@clazic/kordoc 2.3.2 → 2.3.3

package/dist/cli.js

@@ -4,11 +4,11 @@ import {
   markdownToHwpx,
   markdownToXlsx,
   parse
-} from "./chunk-2GFJFTKS.js";
+} from "./chunk-UDFKY7CH.js";
 import {
   VERSION,
   toArrayBuffer
-} from "./chunk-STIKJGEA.js";
+} from "./chunk-NU3KFVVZ.js";
 import "./chunk-MOL7MDBG.js";
 import "./chunk-7FMKAV4P.js";
 import "./chunk-JOGAFNIL.js";
@@ -137,7 +137,7 @@ async function runParse(files, opts) {
         saveImages(absPath);
       }
     } catch (err) {
-      const { sanitizeError } = await import("./utils-FFUQJTTI.js");
+      const { sanitizeError } = await import("./utils-STJT6CFC.js");
       process.stderr.write(`
 [kordoc] ERROR: ${fileName} \u2014 ${sanitizeError(err)}
 `);
@@ -221,7 +221,7 @@ program.command("convert <input>").description("\uB9C8\uD06C\uB2E4\uC6B4 \uD30C\
 `));
     }
   } catch (err) {
-    const { sanitizeError } = await import("./utils-FFUQJTTI.js");
+    const { sanitizeError } = await import("./utils-STJT6CFC.js");
     process.stderr.write(` FAIL
 `);
     process.stderr.write(`  \u2192 ${sanitizeError(err)}
@@ -230,7 +230,7 @@ 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-2O32L6IF.js");
+  const { watchDirectory } = await import("./watch-PRQGLOW3.js");
   await watchDirectory({
     dir,
     outDir: opts.outDir,

package/dist/index.cjs

@@ -2422,7 +2422,7 @@ async function resolveOcrProvider(mode, warnings, concurrency, batchSize) {
     return createCliOcrProvider(mode);
   }
   const detected = detectAvailableOcr();
-  if (detected !== "gemini") {
+  if (detected !== "codex") {
     if (detected === "tesseract") {
       warnings?.push({
         message: getTesseractFallbackMessage(),
@@ -2430,7 +2430,7 @@ async function resolveOcrProvider(mode, warnings, concurrency, batchSize) {
       });
     } else {
       warnings?.push({
-        message: `OCR: '${detected}' \uC0AC\uC6A9 \uC911 (gemini CLI\uAC00 \uC5C6\uC5B4 fallback). \uB354 \uB098\uC740 \uD488\uC9C8\uC744 \uC704\uD574 gemini CLI \uC124\uCE58\uB97C \uAD8C\uC7A5\uD569\uB2C8\uB2E4.`,
+        message: `OCR: '${detected}' \uC0AC\uC6A9 \uC911 (codex CLI\uAC00 \uC5C6\uC5B4 fallback). \uB354 \uB098\uC740 \uD488\uC9C8\uC744 \uC704\uD574 codex CLI \uC124\uCE58\uB97C \uAD8C\uC7A5\uD569\uB2C8\uB2E4.`,
         code: "OCR_CLI_FALLBACK"
       });
     }
@@ -2810,7 +2810,7 @@ var import_jszip2 = __toESM(require("jszip"), 1);
 var import_xmldom = require("@xmldom/xmldom");
 
 // src/utils.ts
-var VERSION = true ? "2.3.2" : "0.0.0-dev";
+var VERSION = true ? "2.3.3" : "0.0.0-dev";
 function toArrayBuffer(buf) {
   if (buf.byteOffset === 0 && buf.byteLength === buf.buffer.byteLength) {
     return buf.buffer;
@@ -6338,15 +6338,26 @@ async function parsePdfDocument(buffer, options) {
         warnings.push({ page: i, message: `\uD398\uC774\uC9C0 ${i} \uD30C\uC2F1 \uC2E4\uD328: ${pageErr instanceof Error ? pageErr.message : "\uC54C \uC218 \uC5C6\uB294 \uC624\uB958"}`, code: "PARTIAL_PARSE" });
       }
     };
-    const sampleCount = Math.min(5, targetPageNums.length);
-    for (let si = 0; si < sampleCount; si++) {
+    const SAMPLE_SIZE = Math.min(10, targetPageNums.length);
+    const sampledIndices = /* @__PURE__ */ new Set();
+    if (targetPageNums.length <= SAMPLE_SIZE) {
+      for (let i = 0; i < targetPageNums.length; i++) sampledIndices.add(i);
+    } else {
+      for (let i = 0; i < SAMPLE_SIZE; i++) {
+        const idx = Math.round(i * (targetPageNums.length - 1) / (SAMPLE_SIZE - 1));
+        sampledIndices.add(idx);
+      }
+    }
+    for (const si of sampledIndices) {
       await parseSinglePage(targetPageNums[si]);
     }
-    const sampleParsed = parsedPages || sampleCount;
+    const sampleParsed = parsedPages || sampledIndices.size;
     const isImageBased = totalChars / Math.max(sampleParsed, 1) < 10;
     if (!isImageBased) {
-      for (let si = sampleCount; si < targetPageNums.length; si++) {
-        await parseSinglePage(targetPageNums[si]);
+      for (let si = 0; si < targetPageNums.length; si++) {
+        if (!sampledIndices.has(si)) {
+          await parseSinglePage(targetPageNums[si]);
+        }
       }
     }
     const parsedPageCount = parsedPages || (pageFilter ? pageFilter.size : effectivePageCount);