kordoc 4.14.2 → 4.14.3

package/dist/{chunk-ZWD734EY.js → chunk-D6KQKMU3.js}

@@ -3,7 +3,7 @@ import {
   ensureModelsIn,
   getModelStatusIn,
   getModelsDir
-} from "./chunk-NS55SM7A.js";
+} from "./chunk-O2PGQNYE.js";
 
 // src/ocr/models.ts
 import { join } from "path";
@@ -80,4 +80,4 @@ export {
   parseCharacterDict,
   ocrModelPath
 };
-//# sourceMappingURL=chunk-ZWD734EY.js.map
+//# sourceMappingURL=chunk-D6KQKMU3.js.map

package/dist/{chunk-QOMNEWCQ.cjs → chunk-EBZUZC6F.cjs}

@@ -1,6 +1,237 @@
 "use strict";Object.defineProperty(exports, "__esModule", {value: true}); function _nullishCoalesce(lhs, rhsFn) { if (lhs != null) { return lhs; } else { return rhsFn(); } } function _optionalChain(ops) { let lastAccessLHS = undefined; let value = ops[0]; let i = 1; while (i < ops.length) { const op = ops[i]; const fn = ops[i + 1]; i += 2; if ((op === 'optionalAccess' || op === 'optionalCall') && value == null) { return undefined; } if (op === 'access' || op === 'optionalAccess') { lastAccessLHS = value; value = fn(value); } else if (op === 'call' || op === 'optionalCall') { value = fn((...args) => value.call(lastAccessLHS, ...args)); lastAccessLHS = undefined; } } return value; }
 
-var _chunkHR2XNERBcjs = require('./chunk-HR2XNERB.cjs');
+var _chunkEMGF6EEVcjs = require('./chunk-EMGF6EEV.cjs');
+
+// src/shared/symbol-fonts.ts
+var WINGDINGS = [
+  "\u{1F589}",
+  "\u2702",
+  "\u2701",
+  "\u{1F453}",
+  "\u{1F56D}",
+  "\u{1F56E}",
+  "\u{1F56F}",
+  "\u{1F57F}",
+  "\u2706",
+  "\u{1F582}",
+  "\u{1F583}",
+  "\u{1F4EA}",
+  "\u{1F4EB}",
+  "\u{1F4EC}",
+  "\u{1F4ED}",
+  "\u{1F4C1}",
+  "\u{1F4C2}",
+  "\u{1F4C4}",
+  "\u{1F5CF}",
+  "\u{1F5D0}",
+  "\u{1F5C4}",
+  "\u231B",
+  "\u{1F5AE}",
+  "\u{1F5B0}",
+  "\u{1F5B2}",
+  "\u{1F5B3}",
+  "\u{1F5B4}",
+  "\u{1F5AB}",
+  "\u{1F5AC}",
+  "\u2707",
+  "\u270D",
+  "\u{1F58E}",
+  "\u270C",
+  "\u{1F44C}",
+  "\u{1F44D}",
+  "\u{1F44E}",
+  "\u261C",
+  "\u261E",
+  "\u261D",
+  "\u261F",
+  "\u{1F590}",
+  "\u263A",
+  "\u{1F610}",
+  "\u2639",
+  "\u{1F4A3}",
+  "\u2620",
+  "\u{1F3F3}",
+  "\u{1F3F1}",
+  "\u2708",
+  "\u263C",
+  "\u{1F4A7}",
+  "\u2744",
+  "\u{1F546}",
+  "\u271E",
+  "\u{1F548}",
+  "\u2720",
+  "\u2721",
+  "\u262A",
+  "\u262F",
+  "\u0950",
+  "\u2638",
+  "\u2648",
+  "\u2649",
+  "\u264A",
+  "\u264B",
+  "\u264C",
+  "\u264D",
+  "\u264E",
+  "\u264F",
+  "\u2650",
+  "\u2651",
+  "\u2652",
+  "\u2653",
+  "\u{1F670}",
+  "\u{1F675}",
+  "\u25CF",
+  "\u25CB",
+  "\u25A0",
+  "\u25A1",
+  "\u25A1",
+  "\u2751",
+  "\u2752",
+  "\u2B27",
+  "\u29EB",
+  "\u25C6",
+  "\u2756",
+  "\u2B25",
+  "\u2327",
+  "\u2BB9",
+  "\u2318",
+  "\u{1F3F5}",
+  "\u{1F3F6}",
+  "\u{1F676}",
+  "\u{1F677}",
+  "",
+  "\u24EA",
+  "\u2460",
+  "\u2461",
+  "\u2462",
+  "\u2463",
+  "\u2464",
+  "\u2465",
+  "\u2466",
+  "\u2467",
+  "\u2468",
+  "\u2469",
+  "\u24FF",
+  "\u2776",
+  "\u2777",
+  "\u2778",
+  "\u2779",
+  "\u277A",
+  "\u277B",
+  "\u277C",
+  "\u277D",
+  "\u277E",
+  "\u277F",
+  "\u{1F662}",
+  "\u{1F660}",
+  "\u{1F661}",
+  "\u{1F663}",
+  "\u{1F65E}",
+  "\u{1F65C}",
+  "\u{1F65D}",
+  "\u{1F65F}",
+  "\xB7",
+  "\u2022",
+  "\u25AA",
+  "\u26AA",
+  "\u25CB",
+  "\u25EF",
+  "\u25C9",
+  "\u25CE",
+  "\u{1F53F}",
+  "\u25AA",
+  "\u25FB",
+  "\u{1F7C2}",
+  "\u2726",
+  "\u2605",
+  "\u2736",
+  "\u2734",
+  "\u2739",
+  "\u2735",
+  "\u2BD0",
+  "\u2316",
+  "\u27E1",
+  "\u2311",
+  "\u2BD1",
+  "\u272A",
+  "\u2730",
+  "\u{1F550}",
+  "\u{1F551}",
+  "\u{1F552}",
+  "\u{1F553}",
+  "\u{1F554}",
+  "\u{1F555}",
+  "\u{1F556}",
+  "\u{1F557}",
+  "\u{1F558}",
+  "\u{1F559}",
+  "\u{1F55A}",
+  "\u{1F55B}",
+  "\u2BB0",
+  "\u2BB1",
+  "\u2BB2",
+  "\u2BB3",
+  "\u2BB4",
+  "\u2BB5",
+  "\u2BB6",
+  "\u2BB7",
+  "\u{1F66A}",
+  "\u{1F66B}",
+  "\u{1F655}",
+  "\u{1F654}",
+  "\u{1F657}",
+  "\u{1F656}",
+  "\u{1F650}",
+  "\u{1F651}",
+  "\u{1F652}",
+  "\u{1F653}",
+  "\u232B",
+  "\u2326",
+  "\u2B98",
+  "\u2B9A",
+  "\u2B99",
+  "\u2B9B",
+  "\u2B88",
+  "\u2B8A",
+  "\u2B89",
+  "\u2B8B",
+  "\u2190",
+  "\u2192",
+  "\u2191",
+  "\u2193",
+  "\u2196",
+  "\u2197",
+  "\u2199",
+  "\u2198",
+  "\u2B05",
+  "\u2794",
+  "\u2B06",
+  "\u2B07",
+  "\u2B09",
+  "\u2B08",
+  "\u2B0B",
+  "\u2B0A",
+  "\u21E6",
+  "\u21E8",
+  "\u21E7",
+  "\u21E9",
+  "\u2B04",
+  "\u21F3",
+  "\u2B00",
+  "\u2B01",
+  "\u2B03",
+  "\u2B02",
+  "\u{1F8AC}",
+  "\u{1F8AD}",
+  "\u2717",
+  "\u2714",
+  "\u2612",
+  "\u2611",
+  ""
+];
+function wingdingsChar(code) {
+  if (code < 33 || code > 255) return void 0;
+  return WINGDINGS[code - 33] || void 0;
+}
 
 // src/shared/pua.ts
 var BMP_SYMBOL_MAP = {
@@ -188,7 +419,7 @@ var BOXED_NUMBER_START = 983729;
 var BOXED_NUMBER_END = 983748;
 function mapPuaChar(code) {
   if (code >= 61472 && code <= 61695) {
-    return BMP_SYMBOL_MAP[code - 61440];
+    return _nullishCoalesce(BMP_SYMBOL_MAP[code - 61440], () => ( wingdingsChar(code - 61440)));
   }
   if (code >= BOXED_NUMBER_START && code <= BOXED_NUMBER_END) {
     return String.fromCodePoint(9312 + (code - BOXED_NUMBER_START));
@@ -264,32 +495,54 @@ function buildTable(rows, options) {
 }
 function buildTableDirect(rows, numRows, options) {
   let maxCols = 0;
+  const rowCap = Math.min(MAX_ROWS, numRows + rows.reduce((n, r) => n + r.length, 0) + 1);
   for (const row of rows) {
     for (const cell of row) {
       const end = (_nullishCoalesce(cell.colAddr, () => ( 0))) + cell.colSpan;
       if (end > maxCols) maxCols = end;
+      if (cell.rowAddr !== void 0 && cell.rowAddr >= numRows) numRows = Math.min(cell.rowAddr + 1, rowCap);
     }
   }
   if (maxCols > MAX_COLS) maxCols = MAX_COLS;
-  if (maxCols === 0) return { rows: 0, cols: 0, cells: [], hasHeader: false };
+  if (maxCols === 0 || numRows === 0) return { rows: 0, cols: 0, cells: [], hasHeader: false };
   const grid = Array.from(
     { length: numRows },
     () => Array.from({ length: maxCols }, () => ({ text: "", colSpan: 1, rowSpan: 1 }))
   );
+  const owner = Array.from({ length: numRows }, () => new Array(maxCols));
   const anchorCols = /* @__PURE__ */ new Set();
-  for (const row of rows) {
-    for (const cell of row) {
-      const r = _nullishCoalesce(cell.rowAddr, () => ( 0));
-      const c = _nullishCoalesce(cell.colAddr, () => ( 0));
-      if (r >= numRows || c >= maxCols || r < 0 || c < 0) continue;
+  for (let ri = 0; ri < rows.length; ri++) {
+    for (const cell of rows[ri]) {
+      const text = _optionalChain([options, 'optionalAccess', _3 => _3.keepEmptyParagraphs]) ? cell.text : cell.text.trim();
+      const r = Math.min(_nullishCoalesce(cell.rowAddr, () => ( ri)), numRows - 1);
+      let c = _nullishCoalesce(cell.colAddr, () => ( 0));
+      if (r < 0 || c < 0) continue;
+      if (cell.colAddr === void 0) while (c < maxCols - 1 && owner[r][c]) c++;
+      if (c >= maxCols) c = maxCols - 1;
+      const own = owner[r][c];
+      if (own) {
+        if (text.trim()) own.text = own.text ? `${own.text}
+${text}` : text;
+        continue;
+      }
+      let colSpan = Math.max(1, Math.min(cell.colSpan, maxCols - c));
+      for (let dc = 1; dc < colSpan; dc++) if (owner[r][c + dc]) {
+        colSpan = dc;
+        break;
+      }
+      let rowSpan = Math.max(1, Math.min(cell.rowSpan, numRows - r));
+      for (let dr = 1; dr < rowSpan; dr++) {
+        if (owner[r + dr].slice(c, c + colSpan).some(Boolean)) {
+          rowSpan = dr;
+          break;
+        }
+      }
+      const ir = { text, colSpan, rowSpan };
       anchorCols.add(c);
-      grid[r][c] = { text: _optionalChain([options, 'optionalAccess', _3 => _3.keepEmptyParagraphs]) ? cell.text : cell.text.trim(), colSpan: cell.colSpan, rowSpan: cell.rowSpan };
-      for (let dr = 0; dr < cell.rowSpan; dr++) {
-        for (let dc = 0; dc < cell.colSpan; dc++) {
-          if (dr === 0 && dc === 0) continue;
-          if (r + dr < numRows && c + dc < maxCols) {
-            grid[r + dr][c + dc] = { text: "", colSpan: 1, rowSpan: 1 };
-          }
+      for (let dr = 0; dr < rowSpan; dr++) {
+        for (let dc = 0; dc < colSpan; dc++) {
+          owner[r + dr][c + dc] = ir;
+          grid[r + dr][c + dc] = dr === 0 && dc === 0 ? ir : { text: "", colSpan: 1, rowSpan: 1 };
         }
       }
     }
@@ -306,6 +559,11 @@ function trimAndReturn(grid, numRows, maxCols, anchorCols, options) {
   }
   if (effectiveCols < maxCols && effectiveCols > 0) {
     const trimmed = grid.map((row) => row.slice(0, effectiveCols));
+    for (const row of trimmed) {
+      for (let c = 0; c < row.length; c++) {
+        if (c + row[c].colSpan > effectiveCols) row[c].colSpan = effectiveCols - c;
+      }
+    }
     return { rows: numRows, cols: effectiveCols, cells: trimmed, hasHeader: numRows > 1 };
   }
   return { rows: numRows, cols: maxCols, cells: grid, hasHeader: numRows > 1 };
@@ -318,13 +576,16 @@ function convertTableToText(rows) {
 function escapeGfm(text) {
   const NUL = String.fromCharCode(0);
   const spans = [];
-  const masked = text.replace(/!\[[^\]]*\]\([^)\n]*\)|\]\((?:https?:|mailto:|tel:|#)[^)\n]*\)|\$\$[^$]*\$\$|\$[^$\n]*\$/gi, (m) => {
+  const masked = text.replace(/!\[[^\]]*\]\([^)\n]*\)|\]\((?:https?:|mailto:|tel:|#)[^)\n]*\)|(?<!\\)\$\$(?:\\[\s\S]|[^\\$])*\$\$|(?<!\\)\$(?:\\[^\n]|[^\\$\n])*\$/gi, (m) => {
     spans.push(m);
     return NUL + (spans.length - 1) + NUL;
   });
-  const escaped = masked.replace(/([~*_`])/g, "\\$1");
+  const escaped = masked.replace(/\\(?=[!-#%-\/:-@\[-\x60{}~])/g, "\\\\").replace(/([~*_`])/g, "\\$1").replace(/(?<!\\)\|/g, "\\|").replace(/^([ \t]*)(?=#{1,6}(?:[ \t]|$))/gm, "$1\\").replace(/<(?!\/?u>)(?=[A-Za-z/!?])/g, "\\<");
   return escaped.replace(new RegExp(NUL + "(\\d+)" + NUL, "g"), (_, n) => spans[Number(n)]);
 }
+function escapeLiteralDollar(text) {
+  return text.includes("$") ? text.replace(/\$/g, "\\$") : text;
+}
 var HWP_SHAPE_ALT_TEXT_RE = /^(?:모서리가 둥근 |둥근 )?(?:사각형|직사각형|정사각형|원|타원|삼각형|이등변 삼각형|직각 삼각형|선|직선|곡선|화살표|굵은 화살표|이중 화살표|오각형|육각형|팔각형|별|[4-8]점별|십자|십자형|구름|구름형|마름모|도넛|평행사변형|사다리꼴|부채꼴|호|반원|물결|번개|하트|빗금|블록 화살표|수식|표|그림|개체|그리기\s?개체|묶음\s?개체|글상자|수식\s?개체|OLE\s?개체)\s?입니다\.?$/gm;
 function sanitizePua(text) {
   return mapPuaText(text).replace(/[\u{F0000}-\u{FFFFD}]/gu, "");
@@ -423,6 +684,7 @@ function dedupeRunningHeaders(blocks) {
 function spansToMarkdown(spans) {
   let out = "";
   for (const s of spans) {
+    if (s.placeholder) continue;
     const text = sanitizePua(_nullishCoalesce(s.text, () => ( "")));
     if (!text) continue;
     let marker = s.code ? "`" : s.bold && s.italic ? "***" : s.bold ? "**" : s.italic ? "*" : "";
@@ -455,7 +717,8 @@ function blocksToMarkdown(blocks) {
     if (block.type === "heading" && block.text) {
       const prefix = "#".repeat(Math.min(block.level || 2, 6));
       const headingText = sanitizeText(block.text);
-      if (headingText) lines.push("", `${prefix} ${escapeGfm(headingText)}`, "");
+      const note = block.footnoteText ? ` (\uC8FC: ${block.footnoteText})` : "";
+      if (headingText) lines.push("", `${prefix} ${escapeGfm(headingText + note)}`, "");
       continue;
     }
     if (block.type === "image" && block.text) {
@@ -501,7 +764,7 @@ function blocksToMarkdown(blocks) {
       if (_optionalChain([block, 'access', _26 => _26.spans, 'optionalAccess', _27 => _27.length])) {
         let rendered = spansToMarkdown(block.spans);
         if (block.href) {
-          const href = _chunkHR2XNERBcjs.sanitizeHref.call(void 0, block.href);
+          const href = _chunkEMGF6EEVcjs.sanitizeHref.call(void 0, block.href);
           if (href) rendered = `[${rendered}](${href})`;
         }
         if (block.footnoteText) rendered += ` (\uC8FC: ${block.footnoteText})`;
@@ -509,7 +772,7 @@ function blocksToMarkdown(blocks) {
         continue;
       }
       if (block.href) {
-        const href = _chunkHR2XNERBcjs.sanitizeHref.call(void 0, block.href);
+        const href = _chunkEMGF6EEVcjs.sanitizeHref.call(void 0, block.href);
         if (href) text = `[${text}](${href})`;
       }
       if (block.footnoteText) {
@@ -549,16 +812,22 @@ function hasStructuredCellContent(table) {
   }
   return false;
 }
+function noteSuffix(b) {
+  return b.footnoteText && b.text ? ` (\uC8FC: ${b.footnoteText})` : "";
+}
+function visibleText(b) {
+  return _optionalChain([b, 'access', _31 => _31.spans, 'optionalAccess', _32 => _32.some, 'call', _33 => _33((s) => s.placeholder)]) ? b.spans.filter((s) => !s.placeholder).map((s) => s.text).join("") : _nullishCoalesce(b.text, () => ( ""));
+}
 function cellInnerHtml(cell) {
-  if (_optionalChain([cell, 'access', _31 => _31.blocks, 'optionalAccess', _32 => _32.length])) {
+  if (_optionalChain([cell, 'access', _34 => _34.blocks, 'optionalAccess', _35 => _35.length])) {
     return cell.blocks.map((b) => {
       if (b.type === "table" && b.table) {
         const cap = b.table.caption ? sanitizeText(b.table.caption) : "";
         return (cap ? cap + "<br>" : "") + tableToHtml(b.table);
       }
       if (b.type === "image" && b.text) return `<img src="${b.text}" alt="image">`;
-      const t = sanitizeText(_nullishCoalesce(b.text, () => ( "")));
-      return t ? t.replace(/\n/g, "<br>") : "";
+      const t = sanitizeText(visibleText(b));
+      return t ? (t + noteSuffix(b)).replace(/\n/g, "<br>") : "";
     }).filter(Boolean).join("<br>");
   }
   return sanitizeText(cell.text).replace(/\n/g, "<br>");
@@ -583,7 +852,7 @@ function tableToHtml(table) {
     const rowHtml = [];
     for (let c = 0; c < numCols; c++) {
       if (skip.has(`${r},${c}`)) continue;
-      const cell = _optionalChain([cells, 'access', _33 => _33[r], 'optionalAccess', _34 => _34[c]]);
+      const cell = _optionalChain([cells, 'access', _36 => _36[r], 'optionalAccess', _37 => _37[c]]);
       if (!cell) continue;
       for (let dr = 0; dr < cell.rowSpan; dr++) {
         for (let dc = 0; dc < cell.colSpan; dc++) {
@@ -598,7 +867,7 @@ function tableToHtml(table) {
       const attrStr = attrs.length ? " " + attrs.join(" ") : "";
       rowHtml.push(`<${tag}${attrStr}>${text}</${tag}>`);
     }
-    if (rowHtml.length) lines.push(`<tr>${rowHtml.join("")}</tr>`);
+    lines.push(`<tr>${rowHtml.join("")}</tr>`);
   }
   lines.push("</table>");
   return lines.join("\n");
@@ -606,7 +875,8 @@ function tableToHtml(table) {
 function tableToMarkdown(table) {
   if (table.rows === 0 || table.cols === 0) return "";
   const { cells, rows: numRows, cols: numCols } = table;
-  if ((hasMergedCells(table) || hasStructuredCellContent(table)) && !tableContainsInlineMath(table)) {
+  if (hasStructuredCellContent(table)) return tableToHtml(table);
+  if (hasMergedCells(table) && !tableContainsInlineMath(table)) {
     return tableToHtml(table);
   }
   if (numRows === 1 && numCols === 1) {
@@ -628,9 +898,9 @@ function tableToMarkdown(table) {
   for (let r = 0; r < numRows; r++) {
     for (let c = 0; c < numCols; c++) {
       if (skip.has(`${r},${c}`)) continue;
-      const cell = _optionalChain([cells, 'access', _35 => _35[r], 'optionalAccess', _36 => _36[c]]);
+      const cell = _optionalChain([cells, 'access', _38 => _38[r], 'optionalAccess', _39 => _39[c]]);
       if (!cell) continue;
-      display[r][c] = (_optionalChain([cell, 'access', _37 => _37.blocks, 'optionalAccess', _38 => _38.some, 'call', _39 => _39((b) => b.spans || b.type === "image" && b.text)]) ? cell.blocks.map((b) => b.type === "image" && b.text ? `![image](${b.text})` : b.spans ? spansToMarkdown(b.spans) : escapeGfm(sanitizeText(_nullishCoalesce(b.text, () => ( ""))))).filter(Boolean).join("<br>") : escapeGfm(sanitizeText(cell.text)).replace(/\n/g, "<br>")).replace(/\|/g, "\\|");
+      display[r][c] = (_optionalChain([cell, 'access', _40 => _40.blocks, 'optionalAccess', _41 => _41.some, 'call', _42 => _42((b) => b.spans || b.type === "image" && b.text)]) ? cell.blocks.map((b) => b.type === "image" && b.text ? `![image](${b.text})` : b.spans ? spansToMarkdown(b.spans) + escapeGfm(noteSuffix(b)) : escapeGfm(sanitizeText(_nullishCoalesce(b.text, () => ( ""))) + noteSuffix(b))).filter(Boolean).join("<br>") : escapeGfm(sanitizeText(cell.text)).replace(/\n/g, "<br>")).replace(/(?<!\\)\|/g, "\\|");
       for (let dr = 0; dr < cell.rowSpan; dr++) {
         for (let dc = 0; dc < cell.colSpan; dc++) {
           if (dr === 0 && dc === 0) continue;
@@ -791,7 +1061,7 @@ function encodePng(width, height, rgba) {
   return out;
 }
 function inlineImagesIntoMarkdown(markdown, images, opts) {
-  const compress = _optionalChain([opts, 'optionalAccess', _40 => _40.compress]) !== false;
+  const compress = _optionalChain([opts, 'optionalAccess', _43 => _43.compress]) !== false;
   const uris = /* @__PURE__ */ new Map();
   for (const img of images) {
     let bytes = img.data;
@@ -829,5 +1099,9 @@ function inlineImagesIntoMarkdown(markdown, images, opts) {
 
 
 
-exports.mapPuaText = mapPuaText; exports.MAX_COLS = MAX_COLS; exports.MAX_ROWS = MAX_ROWS; exports.buildTable = buildTable; exports.convertTableToText = convertTableToText; exports.flattenLayoutTables = flattenLayoutTables; exports.dedupeRunningHeaders = dedupeRunningHeaders; exports.blocksToMarkdown = blocksToMarkdown; exports.hasStructuredCellContent = hasStructuredCellContent; exports.encodePng = encodePng; exports.inlineImagesIntoMarkdown = inlineImagesIntoMarkdown;
-//# sourceMappingURL=chunk-QOMNEWCQ.cjs.map
+
+
+
+
+exports.WINGDINGS = WINGDINGS; exports.mapPuaText = mapPuaText; exports.MAX_COLS = MAX_COLS; exports.MAX_ROWS = MAX_ROWS; exports.buildTable = buildTable; exports.convertTableToText = convertTableToText; exports.escapeGfm = escapeGfm; exports.escapeLiteralDollar = escapeLiteralDollar; exports.flattenLayoutTables = flattenLayoutTables; exports.dedupeRunningHeaders = dedupeRunningHeaders; exports.blocksToMarkdown = blocksToMarkdown; exports.hasStructuredCellContent = hasStructuredCellContent; exports.noteSuffix = noteSuffix; exports.encodePng = encodePng; exports.inlineImagesIntoMarkdown = inlineImagesIntoMarkdown;
+//# sourceMappingURL=chunk-EBZUZC6F.cjs.map