@clazic/kordoc 2.4.9 → 2.4.11

package/dist/{batch-provider-VX7CY6UH.js → batch-provider-5BFJRKAZ.js}

@@ -135,7 +135,9 @@ ${fileRefs}`;
     const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`;
     throw new Error(`${mode} \uBC30\uCE58 OCR \uC2E4\uD328: ${errMsg}`);
   }
-  return result.stdout || "";
+  const output = result.stdout || "";
+  checkForLimitError(output, mode);
+  return output;
 }
 async function callBatchCodexCli(imagePaths) {
   const outPath = join(tmpdir(), `kordoc-codex-batch-${Date.now()}-${Math.random().toString(36).slice(2)}.txt`);
@@ -156,11 +158,14 @@ async function callBatchCodexCli(imagePaths) {
       const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`;
       throw new Error(`codex \uBC30\uCE58 OCR \uC2E4\uD328: ${errMsg}`);
     }
+    let text;
     try {
-      return readFileSync(outPath, "utf-8");
+      text = readFileSync(outPath, "utf-8");
     } catch {
-      return result.stdout || "";
+      text = result.stdout || "";
     }
+    checkForLimitError(text, "codex");
+    return text;
   } finally {
     try {
       unlinkSync(outPath);
@@ -168,6 +173,12 @@ async function callBatchCodexCli(imagePaths) {
     }
   }
 }
+function checkForLimitError(output, mode) {
+  const lower = output.toLowerCase();
+  if (lower.includes("usage limit") || lower.includes("rate limit")) {
+    throw new Error(`${mode} \uC0AC\uC6A9\uB7C9/\uC18D\uB3C4 \uC81C\uD55C: ${output.trim().slice(0, 200)}`);
+  }
+}
 function stripCodeFence(text) {
   const match = text.match(/^```(?:markdown|md)?\s*\n([\s\S]*?)\n```\s*$/m);
   return match ? match[1].trim() : text;
@@ -176,4 +187,4 @@ export {
   DEFAULT_BATCH_SIZES,
   createBatchCliProvider
 };
-//# sourceMappingURL=batch-provider-VX7CY6UH.js.map
+//# sourceMappingURL=batch-provider-5BFJRKAZ.js.map

package/dist/{batch-provider-VX7CY6UH.js.map → batch-provider-5BFJRKAZ.js.map}

@@ -1 +1 @@
-{"version":3,"sources":["../src/ocr/batch-provider.ts"],"sourcesContent":["/**\n * CLI 배치 OCR 프로바이더\n *\n * 여러 페이지 이미지를 단일 CLI 호출로 처리하여 API 호출 수를 대폭 감소.\n * gemini/claude: @file 멀티 참조, codex: --image 멀티 플래그\n *\n * 299페이지 기준:\n * - 기존: CLI 299회 호출 (~30분)\n * - 배치: CLI 3~6회 호출 (~3분)\n */\n\nimport { spawn, execSync } from \"child_process\"\nimport { writeFileSync, readFileSync, unlinkSync, mkdirSync } from \"fs\"\nimport { join } from \"path\"\nimport { tmpdir } from \"os\"\nimport type { StructuredOcrResult, BatchOcrProvider } from \"../types.js\"\n\n/** 배치 OCR 프롬프트 */\nconst BATCH_OCR_PROMPT =\n  \"다음 문서 페이지 이미지들을 OCR하여 순수 Markdown으로 변환하세요.\\n\\n\" +\n  \"규칙:\\n\" +\n  \"- 각 페이지 결과 사이에 반드시 이 구분자를 삽입: <!-- PAGE_BREAK -->\\n\" +\n  \"- 테이블은 Markdown 테이블 문법 사용 (| 구분, |---|---| 헤더 구분선 포함)\\n\" +\n  \"- 병합된 셀은 해당 위치에 내용 기재\\n\" +\n  \"- 헤딩은 글자 크기에 따라 ## ~ ###### 사용\\n\" +\n  \"- 리스트는 - 또는 1. 사용\\n\" +\n  \"- 이미지, 도형 등 비텍스트 요소는 무시\\n\" +\n  \"- 원문의 읽기 순서와 구조를 유지\\n\" +\n  \"- ```로 감싸지 말고 순수 Markdown만 출력\"\n\n/** 모드별 기본 배치 크기 (CLI 내부 타임아웃 + 실측 기반)\n *\n * gemini CLI: 10장 이상에서 AbortError 발생 (내부 타임아웃).\n * 5장 배치가 안정적으로 동작 확인 (35초/배치).\n * 299페이지 = 60배치 = 기존 299회 대비 80% 감소.\n */\nexport const DEFAULT_BATCH_SIZES: Record<string, number> = {\n  gemini: 5,\n  claude: 5,\n  codex: 10,\n}\n\n/**\n * 임시 디렉토리 — gemini CLI는 cwd 하위 + gitignore 밖만 @참조 가능\n *\n * 숨김 처리:\n * - macOS/Linux: '.' 접두사로 기본 숨김 (ls -a 로만 표시)\n * - Windows: '.' 접두사 + attrib +h 로 숨김 속성 부여\n */\nlet _batchTempDir: string | null = null\nfunction getBatchTempDir(): string {\n  if (!_batchTempDir) {\n    _batchTempDir = join(process.cwd(), \".kordoc_ocr_tmp\")\n    mkdirSync(_batchTempDir, { recursive: true })\n    // Windows: dot-prefix만으로 숨김 처리 불충분 → attrib +h 추가\n    if (process.platform === \"win32\") {\n      try { execSync(`attrib +h \"${_batchTempDir}\"`, { stdio: \"ignore\" }) } catch { /* ignore */ }\n    }\n  }\n  return _batchTempDir\n}\n\n/**\n * 배치 CLI 프로바이더 생성\n */\nexport function createBatchCliProvider(\n  mode: \"gemini\" | \"claude\" | \"codex\",\n  batchSize: number\n): BatchOcrProvider {\n  return {\n    __batch: true as const,\n    batchSize,\n    async processBatch(pages) {\n      const results = new Map<number, StructuredOcrResult>()\n      const tempDir = getBatchTempDir()\n      const tempFiles: string[] = []\n\n      try {\n        // 1. Write all page images to temp files\n        for (const { image, pageNum } of pages) {\n          const path = join(tempDir, `batch-p${pageNum}.png`)\n          writeFileSync(path, image)\n          tempFiles.push(path)\n        }\n\n        // 2. Call CLI with all file references (비동기 — 병렬 배치 실행 가능)\n        let output: string\n        if (mode === \"codex\") {\n          output = await callBatchCodexCli(tempFiles)\n        } else {\n          output = await callBatchCli(mode, tempFiles)\n        }\n\n        // 3. Parse response by PAGE_BREAK separator\n        const cleaned = stripCodeFence(output.trim())\n        const parts = cleaned.split(/<!--\\s*PAGE_BREAK\\s*-->/)\n          .map(p => p.trim())\n          .filter(p => p.length > 0)\n\n        // 4. Map results to page numbers (best-effort if count mismatch)\n        for (let i = 0; i < pages.length; i++) {\n          const pageNum = pages[i].pageNum\n          if (i < parts.length) {\n            results.set(pageNum, { markdown: parts[i] })\n          }\n          // If fewer parts than pages, remaining pages get no result\n        }\n      } finally {\n        // 5. Clean up temp files\n        for (const f of tempFiles) {\n          try { unlinkSync(f) } catch { /* ignore */ }\n        }\n      }\n\n      return results\n    },\n  }\n}\n\n/**\n * 비동기 CLI 실행 헬퍼 — spawn + Promise 래핑.\n * spawnSync는 이벤트 루프를 차단하여 병렬 배치 실행 불가.\n */\nfunction spawnAsync(\n  cmd: string,\n  args: string[],\n  opts: { timeoutMs: number; cwd?: string; stdin?: string }\n): Promise<{ stdout: string; stderr: string; exitCode: number }> {\n  return new Promise((resolve, reject) => {\n    const child = spawn(cmd, args, {\n      cwd: opts.cwd,\n      env: process.env,\n      stdio: [\"pipe\", \"pipe\", \"pipe\"],\n      shell: process.platform === \"win32\",\n    })\n\n    let stdout = \"\"\n    let stderr = \"\"\n    let killed = false\n\n    child.stdout.setEncoding(\"utf-8\")\n    child.stderr.setEncoding(\"utf-8\")\n    child.stdout.on(\"data\", (d: string) => { stdout += d })\n    child.stderr.on(\"data\", (d: string) => { stderr += d })\n\n    const timer = setTimeout(() => {\n      killed = true\n      if (process.platform === \"win32\") {\n        child.kill()\n      } else {\n        child.kill(\"SIGTERM\")\n      }\n    }, opts.timeoutMs)\n\n    if (opts.stdin !== undefined) {\n      child.stdin.end(opts.stdin)\n    } else {\n      child.stdin.end()\n    }\n\n    child.on(\"close\", (code) => {\n      clearTimeout(timer)\n      if (killed) {\n        reject(new Error(`타임아웃 (${Math.round(opts.timeoutMs / 1000)}초)`))\n      } else {\n        resolve({ stdout, stderr, exitCode: code ?? 1 })\n      }\n    })\n    child.on(\"error\", (err) => {\n      clearTimeout(timer)\n      reject(err)\n    })\n  })\n}\n\n/** gemini/claude 배치 호출 (비동기) */\nasync function callBatchCli(mode: \"gemini\" | \"claude\", imagePaths: string[]): Promise<string> {\n  const fileRefs = imagePaths.map(p => `@${p.replace(/\\\\/g, \"/\")}`).join(\"\\n\")\n  const prompt = `${BATCH_OCR_PROMPT}\\n\\n${fileRefs}`\n\n  let args: string[]\n  if (mode === \"gemini\") {\n    const model = process.env.KORDOC_GEMINI_MODEL ?? \"gemini-2.5-flash\"\n    args = [\"--prompt\", prompt, \"--yolo\", \"--model\", model]\n  } else {\n    args = [\"--print\", prompt]\n    const model = process.env.KORDOC_CLAUDE_MODEL\n    if (model) args.push(\"--model\", model)\n  }\n\n  const timeoutMs = 60_000 + imagePaths.length * 20_000\n  const result = await spawnAsync(mode, args, {\n    timeoutMs,\n    ...(mode === \"claude\" ? { cwd: tmpdir() } : {}),\n  })\n\n  if (result.exitCode !== 0) {\n    const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`\n    throw new Error(`${mode} 배치 OCR 실패: ${errMsg}`)\n  }\n\n  return result.stdout || \"\"\n}\n\n/** codex 배치 호출 (비동기) — --image를 여러 번 지정 */\nasync function callBatchCodexCli(imagePaths: string[]): Promise<string> {\n  const outPath = join(tmpdir(), `kordoc-codex-batch-${Date.now()}-${Math.random().toString(36).slice(2)}.txt`)\n  try {\n    const args = [\"exec\", BATCH_OCR_PROMPT]\n    for (const p of imagePaths) {\n      args.push(\"--image\", p)\n    }\n    args.push(\"--output-last-message\", outPath)\n    const model = process.env.KORDOC_CODEX_MODEL\n    if (model) args.push(\"--model\", model)\n\n    const timeoutMs = 60_000 + imagePaths.length * 20_000\n    const result = await spawnAsync(\"codex\", args, {\n      timeoutMs,\n      stdin: \"\",\n    })\n\n    if (result.exitCode !== 0) {\n      const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`\n      throw new Error(`codex 배치 OCR 실패: ${errMsg}`)\n    }\n\n    try {\n      return readFileSync(outPath, \"utf-8\")\n    } catch {\n      return result.stdout || \"\"\n    }\n  } finally {\n    try { unlinkSync(outPath) } catch { /* ignore */ }\n  }\n}\n\n/** LLM 출력에서 코드 펜스 제거 (cli-provider.ts와 동일 로직) */\nfunction stripCodeFence(text: string): string {\n  const match = text.match(/^```(?:markdown|md)?\\s*\\n([\\s\\S]*?)\\n```\\s*$/m)\n  return match ? match[1].trim() : text\n}\n"],"mappings":";;;;AAWA,SAAS,OAAO,gBAAgB;AAChC,SAAS,eAAe,cAAc,YAAY,iBAAiB;AACnE,SAAS,YAAY;AACrB,SAAS,cAAc;AAIvB,IAAM,mBACJ;AAiBK,IAAM,sBAA8C;AAAA,EACzD,QAAQ;AAAA,EACR,QAAQ;AAAA,EACR,OAAO;AACT;AASA,IAAI,gBAA+B;AACnC,SAAS,kBAA0B;AACjC,MAAI,CAAC,eAAe;AAClB,oBAAgB,KAAK,QAAQ,IAAI,GAAG,iBAAiB;AACrD,cAAU,eAAe,EAAE,WAAW,KAAK,CAAC;AAE5C,QAAI,QAAQ,aAAa,SAAS;AAChC,UAAI;AAAE,iBAAS,cAAc,aAAa,KAAK,EAAE,OAAO,SAAS,CAAC;AAAA,MAAE,QAAQ;AAAA,MAAe;AAAA,IAC7F;AAAA,EACF;AACA,SAAO;AACT;AAKO,SAAS,uBACd,MACA,WACkB;AAClB,SAAO;AAAA,IACL,SAAS;AAAA,IACT;AAAA,IACA,MAAM,aAAa,OAAO;AACxB,YAAM,UAAU,oBAAI,IAAiC;AACrD,YAAM,UAAU,gBAAgB;AAChC,YAAM,YAAsB,CAAC;AAE7B,UAAI;AAEF,mBAAW,EAAE,OAAO,QAAQ,KAAK,OAAO;AACtC,gBAAM,OAAO,KAAK,SAAS,UAAU,OAAO,MAAM;AAClD,wBAAc,MAAM,KAAK;AACzB,oBAAU,KAAK,IAAI;AAAA,QACrB;AAGA,YAAI;AACJ,YAAI,SAAS,SAAS;AACpB,mBAAS,MAAM,kBAAkB,SAAS;AAAA,QAC5C,OAAO;AACL,mBAAS,MAAM,aAAa,MAAM,SAAS;AAAA,QAC7C;AAGA,cAAM,UAAU,eAAe,OAAO,KAAK,CAAC;AAC5C,cAAM,QAAQ,QAAQ,MAAM,yBAAyB,EAClD,IAAI,OAAK,EAAE,KAAK,CAAC,EACjB,OAAO,OAAK,EAAE,SAAS,CAAC;AAG3B,iBAAS,IAAI,GAAG,IAAI,MAAM,QAAQ,KAAK;AACrC,gBAAM,UAAU,MAAM,CAAC,EAAE;AACzB,cAAI,IAAI,MAAM,QAAQ;AACpB,oBAAQ,IAAI,SAAS,EAAE,UAAU,MAAM,CAAC,EAAE,CAAC;AAAA,UAC7C;AAAA,QAEF;AAAA,MACF,UAAE;AAEA,mBAAW,KAAK,WAAW;AACzB,cAAI;AAAE,uBAAW,CAAC;AAAA,UAAE,QAAQ;AAAA,UAAe;AAAA,QAC7C;AAAA,MACF;AAEA,aAAO;AAAA,IACT;AAAA,EACF;AACF;AAMA,SAAS,WACP,KACA,MACA,MAC+D;AAC/D,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,QAAQ,MAAM,KAAK,MAAM;AAAA,MAC7B,KAAK,KAAK;AAAA,MACV,KAAK,QAAQ;AAAA,MACb,OAAO,CAAC,QAAQ,QAAQ,MAAM;AAAA,MAC9B,OAAO,QAAQ,aAAa;AAAA,IAC9B,CAAC;AAED,QAAI,SAAS;AACb,QAAI,SAAS;AACb,QAAI,SAAS;AAEb,UAAM,OAAO,YAAY,OAAO;AAChC,UAAM,OAAO,YAAY,OAAO;AAChC,UAAM,OAAO,GAAG,QAAQ,CAAC,MAAc;AAAE,gBAAU;AAAA,IAAE,CAAC;AACtD,UAAM,OAAO,GAAG,QAAQ,CAAC,MAAc;AAAE,gBAAU;AAAA,IAAE,CAAC;AAEtD,UAAM,QAAQ,WAAW,MAAM;AAC7B,eAAS;AACT,UAAI,QAAQ,aAAa,SAAS;AAChC,cAAM,KAAK;AAAA,MACb,OAAO;AACL,cAAM,KAAK,SAAS;AAAA,MACtB;AAAA,IACF,GAAG,KAAK,SAAS;AAEjB,QAAI,KAAK,UAAU,QAAW;AAC5B,YAAM,MAAM,IAAI,KAAK,KAAK;AAAA,IAC5B,OAAO;AACL,YAAM,MAAM,IAAI;AAAA,IAClB;AAEA,UAAM,GAAG,SAAS,CAAC,SAAS;AAC1B,mBAAa,KAAK;AAClB,UAAI,QAAQ;AACV,eAAO,IAAI,MAAM,6BAAS,KAAK,MAAM,KAAK,YAAY,GAAI,CAAC,SAAI,CAAC;AAAA,MAClE,OAAO;AACL,gBAAQ,EAAE,QAAQ,QAAQ,UAAU,QAAQ,EAAE,CAAC;AAAA,MACjD;AAAA,IACF,CAAC;AACD,UAAM,GAAG,SAAS,CAAC,QAAQ;AACzB,mBAAa,KAAK;AAClB,aAAO,GAAG;AAAA,IACZ,CAAC;AAAA,EACH,CAAC;AACH;AAGA,eAAe,aAAa,MAA2B,YAAuC;AAC5F,QAAM,WAAW,WAAW,IAAI,OAAK,IAAI,EAAE,QAAQ,OAAO,GAAG,CAAC,EAAE,EAAE,KAAK,IAAI;AAC3E,QAAM,SAAS,GAAG,gBAAgB;AAAA;AAAA,EAAO,QAAQ;AAEjD,MAAI;AACJ,MAAI,SAAS,UAAU;AACrB,UAAM,QAAQ,QAAQ,IAAI,uBAAuB;AACjD,WAAO,CAAC,YAAY,QAAQ,UAAU,WAAW,KAAK;AAAA,EACxD,OAAO;AACL,WAAO,CAAC,WAAW,MAAM;AACzB,UAAM,QAAQ,QAAQ,IAAI;AAC1B,QAAI,MAAO,MAAK,KAAK,WAAW,KAAK;AAAA,EACvC;AAEA,QAAM,YAAY,MAAS,WAAW,SAAS;AAC/C,QAAM,SAAS,MAAM,WAAW,MAAM,MAAM;AAAA,IAC1C;AAAA,IACA,GAAI,SAAS,WAAW,EAAE,KAAK,OAAO,EAAE,IAAI,CAAC;AAAA,EAC/C,CAAC;AAED,MAAI,OAAO,aAAa,GAAG;AACzB,UAAM,SAAS,OAAO,QAAQ,KAAK,KAAK,aAAa,OAAO,QAAQ;AACpE,UAAM,IAAI,MAAM,GAAG,IAAI,mCAAe,MAAM,EAAE;AAAA,EAChD;AAEA,SAAO,OAAO,UAAU;AAC1B;AAGA,eAAe,kBAAkB,YAAuC;AACtE,QAAM,UAAU,KAAK,OAAO,GAAG,sBAAsB,KAAK,IAAI,CAAC,IAAI,KAAK,OAAO,EAAE,SAAS,EAAE,EAAE,MAAM,CAAC,CAAC,MAAM;AAC5G,MAAI;AACF,UAAM,OAAO,CAAC,QAAQ,gBAAgB;AACtC,eAAW,KAAK,YAAY;AAC1B,WAAK,KAAK,WAAW,CAAC;AAAA,IACxB;AACA,SAAK,KAAK,yBAAyB,OAAO;AAC1C,UAAM,QAAQ,QAAQ,IAAI;AAC1B,QAAI,MAAO,MAAK,KAAK,WAAW,KAAK;AAErC,UAAM,YAAY,MAAS,WAAW,SAAS;AAC/C,UAAM,SAAS,MAAM,WAAW,SAAS,MAAM;AAAA,MAC7C;AAAA,MACA,OAAO;AAAA,IACT,CAAC;AAED,QAAI,OAAO,aAAa,GAAG;AACzB,YAAM,SAAS,OAAO,QAAQ,KAAK,KAAK,aAAa,OAAO,QAAQ;AACpE,YAAM,IAAI,MAAM,wCAAoB,MAAM,EAAE;AAAA,IAC9C;AAEA,QAAI;AACF,aAAO,aAAa,SAAS,OAAO;AAAA,IACtC,QAAQ;AACN,aAAO,OAAO,UAAU;AAAA,IAC1B;AAAA,EACF,UAAE;AACA,QAAI;AAAE,iBAAW,OAAO;AAAA,IAAE,QAAQ;AAAA,IAAe;AAAA,EACnD;AACF;AAGA,SAAS,eAAe,MAAsB;AAC5C,QAAM,QAAQ,KAAK,MAAM,+CAA+C;AACxE,SAAO,QAAQ,MAAM,CAAC,EAAE,KAAK,IAAI;AACnC;","names":[]}
+{"version":3,"sources":["../src/ocr/batch-provider.ts"],"sourcesContent":["/**\n * CLI 배치 OCR 프로바이더\n *\n * 여러 페이지 이미지를 단일 CLI 호출로 처리하여 API 호출 수를 대폭 감소.\n * gemini/claude: @file 멀티 참조, codex: --image 멀티 플래그\n *\n * 299페이지 기준:\n * - 기존: CLI 299회 호출 (~30분)\n * - 배치: CLI 3~6회 호출 (~3분)\n */\n\nimport { spawn, execSync } from \"child_process\"\nimport { writeFileSync, readFileSync, unlinkSync, mkdirSync } from \"fs\"\nimport { join } from \"path\"\nimport { tmpdir } from \"os\"\nimport type { StructuredOcrResult, BatchOcrProvider } from \"../types.js\"\n\n/** 배치 OCR 프롬프트 */\nconst BATCH_OCR_PROMPT =\n  \"다음 문서 페이지 이미지들을 OCR하여 순수 Markdown으로 변환하세요.\\n\\n\" +\n  \"규칙:\\n\" +\n  \"- 각 페이지 결과 사이에 반드시 이 구분자를 삽입: <!-- PAGE_BREAK -->\\n\" +\n  \"- 테이블은 Markdown 테이블 문법 사용 (| 구분, |---|---| 헤더 구분선 포함)\\n\" +\n  \"- 병합된 셀은 해당 위치에 내용 기재\\n\" +\n  \"- 헤딩은 글자 크기에 따라 ## ~ ###### 사용\\n\" +\n  \"- 리스트는 - 또는 1. 사용\\n\" +\n  \"- 이미지, 도형 등 비텍스트 요소는 무시\\n\" +\n  \"- 원문의 읽기 순서와 구조를 유지\\n\" +\n  \"- ```로 감싸지 말고 순수 Markdown만 출력\"\n\n/** 모드별 기본 배치 크기 (CLI 내부 타임아웃 + 실측 기반)\n *\n * gemini CLI: 10장 이상에서 AbortError 발생 (내부 타임아웃).\n * 5장 배치가 안정적으로 동작 확인 (35초/배치).\n * 299페이지 = 60배치 = 기존 299회 대비 80% 감소.\n */\nexport const DEFAULT_BATCH_SIZES: Record<string, number> = {\n  gemini: 5,\n  claude: 5,\n  codex: 10,\n}\n\n/**\n * 임시 디렉토리 — gemini CLI는 cwd 하위 + gitignore 밖만 @참조 가능\n *\n * 숨김 처리:\n * - macOS/Linux: '.' 접두사로 기본 숨김 (ls -a 로만 표시)\n * - Windows: '.' 접두사 + attrib +h 로 숨김 속성 부여\n */\nlet _batchTempDir: string | null = null\nfunction getBatchTempDir(): string {\n  if (!_batchTempDir) {\n    _batchTempDir = join(process.cwd(), \".kordoc_ocr_tmp\")\n    mkdirSync(_batchTempDir, { recursive: true })\n    // Windows: dot-prefix만으로 숨김 처리 불충분 → attrib +h 추가\n    if (process.platform === \"win32\") {\n      try { execSync(`attrib +h \"${_batchTempDir}\"`, { stdio: \"ignore\" }) } catch { /* ignore */ }\n    }\n  }\n  return _batchTempDir\n}\n\n/**\n * 배치 CLI 프로바이더 생성\n */\nexport function createBatchCliProvider(\n  mode: \"gemini\" | \"claude\" | \"codex\",\n  batchSize: number\n): BatchOcrProvider {\n  return {\n    __batch: true as const,\n    batchSize,\n    async processBatch(pages) {\n      const results = new Map<number, StructuredOcrResult>()\n      const tempDir = getBatchTempDir()\n      const tempFiles: string[] = []\n\n      try {\n        // 1. Write all page images to temp files\n        for (const { image, pageNum } of pages) {\n          const path = join(tempDir, `batch-p${pageNum}.png`)\n          writeFileSync(path, image)\n          tempFiles.push(path)\n        }\n\n        // 2. Call CLI with all file references (비동기 — 병렬 배치 실행 가능)\n        let output: string\n        if (mode === \"codex\") {\n          output = await callBatchCodexCli(tempFiles)\n        } else {\n          output = await callBatchCli(mode, tempFiles)\n        }\n\n        // 3. Parse response by PAGE_BREAK separator\n        const cleaned = stripCodeFence(output.trim())\n        const parts = cleaned.split(/<!--\\s*PAGE_BREAK\\s*-->/)\n          .map(p => p.trim())\n          .filter(p => p.length > 0)\n\n        // 4. Map results to page numbers (best-effort if count mismatch)\n        for (let i = 0; i < pages.length; i++) {\n          const pageNum = pages[i].pageNum\n          if (i < parts.length) {\n            results.set(pageNum, { markdown: parts[i] })\n          }\n          // If fewer parts than pages, remaining pages get no result\n        }\n      } finally {\n        // 5. Clean up temp files\n        for (const f of tempFiles) {\n          try { unlinkSync(f) } catch { /* ignore */ }\n        }\n      }\n\n      return results\n    },\n  }\n}\n\n/**\n * 비동기 CLI 실행 헬퍼 — spawn + Promise 래핑.\n * spawnSync는 이벤트 루프를 차단하여 병렬 배치 실행 불가.\n */\nfunction spawnAsync(\n  cmd: string,\n  args: string[],\n  opts: { timeoutMs: number; cwd?: string; stdin?: string }\n): Promise<{ stdout: string; stderr: string; exitCode: number }> {\n  return new Promise((resolve, reject) => {\n    const child = spawn(cmd, args, {\n      cwd: opts.cwd,\n      env: process.env,\n      stdio: [\"pipe\", \"pipe\", \"pipe\"],\n      shell: process.platform === \"win32\",\n    })\n\n    let stdout = \"\"\n    let stderr = \"\"\n    let killed = false\n\n    child.stdout.setEncoding(\"utf-8\")\n    child.stderr.setEncoding(\"utf-8\")\n    child.stdout.on(\"data\", (d: string) => { stdout += d })\n    child.stderr.on(\"data\", (d: string) => { stderr += d })\n\n    const timer = setTimeout(() => {\n      killed = true\n      if (process.platform === \"win32\") {\n        child.kill()\n      } else {\n        child.kill(\"SIGTERM\")\n      }\n    }, opts.timeoutMs)\n\n    if (opts.stdin !== undefined) {\n      child.stdin.end(opts.stdin)\n    } else {\n      child.stdin.end()\n    }\n\n    child.on(\"close\", (code) => {\n      clearTimeout(timer)\n      if (killed) {\n        reject(new Error(`타임아웃 (${Math.round(opts.timeoutMs / 1000)}초)`))\n      } else {\n        resolve({ stdout, stderr, exitCode: code ?? 1 })\n      }\n    })\n    child.on(\"error\", (err) => {\n      clearTimeout(timer)\n      reject(err)\n    })\n  })\n}\n\n/** gemini/claude 배치 호출 (비동기) */\nasync function callBatchCli(mode: \"gemini\" | \"claude\", imagePaths: string[]): Promise<string> {\n  const fileRefs = imagePaths.map(p => `@${p.replace(/\\\\/g, \"/\")}`).join(\"\\n\")\n  const prompt = `${BATCH_OCR_PROMPT}\\n\\n${fileRefs}`\n\n  let args: string[]\n  if (mode === \"gemini\") {\n    const model = process.env.KORDOC_GEMINI_MODEL ?? \"gemini-2.5-flash\"\n    args = [\"--prompt\", prompt, \"--yolo\", \"--model\", model]\n  } else {\n    args = [\"--print\", prompt]\n    const model = process.env.KORDOC_CLAUDE_MODEL\n    if (model) args.push(\"--model\", model)\n  }\n\n  const timeoutMs = 60_000 + imagePaths.length * 20_000\n  const result = await spawnAsync(mode, args, {\n    timeoutMs,\n    ...(mode === \"claude\" ? { cwd: tmpdir() } : {}),\n  })\n\n  if (result.exitCode !== 0) {\n    const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`\n    throw new Error(`${mode} 배치 OCR 실패: ${errMsg}`)\n  }\n\n  const output = result.stdout || \"\"\n  checkForLimitError(output, mode)\n  return output\n}\n\n/** codex 배치 호출 (비동기) — --image를 여러 번 지정 */\nasync function callBatchCodexCli(imagePaths: string[]): Promise<string> {\n  const outPath = join(tmpdir(), `kordoc-codex-batch-${Date.now()}-${Math.random().toString(36).slice(2)}.txt`)\n  try {\n    const args = [\"exec\", BATCH_OCR_PROMPT]\n    for (const p of imagePaths) {\n      args.push(\"--image\", p)\n    }\n    args.push(\"--output-last-message\", outPath)\n    const model = process.env.KORDOC_CODEX_MODEL\n    if (model) args.push(\"--model\", model)\n\n    const timeoutMs = 60_000 + imagePaths.length * 20_000\n    const result = await spawnAsync(\"codex\", args, {\n      timeoutMs,\n      stdin: \"\",\n    })\n\n    if (result.exitCode !== 0) {\n      const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`\n      throw new Error(`codex 배치 OCR 실패: ${errMsg}`)\n    }\n\n    let text: string\n    try {\n      text = readFileSync(outPath, \"utf-8\")\n    } catch {\n      text = result.stdout || \"\"\n    }\n    checkForLimitError(text, \"codex\")\n    return text\n  } finally {\n    try { unlinkSync(outPath) } catch { /* ignore */ }\n  }\n}\n\n/**\n * 출력 텍스트에서 사용량·속도 제한 에러 감지.\n * 해당 메시지가 포함된 경우 throw하여 다음 엔진으로 fallback 트리거.\n */\nfunction checkForLimitError(output: string, mode: string): void {\n  const lower = output.toLowerCase()\n  if (lower.includes(\"usage limit\") || lower.includes(\"rate limit\")) {\n    throw new Error(`${mode} 사용량/속도 제한: ${output.trim().slice(0, 200)}`)\n  }\n}\n\n/** LLM 출력에서 코드 펜스 제거 (cli-provider.ts와 동일 로직) */\nfunction stripCodeFence(text: string): string {\n  const match = text.match(/^```(?:markdown|md)?\\s*\\n([\\s\\S]*?)\\n```\\s*$/m)\n  return match ? match[1].trim() : text\n}\n"],"mappings":";;;;AAWA,SAAS,OAAO,gBAAgB;AAChC,SAAS,eAAe,cAAc,YAAY,iBAAiB;AACnE,SAAS,YAAY;AACrB,SAAS,cAAc;AAIvB,IAAM,mBACJ;AAiBK,IAAM,sBAA8C;AAAA,EACzD,QAAQ;AAAA,EACR,QAAQ;AAAA,EACR,OAAO;AACT;AASA,IAAI,gBAA+B;AACnC,SAAS,kBAA0B;AACjC,MAAI,CAAC,eAAe;AAClB,oBAAgB,KAAK,QAAQ,IAAI,GAAG,iBAAiB;AACrD,cAAU,eAAe,EAAE,WAAW,KAAK,CAAC;AAE5C,QAAI,QAAQ,aAAa,SAAS;AAChC,UAAI;AAAE,iBAAS,cAAc,aAAa,KAAK,EAAE,OAAO,SAAS,CAAC;AAAA,MAAE,QAAQ;AAAA,MAAe;AAAA,IAC7F;AAAA,EACF;AACA,SAAO;AACT;AAKO,SAAS,uBACd,MACA,WACkB;AAClB,SAAO;AAAA,IACL,SAAS;AAAA,IACT;AAAA,IACA,MAAM,aAAa,OAAO;AACxB,YAAM,UAAU,oBAAI,IAAiC;AACrD,YAAM,UAAU,gBAAgB;AAChC,YAAM,YAAsB,CAAC;AAE7B,UAAI;AAEF,mBAAW,EAAE,OAAO,QAAQ,KAAK,OAAO;AACtC,gBAAM,OAAO,KAAK,SAAS,UAAU,OAAO,MAAM;AAClD,wBAAc,MAAM,KAAK;AACzB,oBAAU,KAAK,IAAI;AAAA,QACrB;AAGA,YAAI;AACJ,YAAI,SAAS,SAAS;AACpB,mBAAS,MAAM,kBAAkB,SAAS;AAAA,QAC5C,OAAO;AACL,mBAAS,MAAM,aAAa,MAAM,SAAS;AAAA,QAC7C;AAGA,cAAM,UAAU,eAAe,OAAO,KAAK,CAAC;AAC5C,cAAM,QAAQ,QAAQ,MAAM,yBAAyB,EAClD,IAAI,OAAK,EAAE,KAAK,CAAC,EACjB,OAAO,OAAK,EAAE,SAAS,CAAC;AAG3B,iBAAS,IAAI,GAAG,IAAI,MAAM,QAAQ,KAAK;AACrC,gBAAM,UAAU,MAAM,CAAC,EAAE;AACzB,cAAI,IAAI,MAAM,QAAQ;AACpB,oBAAQ,IAAI,SAAS,EAAE,UAAU,MAAM,CAAC,EAAE,CAAC;AAAA,UAC7C;AAAA,QAEF;AAAA,MACF,UAAE;AAEA,mBAAW,KAAK,WAAW;AACzB,cAAI;AAAE,uBAAW,CAAC;AAAA,UAAE,QAAQ;AAAA,UAAe;AAAA,QAC7C;AAAA,MACF;AAEA,aAAO;AAAA,IACT;AAAA,EACF;AACF;AAMA,SAAS,WACP,KACA,MACA,MAC+D;AAC/D,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,QAAQ,MAAM,KAAK,MAAM;AAAA,MAC7B,KAAK,KAAK;AAAA,MACV,KAAK,QAAQ;AAAA,MACb,OAAO,CAAC,QAAQ,QAAQ,MAAM;AAAA,MAC9B,OAAO,QAAQ,aAAa;AAAA,IAC9B,CAAC;AAED,QAAI,SAAS;AACb,QAAI,SAAS;AACb,QAAI,SAAS;AAEb,UAAM,OAAO,YAAY,OAAO;AAChC,UAAM,OAAO,YAAY,OAAO;AAChC,UAAM,OAAO,GAAG,QAAQ,CAAC,MAAc;AAAE,gBAAU;AAAA,IAAE,CAAC;AACtD,UAAM,OAAO,GAAG,QAAQ,CAAC,MAAc;AAAE,gBAAU;AAAA,IAAE,CAAC;AAEtD,UAAM,QAAQ,WAAW,MAAM;AAC7B,eAAS;AACT,UAAI,QAAQ,aAAa,SAAS;AAChC,cAAM,KAAK;AAAA,MACb,OAAO;AACL,cAAM,KAAK,SAAS;AAAA,MACtB;AAAA,IACF,GAAG,KAAK,SAAS;AAEjB,QAAI,KAAK,UAAU,QAAW;AAC5B,YAAM,MAAM,IAAI,KAAK,KAAK;AAAA,IAC5B,OAAO;AACL,YAAM,MAAM,IAAI;AAAA,IAClB;AAEA,UAAM,GAAG,SAAS,CAAC,SAAS;AAC1B,mBAAa,KAAK;AAClB,UAAI,QAAQ;AACV,eAAO,IAAI,MAAM,6BAAS,KAAK,MAAM,KAAK,YAAY,GAAI,CAAC,SAAI,CAAC;AAAA,MAClE,OAAO;AACL,gBAAQ,EAAE,QAAQ,QAAQ,UAAU,QAAQ,EAAE,CAAC;AAAA,MACjD;AAAA,IACF,CAAC;AACD,UAAM,GAAG,SAAS,CAAC,QAAQ;AACzB,mBAAa,KAAK;AAClB,aAAO,GAAG;AAAA,IACZ,CAAC;AAAA,EACH,CAAC;AACH;AAGA,eAAe,aAAa,MAA2B,YAAuC;AAC5F,QAAM,WAAW,WAAW,IAAI,OAAK,IAAI,EAAE,QAAQ,OAAO,GAAG,CAAC,EAAE,EAAE,KAAK,IAAI;AAC3E,QAAM,SAAS,GAAG,gBAAgB;AAAA;AAAA,EAAO,QAAQ;AAEjD,MAAI;AACJ,MAAI,SAAS,UAAU;AACrB,UAAM,QAAQ,QAAQ,IAAI,uBAAuB;AACjD,WAAO,CAAC,YAAY,QAAQ,UAAU,WAAW,KAAK;AAAA,EACxD,OAAO;AACL,WAAO,CAAC,WAAW,MAAM;AACzB,UAAM,QAAQ,QAAQ,IAAI;AAC1B,QAAI,MAAO,MAAK,KAAK,WAAW,KAAK;AAAA,EACvC;AAEA,QAAM,YAAY,MAAS,WAAW,SAAS;AAC/C,QAAM,SAAS,MAAM,WAAW,MAAM,MAAM;AAAA,IAC1C;AAAA,IACA,GAAI,SAAS,WAAW,EAAE,KAAK,OAAO,EAAE,IAAI,CAAC;AAAA,EAC/C,CAAC;AAED,MAAI,OAAO,aAAa,GAAG;AACzB,UAAM,SAAS,OAAO,QAAQ,KAAK,KAAK,aAAa,OAAO,QAAQ;AACpE,UAAM,IAAI,MAAM,GAAG,IAAI,mCAAe,MAAM,EAAE;AAAA,EAChD;AAEA,QAAM,SAAS,OAAO,UAAU;AAChC,qBAAmB,QAAQ,IAAI;AAC/B,SAAO;AACT;AAGA,eAAe,kBAAkB,YAAuC;AACtE,QAAM,UAAU,KAAK,OAAO,GAAG,sBAAsB,KAAK,IAAI,CAAC,IAAI,KAAK,OAAO,EAAE,SAAS,EAAE,EAAE,MAAM,CAAC,CAAC,MAAM;AAC5G,MAAI;AACF,UAAM,OAAO,CAAC,QAAQ,gBAAgB;AACtC,eAAW,KAAK,YAAY;AAC1B,WAAK,KAAK,WAAW,CAAC;AAAA,IACxB;AACA,SAAK,KAAK,yBAAyB,OAAO;AAC1C,UAAM,QAAQ,QAAQ,IAAI;AAC1B,QAAI,MAAO,MAAK,KAAK,WAAW,KAAK;AAErC,UAAM,YAAY,MAAS,WAAW,SAAS;AAC/C,UAAM,SAAS,MAAM,WAAW,SAAS,MAAM;AAAA,MAC7C;AAAA,MACA,OAAO;AAAA,IACT,CAAC;AAED,QAAI,OAAO,aAAa,GAAG;AACzB,YAAM,SAAS,OAAO,QAAQ,KAAK,KAAK,aAAa,OAAO,QAAQ;AACpE,YAAM,IAAI,MAAM,wCAAoB,MAAM,EAAE;AAAA,IAC9C;AAEA,QAAI;AACJ,QAAI;AACF,aAAO,aAAa,SAAS,OAAO;AAAA,IACtC,QAAQ;AACN,aAAO,OAAO,UAAU;AAAA,IAC1B;AACA,uBAAmB,MAAM,OAAO;AAChC,WAAO;AAAA,EACT,UAAE;AACA,QAAI;AAAE,iBAAW,OAAO;AAAA,IAAE,QAAQ;AAAA,IAAe;AAAA,EACnD;AACF;AAMA,SAAS,mBAAmB,QAAgB,MAAoB;AAC9D,QAAM,QAAQ,OAAO,YAAY;AACjC,MAAI,MAAM,SAAS,aAAa,KAAK,MAAM,SAAS,YAAY,GAAG;AACjE,UAAM,IAAI,MAAM,GAAG,IAAI,kDAAe,OAAO,KAAK,EAAE,MAAM,GAAG,GAAG,CAAC,EAAE;AAAA,EACrE;AACF;AAGA,SAAS,eAAe,MAAsB;AAC5C,QAAM,QAAQ,KAAK,MAAM,+CAA+C;AACxE,SAAO,QAAQ,MAAM,CAAC,EAAE,KAAK,IAAI;AACnC;","names":[]}

package/dist/{chunk-YC2MEB7R.js → chunk-34WIGIQC.js}

@@ -17,7 +17,7 @@ var OCR_PROMPT = `\uC774 PDF \uD398\uC774\uC9C0 \uC774\uBBF8\uC9C0\uC5D0\uC11C \
 var _tempDir = null;
 function getTempDir() {
   if (!_tempDir) {
-    _tempDir = join(process.cwd(), "_kordoc_ocr_tmp");
+    _tempDir = join(process.cwd(), ".kordoc_ocr_tmp");
     mkdirSync(_tempDir, { recursive: true });
   }
   return _tempDir;
@@ -42,6 +42,12 @@ function createCliOcrProvider(mode) {
     }
   };
 }
+function checkForLimitError(output, mode) {
+  const lower = output.toLowerCase();
+  if (lower.includes("usage limit") || lower.includes("rate limit")) {
+    throw new Error(`${mode} \uC0AC\uC6A9\uB7C9/\uC18D\uB3C4 \uC81C\uD55C: ${output.trim().slice(0, 200)}`);
+  }
+}
 function callCli(mode, imagePath) {
   if (mode === "codex") {
     return callCodexCli(imagePath);
@@ -62,7 +68,9 @@ function callCli(mode, imagePath) {
     const errMsg = result.stderr?.trim() || `exit code ${result.status}`;
     throw new Error(`${mode} OCR \uC2E4\uD328: ${errMsg}`);
   }
-  return result.stdout || "";
+  const output = result.stdout || "";
+  checkForLimitError(output, mode);
+  return output;
 }
 function callCodexCli(imagePath) {
   const outPath = join(tmpdir(), `kordoc-codex-out-${Date.now()}.txt`);
@@ -85,11 +93,14 @@ function callCodexCli(imagePath) {
       const errMsg = result.stderr?.trim() || `exit code ${result.status}`;
       throw new Error(`codex OCR \uC2E4\uD328: ${errMsg}`);
     }
+    let text;
     try {
-      return readFileSync(outPath, "utf-8");
+      text = readFileSync(outPath, "utf-8");
     } catch {
-      return result.stdout || "";
+      text = result.stdout || "";
     }
+    checkForLimitError(text, "codex");
+    return text;
   } finally {
     try {
       unlinkSync(outPath);
@@ -153,4 +164,4 @@ function stripCodeFence(text) {
 export {
   createCliOcrProvider
 };
-//# sourceMappingURL=chunk-YC2MEB7R.js.map
+//# sourceMappingURL=chunk-34WIGIQC.js.map

package/dist/chunk-34WIGIQC.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/ocr/cli-provider.ts"],"sourcesContent":["/**\n * CLI 기반 OCR 프로바이더\n *\n * gemini / claude / codex / ollama CLI를 subprocess로 호출하여\n * PDF 페이지 이미지를 Markdown으로 변환.\n *\n * 이미지 전달 방식:\n * - gemini:  -p \"프롬프트 @이미지경로\"  (@ 파일 참조)\n * - claude:  -p \"프롬프트 @이미지경로\"  (@ 파일 참조, --print 모드)\n * - codex:   exec -i 이미지경로 \"프롬프트\"  (-i/--image 플래그)\n * - ollama:  REST API (localhost:11434) — CLI는 이미지 입력 미지원\n */\n\nimport { spawnSync } from \"child_process\"\nimport { writeFileSync, readFileSync, unlinkSync, mkdirSync } from \"fs\"\nimport { join } from \"path\"\nimport { tmpdir } from \"os\"\nimport type { OcrMode, StructuredOcrResult } from \"../types.js\"\n\n/** OCR 프롬프트 — 모든 CLI 공통 */\nconst OCR_PROMPT = `이 PDF 페이지 이미지에서 텍스트와 테이블을 추출하여 순수 Markdown으로 변환하세요.\n규칙:\n- 테이블은 Markdown 테이블 문법 사용 (| 구분, |---|---| 헤더 구분선 포함)\n- 병합된 셀은 해당 위치에 내용 기재\n- 헤딩은 글자 크기에 따라 ## ~ ###### 사용\n- 리스트는 - 또는 1. 사용\n- 이미지, 도형 등 비텍스트 요소는 무시\n- 원문의 읽기 순서와 구조를 유지\n- \\`\\`\\`로 감싸지 말고 순수 Markdown만 출력`\n\n/** 임시 디렉토리 (프로세스당 1회 생성)\n *\n * gemini CLI는 /tmp/ 등 시스템 임시 디렉토리를 워크스페이스 외부로 간주하여\n * @파일참조 시 접근을 거부할 수 있음. cwd 하위 폴더를 사용하면 모든 CLI에서 접근 가능.\n *\n * ⚠️ .gitignore에 포함된 경로(예: .kordoc-tmp/)는 gemini CLI가 무시하므로\n * 반드시 gitignore되지 않는 이름 사용. 파일은 try/finally로 즉시 정리.\n */\nlet _tempDir: string | null = null\nfunction getTempDir(): string {\n  if (!_tempDir) {\n    _tempDir = join(process.cwd(), \".kordoc_ocr_tmp\")\n    mkdirSync(_tempDir, { recursive: true })\n  }\n  return _tempDir\n}\n\n/**\n * CLI OcrProvider 생성.\n *\n * @param mode - 사용할 CLI (gemini, claude, codex, ollama)\n * @returns OcrProvider 함수 (StructuredOcrResult 반환)\n */\nexport function createCliOcrProvider(\n  mode: Exclude<OcrMode, \"auto\" | \"off\" | \"tesseract\">\n): (pageImage: Uint8Array, pageNumber: number, mimeType: \"image/png\") => Promise<StructuredOcrResult> {\n  return async (pageImage: Uint8Array, pageNumber: number): Promise<StructuredOcrResult> => {\n    const tempPath = join(getTempDir(), `page-${pageNumber}.png`)\n\n    try {\n      writeFileSync(tempPath, pageImage)\n\n      let output: string\n      if (mode === \"ollama\") {\n        output = await callOllamaApi(tempPath)\n      } else {\n        output = callCli(mode, tempPath)\n      }\n\n      return { markdown: stripCodeFence(output.trim()) }\n    } finally {\n      try { unlinkSync(tempPath) } catch { /* 임시 파일 정리 실패 무시 */ }\n    }\n  }\n}\n\n/**\n * 출력 텍스트에서 사용량·속도 제한 에러 감지.\n * 해당 메시지가 포함된 경우 throw하여 다음 엔진으로 fallback 트리거.\n */\nfunction checkForLimitError(output: string, mode: string): void {\n  const lower = output.toLowerCase()\n  if (lower.includes(\"usage limit\") || lower.includes(\"rate limit\")) {\n    throw new Error(`${mode} 사용량/속도 제한: ${output.trim().slice(0, 200)}`)\n  }\n}\n\n/**\n * CLI 실행 — gemini / claude / codex\n *\n * @throws CLI 실행 실패 또는 타임아웃(180초) 시 Error\n */\nfunction callCli(mode: string, imagePath: string): string {\n  // codex는 --output-last-message로 대화 헤더 없는 깔끔한 출력 사용\n  if (mode === \"codex\") {\n    return callCodexCli(imagePath)\n  }\n\n  const args = buildCliArgs(mode, imagePath)\n\n  const result = spawnSync(mode, args, {\n    encoding: \"utf-8\",\n    timeout: 600_000,\n    maxBuffer: 10 * 1024 * 1024,\n    shell: process.platform === \"win32\",\n    // claude: /tmp에서 실행하여 프로젝트 CLAUDE.md의 규칙 간섭 방지\n    ...(mode === \"claude\" ? { cwd: tmpdir() } : {}),\n  })\n\n  if (result.error) {\n    throw new Error(`${mode} CLI 실행 실패: ${result.error.message}`)\n  }\n  if (result.status !== 0) {\n    const errMsg = result.stderr?.trim() || `exit code ${result.status}`\n    throw new Error(`${mode} OCR 실패: ${errMsg}`)\n  }\n\n  const output = result.stdout || \"\"\n  checkForLimitError(output, mode)\n  return output\n}\n\n/**\n * codex exec 실행 — --output-last-message로 대화 헤더 없는 깔끔한 출력.\n * 인자 순서: `codex exec <prompt> --image <file> --output-last-message <outfile>`\n */\nfunction callCodexCli(imagePath: string): string {\n  // 출력 파일은 /tmp/ 사용 — codex sandbox는 cwd 내 쓰기를 막을 수 있음\n  const outPath = join(tmpdir(), `kordoc-codex-out-${Date.now()}.txt`)\n  try {\n    const args = [\"exec\", OCR_PROMPT, \"--image\", imagePath, \"--output-last-message\", outPath]\n    const model = process.env.KORDOC_CODEX_MODEL\n    if (model) args.push(\"--model\", model)\n\n    const result = spawnSync(\"codex\", args, {\n      encoding: \"utf-8\",\n      timeout: 180_000,\n      maxBuffer: 10 * 1024 * 1024,\n      input: \"\",  // stdin EOF 즉시 전달 (대화형 입력 차단)\n      shell: process.platform === \"win32\",\n    })\n\n    if (result.error) {\n      throw new Error(`codex CLI 실행 실패: ${result.error.message}`)\n    }\n    if (result.status !== 0) {\n      const errMsg = result.stderr?.trim() || `exit code ${result.status}`\n      throw new Error(`codex OCR 실패: ${errMsg}`)\n    }\n\n    // --output-last-message 파일에서 읽기 (없으면 stdout 폴백)\n    let text: string\n    try {\n      text = readFileSync(outPath, \"utf-8\")\n    } catch {\n      text = result.stdout || \"\"\n    }\n    checkForLimitError(text, \"codex\")\n    return text\n  } finally {\n    try { unlinkSync(outPath) } catch { /* 무시 */ }\n  }\n}\n\n/**\n * CLI별 인자 배열 생성.\n *\n * gemini:  [\"--prompt\", \"프롬프트 @이미지경로\", \"--yolo\"]\n *   - -y/--yolo: 자동 승인 (OCR은 도구 사용 없으므로 실질적 영향 없음)\n *   - @ 파일 참조로 이미지를 컨텍스트에 포함\n *\n * claude:  [\"--print\", \"프롬프트 @이미지경로\"]\n *   - --print(-p): 비대화형 출력 모드\n *   - @ 파일 참조로 이미지를 컨텍스트에 포함\n *\n * codex: callCodexCli()에서 별도 처리\n *   - `codex exec <prompt> --image <file> --output-last-message <outfile>`\n *   - 프롬프트가 --image보다 앞에 위치해야 함 (인자 순서 중요)\n *\n * ⚠️ CLI 버전에 따라 문법이 다를 수 있음. 업데이트 시 --help 재확인 필요.\n */\nfunction buildCliArgs(mode: string, imagePath: string): string[] {\n  const normalizedPath = imagePath.replace(/\\\\/g, \"/\")\n  const promptWithImage = `${OCR_PROMPT}\n\n이미지: @${normalizedPath}`\n\n  switch (mode) {\n    case \"gemini\": {\n      const args = [\"--prompt\", promptWithImage, \"--yolo\"]\n      const model = process.env.KORDOC_GEMINI_MODEL\n      if (model) args.push(\"--model\", model)\n      return args\n    }\n\n    case \"claude\": {\n      const args = [\"--print\", promptWithImage]\n      const model = process.env.KORDOC_CLAUDE_MODEL\n      if (model) args.push(\"--model\", model)\n      return args\n    }\n\n    default:\n      throw new Error(`지원하지 않는 CLI: ${mode}`)\n  }\n}\n\n/**\n * Ollama REST API 호출 — CLI는 이미지 입력을 지원하지 않으므로 API 직접 사용.\n *\n * 기본 모델: KORDOC_OLLAMA_MODEL 환경변수 또는 \"gemma4:27b\"\n * 기본 호스트: KORDOC_OLLAMA_HOST 환경변수 또는 \"http://localhost:11434\"\n *\n * @throws Ollama 서버 미실행 또는 응답 오류 시 Error\n */\nasync function callOllamaApi(imagePath: string): Promise<string> {\n  const { readFileSync } = await import(\"fs\")\n  const imageBase64 = readFileSync(imagePath).toString(\"base64\")\n\n  const model = process.env.KORDOC_OLLAMA_MODEL || \"qwen3-vl:8b\"\n  const host = process.env.KORDOC_OLLAMA_HOST || \"http://localhost:11434\"\n  const timeoutMs = Number(process.env.KORDOC_OLLAMA_TIMEOUT) || 120_000\n\n  const response = await fetch(`${host}/api/chat`, {\n    method: \"POST\",\n    headers: { \"Content-Type\": \"application/json\" },\n    body: JSON.stringify({\n      model,\n      messages: [{\n        role: \"user\",\n        content: OCR_PROMPT,\n        images: [imageBase64],\n      }],\n      stream: false,\n    }),\n    signal: AbortSignal.timeout(timeoutMs),\n  })\n\n  if (!response.ok) {\n    throw new Error(`Ollama API 오류: ${response.status} ${response.statusText}`)\n  }\n\n  const data = await response.json() as { message?: { content?: string } }\n  return data.message?.content || \"\"\n}\n\n/**\n * LLM 출력에서 코드 펜스 제거.\n * LLM이 가끔 결과를 ```markdown ... ``` 으로 감싸는 경우 처리.\n */\nfunction stripCodeFence(text: string): string {\n  const match = text.match(/^```(?:markdown|md)?\\s*([\\s\\S]*?)```\\s*$/m)\n  return match ? match[1].trim() : text\n}\n"],"mappings":";;;AAaA,SAAS,iBAAiB;AAC1B,SAAS,eAAe,cAAc,YAAY,iBAAiB;AACnE,SAAS,YAAY;AACrB,SAAS,cAAc;AAIvB,IAAM,aAAa;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAkBnB,IAAI,WAA0B;AAC9B,SAAS,aAAqB;AAC5B,MAAI,CAAC,UAAU;AACb,eAAW,KAAK,QAAQ,IAAI,GAAG,iBAAiB;AAChD,cAAU,UAAU,EAAE,WAAW,KAAK,CAAC;AAAA,EACzC;AACA,SAAO;AACT;AAQO,SAAS,qBACd,MACoG;AACpG,SAAO,OAAO,WAAuB,eAAqD;AACxF,UAAM,WAAW,KAAK,WAAW,GAAG,QAAQ,UAAU,MAAM;AAE5D,QAAI;AACF,oBAAc,UAAU,SAAS;AAEjC,UAAI;AACJ,UAAI,SAAS,UAAU;AACrB,iBAAS,MAAM,cAAc,QAAQ;AAAA,MACvC,OAAO;AACL,iBAAS,QAAQ,MAAM,QAAQ;AAAA,MACjC;AAEA,aAAO,EAAE,UAAU,eAAe,OAAO,KAAK,CAAC,EAAE;AAAA,IACnD,UAAE;AACA,UAAI;AAAE,mBAAW,QAAQ;AAAA,MAAE,QAAQ;AAAA,MAAuB;AAAA,IAC5D;AAAA,EACF;AACF;AAMA,SAAS,mBAAmB,QAAgB,MAAoB;AAC9D,QAAM,QAAQ,OAAO,YAAY;AACjC,MAAI,MAAM,SAAS,aAAa,KAAK,MAAM,SAAS,YAAY,GAAG;AACjE,UAAM,IAAI,MAAM,GAAG,IAAI,kDAAe,OAAO,KAAK,EAAE,MAAM,GAAG,GAAG,CAAC,EAAE;AAAA,EACrE;AACF;AAOA,SAAS,QAAQ,MAAc,WAA2B;AAExD,MAAI,SAAS,SAAS;AACpB,WAAO,aAAa,SAAS;AAAA,EAC/B;AAEA,QAAM,OAAO,aAAa,MAAM,SAAS;AAEzC,QAAM,SAAS,UAAU,MAAM,MAAM;AAAA,IACnC,UAAU;AAAA,IACV,SAAS;AAAA,IACT,WAAW,KAAK,OAAO;AAAA,IACvB,OAAO,QAAQ,aAAa;AAAA;AAAA,IAE5B,GAAI,SAAS,WAAW,EAAE,KAAK,OAAO,EAAE,IAAI,CAAC;AAAA,EAC/C,CAAC;AAED,MAAI,OAAO,OAAO;AAChB,UAAM,IAAI,MAAM,GAAG,IAAI,mCAAe,OAAO,MAAM,OAAO,EAAE;AAAA,EAC9D;AACA,MAAI,OAAO,WAAW,GAAG;AACvB,UAAM,SAAS,OAAO,QAAQ,KAAK,KAAK,aAAa,OAAO,MAAM;AAClE,UAAM,IAAI,MAAM,GAAG,IAAI,sBAAY,MAAM,EAAE;AAAA,EAC7C;AAEA,QAAM,SAAS,OAAO,UAAU;AAChC,qBAAmB,QAAQ,IAAI;AAC/B,SAAO;AACT;AAMA,SAAS,aAAa,WAA2B;AAE/C,QAAM,UAAU,KAAK,OAAO,GAAG,oBAAoB,KAAK,IAAI,CAAC,MAAM;AACnE,MAAI;AACF,UAAM,OAAO,CAAC,QAAQ,YAAY,WAAW,WAAW,yBAAyB,OAAO;AACxF,UAAM,QAAQ,QAAQ,IAAI;AAC1B,QAAI,MAAO,MAAK,KAAK,WAAW,KAAK;AAErC,UAAM,SAAS,UAAU,SAAS,MAAM;AAAA,MACtC,UAAU;AAAA,MACV,SAAS;AAAA,MACT,WAAW,KAAK,OAAO;AAAA,MACvB,OAAO;AAAA;AAAA,MACP,OAAO,QAAQ,aAAa;AAAA,IAC9B,CAAC;AAED,QAAI,OAAO,OAAO;AAChB,YAAM,IAAI,MAAM,wCAAoB,OAAO,MAAM,OAAO,EAAE;AAAA,IAC5D;AACA,QAAI,OAAO,WAAW,GAAG;AACvB,YAAM,SAAS,OAAO,QAAQ,KAAK,KAAK,aAAa,OAAO,MAAM;AAClE,YAAM,IAAI,MAAM,2BAAiB,MAAM,EAAE;AAAA,IAC3C;AAGA,QAAI;AACJ,QAAI;AACF,aAAO,aAAa,SAAS,OAAO;AAAA,IACtC,QAAQ;AACN,aAAO,OAAO,UAAU;AAAA,IAC1B;AACA,uBAAmB,MAAM,OAAO;AAChC,WAAO;AAAA,EACT,UAAE;AACA,QAAI;AAAE,iBAAW,OAAO;AAAA,IAAE,QAAQ;AAAA,IAAW;AAAA,EAC/C;AACF;AAmBA,SAAS,aAAa,MAAc,WAA6B;AAC/D,QAAM,iBAAiB,UAAU,QAAQ,OAAO,GAAG;AACnD,QAAM,kBAAkB,GAAG,UAAU;AAAA;AAAA,uBAE/B,cAAc;AAEpB,UAAQ,MAAM;AAAA,IACZ,KAAK,UAAU;AACb,YAAM,OAAO,CAAC,YAAY,iBAAiB,QAAQ;AACnD,YAAM,QAAQ,QAAQ,IAAI;AAC1B,UAAI,MAAO,MAAK,KAAK,WAAW,KAAK;AACrC,aAAO;AAAA,IACT;AAAA,IAEA,KAAK,UAAU;AACb,YAAM,OAAO,CAAC,WAAW,eAAe;AACxC,YAAM,QAAQ,QAAQ,IAAI;AAC1B,UAAI,MAAO,MAAK,KAAK,WAAW,KAAK;AACrC,aAAO;AAAA,IACT;AAAA,IAEA;AACE,YAAM,IAAI,MAAM,8CAAgB,IAAI,EAAE;AAAA,EAC1C;AACF;AAUA,eAAe,cAAc,WAAoC;AAC/D,QAAM,EAAE,cAAAA,cAAa,IAAI,MAAM,OAAO,IAAI;AAC1C,QAAM,cAAcA,cAAa,SAAS,EAAE,SAAS,QAAQ;AAE7D,QAAM,QAAQ,QAAQ,IAAI,uBAAuB;AACjD,QAAM,OAAO,QAAQ,IAAI,sBAAsB;AAC/C,QAAM,YAAY,OAAO,QAAQ,IAAI,qBAAqB,KAAK;AAE/D,QAAM,WAAW,MAAM,MAAM,GAAG,IAAI,aAAa;AAAA,IAC/C,QAAQ;AAAA,IACR,SAAS,EAAE,gBAAgB,mBAAmB;AAAA,IAC9C,MAAM,KAAK,UAAU;AAAA,MACnB;AAAA,MACA,UAAU,CAAC;AAAA,QACT,MAAM;AAAA,QACN,SAAS;AAAA,QACT,QAAQ,CAAC,WAAW;AAAA,MACtB,CAAC;AAAA,MACD,QAAQ;AAAA,IACV,CAAC;AAAA,IACD,QAAQ,YAAY,QAAQ,SAAS;AAAA,EACvC,CAAC;AAED,MAAI,CAAC,SAAS,IAAI;AAChB,UAAM,IAAI,MAAM,4BAAkB,SAAS,MAAM,IAAI,SAAS,UAAU,EAAE;AAAA,EAC5E;AAEA,QAAM,OAAO,MAAM,SAAS,KAAK;AACjC,SAAO,KAAK,SAAS,WAAW;AAClC;AAMA,SAAS,eAAe,MAAsB;AAC5C,QAAM,QAAQ,KAAK,MAAM,2CAA2C;AACpE,SAAO,QAAQ,MAAM,CAAC,EAAE,KAAK,IAAI;AACnC;","names":["readFileSync"]}

package/dist/{chunk-MPMKWVV2.js → chunk-JGMLDBW5.js}

@@ -9,7 +9,7 @@ import {
   precheckZipSize,
   sanitizeHref,
   toArrayBuffer
-} from "./chunk-SHM3PYVA.js";
+} from "./chunk-PJSXZBZB.js";
 import {
   parsePageRange
 } from "./chunk-MOL7MDBG.js";
@@ -18,7 +18,7 @@ import {
 } from "./chunk-7FMKAV4P.js";
 import {
   createCliOcrProvider
-} from "./chunk-YC2MEB7R.js";
+} from "./chunk-34WIGIQC.js";
 import {
   __commonJS,
   __require,
@@ -5462,7 +5462,7 @@ async function parsePdfDocument(buffer, options) {
       if (ocrMode === "off") {
         throw Object.assign(new KordocError(`\uC774\uBBF8\uC9C0 \uAE30\uBC18 PDF (${pageCount}\uD398\uC774\uC9C0, ${totalChars}\uC790)`), { isImageBased: true });
       }
-      const { resolveOcrProvider } = await import("./resolve-BGOGWG6E.js");
+      const { resolveOcrProvider } = await import("./resolve-4I65IGMM.js");
       const { ocrPages } = await import("./provider-PYZL2VNN.js");
       const tryProvider = async (provider, filter) => {
         try {
@@ -9813,4 +9813,4 @@ export {
 cfb/cfb.js:
   (*! crc32.js (C) 2014-present SheetJS -- http://sheetjs.com *)
 */
-//# sourceMappingURL=chunk-MPMKWVV2.js.map
+//# sourceMappingURL=chunk-JGMLDBW5.js.map

package/dist/{chunk-SHM3PYVA.js → chunk-PJSXZBZB.js}

@@ -1,7 +1,7 @@
 #!/usr/bin/env node
 
 // src/utils.ts
-var VERSION = true ? "2.4.9" : "0.0.0-dev";
+var VERSION = true ? "2.4.11" : "0.0.0-dev";
 function toArrayBuffer(buf) {
   if (buf.byteOffset === 0 && buf.byteLength === buf.buffer.byteLength) {
     return buf.buffer;
@@ -90,4 +90,4 @@ export {
   sanitizeHref,
   classifyError
 };
-//# sourceMappingURL=chunk-SHM3PYVA.js.map
+//# sourceMappingURL=chunk-PJSXZBZB.js.map

package/dist/{chunk-SHM3PYVA.js.map → chunk-PJSXZBZB.js.map}

@@ -1 +1 @@
-{"version":3,"sources":["../src/utils.ts"],"sourcesContent":["/** kordoc 공용 유틸리티 */\n\n/** 빌드 타임에 tsup define으로 주입되는 버전 */\ndeclare const __KORDOC_VERSION__: string\nexport const VERSION: string = typeof __KORDOC_VERSION__ !== \"undefined\" ? __KORDOC_VERSION__ : \"0.0.0-dev\"\n\n/**\n * Node.js Buffer → ArrayBuffer 변환\n * pool Buffer의 공유 ArrayBuffer 문제를 안전하게 처리.\n * offset=0이고 전체 ArrayBuffer를 차지하면 복사 없이 직접 반환.\n */\nexport function toArrayBuffer(buf: Buffer): ArrayBuffer {\n  if (buf.byteOffset === 0 && buf.byteLength === buf.buffer.byteLength) {\n    return buf.buffer as ArrayBuffer\n  }\n  return buf.buffer.slice(buf.byteOffset, buf.byteOffset + buf.byteLength) as ArrayBuffer\n}\n\n/**\n * kordoc 내부 에러 클래스 — 사용자에게 노출해도 안전한 메시지만 포함.\n * MCP 에러 정제에서 instanceof로 판별하여 allowlist 패턴 매칭 없이 안전하게 통과.\n */\nexport class KordocError extends Error {\n  constructor(message: string) {\n    super(message)\n    this.name = \"KordocError\"\n  }\n}\n\n/**\n * 에러 메시지 정제 — KordocError는 그대로, 나머지는 일반 메시지로 대체.\n * 파일시스템 경로, 스택 트레이스 등 내부 정보 노출 방지.\n */\nexport function sanitizeError(err: unknown): string {\n  if (err instanceof KordocError) return err.message\n  return \"문서 처리 중 오류가 발생했습니다\"\n}\n\n/**\n * ZIP 엔트리 경로의 경로 순회 여부 판별.\n * 백슬래시 정규화, .., 절대경로, Windows 드라이브 문자 모두 차단.\n */\nexport function isPathTraversal(name: string): boolean {\n  if (name.includes(\"\\x00\")) return true\n  const normalized = name.replace(/\\\\/g, \"/\")\n  return normalized.includes(\"..\") || normalized.startsWith(\"/\") || /^[A-Za-z]:/.test(normalized)\n}\n\n// ─── ZIP 안전 로딩 (ZIP bomb 방지) ────────────────────\n\n/**\n * ZIP bomb 사전 검사 — Central Directory에서 비압축 합계와 엔트리 수 확인.\n * HWPX/XLSX/DOCX 등 모든 ZIP 기반 포맷에서 공통 사용.\n */\nexport function precheckZipSize(\n  buffer: ArrayBuffer,\n  maxUncompressedSize = 100 * 1024 * 1024,\n  maxEntries = 500,\n): { totalUncompressed: number; entryCount: number } {\n  try {\n    const data = new DataView(buffer)\n    const len = buffer.byteLength\n    // EOCD 시그니처 역방향 스캔\n    let eocdOffset = -1\n    for (let i = len - 22; i >= Math.max(0, len - 65557); i--) {\n      if (data.getUint32(i, true) === 0x06054b50) { eocdOffset = i; break }\n    }\n    if (eocdOffset < 0) return { totalUncompressed: 0, entryCount: 0 }\n\n    const entryCount = data.getUint16(eocdOffset + 10, true)\n    if (entryCount > maxEntries) {\n      throw new KordocError(`ZIP 엔트리 수 초과: ${entryCount} (최대 ${maxEntries})`)\n    }\n\n    const cdSize = data.getUint32(eocdOffset + 12, true)\n    const cdOffset = data.getUint32(eocdOffset + 16, true)\n    if (cdOffset + cdSize > len) return { totalUncompressed: 0, entryCount }\n\n    let totalUncompressed = 0\n    let pos = cdOffset\n    for (let i = 0; i < entryCount && pos + 46 <= cdOffset + cdSize; i++) {\n      if (data.getUint32(pos, true) !== 0x02014b50) break\n      totalUncompressed += data.getUint32(pos + 24, true)\n      const nameLen = data.getUint16(pos + 28, true)\n      const extraLen = data.getUint16(pos + 30, true)\n      const commentLen = data.getUint16(pos + 32, true)\n      pos += 46 + nameLen + extraLen + commentLen\n    }\n\n    if (totalUncompressed > maxUncompressedSize) {\n      throw new KordocError(`ZIP 비압축 크기 초과: ${(totalUncompressed / 1024 / 1024).toFixed(1)}MB (최대 ${maxUncompressedSize / 1024 / 1024}MB)`)\n    }\n\n    return { totalUncompressed, entryCount }\n  } catch (err) {\n    if (err instanceof KordocError) throw err\n    return { totalUncompressed: 0, entryCount: 0 }\n  }\n}\n\n/** 하이퍼링크 URL 살균 — javascript: 등 XSS 위험 스킴 차단 */\nconst SAFE_HREF_RE = /^(?:https?:|mailto:|tel:|#)/i\nexport function sanitizeHref(href: string): string | null {\n  const trimmed = href.trim()\n  if (!trimmed || !SAFE_HREF_RE.test(trimmed)) return null\n  return trimmed\n}\n\n// ─── 에러 분류 ──────────────────────────────────────\n\nimport type { ErrorCode } from \"./types.js\"\n\n/** 에러를 구조화된 ErrorCode로 분류 — KordocError 메시지 패턴 매칭 */\nexport function classifyError(err: unknown): ErrorCode {\n  if (!(err instanceof Error)) return \"PARSE_ERROR\"\n  const msg = err.message\n  if (msg.includes(\"암호화\")) return \"ENCRYPTED\"\n  if (msg.includes(\"DRM\")) return \"DRM_PROTECTED\"\n  if (msg.includes(\"ZIP bomb\") || msg.includes(\"ZIP 비압축 크기 초과\") || msg.includes(\"ZIP 엔트리 수 초과\")) return \"ZIP_BOMB\"\n  if (msg.includes(\"bomb\") || msg.includes(\"크기 초과\") || msg.includes(\"압축 해제\")) return \"DECOMPRESSION_BOMB\"\n  if (msg.includes(\"이미지 기반\")) return \"IMAGE_BASED_PDF\"\n  if (msg.includes(\"섹션\") && (msg.includes(\"찾을 수 없\") || msg.includes(\"없음\"))) return \"NO_SECTIONS\"\n  if (msg.includes(\"시그니처\") || msg.includes(\"복구할 수 없\")) return \"CORRUPTED\"\n  return \"PARSE_ERROR\"\n}\n"],"mappings":";;;AAIO,IAAM,UAAkB,OAA4C,UAAqB;AAOzF,SAAS,cAAc,KAA0B;AACtD,MAAI,IAAI,eAAe,KAAK,IAAI,eAAe,IAAI,OAAO,YAAY;AACpE,WAAO,IAAI;AAAA,EACb;AACA,SAAO,IAAI,OAAO,MAAM,IAAI,YAAY,IAAI,aAAa,IAAI,UAAU;AACzE;AAMO,IAAM,cAAN,cAA0B,MAAM;AAAA,EACrC,YAAY,SAAiB;AAC3B,UAAM,OAAO;AACb,SAAK,OAAO;AAAA,EACd;AACF;AAMO,SAAS,cAAc,KAAsB;AAClD,MAAI,eAAe,YAAa,QAAO,IAAI;AAC3C,SAAO;AACT;AAMO,SAAS,gBAAgB,MAAuB;AACrD,MAAI,KAAK,SAAS,IAAM,EAAG,QAAO;AAClC,QAAM,aAAa,KAAK,QAAQ,OAAO,GAAG;AAC1C,SAAO,WAAW,SAAS,IAAI,KAAK,WAAW,WAAW,GAAG,KAAK,aAAa,KAAK,UAAU;AAChG;AAQO,SAAS,gBACd,QACA,sBAAsB,MAAM,OAAO,MACnC,aAAa,KACsC;AACnD,MAAI;AACF,UAAM,OAAO,IAAI,SAAS,MAAM;AAChC,UAAM,MAAM,OAAO;AAEnB,QAAI,aAAa;AACjB,aAAS,IAAI,MAAM,IAAI,KAAK,KAAK,IAAI,GAAG,MAAM,KAAK,GAAG,KAAK;AACzD,UAAI,KAAK,UAAU,GAAG,IAAI,MAAM,WAAY;AAAE,qBAAa;AAAG;AAAA,MAAM;AAAA,IACtE;AACA,QAAI,aAAa,EAAG,QAAO,EAAE,mBAAmB,GAAG,YAAY,EAAE;AAEjE,UAAM,aAAa,KAAK,UAAU,aAAa,IAAI,IAAI;AACvD,QAAI,aAAa,YAAY;AAC3B,YAAM,IAAI,YAAY,+CAAiB,UAAU,kBAAQ,UAAU,GAAG;AAAA,IACxE;AAEA,UAAM,SAAS,KAAK,UAAU,aAAa,IAAI,IAAI;AACnD,UAAM,WAAW,KAAK,UAAU,aAAa,IAAI,IAAI;AACrD,QAAI,WAAW,SAAS,IAAK,QAAO,EAAE,mBAAmB,GAAG,WAAW;AAEvE,QAAI,oBAAoB;AACxB,QAAI,MAAM;AACV,aAAS,IAAI,GAAG,IAAI,cAAc,MAAM,MAAM,WAAW,QAAQ,KAAK;AACpE,UAAI,KAAK,UAAU,KAAK,IAAI,MAAM,SAAY;AAC9C,2BAAqB,KAAK,UAAU,MAAM,IAAI,IAAI;AAClD,YAAM,UAAU,KAAK,UAAU,MAAM,IAAI,IAAI;AAC7C,YAAM,WAAW,KAAK,UAAU,MAAM,IAAI,IAAI;AAC9C,YAAM,aAAa,KAAK,UAAU,MAAM,IAAI,IAAI;AAChD,aAAO,KAAK,UAAU,WAAW;AAAA,IACnC;AAEA,QAAI,oBAAoB,qBAAqB;AAC3C,YAAM,IAAI,YAAY,sDAAmB,oBAAoB,OAAO,MAAM,QAAQ,CAAC,CAAC,oBAAU,sBAAsB,OAAO,IAAI,KAAK;AAAA,IACtI;AAEA,WAAO,EAAE,mBAAmB,WAAW;AAAA,EACzC,SAAS,KAAK;AACZ,QAAI,eAAe,YAAa,OAAM;AACtC,WAAO,EAAE,mBAAmB,GAAG,YAAY,EAAE;AAAA,EAC/C;AACF;AAGA,IAAM,eAAe;AACd,SAAS,aAAa,MAA6B;AACxD,QAAM,UAAU,KAAK,KAAK;AAC1B,MAAI,CAAC,WAAW,CAAC,aAAa,KAAK,OAAO,EAAG,QAAO;AACpD,SAAO;AACT;AAOO,SAAS,cAAc,KAAyB;AACrD,MAAI,EAAE,eAAe,OAAQ,QAAO;AACpC,QAAM,MAAM,IAAI;AAChB,MAAI,IAAI,SAAS,oBAAK,EAAG,QAAO;AAChC,MAAI,IAAI,SAAS,KAAK,EAAG,QAAO;AAChC,MAAI,IAAI,SAAS,UAAU,KAAK,IAAI,SAAS,kDAAe,KAAK,IAAI,SAAS,4CAAc,EAAG,QAAO;AACtG,MAAI,IAAI,SAAS,MAAM,KAAK,IAAI,SAAS,2BAAO,KAAK,IAAI,SAAS,2BAAO,EAAG,QAAO;AACnF,MAAI,IAAI,SAAS,iCAAQ,EAAG,QAAO;AACnC,MAAI,IAAI,SAAS,cAAI,MAAM,IAAI,SAAS,4BAAQ,KAAK,IAAI,SAAS,cAAI,GAAI,QAAO;AACjF,MAAI,IAAI,SAAS,0BAAM,KAAK,IAAI,SAAS,kCAAS,EAAG,QAAO;AAC5D,SAAO;AACT;","names":[]}
+{"version":3,"sources":["../src/utils.ts"],"sourcesContent":["/** kordoc 공용 유틸리티 */\n\n/** 빌드 타임에 tsup define으로 주입되는 버전 */\ndeclare const __KORDOC_VERSION__: string\nexport const VERSION: string = typeof __KORDOC_VERSION__ !== \"undefined\" ? __KORDOC_VERSION__ : \"0.0.0-dev\"\n\n/**\n * Node.js Buffer → ArrayBuffer 변환\n * pool Buffer의 공유 ArrayBuffer 문제를 안전하게 처리.\n * offset=0이고 전체 ArrayBuffer를 차지하면 복사 없이 직접 반환.\n */\nexport function toArrayBuffer(buf: Buffer): ArrayBuffer {\n  if (buf.byteOffset === 0 && buf.byteLength === buf.buffer.byteLength) {\n    return buf.buffer as ArrayBuffer\n  }\n  return buf.buffer.slice(buf.byteOffset, buf.byteOffset + buf.byteLength) as ArrayBuffer\n}\n\n/**\n * kordoc 내부 에러 클래스 — 사용자에게 노출해도 안전한 메시지만 포함.\n * MCP 에러 정제에서 instanceof로 판별하여 allowlist 패턴 매칭 없이 안전하게 통과.\n */\nexport class KordocError extends Error {\n  constructor(message: string) {\n    super(message)\n    this.name = \"KordocError\"\n  }\n}\n\n/**\n * 에러 메시지 정제 — KordocError는 그대로, 나머지는 일반 메시지로 대체.\n * 파일시스템 경로, 스택 트레이스 등 내부 정보 노출 방지.\n */\nexport function sanitizeError(err: unknown): string {\n  if (err instanceof KordocError) return err.message\n  return \"문서 처리 중 오류가 발생했습니다\"\n}\n\n/**\n * ZIP 엔트리 경로의 경로 순회 여부 판별.\n * 백슬래시 정규화, .., 절대경로, Windows 드라이브 문자 모두 차단.\n */\nexport function isPathTraversal(name: string): boolean {\n  if (name.includes(\"\\x00\")) return true\n  const normalized = name.replace(/\\\\/g, \"/\")\n  return normalized.includes(\"..\") || normalized.startsWith(\"/\") || /^[A-Za-z]:/.test(normalized)\n}\n\n// ─── ZIP 안전 로딩 (ZIP bomb 방지) ────────────────────\n\n/**\n * ZIP bomb 사전 검사 — Central Directory에서 비압축 합계와 엔트리 수 확인.\n * HWPX/XLSX/DOCX 등 모든 ZIP 기반 포맷에서 공통 사용.\n */\nexport function precheckZipSize(\n  buffer: ArrayBuffer,\n  maxUncompressedSize = 100 * 1024 * 1024,\n  maxEntries = 500,\n): { totalUncompressed: number; entryCount: number } {\n  try {\n    const data = new DataView(buffer)\n    const len = buffer.byteLength\n    // EOCD 시그니처 역방향 스캔\n    let eocdOffset = -1\n    for (let i = len - 22; i >= Math.max(0, len - 65557); i--) {\n      if (data.getUint32(i, true) === 0x06054b50) { eocdOffset = i; break }\n    }\n    if (eocdOffset < 0) return { totalUncompressed: 0, entryCount: 0 }\n\n    const entryCount = data.getUint16(eocdOffset + 10, true)\n    if (entryCount > maxEntries) {\n      throw new KordocError(`ZIP 엔트리 수 초과: ${entryCount} (최대 ${maxEntries})`)\n    }\n\n    const cdSize = data.getUint32(eocdOffset + 12, true)\n    const cdOffset = data.getUint32(eocdOffset + 16, true)\n    if (cdOffset + cdSize > len) return { totalUncompressed: 0, entryCount }\n\n    let totalUncompressed = 0\n    let pos = cdOffset\n    for (let i = 0; i < entryCount && pos + 46 <= cdOffset + cdSize; i++) {\n      if (data.getUint32(pos, true) !== 0x02014b50) break\n      totalUncompressed += data.getUint32(pos + 24, true)\n      const nameLen = data.getUint16(pos + 28, true)\n      const extraLen = data.getUint16(pos + 30, true)\n      const commentLen = data.getUint16(pos + 32, true)\n      pos += 46 + nameLen + extraLen + commentLen\n    }\n\n    if (totalUncompressed > maxUncompressedSize) {\n      throw new KordocError(`ZIP 비압축 크기 초과: ${(totalUncompressed / 1024 / 1024).toFixed(1)}MB (최대 ${maxUncompressedSize / 1024 / 1024}MB)`)\n    }\n\n    return { totalUncompressed, entryCount }\n  } catch (err) {\n    if (err instanceof KordocError) throw err\n    return { totalUncompressed: 0, entryCount: 0 }\n  }\n}\n\n/** 하이퍼링크 URL 살균 — javascript: 등 XSS 위험 스킴 차단 */\nconst SAFE_HREF_RE = /^(?:https?:|mailto:|tel:|#)/i\nexport function sanitizeHref(href: string): string | null {\n  const trimmed = href.trim()\n  if (!trimmed || !SAFE_HREF_RE.test(trimmed)) return null\n  return trimmed\n}\n\n// ─── 에러 분류 ──────────────────────────────────────\n\nimport type { ErrorCode } from \"./types.js\"\n\n/** 에러를 구조화된 ErrorCode로 분류 — KordocError 메시지 패턴 매칭 */\nexport function classifyError(err: unknown): ErrorCode {\n  if (!(err instanceof Error)) return \"PARSE_ERROR\"\n  const msg = err.message\n  if (msg.includes(\"암호화\")) return \"ENCRYPTED\"\n  if (msg.includes(\"DRM\")) return \"DRM_PROTECTED\"\n  if (msg.includes(\"ZIP bomb\") || msg.includes(\"ZIP 비압축 크기 초과\") || msg.includes(\"ZIP 엔트리 수 초과\")) return \"ZIP_BOMB\"\n  if (msg.includes(\"bomb\") || msg.includes(\"크기 초과\") || msg.includes(\"압축 해제\")) return \"DECOMPRESSION_BOMB\"\n  if (msg.includes(\"이미지 기반\")) return \"IMAGE_BASED_PDF\"\n  if (msg.includes(\"섹션\") && (msg.includes(\"찾을 수 없\") || msg.includes(\"없음\"))) return \"NO_SECTIONS\"\n  if (msg.includes(\"시그니처\") || msg.includes(\"복구할 수 없\")) return \"CORRUPTED\"\n  return \"PARSE_ERROR\"\n}\n"],"mappings":";;;AAIO,IAAM,UAAkB,OAA4C,WAAqB;AAOzF,SAAS,cAAc,KAA0B;AACtD,MAAI,IAAI,eAAe,KAAK,IAAI,eAAe,IAAI,OAAO,YAAY;AACpE,WAAO,IAAI;AAAA,EACb;AACA,SAAO,IAAI,OAAO,MAAM,IAAI,YAAY,IAAI,aAAa,IAAI,UAAU;AACzE;AAMO,IAAM,cAAN,cAA0B,MAAM;AAAA,EACrC,YAAY,SAAiB;AAC3B,UAAM,OAAO;AACb,SAAK,OAAO;AAAA,EACd;AACF;AAMO,SAAS,cAAc,KAAsB;AAClD,MAAI,eAAe,YAAa,QAAO,IAAI;AAC3C,SAAO;AACT;AAMO,SAAS,gBAAgB,MAAuB;AACrD,MAAI,KAAK,SAAS,IAAM,EAAG,QAAO;AAClC,QAAM,aAAa,KAAK,QAAQ,OAAO,GAAG;AAC1C,SAAO,WAAW,SAAS,IAAI,KAAK,WAAW,WAAW,GAAG,KAAK,aAAa,KAAK,UAAU;AAChG;AAQO,SAAS,gBACd,QACA,sBAAsB,MAAM,OAAO,MACnC,aAAa,KACsC;AACnD,MAAI;AACF,UAAM,OAAO,IAAI,SAAS,MAAM;AAChC,UAAM,MAAM,OAAO;AAEnB,QAAI,aAAa;AACjB,aAAS,IAAI,MAAM,IAAI,KAAK,KAAK,IAAI,GAAG,MAAM,KAAK,GAAG,KAAK;AACzD,UAAI,KAAK,UAAU,GAAG,IAAI,MAAM,WAAY;AAAE,qBAAa;AAAG;AAAA,MAAM;AAAA,IACtE;AACA,QAAI,aAAa,EAAG,QAAO,EAAE,mBAAmB,GAAG,YAAY,EAAE;AAEjE,UAAM,aAAa,KAAK,UAAU,aAAa,IAAI,IAAI;AACvD,QAAI,aAAa,YAAY;AAC3B,YAAM,IAAI,YAAY,+CAAiB,UAAU,kBAAQ,UAAU,GAAG;AAAA,IACxE;AAEA,UAAM,SAAS,KAAK,UAAU,aAAa,IAAI,IAAI;AACnD,UAAM,WAAW,KAAK,UAAU,aAAa,IAAI,IAAI;AACrD,QAAI,WAAW,SAAS,IAAK,QAAO,EAAE,mBAAmB,GAAG,WAAW;AAEvE,QAAI,oBAAoB;AACxB,QAAI,MAAM;AACV,aAAS,IAAI,GAAG,IAAI,cAAc,MAAM,MAAM,WAAW,QAAQ,KAAK;AACpE,UAAI,KAAK,UAAU,KAAK,IAAI,MAAM,SAAY;AAC9C,2BAAqB,KAAK,UAAU,MAAM,IAAI,IAAI;AAClD,YAAM,UAAU,KAAK,UAAU,MAAM,IAAI,IAAI;AAC7C,YAAM,WAAW,KAAK,UAAU,MAAM,IAAI,IAAI;AAC9C,YAAM,aAAa,KAAK,UAAU,MAAM,IAAI,IAAI;AAChD,aAAO,KAAK,UAAU,WAAW;AAAA,IACnC;AAEA,QAAI,oBAAoB,qBAAqB;AAC3C,YAAM,IAAI,YAAY,sDAAmB,oBAAoB,OAAO,MAAM,QAAQ,CAAC,CAAC,oBAAU,sBAAsB,OAAO,IAAI,KAAK;AAAA,IACtI;AAEA,WAAO,EAAE,mBAAmB,WAAW;AAAA,EACzC,SAAS,KAAK;AACZ,QAAI,eAAe,YAAa,OAAM;AACtC,WAAO,EAAE,mBAAmB,GAAG,YAAY,EAAE;AAAA,EAC/C;AACF;AAGA,IAAM,eAAe;AACd,SAAS,aAAa,MAA6B;AACxD,QAAM,UAAU,KAAK,KAAK;AAC1B,MAAI,CAAC,WAAW,CAAC,aAAa,KAAK,OAAO,EAAG,QAAO;AACpD,SAAO;AACT;AAOO,SAAS,cAAc,KAAyB;AACrD,MAAI,EAAE,eAAe,OAAQ,QAAO;AACpC,QAAM,MAAM,IAAI;AAChB,MAAI,IAAI,SAAS,oBAAK,EAAG,QAAO;AAChC,MAAI,IAAI,SAAS,KAAK,EAAG,QAAO;AAChC,MAAI,IAAI,SAAS,UAAU,KAAK,IAAI,SAAS,kDAAe,KAAK,IAAI,SAAS,4CAAc,EAAG,QAAO;AACtG,MAAI,IAAI,SAAS,MAAM,KAAK,IAAI,SAAS,2BAAO,KAAK,IAAI,SAAS,2BAAO,EAAG,QAAO;AACnF,MAAI,IAAI,SAAS,iCAAQ,EAAG,QAAO;AACnC,MAAI,IAAI,SAAS,cAAI,MAAM,IAAI,SAAS,4BAAQ,KAAK,IAAI,SAAS,cAAI,GAAI,QAAO;AACjF,MAAI,IAAI,SAAS,0BAAM,KAAK,IAAI,SAAS,kCAAS,EAAG,QAAO;AAC5D,SAAO;AACT;","names":[]}

package/dist/cli.js

@@ -4,15 +4,15 @@ import {
   markdownToHwpx,
   markdownToXlsx,
   parse
-} from "./chunk-MPMKWVV2.js";
+} from "./chunk-JGMLDBW5.js";
 import "./chunk-YW5G6BCJ.js";
 import {
   VERSION,
   toArrayBuffer
-} from "./chunk-SHM3PYVA.js";
+} from "./chunk-PJSXZBZB.js";
 import "./chunk-MOL7MDBG.js";
 import "./chunk-7FMKAV4P.js";
-import "./chunk-YC2MEB7R.js";
+import "./chunk-34WIGIQC.js";
 import "./chunk-ZWE3DS7E.js";
 
 // src/cli.ts
@@ -141,7 +141,7 @@ async function runParse(files, opts) {
         saveImages(absPath);
       }
     } catch (err) {
-      const { sanitizeError } = await import("./utils-IUWGWQWV.js");
+      const { sanitizeError } = await import("./utils-HKVOS2O3.js");
       process.stderr.write(`
 [kordoc] ERROR: ${fileName} \u2014 ${sanitizeError(err)}
 `);
@@ -225,7 +225,7 @@ program.command("convert <input>").description("\uB9C8\uD06C\uB2E4\uC6B4 \uD30C\
 `));
     }
   } catch (err) {
-    const { sanitizeError } = await import("./utils-IUWGWQWV.js");
+    const { sanitizeError } = await import("./utils-HKVOS2O3.js");
     process.stderr.write(` FAIL
 `);
     process.stderr.write(`  \u2192 ${sanitizeError(err)}
@@ -234,7 +234,7 @@ program.command("convert <input>").description("\uB9C8\uD06C\uB2E4\uC6B4 \uD30C\
   }
 });
 program.command("watch <dir>").description("\uB514\uB809\uD1A0\uB9AC \uAC10\uC2DC \u2014 \uC0C8 \uBB38\uC11C \uC790\uB3D9 \uBCC0\uD658").option("--webhook <url>", "\uACB0\uACFC \uC804\uC1A1 \uC6F9\uD6C5 URL").option("-d, --out-dir <dir>", "\uBCC0\uD658 \uACB0\uACFC \uCD9C\uB825 \uB514\uB809\uD1A0\uB9AC").option("-p, --pages <range>", "\uD398\uC774\uC9C0/\uC139\uC158 \uBC94\uC704").option("--format <type>", "\uCD9C\uB825 \uD615\uC2DD: markdown \uB610\uB294 json", "markdown").option("--silent", "\uC9C4\uD589 \uBA54\uC2DC\uC9C0 \uC228\uAE30\uAE30").action(async (dir, opts) => {
-  const { watchDirectory } = await import("./watch-CRENPZU5.js");
+  const { watchDirectory } = await import("./watch-EYOGF3HY.js");
   await watchDirectory({
     dir,
     outDir: opts.outDir,

package/dist/index.cjs

@@ -2026,7 +2026,7 @@ var init_auto_detect = __esm({
 // src/ocr/cli-provider.ts
 function getTempDir() {
   if (!_tempDir) {
-    _tempDir = (0, import_path.join)(process.cwd(), "_kordoc_ocr_tmp");
+    _tempDir = (0, import_path.join)(process.cwd(), ".kordoc_ocr_tmp");
     (0, import_fs.mkdirSync)(_tempDir, { recursive: true });
   }
   return _tempDir;
@@ -2051,6 +2051,12 @@ function createCliOcrProvider(mode) {
     }
   };
 }
+function checkForLimitError(output, mode) {
+  const lower = output.toLowerCase();
+  if (lower.includes("usage limit") || lower.includes("rate limit")) {
+    throw new Error(`${mode} \uC0AC\uC6A9\uB7C9/\uC18D\uB3C4 \uC81C\uD55C: ${output.trim().slice(0, 200)}`);
+  }
+}
 function callCli(mode, imagePath) {
   if (mode === "codex") {
     return callCodexCli(imagePath);
@@ -2071,7 +2077,9 @@ function callCli(mode, imagePath) {
     const errMsg = result.stderr?.trim() || `exit code ${result.status}`;
     throw new Error(`${mode} OCR \uC2E4\uD328: ${errMsg}`);
   }
-  return result.stdout || "";
+  const output = result.stdout || "";
+  checkForLimitError(output, mode);
+  return output;
 }
 function callCodexCli(imagePath) {
   const outPath = (0, import_path.join)((0, import_os.tmpdir)(), `kordoc-codex-out-${Date.now()}.txt`);
@@ -2094,11 +2102,14 @@ function callCodexCli(imagePath) {
       const errMsg = result.stderr?.trim() || `exit code ${result.status}`;
       throw new Error(`codex OCR \uC2E4\uD328: ${errMsg}`);
     }
+    let text;
     try {
-      return (0, import_fs.readFileSync)(outPath, "utf-8");
+      text = (0, import_fs.readFileSync)(outPath, "utf-8");
     } catch {
-      return result.stdout || "";
+      text = result.stdout || "";
     }
+    checkForLimitError(text, "codex");
+    return text;
   } finally {
     try {
       (0, import_fs.unlinkSync)(outPath);
@@ -2367,7 +2378,9 @@ ${fileRefs}`;
     const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`;
     throw new Error(`${mode} \uBC30\uCE58 OCR \uC2E4\uD328: ${errMsg}`);
   }
-  return result.stdout || "";
+  const output = result.stdout || "";
+  checkForLimitError2(output, mode);
+  return output;
 }
 async function callBatchCodexCli(imagePaths) {
   const outPath = (0, import_path2.join)((0, import_os2.tmpdir)(), `kordoc-codex-batch-${Date.now()}-${Math.random().toString(36).slice(2)}.txt`);
@@ -2388,11 +2401,14 @@ async function callBatchCodexCli(imagePaths) {
       const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`;
       throw new Error(`codex \uBC30\uCE58 OCR \uC2E4\uD328: ${errMsg}`);
     }
+    let text;
     try {
-      return (0, import_fs2.readFileSync)(outPath, "utf-8");
+      text = (0, import_fs2.readFileSync)(outPath, "utf-8");
     } catch {
-      return result.stdout || "";
+      text = result.stdout || "";
     }
+    checkForLimitError2(text, "codex");
+    return text;
   } finally {
     try {
       (0, import_fs2.unlinkSync)(outPath);
@@ -2400,6 +2416,12 @@ async function callBatchCodexCli(imagePaths) {
     }
   }
 }
+function checkForLimitError2(output, mode) {
+  const lower = output.toLowerCase();
+  if (lower.includes("usage limit") || lower.includes("rate limit")) {
+    throw new Error(`${mode} \uC0AC\uC6A9\uB7C9/\uC18D\uB3C4 \uC81C\uD55C: ${output.trim().slice(0, 200)}`);
+  }
+}
 function stripCodeFence2(text) {
   const match = text.match(/^```(?:markdown|md)?\s*\n([\s\S]*?)\n```\s*$/m);
   return match ? match[1].trim() : text;
@@ -2854,7 +2876,7 @@ var import_jszip2 = __toESM(require("jszip"), 1);
 var import_xmldom = require("@xmldom/xmldom");
 
 // src/utils.ts
-var VERSION = true ? "2.4.9" : "0.0.0-dev";
+var VERSION = true ? "2.4.11" : "0.0.0-dev";
 function toArrayBuffer(buf) {
   if (buf.byteOffset === 0 && buf.byteLength === buf.buffer.byteLength) {
     return buf.buffer;