@clazic/kordoc 2.3.0 → 2.3.1

package/dist/index.js

@@ -2222,7 +2222,7 @@ __export(batch_provider_exports, {
   DEFAULT_BATCH_SIZES: () => DEFAULT_BATCH_SIZES,
   createBatchCliProvider: () => createBatchCliProvider
 });
-import { spawnSync as spawnSync2 } from "child_process";
+import { spawn } from "child_process";
 import { writeFileSync as writeFileSync2, readFileSync as readFileSync2, unlinkSync as unlinkSync2, mkdirSync as mkdirSync2 } from "fs";
 import { join as join2 } from "path";
 import { tmpdir as tmpdir2 } from "os";
@@ -2249,9 +2249,9 @@ function createBatchCliProvider(mode, batchSize) {
         }
         let output;
         if (mode === "codex") {
-          output = callBatchCodexCli(tempFiles);
+          output = await callBatchCodexCli(tempFiles);
         } else {
-          output = callBatchCli(mode, tempFiles);
+          output = await callBatchCli(mode, tempFiles);
         }
         const cleaned = stripCodeFence2(output.trim());
         const parts = cleaned.split(/<!--\s*PAGE_BREAK\s*-->/).map((p) => p.trim()).filter((p) => p.length > 0);
@@ -2273,7 +2273,48 @@ function createBatchCliProvider(mode, batchSize) {
     }
   };
 }
-function callBatchCli(mode, imagePaths) {
+function spawnAsync(cmd, args, opts) {
+  return new Promise((resolve, reject) => {
+    const child = spawn(cmd, args, {
+      cwd: opts.cwd,
+      env: process.env,
+      stdio: ["pipe", "pipe", "pipe"]
+    });
+    let stdout = "";
+    let stderr = "";
+    let killed = false;
+    child.stdout.setEncoding("utf-8");
+    child.stderr.setEncoding("utf-8");
+    child.stdout.on("data", (d) => {
+      stdout += d;
+    });
+    child.stderr.on("data", (d) => {
+      stderr += d;
+    });
+    const timer = setTimeout(() => {
+      killed = true;
+      child.kill("SIGTERM");
+    }, opts.timeoutMs);
+    if (opts.stdin !== void 0) {
+      child.stdin.end(opts.stdin);
+    } else {
+      child.stdin.end();
+    }
+    child.on("close", (code) => {
+      clearTimeout(timer);
+      if (killed) {
+        reject(new Error(`\uD0C0\uC784\uC544\uC6C3 (${Math.round(opts.timeoutMs / 1e3)}\uCD08)`));
+      } else {
+        resolve({ stdout, stderr, exitCode: code ?? 1 });
+      }
+    });
+    child.on("error", (err) => {
+      clearTimeout(timer);
+      reject(err);
+    });
+  });
+}
+async function callBatchCli(mode, imagePaths) {
   const fileRefs = imagePaths.map((p) => `@${p}`).join("\n");
   const prompt = `${BATCH_OCR_PROMPT}
 
@@ -2289,24 +2330,18 @@ ${fileRefs}`;
     if (model) args.push("--model", model);
   }
   const timeoutMs = 6e4 + imagePaths.length * 2e4;
-  const result = spawnSync2(mode, args, {
-    encoding: "utf-8",
-    timeout: timeoutMs,
-    maxBuffer: 50 * 1024 * 1024,
-    // 50MB (large batch output)
+  const result = await spawnAsync(mode, args, {
+    timeoutMs,
     ...mode === "claude" ? { cwd: tmpdir2() } : {}
   });
-  if (result.error) {
-    throw new Error(`${mode} \uBC30\uCE58 OCR \uC2E4\uD328: ${result.error.message}`);
-  }
-  if (result.status !== 0) {
-    const errMsg = result.stderr?.trim() || `exit code ${result.status}`;
+  if (result.exitCode !== 0) {
+    const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`;
     throw new Error(`${mode} \uBC30\uCE58 OCR \uC2E4\uD328: ${errMsg}`);
   }
   return result.stdout || "";
 }
-function callBatchCodexCli(imagePaths) {
-  const outPath = join2(tmpdir2(), `kordoc-codex-batch-${Date.now()}.txt`);
+async function callBatchCodexCli(imagePaths) {
+  const outPath = join2(tmpdir2(), `kordoc-codex-batch-${Date.now()}-${Math.random().toString(36).slice(2)}.txt`);
   try {
     const args = ["exec", BATCH_OCR_PROMPT];
     for (const p of imagePaths) {
@@ -2316,17 +2351,12 @@ function callBatchCodexCli(imagePaths) {
     const model = process.env.KORDOC_CODEX_MODEL;
     if (model) args.push("--model", model);
     const timeoutMs = 6e4 + imagePaths.length * 2e4;
-    const result = spawnSync2("codex", args, {
-      encoding: "utf-8",
-      timeout: timeoutMs,
-      maxBuffer: 50 * 1024 * 1024,
-      input: ""
+    const result = await spawnAsync("codex", args, {
+      timeoutMs,
+      stdin: ""
     });
-    if (result.error) {
-      throw new Error(`codex \uBC30\uCE58 OCR \uC2E4\uD328: ${result.error.message}`);
-    }
-    if (result.status !== 0) {
-      const errMsg = result.stderr?.trim() || `exit code ${result.status}`;
+    if (result.exitCode !== 0) {
+      const errMsg = result.stderr?.trim() || `exit code ${result.exitCode}`;
       throw new Error(`codex \uBC30\uCE58 OCR \uC2E4\uD328: ${errMsg}`);
     }
     try {
@@ -2584,7 +2614,7 @@ function isBatchProvider(p) {
 async function ocrPages(doc, provider, pageFilter, effectivePageCount, warnings, concurrency = 1, onProgress) {
   const blocks = [];
   if (isBatchProvider(provider)) {
-    return ocrPagesBatch(doc, provider, pageFilter, effectivePageCount, warnings, onProgress);
+    return ocrPagesBatch(doc, provider, pageFilter, effectivePageCount, warnings, concurrency, onProgress);
   }
   if (concurrency <= 1) {
     for (let i = 1; i <= effectivePageCount; i++) {
@@ -2631,8 +2661,7 @@ async function ocrPages(doc, provider, pageFilter, effectivePageCount, warnings,
   }
   return blocks;
 }
-async function ocrPagesBatch(doc, provider, pageFilter, effectivePageCount, warnings, onProgress) {
-  const blocks = [];
+async function ocrPagesBatch(doc, provider, pageFilter, effectivePageCount, warnings, concurrency = 1, onProgress) {
   const pageNumbers = [];
   for (let i = 1; i <= effectivePageCount; i++) {
     if (pageFilter && !pageFilter.has(i)) continue;
@@ -2649,16 +2678,16 @@ async function ocrPagesBatch(doc, provider, pageFilter, effectivePageCount, warn
     batches.push(pageImages.slice(i, i + provider.batchSize));
   }
   let processed = 0;
-  for (const batch of batches) {
+  const batchTasks = batches.map((batch, batchIdx) => async () => {
+    const pageBlocks = [];
     try {
       const results = await provider.processBatch(batch);
       for (const { pageNum } of batch) {
         const result = results.get(pageNum);
-        if (result) {
-          for (const b of ocrResultToBlocks(result, pageNum)) blocks.push(b);
-        }
-        processed++;
-        onProgress?.(processed, pageNumbers.length);
+        pageBlocks.push({
+          pageNum,
+          blocks: result ? ocrResultToBlocks(result, pageNum) : []
+        });
       }
     } catch (err) {
       const range = `${batch[0].pageNum}-${batch[batch.length - 1].pageNum}`;
@@ -2666,8 +2695,20 @@ async function ocrPagesBatch(doc, provider, pageFilter, effectivePageCount, warn
         message: `\uBC30\uCE58 OCR \uC2E4\uD328 (\uD398\uC774\uC9C0 ${range}): ${err instanceof Error ? err.message : "\uC54C \uC218 \uC5C6\uB294 \uC624\uB958"}`,
         code: "OCR_PAGE_FAILED"
       });
-      processed += batch.length;
-      onProgress?.(processed, pageNumbers.length);
+      for (const { pageNum } of batch) {
+        pageBlocks.push({ pageNum, blocks: [] });
+      }
+    }
+    processed += batch.length;
+    onProgress?.(processed, pageNumbers.length);
+    return { batchIdx, pageBlocks };
+  });
+  const effectiveConcurrency = Math.max(1, concurrency);
+  const batchResults = await runWithConcurrency(batchTasks, effectiveConcurrency);
+  const blocks = [];
+  for (const result of batchResults) {
+    for (const { blocks: pageBlks } of result.pageBlocks) {
+      for (const b of pageBlks) blocks.push(b);
     }
   }
   return blocks;
@@ -2737,7 +2778,7 @@ import JSZip2 from "jszip";
 import { DOMParser } from "@xmldom/xmldom";
 
 // src/utils.ts
-var VERSION = true ? "2.3.0" : "0.0.0-dev";
+var VERSION = true ? "2.3.1" : "0.0.0-dev";
 function toArrayBuffer(buf) {
   if (buf.byteOffset === 0 && buf.byteLength === buf.buffer.byteLength) {
     return buf.buffer;