw new InvalidArgumentError('Expected a positive integer.');\n  }\n  return parsed;\n}\n"],"mappings":";8oBAiEK,EAAK,CAAC,CAAC,MAAO,GAAmB,CACpC,EAAY,UAAU,EAAgB,CAAK,EAAE,GAAG,EAChD,QAAQ,SAAW,CACrB,CAAC,EAED,eAAe,GAAsB,CACnC,IAAM,EAAU,IAAI,EAAQ,CAAC,CAC1B,KAAK,YAAY,CAAC,CAClB,YAAY,8FAA8F,EAE7G,EACE,EACG,QAAQ,OAAQ,CAAE,UAAW,EAAK,CAAC,CAAC,CACpC,YAAY,qDAAqD,CAAC,CAClE,SAAS,WAAY,+BAAgC,GAAG,CAC7D,CAAC,CACE,OAAO,iBAAkB,qDAAsD,CAA0B,CAAC,CAC1G,OAAO,kBAAmB,8DAA8D,CAAC,CACzF,OAAO,CAAc,EAExB,IAAM,EAAQ,EACZ,EACG,QAAQ,OAAO,CAAC,CAChB,YACC,4JACF,CAAC,CACA,SAAS,WAAY,6BAA8B,GAAG,CAC3D,CAAC,CAAC,OAAO,eAAgB,wFAAwF,EACjH,IAAK,IAAM,KAAS,EAClB,IAAK,GAAM,CAAE,MAAK,mBAAmB,EAAY,CAC/C,IAAM,EAAe,EAAc,GACnC,EAAM,OACJ,KAAK,EAAM,GAAG,EAAY,CAAG,EAAE,WAC/B,GAAG,EAAc,EAAK,CAAK,EAAE,YAAY,OAAO,SAAS,CAAY,EAAI,EAAe,MAAM,sBAC9F,CACF,CACF,CAGF,EAAM,aAAc,GAAU,QAAQ,KAAK,EAAM,WAAa,EAAI,EAAI,CAAC,CAAC,EACxE,EAAM,OAAO,MAAO,EAAgB,IAA0D,CAC5F,IAAM,EAAc,GAClB,OAAO,YACL,EAAW,SAAS,CAAE,SAAU,CAC9B,IAAM,EAAQ,EAAW,GAAG,IAAQ,EAAI,OAAO,CAAC,CAAC,CAAC,YAAY,IAAI,EAAI,MAAM,CAAC,KAC7E,OAAO,IAAU,IAAA,GAAY,CAAC,EAAI,CAAC,CAAC,EAAK,CAAe,CAAC,CAC3D,CAAC,CACH,EACF,MAAM,EAAgB,EAAQ,EAAY,CAAE,QAAS,EAAW,SAAS,EAAG,MAAO,EAAW,OAAO,CAAE,CAAC,CAC1G,CAAC,EAED,MAAM,EAAQ,WAAW,CAC3B,CAGA,SAAS,EAAiB,EAA2B,CACnD,OAAO,EACJ,OAAO,kBAAmB,mDAAmD,GAAgB,CAAC,CAC9F,OACC,oCACA,qEACA,CACF,CAAC,CACA,OACC,wCACA,wFACA,CACF,CAAC,CACA,OACC,gDACA,sHACA,CACF,CAAC,CACA,OAAO,kBAAmB,yCAAyC,CAAC,CACpE,OAAO,SAAU,mBAAmB,CACzC,CAEA,SAAS,EAAc,EAAa,EAAsB,CACxD,GAAI,IAAQ,EAAqB,IAC/B,MAAO,+CAA+C,IAAU,QAAU,WAAa,cAEzF,IAAM,EAAgB,EAAe,KAAM,GAAc,EAAU,MAAQ,CAAG,EAI9E,MAAO,WAHS,EACZ,EAAc,MACd,GAAG,EAAmB,KAAM,GAAc,EAAU,MAAQ,CAAG,CAAC,EAAE,MAAM,gBAClD,WAAW,IAAU,QAAU,WAAa,aACxE,CAEA,SAAS,EAAY,EAAqB,CACxC,OAAO,EAAI,WAAW,UAAY,GAAW,IAAI,EAAO,YAAY,GAAG,CACzE,CAEA,eAAe,EAAe,EAAgB,EAAuC,CACnF,IAAM,EAAiB,EAAc,CAAM,EACrC,EAAkB,MAAM,EAAsB,CAAc,EAC5D,EAAS,MAAM,EAAW,EAAW,OAAQ,CAAe,EAC5D,EAAU,EAAe,EAAY,CAAM,EAC3C,EAAS,MAAM,EAAW,EAAgB,CAC9C,GAAG,EACH,cAAgB,GAAkB,EAAc,EAAiB,EAAQ,QAAS,CAAa,CACjG,CAAC,EACD,EAAwB,EAAQ,CAAO,EACvC,IAAM,EAAc,EAAU,EAAQ,EAAQ,GAAG,EAE7C,EAAQ,KACV,EAAU,EAAQ,EAAa,CAAO,EAEtC,EAAgB,EAAgB,EAAQ,EAAa,CAAO,GAG1D,EAAO,YAAe,EAAQ,aAAe,EAAO,OAAO,OAAS,KACtE,QAAQ,SAAW,EAEvB,CAEA,SAAS,EAAU,EAAoB,EAA2B,CAChE,IAAM,EAAa,EAAO,MAAM,KAAK,CAAE,OAAM,aAAc,CACzD,IAAM,EAAgB,EAAW,EAAM,EAAO,WAAW,EACnD,EAAkB,EAA6B,EAAS,EAAO,qBAAsB,CAAa,EACxG,MAAO,CACL,KAAM,EACN,cAAe,EAAkB,EAAQ,SAAS,EAClD,oBAAqB,EAAgB,KACrC,oBAAqB,EAAQ,MAAM,OAAS,EAAI,EAAI,EAAgB,KAAO,EAAQ,MAAM,KACzF,kBAAmB,CAAC,EACpB,KAAM,EAAQ,UACd,UAAW,EAAQ,MAAM,IAC3B,CACF,CAAC,EAEK,EAAsB,EAAe,EAAW,IAAK,GAAM,EAAE,eAAe,qBAAuB,CAAC,CAAC,EACrG,EAAwB,EAAe,EAAW,IAAK,GAAM,EAAE,mBAAmB,CAAC,EACnF,EAAiB,EAAe,EAAW,IAAK,GAAM,EAAE,IAAI,CAAC,EAE7D,EAAS,EACZ,IAAK,IAAe,CACnB,GAAG,EACH,MACE,EAAoB,EAAU,eAAe,qBAAuB,CAAC,EACrE,EAAsB,EAAU,mBAAmB,EACnD,EAAe,EAAU,IAAI,CACjC,EAAE,CAAC,CACF,UACE,EAAM,IAAU,EAAM,MAAQ,EAAK,OAAS,EAAM,KAAO,EAAK,MAAQ,EAAK,KAAK,cAAc,EAAM,IAAI,CAC3G,EAKF,OADA,EAAwB,EAAO,MAAM,EAAG,CAAG,EAAG,EAAO,oBAAoB,EAClE,CACT,CAGA,SAAS,EACP,EACA,EACM,CACN,GAAI,CAAC,EACH,OAEF,IAAM,EAAiB,IAAI,IAAI,EAAc,IAAK,GAAW,CAAC,EAAO,qBAAM,IAAI,GAAa,CAAC,CAAC,EAC9F,IAAK,IAAM,KAAS,EAAqB,OACvC,IAAK,IAAM,KAAQ,EAAM,MAAO,CAC9B,IAAM,EAAW,EAAe,IAAI,CAAI,EACnC,KAGL,IAAK,IAAM,KAAW,EAAM,MACtB,IAAY,GACd,EAAS,IAAI,CAAO,CAG1B,CAEF,IAAK,IAAM,KAAU,EACnB,EAAO,kBAAoB,CAAC,GAAI,EAAe,IAAI,EAAO,IAAI,GAAK,CAAC,CAAE,CAAC,CAAC,SAAS,CAErF,CAQA,SAAS,EAAe,EAA6C,CACnE,IAAM,EAAS,EAAO,UAAU,EAAM,IAAU,EAAO,CAAK,EAC5D,MAAQ,IAAkB,CACxB,IAAI,EAAM,EACN,EAAO,EAAO,OAClB,KAAO,EAAM,GAAM,CACjB,IAAM,EAAU,EAAM,GAAS,EAC1B,EAAO,GAAqB,EAC/B,EAAM,EAAS,EAEf,EAAO,CAEX,CACA,OAAO,EAAO,SAAW,EAAI,EAAI,EAAM,EAAO,MAChD,CACF,CAGA,SAAS,EAAkB,EAAyD,CAClF,IAAI,EACJ,IAAK,IAAM,KAAM,GAEb,CAAC,GACD,EAAG,oBAAsB,EAAM,qBAC9B,EAAG,sBAAwB,EAAM,qBAAuB,EAAG,KAAO,EAAM,QAEzE,EAAQ,GAGP,KAGL,MAAO,CACL,KAAM,EAAM,MAAQ,cACpB,UAAW,EAAM,UACjB,QAAS,EAAM,QACf,oBAAqB,EAAM,oBAC3B,KAAM,EAAM,KACZ,aAAc,EAAM,YACtB,CACF,CAEA,SAAS,EAAU,EAAoB,EAA2B,EAAgC,CAChG,IAAM,EAAgB,EAAY,MAAM,EAAG,EAAQ,GAAG,EACtD,EACE,KAAK,UACH,CACE,QAAS,EAAU,EAAO,KAAK,EAC/B,iBAAkB,EAAY,OAC9B,UAAW,EAAc,OAAS,EAAY,OAC9C,MAAO,EACP,OAAQ,EAAO,OACf,SAAU,EAAO,QACnB,EACA,IAAA,GACA,CACF,EAAI;CACN,CACF,CAEA,SAAS,EACP,EACA,EACA,EACA,EACM,CACN,GAAI,EAAO,WAAY,CACrB,EAAY,UAAU,EAAO,WAAW,GAAG,EAC3C,MACF,CAEA,IAAM,EAAU,EAAU,EAAO,KAAK,EAKtC,GAJA,EACE,YAAY,EAAQ,UAAU,eAAe,EAAO,aAAa,EAAQ,YAAY,SAAS,EAAQ,UAAU,cAAc,EAAQ,cAAc,IACtJ,EAEI,EAAY,SAAW,EACzB,EAAY;CAA8B,MACrC,CACL,IAAM,EAAgB,EAAY,MAAM,EAAG,EAAQ,GAAG,EAChD,EAAc,EAAY,OAAS,EAAc,OAAS,OAAO,EAAY,SAAW,GAC9F,EAAY,iCAAiC,EAAc,SAAS,EAAY,KAAK,EACrF,IAAK,GAAM,CAAC,EAAO,KAAW,EAAc,QAAQ,EAClD,EAAY,GAAG,EAAQ,EAAE,IAAI,EAAiB,CAAM,EAAE,GAAG,CAE7D,CAEA,GAAI,EAAO,SAAS,OAAS,EAAG,CAC9B,EAAY,cAAc,EAAO,SAAS,OAAO,4DAA4D,EAC7G,IAAK,IAAM,KAAW,EAAO,SAAS,MAAM,EAAG,EAAE,EAC/C,EAAY,KAAK,EAAQ,GAAG,EAE1B,EAAO,SAAS,OAAS,IAC3B,EAAY,SAAS,EAAO,SAAS,OAAS,GAAG,QAAQ,CAE7D,CAEA,GAAI,EAAO,OAAO,OAAS,EAAG,CAC5B,EAAY,aAAa,EAAO,OAAO,OAAO,yBAAyB,EACvE,IAAK,IAAM,KAAS,EAAO,OAAO,MAAM,EAAG,EAAE,EAC3C,EAAY,KAAK,EAAM,GAAG,EAExB,EAAO,OAAO,OAAS,IACzB,EAAY,SAAS,EAAO,OAAO,OAAS,GAAG,QAAQ,CAE3D,CACF,CAGA,SAAS,EAAiB,EAA4B,CACpD,IAAM,EAAoB,CAAC,EAC3B,GAAI,EAAO,cAAe,CACxB,IAAM,EAAK,EAAO,cAClB,EAAQ,KACN,kBAAkB,EAAG,KAAK,KAAK,EAAG,UAAU,GAAG,EAAG,QAAQ,cAAc,EAAG,oBAAoB,SAAS,EAAG,KAAK,YAAY,EAAG,cACjI,CACF,CACA,GAAI,EAAO,oBAAsB,EAAG,CAClC,IAAM,EACJ,EAAO,kBAAkB,OAAS,EAC9B,iBAAiB,EAAO,kBAAkB,MAAM,EAAG,CAAoB,CAAC,CAAC,KAAK,IAAI,IAAI,EAAO,kBAAkB,OAAS,EAAuB,QAAU,KACzJ,GACN,EAAQ,KACN,oBAAoB,EAAO,oBAAoB,IAAI,KAAK,MAAM,EAAO,oBAAsB,GAAG,EAAE,GAAG,EAAe,EACpH,CACF,CAEA,OADA,EAAQ,KAAK,aAAa,EAAO,MAAM,EAChC,GAAG,EAAO,KAAK,UAAU,EAAO,MAAM,QAAQ,CAAC,EAAE,KAAK,EAAQ,KAAK,IAAI,GAChF,CAEA,SAAS,EAAU,EAMjB,CACA,IAAI,EAAgB,EAChB,EAAc,EACd,EAAyB,EACzB,EAAY,EAEhB,IAAK,IAAM,KAAQ,EACjB,GAAiB,EAAK,QAAQ,UAAU,OACxC,GAAe,EAAK,QAAQ,MAAM,KAClC,EAAyB,KAAK,IAAI,EAAwB,EAAK,QAAQ,sBAAsB,EAC7F,GAAa,EAAK,QAAQ,UAG5B,MAAO,CAAE,UAAW,EAAM,OAAQ,gBAAe,cAAa,yBAAwB,WAAU,CAClG,CAEA,SAAS,EAAW,EAAuB,CACzC,GAAI,IAAU,MACZ,MAAO,KAET,IAAM,EAAS,EAAM,KAAK,IAAM,GAAK,IAAa,OAAO,CAAK,EAC9D,GAAI,CAAC,OAAO,SAAS,CAAM,GAAK,EAAS,EACvC,MAAM,IAAI,EAAqB,0CAA0C,EAE3E,OAAO,CACT,CAEA,SAAS,EAAoB,EAAuB,CAClD,IAAM,EAAS,EAA2B,CAAK,EAC/C,GAAI,EAAS,IACX,MAAM,IAAI,EAAqB,wCAAwC,EAEzE,OAAO,CACT,CAEA,SAAS,EAAwB,EAAuB,CACtD,GAAI,CAAC,SAAS,KAAK,CAAK,EACtB,MAAM,IAAI,EAAqB,kCAAkC,EAGnE,IAAM,EAAS,OAAO,CAAK,EAC3B,GAAI,CAAC,OAAO,cAAc,CAAM,GAAK,EAAS,EAC5C,MAAM,IAAI,EAAqB,kCAAkC,EAEnE,OAAO,CACT,CAEA,SAAS,EAA2B,EAAuB,CACzD,IAAM,EAAS,EAAqB,CAAK,EACzC,GAAI,IAAW,IAAA,GACb,MAAM,IAAI,EAAqB,8BAA8B,EAE/D,OAAO,CACT"}
+{"version":3,"file":"cli.js","names":[],"sources":["../src/cli.ts"],"sourcesContent":["#!/usr/bin/env node\n\nimport { Command, InvalidArgumentError } from 'commander';\nimport { getErrorMessage, parsePositiveInteger } from '@willbooster/shared-lib';\nimport { type CliOptions, configFileName, loadConfig, type ResolvedOptions, resolveOptions } from './cliConfig.js';\nimport type { CrossFileDuplicationMetrics } from './crossFileDuplication.js';\nimport { runCheckCommand, type CheckCliOptions } from './checkCommand.js';\nimport { loadExclusion } from './exclusion.js';\nimport {\n  addCrossFileDuplication,\n  collectDuplicatedLineNumbers,\n  configSearchDirectory,\n  formatPath,\n  resolveTarget,\n  scanTarget,\n  writeStderr,\n  writeStdout,\n  type FileMetrics,\n  type ScanResult,\n} from './scan.js';\nimport {\n  duplicationThreshold,\n  fileThresholds,\n  functionThresholds,\n  levels,\n  thresholds,\n  type Level,\n  type Limits,\n} from './thresholds.js';\nimport { qualifyFunctionName } from './thresholdCheck.js';\nimport type { FunctionMetrics } from './types.js';\n\n/** The worst (highest-cognitive-complexity) function of a file, reported as the ranking evidence. */\ninterface WorstFunction {\n  name: string;\n  startLine: number;\n  endLine: number;\n  cognitiveComplexity: number;\n  ncss: number;\n  nestingDepth: number;\n}\n\n/**\n * One ranked refactoring candidate. The score is the sum of the file's repo-relative percentile\n * ranks (each in [0, 1)) over three dimensions: worst-function cognitive complexity, duplicated\n * lines (within-file and cross-file combined), and file NCSS. Ranking is relative to the scanned\n * project, so no absolute threshold is involved.\n */\ninterface RankedFile {\n  file: string;\n  score: number;\n  worstFunction?: WorstFunction;\n  /** Distinct lines covered by within-file duplicate blocks or cross-file duplicate occurrences. */\n  duplicatedLineCount: number;\n  /** duplicatedLineCount / code lines (0 when the file has no code). */\n  duplicatedLineRatio: number;\n  /** Other files sharing cross-file duplicate blocks; filled only for the reported top files. */\n  crossFilePartners: string[];\n  ncss: number;\n  codeLines: number;\n}\n\n/** Caps how many cross-file partners a ranked file lists so the report stays scannable. */\nconst maxCrossFilePartners = 3;\n\n// oxlint-disable-next-line unicorn/prefer-top-level-await -- CommonJS build output cannot preserve top-level await.\nvoid main().catch((error: unknown) => {\n  writeStderr(`Error: ${getErrorMessage(error)}\\n`);\n  process.exitCode = 1;\n});\n\nasync function main(): Promise<void> {\n  const program = new Command()\n    .name('code-gauge')\n    .description('Measure code metrics: rank refactoring candidates (the default command) or check thresholds.');\n\n  addSharedOptions(\n    program\n      .command('rank', { isDefault: true })\n      .description('Rank the files of a project by refactoring priority')\n      .argument('[target]', 'file or directory to measure', '.')\n  )\n    .option('--top <number>', 'number of top-ranked files to report (default: 10)', parsePositiveIntegerOption)\n    .option('--fail-on-error', 'exit with code 1 when files or directories cannot be scanned')\n    .action(runRankCommand);\n\n  const check = addSharedOptions(\n    program\n      .command('check')\n      .description(\n        'Report every exceeded threshold as a warning or an error (exit 1 on errors, 2 when files cannot be measured); with --base, only in what the change touches'\n      )\n      .argument('[target]', 'file or directory to check', '.')\n  ).option('--base <ref>', 'check only what the working tree changed since the merge-base of this git ref and HEAD');\n  for (const level of levels) {\n    for (const { key, defaultLimits } of thresholds) {\n      const defaultLimit = defaultLimits[level];\n      check.option(\n        `--${level}-${toKebabCase(key)} <number>`,\n        `${describeLimit(key, level)} (default ${Number.isFinite(defaultLimit) ? defaultLimit : 'off'}); \"off\" disables it`,\n        parseLimit\n      );\n    }\n  }\n  // Exit code 1 means error-level violations, so a rejected command line must not use it.\n  check.exitOverride((error) => process.exit(error.exitCode === 0 ? 0 : 2));\n  check.action(async (target: string, cliOptions: CheckCliOptions & Record<string, unknown>) => {\n    const readLimits = (level: Level): Limits =>\n      Object.fromEntries(\n        thresholds.flatMap(({ key }) => {\n          const limit = cliOptions[`${level}${key.charAt(0).toUpperCase()}${key.slice(1)}`];\n          return limit === undefined ? [] : [[key, limit as number]];\n        })\n      );\n    await runCheckCommand(target, cliOptions, { warning: readLimits('warning'), error: readLimits('error') });\n  });\n\n  await program.parseAsync();\n}\n\n/** Registers the options shared by the ranking and check commands. */\nfunction addSharedOptions(command: Command): Command {\n  return command\n    .option('--config <path>', `config file to use instead of the auto-detected ${configFileName}`)\n    .option(\n      '--duplication-min-tokens <number>',\n      'minimum normalized token count for a duplicate region (default 40)',\n      parsePositiveIntegerOption\n    )\n    .option(\n      '--duplication-max-gap-tokens <number>',\n      'maximum token gap merged into one gapped clone group; 0 disables merging (default 30)',\n      parseNonNegativeInteger\n    )\n    .option(\n      '--duplication-min-similarity-percent <number>',\n      'minimum similarity percent (1-100) for near-miss (Type-3) clone blocks; 100 reports exact matches only (default 70)',\n      parsePercentInteger\n    )\n    .option('--include-tests', 'include test files and test directories')\n    .option('--json', 'print JSON output');\n}\n\nfunction describeLimit(key: string, level: Level): string {\n  if (key === duplicationThreshold.key) {\n    return `code-line count from which a duplicated block is ${level === 'error' ? 'an error' : 'a warning'}`;\n  }\n  const fileThreshold = fileThresholds.find((threshold) => threshold.key === key);\n  const subject = fileThreshold\n    ? fileThreshold.label\n    : `${functionThresholds.find((threshold) => threshold.key === key)?.label} of a function`;\n  return `maximum ${subject} without ${level === 'error' ? 'an error' : 'a warning'}`;\n}\n\nfunction toKebabCase(key: string): string {\n  return key.replaceAll(/[A-Z]/gu, (letter) => `-${letter.toLowerCase()}`);\n}\n\nasync function runRankCommand(target: string, cliOptions: CliOptions): Promise<void> {\n  const resolvedTarget = resolveTarget(target);\n  const searchDirectory = await configSearchDirectory(resolvedTarget);\n  const config = await loadConfig(cliOptions.config, searchDirectory);\n  const options = resolveOptions(cliOptions, config);\n  const result = await scanTarget(resolvedTarget, {\n    ...options,\n    loadExclusion: (absolutePaths) => loadExclusion(searchDirectory, options.exclude, absolutePaths),\n  });\n  addCrossFileDuplication(result, options);\n  const rankedFiles = rankFiles(result, options.top);\n\n  if (options.json) {\n    printJson(result, rankedFiles, options);\n  } else {\n    printTextReport(resolvedTarget, result, rankedFiles, options);\n  }\n\n  if (result.fatalError || (options.failOnError && result.errors.length > 0)) {\n    process.exitCode = 1;\n  }\n}\n\nfunction rankFiles(result: ScanResult, top: number): RankedFile[] {\n  const candidates = result.files.map(({ file, metrics }) => {\n    const formattedFile = formatPath(file, result.displayRoot);\n    const duplicatedLines = collectDuplicatedLineNumbers(metrics, result.crossFileDuplication, formattedFile);\n    return {\n      file: formattedFile,\n      worstFunction: findWorstFunction(metrics.functions),\n      duplicatedLineCount: duplicatedLines.size,\n      duplicatedLineRatio: metrics.lines.code === 0 ? 0 : duplicatedLines.size / metrics.lines.code,\n      crossFilePartners: [] as string[],\n      ncss: metrics.ncssCount,\n      codeLines: metrics.lines.code,\n    };\n  });\n\n  const cognitivePercentile = makePercentile(candidates.map((c) => c.worstFunction?.cognitiveComplexity ?? 0));\n  const duplicationPercentile = makePercentile(candidates.map((c) => c.duplicatedLineCount));\n  const ncssPercentile = makePercentile(candidates.map((c) => c.ncss));\n\n  const ranked = candidates\n    .map((candidate) => ({\n      ...candidate,\n      score:\n        cognitivePercentile(candidate.worstFunction?.cognitiveComplexity ?? 0) +\n        duplicationPercentile(candidate.duplicatedLineCount) +\n        ncssPercentile(candidate.ncss),\n    }))\n    .toSorted(\n      (left, right) => right.score - left.score || right.ncss - left.ncss || left.file.localeCompare(right.file)\n    );\n  // Partner evidence is attached only to the files that will be reported: expanding partners for\n  // every scanned file first would retain TH(F^2) strings when one clone group spans F files, and\n  // rescanning all groups per file would be O(files x groups) on the post-scan ranking step.\n  attachCrossFilePartners(ranked.slice(0, top), result.crossFileDuplication);\n  return ranked;\n}\n\n/** One pass over the groups fills the reported files' partner lists (other files sharing a group). */\nfunction attachCrossFilePartners(\n  reportedFiles: RankedFile[],\n  crossFileDuplication: CrossFileDuplicationMetrics | undefined\n): void {\n  if (!crossFileDuplication) {\n    return;\n  }\n  const partnersByFile = new Map(reportedFiles.map((ranked) => [ranked.file, new Set<string>()]));\n  for (const group of crossFileDuplication.groups) {\n    for (const file of group.files) {\n      const partners = partnersByFile.get(file);\n      if (!partners) {\n        continue;\n      }\n      for (const partner of group.files) {\n        if (partner !== file) {\n          partners.add(partner);\n        }\n      }\n    }\n  }\n  for (const ranked of reportedFiles) {\n    ranked.crossFilePartners = [...(partnersByFile.get(ranked.file) ?? [])].toSorted();\n  }\n}\n\n/**\n * Percentile rank within the scanned project: the fraction of files with a strictly smaller\n * value, in [0, 1). Relative ranking needs no absolute threshold, which sidesteps the metric\n * calibration problem entirely — the top of the list is worth refactoring first regardless of\n * where any cutoff would sit.\n */\nfunction makePercentile(values: number[]): (value: number) => number {\n  const sorted = values.toSorted((left, right) => left - right);\n  return (value: number) => {\n    let low = 0;\n    let high = sorted.length;\n    while (low < high) {\n      const middle = (low + high) >> 1;\n      if ((sorted[middle] as number) < value) {\n        low = middle + 1;\n      } else {\n        high = middle;\n      }\n    }\n    return sorted.length === 0 ? 0 : low / sorted.length;\n  };\n}\n\n/** The highest-cognitive-complexity function; NCSS breaks ties so the larger body is reported. */\nfunction findWorstFunction(functions: FunctionMetrics[]): WorstFunction | undefined {\n  let worst: FunctionMetrics | undefined;\n  for (const fn of functions) {\n    if (\n      !worst ||\n      fn.cognitiveComplexity > worst.cognitiveComplexity ||\n      (fn.cognitiveComplexity === worst.cognitiveComplexity && fn.ncss > worst.ncss)\n    ) {\n      worst = fn;\n    }\n  }\n  if (!worst) {\n    return undefined;\n  }\n  return {\n    name: qualifyFunctionName(worst),\n    startLine: worst.startLine,\n    endLine: worst.endLine,\n    cognitiveComplexity: worst.cognitiveComplexity,\n    ncss: worst.ncss,\n    nestingDepth: worst.nestingDepth,\n  };\n}\n\nfunction printJson(result: ScanResult, rankedFiles: RankedFile[], options: ResolvedOptions): void {\n  const reportedFiles = rankedFiles.slice(0, options.top);\n  writeStdout(\n    JSON.stringify(\n      {\n        summary: summarize(result.files),\n        totalRankedFiles: rankedFiles.length,\n        truncated: reportedFiles.length < rankedFiles.length,\n        files: reportedFiles,\n        errors: result.errors,\n        warnings: result.warnings,\n      },\n      undefined,\n      2\n    ) + '\\n'\n  );\n}\n\nfunction printTextReport(\n  target: string,\n  result: ScanResult,\n  rankedFiles: RankedFile[],\n  options: ResolvedOptions\n): void {\n  if (result.fatalError) {\n    writeStderr(`Error: ${result.fatalError}\\n`);\n    return;\n  }\n\n  const summary = summarize(result.files);\n  writeStdout(\n    `Measured ${summary.fileCount} files under ${target} (code LOC ${summary.linesOfCode}, NCSS ${summary.ncssCount}, functions ${summary.functionCount})\\n`\n  );\n\n  if (rankedFiles.length === 0) {\n    writeStdout('No measurable files found.\\n');\n  } else {\n    const reportedFiles = rankedFiles.slice(0, options.top);\n    const totalSuffix = rankedFiles.length > reportedFiles.length ? ` of ${rankedFiles.length}` : '';\n    writeStdout(`\\nRefactoring candidates (top ${reportedFiles.length}${totalSuffix}):\\n`);\n    for (const [index, ranked] of reportedFiles.entries()) {\n      writeStdout(`${index + 1}. ${formatRankedFile(ranked)}\\n`);\n    }\n  }\n\n  if (result.warnings.length > 0) {\n    writeStderr(`\\nDegraded ${result.warnings.length} files (measured, but excluded from cross-file matching):\\n`);\n    for (const warning of result.warnings.slice(0, 10)) {\n      writeStderr(`- ${warning}\\n`);\n    }\n    if (result.warnings.length > 10) {\n      writeStderr(`- ... ${result.warnings.length - 10} more\\n`);\n    }\n  }\n\n  if (result.errors.length > 0) {\n    writeStderr(`\\nSkipped ${result.errors.length} files or directories:\\n`);\n    for (const error of result.errors.slice(0, 10)) {\n      writeStderr(`- ${error}\\n`);\n    }\n    if (result.errors.length > 10) {\n      writeStderr(`- ... ${result.errors.length - 10} more\\n`);\n    }\n  }\n}\n\n/** One ranked file as a single line: the location, the evidence, and where the duplication points. */\nfunction formatRankedFile(ranked: RankedFile): string {\n  const reasons: string[] = [];\n  if (ranked.worstFunction) {\n    const fn = ranked.worstFunction;\n    reasons.push(\n      `worst function ${fn.name} (L${fn.startLine}-${fn.endLine}) cognitive ${fn.cognitiveComplexity}, NCSS ${fn.ncss}, nesting ${fn.nestingDepth}`\n    );\n  }\n  if (ranked.duplicatedLineCount > 0) {\n    const partnersSuffix =\n      ranked.crossFilePartners.length > 0\n        ? `, shared with ${ranked.crossFilePartners.slice(0, maxCrossFilePartners).join(', ')}${ranked.crossFilePartners.length > maxCrossFilePartners ? ', ...' : ''}`\n        : '';\n    reasons.push(\n      `duplicated lines ${ranked.duplicatedLineCount} (${Math.round(ranked.duplicatedLineRatio * 100)}%${partnersSuffix})`\n    );\n  }\n  reasons.push(`file NCSS ${ranked.ncss}`);\n  return `${ranked.file} (score ${ranked.score.toFixed(2)}): ${reasons.join('; ')}`;\n}\n\nfunction summarize(files: FileMetrics[]): {\n  fileCount: number;\n  functionCount: number;\n  linesOfCode: number;\n  maxCognitiveComplexity: number;\n  ncssCount: number;\n} {\n  let functionCount = 0;\n  let linesOfCode = 0;\n  let maxCognitiveComplexity = 0;\n  let ncssCount = 0;\n\n  for (const file of files) {\n    functionCount += file.metrics.functions.length;\n    linesOfCode += file.metrics.lines.code;\n    maxCognitiveComplexity = Math.max(maxCognitiveComplexity, file.metrics.maxCognitiveComplexity);\n    ncssCount += file.metrics.ncssCount;\n  }\n\n  return { fileCount: files.length, functionCount, linesOfCode, maxCognitiveComplexity, ncssCount };\n}\n\nfunction parseLimit(value: string): number {\n  if (value === 'off') {\n    return Infinity;\n  }\n  const parsed = value.trim() === '' ? Number.NaN : Number(value);\n  if (!Number.isFinite(parsed) || parsed < 0) {\n    throw new InvalidArgumentError('Expected a non-negative number or \"off\".');\n  }\n  return parsed;\n}\n\nfunction parsePercentInteger(value: string): number {\n  const parsed = parsePositiveIntegerOption(value);\n  if (parsed > 100) {\n    throw new InvalidArgumentError('Expected an integer between 1 and 100.');\n  }\n  return parsed;\n}\n\nfunction parseNonNegativeInteger(value: string): number {\n  if (!/^\\d+$/u.test(value)) {\n    throw new InvalidArgumentError('Expected a non-negative integer.');\n  }\n\n  const parsed = Number(value);\n  if (!Number.isSafeInteger(parsed) || parsed < 0) {\n    throw new InvalidArgumentError('Expected a non-negative integer.');\n  }\n  return parsed;\n}\n\nfunction parsePositiveIntegerOption(value: string): number {\n  const parsed = parsePositiveInteger(value);\n  if (parsed === undefined) {\n    throw new InvalidArgumentError('Expected a positive integer.');\n  }\n  return parsed;\n}\n"],"mappings":";wsBAkEK,EAAK,CAAC,CAAC,MAAO,GAAmB,CACpC,EAAY,UAAU,EAAgB,CAAK,EAAE,GAAG,EAChD,QAAQ,SAAW,CACrB,CAAC,EAED,eAAe,GAAsB,CACnC,IAAM,EAAU,IAAI,EAAQ,CAAC,CAC1B,KAAK,YAAY,CAAC,CAClB,YAAY,8FAA8F,EAE7G,EACE,EACG,QAAQ,OAAQ,CAAE,UAAW,EAAK,CAAC,CAAC,CACpC,YAAY,qDAAqD,CAAC,CAClE,SAAS,WAAY,+BAAgC,GAAG,CAC7D,CAAC,CACE,OAAO,iBAAkB,qDAAsD,CAA0B,CAAC,CAC1G,OAAO,kBAAmB,8DAA8D,CAAC,CACzF,OAAO,CAAc,EAExB,IAAM,EAAQ,EACZ,EACG,QAAQ,OAAO,CAAC,CAChB,YACC,4JACF,CAAC,CACA,SAAS,WAAY,6BAA8B,GAAG,CAC3D,CAAC,CAAC,OAAO,eAAgB,wFAAwF,EACjH,IAAK,IAAM,KAAS,EAClB,IAAK,GAAM,CAAE,MAAK,mBAAmB,EAAY,CAC/C,IAAM,EAAe,EAAc,GACnC,EAAM,OACJ,KAAK,EAAM,GAAG,EAAY,CAAG,EAAE,WAC/B,GAAG,EAAc,EAAK,CAAK,EAAE,YAAY,OAAO,SAAS,CAAY,EAAI,EAAe,MAAM,sBAC9F,CACF,CACF,CAGF,EAAM,aAAc,GAAU,QAAQ,KAAK,EAAM,WAAa,EAAI,EAAI,CAAC,CAAC,EACxE,EAAM,OAAO,MAAO,EAAgB,IAA0D,CAC5F,IAAM,EAAc,GAClB,OAAO,YACL,EAAW,SAAS,CAAE,SAAU,CAC9B,IAAM,EAAQ,EAAW,GAAG,IAAQ,EAAI,OAAO,CAAC,CAAC,CAAC,YAAY,IAAI,EAAI,MAAM,CAAC,KAC7E,OAAO,IAAU,IAAA,GAAY,CAAC,EAAI,CAAC,CAAC,EAAK,CAAe,CAAC,CAC3D,CAAC,CACH,EACF,MAAM,EAAgB,EAAQ,EAAY,CAAE,QAAS,EAAW,SAAS,EAAG,MAAO,EAAW,OAAO,CAAE,CAAC,CAC1G,CAAC,EAED,MAAM,EAAQ,WAAW,CAC3B,CAGA,SAAS,EAAiB,EAA2B,CACnD,OAAO,EACJ,OAAO,kBAAmB,mDAAmD,GAAgB,CAAC,CAC9F,OACC,oCACA,qEACA,CACF,CAAC,CACA,OACC,wCACA,wFACA,CACF,CAAC,CACA,OACC,gDACA,sHACA,CACF,CAAC,CACA,OAAO,kBAAmB,yCAAyC,CAAC,CACpE,OAAO,SAAU,mBAAmB,CACzC,CAEA,SAAS,EAAc,EAAa,EAAsB,CACxD,GAAI,IAAQ,EAAqB,IAC/B,MAAO,oDAAoD,IAAU,QAAU,WAAa,cAE9F,IAAM,EAAgB,EAAe,KAAM,GAAc,EAAU,MAAQ,CAAG,EAI9E,MAAO,WAHS,EACZ,EAAc,MACd,GAAG,EAAmB,KAAM,GAAc,EAAU,MAAQ,CAAG,CAAC,EAAE,MAAM,gBAClD,WAAW,IAAU,QAAU,WAAa,aACxE,CAEA,SAAS,EAAY,EAAqB,CACxC,OAAO,EAAI,WAAW,UAAY,GAAW,IAAI,EAAO,YAAY,GAAG,CACzE,CAEA,eAAe,EAAe,EAAgB,EAAuC,CACnF,IAAM,EAAiB,EAAc,CAAM,EACrC,EAAkB,MAAM,EAAsB,CAAc,EAC5D,EAAS,MAAM,EAAW,EAAW,OAAQ,CAAe,EAC5D,EAAU,EAAe,EAAY,CAAM,EAC3C,EAAS,MAAM,EAAW,EAAgB,CAC9C,GAAG,EACH,cAAgB,GAAkB,EAAc,EAAiB,EAAQ,QAAS,CAAa,CACjG,CAAC,EACD,EAAwB,EAAQ,CAAO,EACvC,IAAM,EAAc,EAAU,EAAQ,EAAQ,GAAG,EAE7C,EAAQ,KACV,EAAU,EAAQ,EAAa,CAAO,EAEtC,EAAgB,EAAgB,EAAQ,EAAa,CAAO,GAG1D,EAAO,YAAe,EAAQ,aAAe,EAAO,OAAO,OAAS,KACtE,QAAQ,SAAW,EAEvB,CAEA,SAAS,EAAU,EAAoB,EAA2B,CAChE,IAAM,EAAa,EAAO,MAAM,KAAK,CAAE,OAAM,aAAc,CACzD,IAAM,EAAgB,EAAW,EAAM,EAAO,WAAW,EACnD,EAAkB,EAA6B,EAAS,EAAO,qBAAsB,CAAa,EACxG,MAAO,CACL,KAAM,EACN,cAAe,EAAkB,EAAQ,SAAS,EAClD,oBAAqB,EAAgB,KACrC,oBAAqB,EAAQ,MAAM,OAAS,EAAI,EAAI,EAAgB,KAAO,EAAQ,MAAM,KACzF,kBAAmB,CAAC,EACpB,KAAM,EAAQ,UACd,UAAW,EAAQ,MAAM,IAC3B,CACF,CAAC,EAEK,EAAsB,EAAe,EAAW,IAAK,GAAM,EAAE,eAAe,qBAAuB,CAAC,CAAC,EACrG,EAAwB,EAAe,EAAW,IAAK,GAAM,EAAE,mBAAmB,CAAC,EACnF,EAAiB,EAAe,EAAW,IAAK,GAAM,EAAE,IAAI,CAAC,EAE7D,EAAS,EACZ,IAAK,IAAe,CACnB,GAAG,EACH,MACE,EAAoB,EAAU,eAAe,qBAAuB,CAAC,EACrE,EAAsB,EAAU,mBAAmB,EACnD,EAAe,EAAU,IAAI,CACjC,EAAE,CAAC,CACF,UACE,EAAM,IAAU,EAAM,MAAQ,EAAK,OAAS,EAAM,KAAO,EAAK,MAAQ,EAAK,KAAK,cAAc,EAAM,IAAI,CAC3G,EAKF,OADA,EAAwB,EAAO,MAAM,EAAG,CAAG,EAAG,EAAO,oBAAoB,EAClE,CACT,CAGA,SAAS,EACP,EACA,EACM,CACN,GAAI,CAAC,EACH,OAEF,IAAM,EAAiB,IAAI,IAAI,EAAc,IAAK,GAAW,CAAC,EAAO,qBAAM,IAAI,GAAa,CAAC,CAAC,EAC9F,IAAK,IAAM,KAAS,EAAqB,OACvC,IAAK,IAAM,KAAQ,EAAM,MAAO,CAC9B,IAAM,EAAW,EAAe,IAAI,CAAI,EACnC,KAGL,IAAK,IAAM,KAAW,EAAM,MACtB,IAAY,GACd,EAAS,IAAI,CAAO,CAG1B,CAEF,IAAK,IAAM,KAAU,EACnB,EAAO,kBAAoB,CAAC,GAAI,EAAe,IAAI,EAAO,IAAI,GAAK,CAAC,CAAE,CAAC,CAAC,SAAS,CAErF,CAQA,SAAS,EAAe,EAA6C,CACnE,IAAM,EAAS,EAAO,UAAU,EAAM,IAAU,EAAO,CAAK,EAC5D,MAAQ,IAAkB,CACxB,IAAI,EAAM,EACN,EAAO,EAAO,OAClB,KAAO,EAAM,GAAM,CACjB,IAAM,EAAU,EAAM,GAAS,EAC1B,EAAO,GAAqB,EAC/B,EAAM,EAAS,EAEf,EAAO,CAEX,CACA,OAAO,EAAO,SAAW,EAAI,EAAI,EAAM,EAAO,MAChD,CACF,CAGA,SAAS,EAAkB,EAAyD,CAClF,IAAI,EACJ,IAAK,IAAM,KAAM,GAEb,CAAC,GACD,EAAG,oBAAsB,EAAM,qBAC9B,EAAG,sBAAwB,EAAM,qBAAuB,EAAG,KAAO,EAAM,QAEzE,EAAQ,GAGP,KAGL,MAAO,CACL,KAAM,EAAoB,CAAK,EAC/B,UAAW,EAAM,UACjB,QAAS,EAAM,QACf,oBAAqB,EAAM,oBAC3B,KAAM,EAAM,KACZ,aAAc,EAAM,YACtB,CACF,CAEA,SAAS,EAAU,EAAoB,EAA2B,EAAgC,CAChG,IAAM,EAAgB,EAAY,MAAM,EAAG,EAAQ,GAAG,EACtD,EACE,KAAK,UACH,CACE,QAAS,EAAU,EAAO,KAAK,EAC/B,iBAAkB,EAAY,OAC9B,UAAW,EAAc,OAAS,EAAY,OAC9C,MAAO,EACP,OAAQ,EAAO,OACf,SAAU,EAAO,QACnB,EACA,IAAA,GACA,CACF,EAAI;CACN,CACF,CAEA,SAAS,EACP,EACA,EACA,EACA,EACM,CACN,GAAI,EAAO,WAAY,CACrB,EAAY,UAAU,EAAO,WAAW,GAAG,EAC3C,MACF,CAEA,IAAM,EAAU,EAAU,EAAO,KAAK,EAKtC,GAJA,EACE,YAAY,EAAQ,UAAU,eAAe,EAAO,aAAa,EAAQ,YAAY,SAAS,EAAQ,UAAU,cAAc,EAAQ,cAAc,IACtJ,EAEI,EAAY,SAAW,EACzB,EAAY;CAA8B,MACrC,CACL,IAAM,EAAgB,EAAY,MAAM,EAAG,EAAQ,GAAG,EAChD,EAAc,EAAY,OAAS,EAAc,OAAS,OAAO,EAAY,SAAW,GAC9F,EAAY,iCAAiC,EAAc,SAAS,EAAY,KAAK,EACrF,IAAK,GAAM,CAAC,EAAO,KAAW,EAAc,QAAQ,EAClD,EAAY,GAAG,EAAQ,EAAE,IAAI,EAAiB,CAAM,EAAE,GAAG,CAE7D,CAEA,GAAI,EAAO,SAAS,OAAS,EAAG,CAC9B,EAAY,cAAc,EAAO,SAAS,OAAO,4DAA4D,EAC7G,IAAK,IAAM,KAAW,EAAO,SAAS,MAAM,EAAG,EAAE,EAC/C,EAAY,KAAK,EAAQ,GAAG,EAE1B,EAAO,SAAS,OAAS,IAC3B,EAAY,SAAS,EAAO,SAAS,OAAS,GAAG,QAAQ,CAE7D,CAEA,GAAI,EAAO,OAAO,OAAS,EAAG,CAC5B,EAAY,aAAa,EAAO,OAAO,OAAO,yBAAyB,EACvE,IAAK,IAAM,KAAS,EAAO,OAAO,MAAM,EAAG,EAAE,EAC3C,EAAY,KAAK,EAAM,GAAG,EAExB,EAAO,OAAO,OAAS,IACzB,EAAY,SAAS,EAAO,OAAO,OAAS,GAAG,QAAQ,CAE3D,CACF,CAGA,SAAS,EAAiB,EAA4B,CACpD,IAAM,EAAoB,CAAC,EAC3B,GAAI,EAAO,cAAe,CACxB,IAAM,EAAK,EAAO,cAClB,EAAQ,KACN,kBAAkB,EAAG,KAAK,KAAK,EAAG,UAAU,GAAG,EAAG,QAAQ,cAAc,EAAG,oBAAoB,SAAS,EAAG,KAAK,YAAY,EAAG,cACjI,CACF,CACA,GAAI,EAAO,oBAAsB,EAAG,CAClC,IAAM,EACJ,EAAO,kBAAkB,OAAS,EAC9B,iBAAiB,EAAO,kBAAkB,MAAM,EAAG,CAAoB,CAAC,CAAC,KAAK,IAAI,IAAI,EAAO,kBAAkB,OAAS,EAAuB,QAAU,KACzJ,GACN,EAAQ,KACN,oBAAoB,EAAO,oBAAoB,IAAI,KAAK,MAAM,EAAO,oBAAsB,GAAG,EAAE,GAAG,EAAe,EACpH,CACF,CAEA,OADA,EAAQ,KAAK,aAAa,EAAO,MAAM,EAChC,GAAG,EAAO,KAAK,UAAU,EAAO,MAAM,QAAQ,CAAC,EAAE,KAAK,EAAQ,KAAK,IAAI,GAChF,CAEA,SAAS,EAAU,EAMjB,CACA,IAAI,EAAgB,EAChB,EAAc,EACd,EAAyB,EACzB,EAAY,EAEhB,IAAK,IAAM,KAAQ,EACjB,GAAiB,EAAK,QAAQ,UAAU,OACxC,GAAe,EAAK,QAAQ,MAAM,KAClC,EAAyB,KAAK,IAAI,EAAwB,EAAK,QAAQ,sBAAsB,EAC7F,GAAa,EAAK,QAAQ,UAG5B,MAAO,CAAE,UAAW,EAAM,OAAQ,gBAAe,cAAa,yBAAwB,WAAU,CAClG,CAEA,SAAS,EAAW,EAAuB,CACzC,GAAI,IAAU,MACZ,MAAO,KAET,IAAM,EAAS,EAAM,KAAK,IAAM,GAAK,IAAa,OAAO,CAAK,EAC9D,GAAI,CAAC,OAAO,SAAS,CAAM,GAAK,EAAS,EACvC,MAAM,IAAI,EAAqB,0CAA0C,EAE3E,OAAO,CACT,CAEA,SAAS,EAAoB,EAAuB,CAClD,IAAM,EAAS,EAA2B,CAAK,EAC/C,GAAI,EAAS,IACX,MAAM,IAAI,EAAqB,wCAAwC,EAEzE,OAAO,CACT,CAEA,SAAS,EAAwB,EAAuB,CACtD,GAAI,CAAC,SAAS,KAAK,CAAK,EACtB,MAAM,IAAI,EAAqB,kCAAkC,EAGnE,IAAM,EAAS,OAAO,CAAK,EAC3B,GAAI,CAAC,OAAO,cAAc,CAAM,GAAK,EAAS,EAC5C,MAAM,IAAI,EAAqB,kCAAkC,EAEnE,OAAO,CACT,CAEA,SAAS,EAA2B,EAAuB,CACzD,IAAM,EAAS,EAAqB,CAAK,EACzC,GAAI,IAAW,IAAA,GACb,MAAM,IAAI,EAAqB,8BAA8B,EAE/D,OAAO,CACT"}

package/dist/crossFileDuplication.cjs

@@ -1,2 +1,2 @@
-"use strict";const e=require("./duplication.cjs"),t=require("./crossFileNearMiss.cjs"),n=require("./duplicateSelection.cjs");function r(t,r){let{minTokens:o,maxGapTokens:d,minSimilarityPercent:f}=e.resolveDuplicationOptions(r),p=t.flatMap(({file:e,candidates:t},n)=>t.map(t=>({...t,regionBucket:n,file:e})));for(let e of a(t,o))p.push(e);let m=n.selectMaximalGroups(p,s,(e,t)=>e.regionBucket-t.regionBucket||e.startIndex-t.startIndex),h=c(t,d),g=l([...m.values()],h,d);for(let e of i(t,g,h,o,f))g.push(e);return u(g,t,h)}function i(e,n,r,i,a){let o=e.map(()=>[]);for(let{fileIndex:e,startTokenIndex:t,endTokenIndex:i}of n.flat()){let n=r[e]??0;o[e]?.push({startTokenIndex:t-n,endTokenIndex:i-n})}return t.collectCrossFileNearMissGroups(e,o,i,a).map(t=>t.map(t=>{let n=r[t.fileIndex]??0;return{...t,file:e[t.fileIndex]?.file??``,segments:t.segments.map(e=>({startTokenIndex:e.startTokenIndex+n,endTokenIndex:e.endTokenIndex+n})),startTokenIndex:t.startTokenIndex+n,endTokenIndex:t.endTokenIndex+n}}))}function a(t,n){let r=[],i=[];for(let[n,{tokens:a,containerStatements:o}]of t.entries())a&&o&&(r.push(n),i.push({tokens:a,literalCountPrefix:e.buildLiteralCountPrefix(a),containers:o}));return i.length<2?[]:e.collectSequenceWindowCandidates(i,n,!0).flatMap(({candidate:e,contextIndex:n})=>{let i=r[n],a=i===void 0?void 0:t[i];return i===void 0||a===void 0?[]:[{...e,regionBucket:i,file:a.file}]})}function o(e){return new Set(e.map(e=>e.file)).size>=2}function s(e){return e.length>=2&&new Set(e.map(e=>e.regionBucket)).size>=2}function c(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 l(t,n,r){let i=t.map(e=>e.map(e=>{let t=e.startTokenIndex+(n[e.regionBucket]??0),r=e.endTokenIndex+(n[e.regionBucket]??0);return{file:e.file,fileIndex:e.regionBucket,spanCountedElsewhere:e.nestedInLargerGroup,nestedInLargerGroup:e.nestedInLargerGroup,segments:[{startTokenIndex:t,endTokenIndex:r}],tokenCount:e.tokenCount,startTokenIndex:t,endTokenIndex:r,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startLine,endLine:e.endLine}}).toSorted((e,t)=>e.startTokenIndex-t.startTokenIndex));return e.mergeAdjacentGroups(i,r,o)}function u(t,n,r){let i=[],a=/* @__PURE__ */ new Map,o=new Map(n.map((e,t)=>[e.file,{tokens:e.tokens,codeLineNumbers:e.codeLineNumbers,offset:r[t]??0}])),s=/* @__PURE__ */ new Map,c=0;for(let n of t){c+=e.countRedundantFragments(n);for(let e of n)d(e,o,s);let t=n.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),r=[...new Set(t.map(({file:e})=>e))];for(let e of r)a.set(e,(a.get(e)??0)+1);i.push({files:r,occurrences:t,tokenCount:Math.min(...n.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 d(t,n,r){let i=n.get(t.file);if(!i?.tokens)return;let a=r.get(t.file);a||(a=/* @__PURE__ */ new Set,r.set(t.file,a));for(let n of t.segments)e.collectSegmentLines({startTokenIndex:n.startTokenIndex-i.offset,endTokenIndex:n.endTokenIndex-i.offset},i.tokens,i.codeLineNumbers,a)}exports.measureCrossFileDuplication=r;
+"use strict";const e=require("./duplication.cjs"),t=require("./crossFileNearMiss.cjs"),n=require("./duplicateSelection.cjs");function r(t,r){let{minTokens:o,maxGapTokens:d,minSimilarityPercent:f}=e.resolveDuplicationOptions(r),p=t.flatMap(({file:e,candidates:t},n)=>t.map(t=>({...t,regionBucket:n,file:e})));for(let e of a(t,o))p.push(e);let m=n.selectMaximalGroups(p,s,(e,t)=>e.regionBucket-t.regionBucket||e.startIndex-t.startIndex),h=c(t,d),g=l([...m.values()],h,d);for(let e of i(t,g,h,o,f))g.push(e);return u(g,t,h)}function i(e,n,r,i,a){let o=e.map(()=>[]);for(let{fileIndex:e,startTokenIndex:t,endTokenIndex:i}of n.flat()){let n=r[e]??0;o[e]?.push({startTokenIndex:t-n,endTokenIndex:i-n})}return t.collectCrossFileNearMissGroups(e,o,i,a).map(t=>t.map(t=>{let n=r[t.fileIndex]??0;return{...t,file:e[t.fileIndex]?.file??``,segments:t.segments.map(e=>({startTokenIndex:e.startTokenIndex+n,endTokenIndex:e.endTokenIndex+n})),startTokenIndex:t.startTokenIndex+n,endTokenIndex:t.endTokenIndex+n}}))}function a(t,n){let r=[],i=[];for(let[n,{tokens:a,containerStatements:o}]of t.entries())a&&o&&(r.push(n),i.push({tokens:a,literalCountPrefix:e.buildLiteralCountPrefix(a),containers:o}));return i.length<2?[]:e.collectSequenceWindowCandidates(i,n,!0).flatMap(({candidate:e,contextIndex:n})=>{let i=r[n],a=i===void 0?void 0:t[i];return i===void 0||a===void 0?[]:[{...e,regionBucket:i,file:a.file}]})}function o(e){return new Set(e.map(e=>e.file)).size>=2}function s(e){return e.length>=2&&new Set(e.map(e=>e.regionBucket)).size>=2}function c(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 l(t,n,r){let i=t.map(e=>e.map(e=>{let t=e.startTokenIndex+(n[e.regionBucket]??0),r=e.endTokenIndex+(n[e.regionBucket]??0);return{file:e.file,fileIndex:e.regionBucket,spanCountedElsewhere:e.nestedInLargerGroup,nestedInLargerGroup:e.nestedInLargerGroup,segments:[{startTokenIndex:t,endTokenIndex:r}],tokenCount:e.tokenCount,startTokenIndex:t,endTokenIndex:r,startIndex:e.startIndex,endIndex:e.endIndex,startLine:e.startLine,endLine:e.endLine}}).toSorted((e,t)=>e.startTokenIndex-t.startTokenIndex));return e.mergeAdjacentGroups(i,r,o)}function u(t,n,r){let i=[],a=/* @__PURE__ */ new Map,o=new Map(n.map((e,t)=>[e.file,{tokens:e.tokens,codeLineNumbers:e.codeLineNumbers,offset:r[t]??0}])),s=/* @__PURE__ */ new Map,c=0;for(let n of t){c+=e.countRedundantFragments(n);let t=n.map(e=>{let{file:t,startLine:n,endLine:r,tokenCount:i}=e,a=d(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),r=[...new Set(t.map(({file:e})=>e))];for(let e of r)a.set(e,(a.get(e)??0)+1);i.push({files:r,occurrences:t,tokenCount:Math.min(...n.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 d(t,n,r){let i=n.get(t.file);if(!i?.tokens)return;let a=r.get(t.file);a||(a=/* @__PURE__ */ new Set,r.set(t.file,a));let o=/* @__PURE__ */ new Set;for(let n of t.segments)e.collectSegmentLines({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)}exports.measureCrossFileDuplication=r;
 //# sourceMappingURL=crossFileDuplication.cjs.map

package/dist/crossFileDuplication.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"crossFileDuplication.cjs","names":["resolveDuplicationOptions","selectMaximalGroups","collectCrossFileNearMissGroups","buildLiteralCountPrefix","collectSequenceWindowCandidates","mergeAdjacentGroups","countRedundantFragments"],"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":"6HAkFA,SAAgB,EACd,EACA,EAC6B,CAC7B,GAAM,CAAE,YAAW,eAAc,wBAAyBA,EAAAA,0BAA0B,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,EAAUC,EAAAA,oBACd,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,OAAOC,EAAAA,+BAA+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,mBAAoBC,EAAAA,wBAAwB,CAAM,EAAG,WAAY,CAAoB,CAAC,GAMlH,OAHI,EAAS,OAAS,EACb,CAAC,EAEHC,EAAAA,gCAAgC,EAAU,EAAW,EAAI,CAAC,CAAC,SAAS,CAAE,YAAW,kBAAmB,CACzG,IAAM,EAAY,EAAmB,GAC/B,EAAO,IAAc,IAAA,GAAY,IAAA,GAAY,EAAM,GACzD,OAAO,IAAc,IAAA,IAAa,IAAS,IAAA,GACvC,CAAC,EACD,CAAC,CAAE,GAAG,EAAW,aAAc,EAAW,KAAM,EAAK,IAAK,CAAC,CACjE,CAAC,CACH,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,OAAOC,EAAAA,oBAAoB,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,GAAuBC,EAAAA,wBAAwB,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,EAAA,oBACE,CACE,gBAAiB,EAAQ,gBAAkB,EAAS,OACpD,cAAe,EAAQ,cAAgB,EAAS,MAClD,EACA,EAAS,OACT,EAAS,gBACT,CACF,CAEJ"}
+{"version":3,"file":"crossFileDuplication.cjs","names":["resolveDuplicationOptions","selectMaximalGroups","collectCrossFileNearMissGroups","buildLiteralCountPrefix","collectSequenceWindowCandidates","mergeAdjacentGroups","countRedundantFragments"],"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":"6HA6FA,SAAgB,EACd,EACA,EAC6B,CAC7B,GAAM,CAAE,YAAW,eAAc,wBAAyBA,EAAAA,0BAA0B,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,EAAUC,EAAAA,oBACd,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,OAAOC,EAAAA,+BAA+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,mBAAoBC,EAAAA,wBAAwB,CAAM,EAAG,WAAY,CAAoB,CAAC,GAMlH,OAHI,EAAS,OAAS,EACb,CAAC,EAEHC,EAAAA,gCAAgC,EAAU,EAAW,EAAI,CAAC,CAAC,SAAS,CAAE,YAAW,kBAAmB,CACzG,IAAM,EAAY,EAAmB,GAC/B,EAAO,IAAc,IAAA,GAAY,IAAA,GAAY,EAAM,GACzD,OAAO,IAAc,IAAA,IAAa,IAAS,IAAA,GACvC,CAAC,EACD,CAAC,CAAE,GAAG,EAAW,aAAc,EAAW,KAAM,EAAK,IAAK,CAAC,CACjE,CAAC,CACH,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,OAAOC,EAAAA,oBAAoB,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,GAAuBC,EAAAA,wBAAwB,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,EAAA,oBACE,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/crossFileDuplication.d.ts

@@ -7,7 +7,18 @@ export interface CrossFileDuplicationSourceFile extends Partial<CrossFileDuplica
 export interface CrossFileDuplicateOccurrence {
     endLine: number;
     file: string;
+    /**
+     * The code lines carrying the occurrence's matched tokens (see
+     * DuplicateBlockOccurrence.lineNumbers). Absent for a file that supplied only candidates (no
+     * `tokens`), whose matched lines are unknowable.
+     */
+    lineNumbers?: number[];
     startLine: number;
+    /**
+     * The tokens of its matched runs, or of a near-miss block or its matched cores with the edited
+     * tokens among them; the group's `tokenCount` is the smallest of its occurrences'.
+     */
+    tokenCount: number;
 }
 export interface CrossFileDuplicateBlockGroup {
     files: string[];

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/index.d.ts

@@ -3,4 +3,4 @@ export type { CrossFileDuplicateBlockGroup, CrossFileDuplicateOccurrence, CrossF
 export type { CrossFileDuplicateCandidate, CrossFileDuplicationFileData } from './duplication.js';
 export { defaultLanguages, detectLanguage, supportedLanguages } from './languages.js';
 export { TreeMeasurer, collectCrossFileDuplicationFileData, collectDuplicationCandidates, collectFunctionTokenSequences, defaultMeasurer, measureCode, } from './metrics.js';
-export type { CodeMetrics, DuplicationMetrics, DuplicationOptions, FunctionMetrics, HalsteadMetrics, LanguageDefinition, LanguageName, LineMetrics, MeasureOptions, SupportedLanguage, } from './types.js';
+export type { CodeMetrics, CognitiveBlock, DuplicationMetrics, DuplicationOptions, FunctionMetrics, HalsteadMetrics, LanguageDefinition, LanguageName, LineMetrics, MeasureOptions, SupportedLanguage, } from './types.js';

package/dist/metrics.cjs

@@ -1,2 +1,2 @@
-"use strict";const e=require("./languages.cjs"),t=require("./nativeMetrics.cjs");var n=class{registry=e.createLanguageRegistry();getSupportedLanguages(){return[...new Set([...this.registry.values()].map(e=>e.name))]}measure(e,n){let r=this.resolveLanguage(n.language),i=n.includeSyntaxTree??!1;return a(t.measureCodeNative(e,r.name,i,n.duplication),i)}measureWithCrossFileData(e,n){let i=this.resolveLanguage(n.language),a=n.includeSyntaxTree??!1;return r(t.measureCodeNative(e,i.name,a,n.duplication,!0),a)}async measureWithCrossFileDataAsync(e,n){let i=this.resolveLanguage(n.language),a=n.includeSyntaxTree??!1;return r(await t.measureCodeNativeAsync(e,i.name,a,n.duplication,!0),a)}collectDuplicationCandidates(e,t){return this.collectCrossFileDuplicationFileData(e,t).candidates}collectCrossFileDuplicationFileData(e,n){let r=this.resolveLanguage(n.language);return i(t.collectCrossFileDataNative(e,r.name,n.duplication?.minTokens))}collectFunctionTokenSequences(e,n){let r=this.resolveLanguage(n.language);return t.collectFunctionTokenSequencesNative(e,r.name)}resolveLanguage(e){let t=this.registry.get(e);if(!t)throw Error(`Unsupported language: ${e}`);return t}};function r(e,t){if(!e.crossFileData)throw Error(`The native addon returned no cross-file duplication data`);return{metrics:a(e,t),crossFileData:i(e.crossFileData)}}function i(e){return{candidates:e.candidates,tokens:e.tokens,containerStatements:e.containerStatements,nearMissBlocks:e.nearMissBlocks,codeLineNumbers:new Set(e.codeLineNumbers)}}function a(e,t){return{language:e.language,bytes:e.bytes,lines:e.lines,functions:e.functions.map(e=>({name:e.name,nodeType:e.nodeType,startLine:e.startLine,startColumn:e.startColumn,endLine:e.endLine,endColumn:e.endColumn,cyclomaticComplexity:e.cyclomaticComplexity,cognitiveComplexity:e.cognitiveComplexity,nestingDepth:e.nestingDepth,ncss:e.ncss,parameterCount:e.parameterCount,halstead:o(e.halsteadCounts),depDegree:e.depDegree})),cyclomaticComplexity:e.cyclomaticComplexity,cognitiveComplexity:e.cognitiveComplexity,maxCognitiveComplexity:e.maxCognitiveComplexity,nestingDepth:e.nestingDepth,ncssCount:e.ncssCount,duplication:e.duplication,halstead:o(e.halsteadCounts),syntaxTree:t?e.syntaxTree:void 0}}function o(e){let{distinctOperators:t,distinctOperands:n,totalOperators:r,totalOperands:i}=e,a=t+n,o=r+i,s=a===0?0:o*Math.log2(a),c=n===0?0:t/2*(i/n);return{distinctOperators:t,distinctOperands:n,totalOperators:r,totalOperands:i,vocabulary:a,length:o,volume:s,difficulty:c,effort:c*s}}const s=new n;function c(e,t){return s.measure(e,t)}function l(e,t){return s.measureWithCrossFileDataAsync(e,t)}function u(e,t){return s.collectDuplicationCandidates(e,t)}function d(e,t){return s.collectFunctionTokenSequences(e,t)}function f(e,t){return s.collectCrossFileDuplicationFileData(e,t)}exports.TreeMeasurer=n,exports.collectCrossFileDuplicationFileData=f,exports.collectDuplicationCandidates=u,exports.collectFunctionTokenSequences=d,exports.defaultMeasurer=s,exports.measureCode=c,exports.measureCodeWithCrossFileDataAsync=l;
+"use strict";const e=require("./languages.cjs"),t=require("./nativeMetrics.cjs");var n=class{registry=e.createLanguageRegistry();getSupportedLanguages(){return[...new Set([...this.registry.values()].map(e=>e.name))]}measure(e,n){let r=this.resolveLanguage(n.language),i=n.includeSyntaxTree??!1;return a(t.measureCodeNative(e,r.name,i,n.duplication),i)}measureWithCrossFileData(e,n){let i=this.resolveLanguage(n.language),a=n.includeSyntaxTree??!1;return r(t.measureCodeNative(e,i.name,a,n.duplication,!0),a)}async measureWithCrossFileDataAsync(e,n){let i=this.resolveLanguage(n.language),a=n.includeSyntaxTree??!1;return r(await t.measureCodeNativeAsync(e,i.name,a,n.duplication,!0),a)}collectDuplicationCandidates(e,t){return this.collectCrossFileDuplicationFileData(e,t).candidates}collectCrossFileDuplicationFileData(e,n){let r=this.resolveLanguage(n.language);return i(t.collectCrossFileDataNative(e,r.name,n.duplication?.minTokens))}collectFunctionTokenSequences(e,n){let r=this.resolveLanguage(n.language);return t.collectFunctionTokenSequencesNative(e,r.name)}resolveLanguage(e){let t=this.registry.get(e);if(!t)throw Error(`Unsupported language: ${e}`);return t}};function r(e,t){if(!e.crossFileData)throw Error(`The native addon returned no cross-file duplication data`);return{metrics:a(e,t),crossFileData:i(e.crossFileData)}}function i(e){return{candidates:e.candidates,tokens:e.tokens,containerStatements:e.containerStatements,nearMissBlocks:e.nearMissBlocks,codeLineNumbers:new Set(e.codeLineNumbers)}}function a(e,t){return{language:e.language,bytes:e.bytes,lines:e.lines,codeLineNumbers:e.codeLineNumbers,functions:e.functions.map(e=>({name:e.name,containerName:e.containerName,nodeType:e.nodeType,startLine:e.startLine,startColumn:e.startColumn,endLine:e.endLine,endColumn:e.endColumn,cyclomaticComplexity:e.cyclomaticComplexity,cognitiveComplexity:e.cognitiveComplexity,cognitiveBlocks:e.cognitiveBlocks??[],nestingDepth:e.nestingDepth,ncss:e.ncss,parameterCount:e.parameterCount,halstead:o(e.halsteadCounts),depDegree:e.depDegree})),cyclomaticComplexity:e.cyclomaticComplexity,cognitiveComplexity:e.cognitiveComplexity,maxCognitiveComplexity:e.maxCognitiveComplexity,nestingDepth:e.nestingDepth,ncssCount:e.ncssCount,duplication:e.duplication,halstead:o(e.halsteadCounts),syntaxTree:t?e.syntaxTree:void 0}}function o(e){let{distinctOperators:t,distinctOperands:n,totalOperators:r,totalOperands:i}=e,a=t+n,o=r+i,s=a===0?0:o*Math.log2(a),c=n===0?0:t/2*(i/n);return{distinctOperators:t,distinctOperands:n,totalOperators:r,totalOperands:i,vocabulary:a,length:o,volume:s,difficulty:c,effort:c*s}}const s=new n;function c(e,t){return s.measure(e,t)}function l(e,t){return s.measureWithCrossFileDataAsync(e,t)}function u(e,t){return s.collectDuplicationCandidates(e,t)}function d(e,t){return s.collectFunctionTokenSequences(e,t)}function f(e,t){return s.collectCrossFileDuplicationFileData(e,t)}exports.TreeMeasurer=n,exports.collectCrossFileDuplicationFileData=f,exports.collectDuplicationCandidates=u,exports.collectFunctionTokenSequences=d,exports.defaultMeasurer=s,exports.measureCode=c,exports.measureCodeWithCrossFileDataAsync=l;
 //# sourceMappingURL=metrics.cjs.map