kordoc 4.12.1 → 4.12.2

package/dist/{chunk-MXZXZTKW.cjs → chunk-ONNAHSPV.cjs}

@@ -6,7 +6,7 @@ var _chunkT7GRHAIWcjs = require('./chunk-T7GRHAIW.cjs');
 
 
 
-var _chunk6MF7JRG3cjs = require('./chunk-6MF7JRG3.cjs');
+var _chunk62IU4NV3cjs = require('./chunk-62IU4NV3.cjs');
 
 // src/pdf/image-regions.ts
 var _pdfmjs = require('pdfjs-dist/legacy/build/pdf.mjs');
@@ -1009,8 +1009,14 @@ function edgeStroked(lines, dir, pos, a1, a2) {
 }
 function buildClipCellGrids(rects, strokedH, strokedV, pageWidth, pageHeight) {
   const pageArea = pageWidth * pageHeight;
-  const cells = rects.filter((r) => r.x2 - r.x1 >= CLIP_MIN_W2 && r.y2 - r.y1 >= CLIP_MIN_H2 && (pageArea <= 0 || (r.x2 - r.x1) * (r.y2 - r.y1) < pageArea * CLIP_MAX_PAGE_FRAC));
-  if (cells.length < CLIP_MIN_GROUP) return { grids: [], containers: [] };
+  const sameRect = (a, b) => Math.abs(a.x1 - b.x1) <= CLIP_EDGE_TOL && Math.abs(a.x2 - b.x2) <= CLIP_EDGE_TOL && Math.abs(a.y1 - b.y1) <= CLIP_EDGE_TOL && Math.abs(a.y2 - b.y2) <= CLIP_EDGE_TOL;
+  const cells = [];
+  for (const r of rects) {
+    if (r.x2 - r.x1 < CLIP_MIN_W2 || r.y2 - r.y1 < CLIP_MIN_H2) continue;
+    if (pageArea > 0 && (r.x2 - r.x1) * (r.y2 - r.y1) >= pageArea * CLIP_MAX_PAGE_FRAC) continue;
+    if (!cells.some((c) => sameRect(c, r))) cells.push(r);
+  }
+  if (cells.length < 1) return { grids: [], containers: [] };
   const parent = new Array(cells.length).fill(-1);
   for (let i = 0; i < cells.length; i++) {
     for (let j = 0; j < cells.length; j++) {
@@ -1045,15 +1051,30 @@ function buildClipCellGrids(rects, strokedH, strokedV, pageWidth, pageHeight) {
     if (g) g.push(i);
     else groups.set(r, [i]);
   }
-  const containers = [];
-  for (let i = 0; i < cells.length; i++) {
-    if (!isContainer[i]) continue;
-    const r = cells[i];
-    if (!containers.some((c) => Math.abs(c.x1 - r.x1) <= CLIP_EDGE_TOL && Math.abs(c.x2 - r.x2) <= CLIP_EDGE_TOL && Math.abs(c.y1 - r.y1) <= CLIP_EDGE_TOL && Math.abs(c.y2 - r.y2) <= CLIP_EDGE_TOL)) containers.push(r);
-  }
+  const containers = cells.filter((_, i) => isContainer[i]);
+  const groupSize = (i) => _nullishCoalesce(_optionalChain([groups, 'access', _4 => _4.get, 'call', _5 => _5(find(i)), 'optionalAccess', _6 => _6.length]), () => ( 0));
+  const isGridMember = (i) => groupSize(i) >= CLIP_MIN_GROUP;
+  const framed = (r) => edgeStroked(strokedH, "h", r.y1, r.x1, r.x2) && edgeStroked(strokedH, "h", r.y2, r.x1, r.x2) && edgeStroked(strokedV, "v", r.x1, r.y1, r.y2) && edgeStroked(strokedV, "v", r.x2, r.y1, r.y2);
+  const loneFrame = (i) => isContainer[i] && !isGridMember(i) && framed(cells[i]);
+  const parentAttachable = (i) => parent[i] >= 0 && (isGridMember(parent[i]) || loneFrame(parent[i]));
+  const onlyChild = (i) => parent[i] >= 0 && parent.filter((p) => p === parent[i]).length === 1;
   const grids = [];
   for (const idxs of groups.values()) {
-    if (idxs.length < CLIP_MIN_GROUP) continue;
+    const first = idxs[0];
+    const parentRect = parentAttachable(first) ? cells[parent[first]] : void 0;
+    if (idxs.length < CLIP_MIN_GROUP) {
+      if (!loneFrame(first) && !((parentRect || onlyChild(first)) && framed(cells[first]))) continue;
+      const r = cells[first];
+      grids.push({
+        rowYs: [r.y2, r.y1],
+        colXs: [r.x1, r.x2],
+        bbox: { x1: r.x1, y1: r.y1, x2: r.x2, y2: r.y2 },
+        vertexRadius: 1,
+        cells: [{ row: 0, col: 0, rowSpan: 1, colSpan: 1, bbox: { x1: r.x1, y1: r.y1, x2: r.x2, y2: r.y2 } }],
+        ...parentRect ? { clipParent: parentRect } : {}
+      });
+      continue;
+    }
     const members = idxs.map((i) => cells[i]);
     let edges = 0, invisible = 0;
     for (const r of members) {
@@ -1094,7 +1115,8 @@ function buildClipCellGrids(rects, strokedH, strokedV, pageWidth, pageHeight) {
       colXs,
       bbox: { x1: colXs[0], y1: rowYs[numRows], x2: colXs[numCols], y2: rowYs[0] },
       vertexRadius: 1,
-      cells: out
+      cells: out,
+      ...parentRect ? { clipParent: parentRect } : {}
     });
   }
   return { grids, containers };
@@ -1573,7 +1595,7 @@ function dominantStyle(items) {
     }
   }
   if (dominantSize === 0) return void 0;
-  const fontName = _optionalChain([items, 'access', _4 => _4.find, 'call', _5 => _5((i) => i.fontSize === dominantSize), 'optionalAccess', _6 => _6.fontName]) || void 0;
+  const fontName = _optionalChain([items, 'access', _7 => _7.find, 'call', _8 => _8((i) => i.fontSize === dominantSize), 'optionalAccess', _9 => _9.fontName]) || void 0;
   return fontName ? { fontSize: dominantSize, fontName } : { fontSize: dominantSize };
 }
 function normalizeItems(rawItems) {
@@ -1700,7 +1722,7 @@ function mergeSuperscriptLines(lines) {
   return result;
 }
 function mergeLineSimple(items) {
-  if (items.length <= 1) return _optionalChain([items, 'access', _7 => _7[0], 'optionalAccess', _8 => _8.text]) || "";
+  if (items.length <= 1) return _optionalChain([items, 'access', _10 => _10[0], 'optionalAccess', _11 => _11.text]) || "";
   const sorted = [...items].sort((a, b) => a.x - b.x);
   const isEvenSpaced = detectEvenSpacedItems(sorted);
   let result = sorted[0].text;
@@ -1749,7 +1771,7 @@ function computeMedianFontSizeFromFreq(freq) {
 }
 function detectHeadings(blocks, medianFontSize) {
   for (const block of blocks) {
-    if (block.type !== "paragraph" || !block.text || !_optionalChain([block, 'access', _9 => _9.style, 'optionalAccess', _10 => _10.fontSize])) continue;
+    if (block.type !== "paragraph" || !block.text || !_optionalChain([block, 'access', _12 => _12.style, 'optionalAccess', _13 => _13.fontSize])) continue;
     const text = block.text.trim();
     if (text.length === 0 || text.length > 200) continue;
     if (/^\d+$/.test(text)) continue;
@@ -1810,7 +1832,7 @@ function detectMarkerHeadings(blocks) {
       block.level = 4;
       continue;
     }
-    if (/^[가-힣]{2,6}$/.test(text) && _optionalChain([block, 'access', _11 => _11.style, 'optionalAccess', _12 => _12.fontSize])) {
+    if (/^[가-힣]{2,6}$/.test(text) && _optionalChain([block, 'access', _14 => _14.style, 'optionalAccess', _15 => _15.fontSize])) {
       const prev = blocks[i - 1];
       const next = blocks[i + 1];
       const prevIsStructural = !prev || prev.type === "table" || prev.type === "heading" || prev.type === "separator";
@@ -2065,7 +2087,7 @@ function removeHeaderFooterBlocks(blocks, pageHeights, warnings) {
   const bottomEntries = [];
   for (let bi = 0; bi < blocks.length; bi++) {
     const b = blocks[bi];
-    if (!b.bbox || !b.pageNumber || !_optionalChain([b, 'access', _13 => _13.text, 'optionalAccess', _14 => _14.trim, 'call', _15 => _15()])) continue;
+    if (!b.bbox || !b.pageNumber || !_optionalChain([b, 'access', _16 => _16.text, 'optionalAccess', _17 => _17.trim, 'call', _18 => _18()])) continue;
     const ph = pageHeights.get(b.bbox.page) || pageHeights.get(b.pageNumber);
     if (!ph) continue;
     const blockTop = ph - (b.bbox.y + b.bbox.height);
@@ -2088,7 +2110,7 @@ function removeHeaderFooterBlocks(blocks, pageHeights, warnings) {
     }
     const repeatedPatterns = /* @__PURE__ */ new Set();
     for (const [p, count] of patternCount) {
-      if (count >= MIN_REPEAT && (_nullishCoalesce(_optionalChain([patternPages, 'access', _16 => _16.get, 'call', _17 => _17(p), 'optionalAccess', _18 => _18.size]), () => ( 0))) >= MIN_REPEAT) {
+      if (count >= MIN_REPEAT && (_nullishCoalesce(_optionalChain([patternPages, 'access', _19 => _19.get, 'call', _20 => _20(p), 'optionalAccess', _21 => _21.size]), () => ( 0))) >= MIN_REPEAT) {
         repeatedPatterns.add(p);
       }
     }
@@ -2102,7 +2124,7 @@ function removeHeaderFooterBlocks(blocks, pageHeights, warnings) {
         if (bi === e.blockIdx) continue;
         const o = blocks[bi];
         if (!o.bbox || o.bbox.page !== cb.page) continue;
-        const oNorm = _optionalChain([o, 'access', _19 => _19.text, 'optionalAccess', _20 => _20.trim, 'call', _21 => _21(), 'access', _22 => _22.replace, 'call', _23 => _23(/\d+/g, "#")]);
+        const oNorm = _optionalChain([o, 'access', _22 => _22.text, 'optionalAccess', _23 => _23.trim, 'call', _24 => _24(), 'access', _25 => _25.replace, 'call', _26 => _26(/\d+/g, "#")]);
         if (oNorm && repeatedPatterns.has(oNorm)) continue;
         const overlap2 = Math.min(cb.y + cb.height, o.bbox.y + o.bbox.height) - Math.max(cb.y, o.bbox.y);
         if (overlap2 >= cb.height * 0.5) {
@@ -2173,12 +2195,12 @@ function isTwoColumnProse(r) {
   if (t.rows < 8) return false;
   const dense = [];
   for (let c = 0; c < t.cols; c++) {
-    const filled = t.cells.filter((row) => _optionalChain([row, 'access', _24 => _24[c], 'optionalAccess', _25 => _25.text, 'access', _26 => _26.trim, 'call', _27 => _27()])).length;
+    const filled = t.cells.filter((row) => _optionalChain([row, 'access', _27 => _27[c], 'optionalAccess', _28 => _28.text, 'access', _29 => _29.trim, 'call', _30 => _30()])).length;
     if (filled / t.rows >= 0.3) dense.push(c);
   }
   if (dense.length !== 2) return false;
   for (const c of dense) {
-    const lens = t.cells.map((row) => _nullishCoalesce(_optionalChain([row, 'access', _28 => _28[c], 'optionalAccess', _29 => _29.text, 'access', _30 => _30.replace, 'call', _31 => _31(/\s+/g, ""), 'access', _32 => _32.length]), () => ( 0))).filter((n) => n > 0);
+    const lens = t.cells.map((row) => _nullishCoalesce(_optionalChain([row, 'access', _31 => _31[c], 'optionalAccess', _32 => _32.text, 'access', _33 => _33.replace, 'call', _34 => _34(/\s+/g, ""), 'access', _35 => _35.length]), () => ( 0))).filter((n) => n > 0);
     const avg = lens.reduce((s, v) => s + v, 0) / (lens.length || 1);
     if (avg < 12) return false;
   }
@@ -2655,7 +2677,7 @@ function buildClusterTable(rows, columns, pageNum) {
     const nonEmptyCols = cells[r].filter((c) => c.text.trim()).length;
     if (nonEmptyCols !== 1) continue;
     if (cells[r][0].text.trim() !== "") continue;
-    const contentText = _optionalChain([cells, 'access', _33 => _33[r], 'access', _34 => _34.find, 'call', _35 => _35((c) => c.text.trim()), 'optionalAccess', _36 => _36.text, 'access', _37 => _37.trim, 'call', _38 => _38()]) || "";
+    const contentText = _optionalChain([cells, 'access', _36 => _36[r], 'access', _37 => _37.find, 'call', _38 => _38((c) => c.text.trim()), 'optionalAccess', _39 => _39.text, 'access', _40 => _40.trim, 'call', _41 => _41()]) || "";
     if (/^[○●▶\-·]/.test(contentText)) continue;
     for (let pr = r - 1; pr >= 0; pr--) {
       if (cells[pr].some((c) => c.text.trim())) {
@@ -3020,14 +3042,14 @@ function isProseSpread(items) {
   for (let i = 1; i < sorted.length; i++) {
     gaps.push(sorted[i].x - (sorted[i - 1].x + sorted[i - 1].w));
   }
-  const maxGap = _chunk6MF7JRG3cjs.safeMax.call(void 0, gaps);
+  const maxGap = _chunk62IU4NV3cjs.safeMax.call(void 0, gaps);
   const avgLen = items.reduce((s, i) => s + i.text.length, 0) / items.length;
   return maxGap < 40 && avgLen < 5;
 }
 function detectColumns(yLines) {
   const allItems = yLines.flat();
   if (allItems.length === 0) return null;
-  const pageWidth = _chunk6MF7JRG3cjs.safeMax.call(void 0, allItems.map((i) => i.x + i.w)) - _chunk6MF7JRG3cjs.safeMin.call(void 0, allItems.map((i) => i.x));
+  const pageWidth = _chunk62IU4NV3cjs.safeMax.call(void 0, allItems.map((i) => i.x + i.w)) - _chunk62IU4NV3cjs.safeMin.call(void 0, allItems.map((i) => i.x));
   if (pageWidth < 100) return null;
   let bigoLineIdx = -1;
   for (let i = 0; i < yLines.length; i++) {
@@ -3463,9 +3485,47 @@ function isProseBoxGrid(grid, verticals, table) {
   if (longCells < PROSEBOX_LONG_CELL_MIN || longCells < texts.length * PROSEBOX_LONG_CELL_RATIO) return false;
   return true;
 }
+function cleanCellText(text) {
+  const stripped = text.replace(/^[\s]*[-–—]\s*\d+\s*[-–—][\s]*$/gm, "").trim();
+  return stripped.split("\n").map((line) => collapseEvenSpacing(line)).join("\n");
+}
+var FRAME_RECT_TOL = 1.5;
+function takePendingNested(pending, cellBox) {
+  const out = [];
+  for (let i = pending.length - 1; i >= 0; i--) {
+    const p = pending[i].parent;
+    if (Math.abs(p.x1 - cellBox.x1) <= FRAME_RECT_TOL && Math.abs(p.x2 - cellBox.x2) <= FRAME_RECT_TOL && Math.abs(p.y1 - cellBox.y1) <= FRAME_RECT_TOL && Math.abs(p.y2 - cellBox.y2) <= FRAME_RECT_TOL) {
+      out.push(pending[i].block);
+      pending.splice(i, 1);
+    }
+  }
+  return out;
+}
+function buildFrameCellBlocks(cellItems, nested, pageNum) {
+  const tables = [...nested].sort((a, b) => b.bbox.y + b.bbox.height - (a.bbox.y + a.bbox.height));
+  const blocks = [];
+  let rest = [...cellItems];
+  const pushParagraphs = (items) => {
+    if (items.length === 0) return;
+    for (const line of cleanCellText(cellTextToString(items)).split("\n")) {
+      const t = line.trim();
+      if (t) blocks.push({ type: "paragraph", text: t, pageNumber: pageNum });
+    }
+  };
+  for (const tb of tables) {
+    const bottom = tb.bbox.y;
+    pushParagraphs(rest.filter((it) => it.y >= bottom));
+    rest = rest.filter((it) => it.y < bottom);
+    blocks.push(tb);
+  }
+  pushParagraphs(rest);
+  const text = blocks.map((b) => b.type === "table" && b.table ? b.table.cells.flat().map((c) => c.text).filter(Boolean).join("\n") : _nullishCoalesce(b.text, () => ( ""))).filter(Boolean).join("\n");
+  return { blocks, text };
+}
 function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
   const blocks = [];
   const usedItems = /* @__PURE__ */ new Set();
+  const pendingNested = [];
   const gridArea = (g) => (g.bbox.x2 - g.bbox.x1) * (g.bbox.y2 - g.bbox.y1);
   const sortedGrids = [...grids].sort((a, b) => (b.cells ? 1 : 0) - (a.cells ? 1 : 0) || (a.cells && b.cells ? gridArea(a) - gridArea(b) : 0) || b.bbox.y2 - a.bbox.y2);
   for (const grid of sortedGrids) {
@@ -3505,13 +3565,18 @@ function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
       { length: numRows },
       () => Array.from({ length: numCols }, () => ({ text: "", colSpan: 1, rowSpan: 1 }))
     );
+    let nestedAttached = false;
     for (const cell of cells) {
       const cellItems = cellTextMap.get(cell) || [];
-      let text = cellTextToString(cellItems);
-      text = text.replace(/^[\s]*[-–—]\s*\d+\s*[-–—][\s]*$/gm, "").trim();
-      text = text.split("\n").map((line) => collapseEvenSpacing(line)).join("\n");
+      const nested = grid.cells ? takePendingNested(pendingNested, cell.bbox) : [];
+      if (nested.length > 0) {
+        nestedAttached = true;
+        const built = buildFrameCellBlocks(cellItems, nested, pageNum);
+        irGrid[cell.row][cell.col] = { text: built.text, colSpan: cell.colSpan, rowSpan: cell.rowSpan, blocks: built.blocks };
+        continue;
+      }
       irGrid[cell.row][cell.col] = {
-        text,
+        text: cleanCellText(cellTextToString(cellItems)),
         colSpan: cell.colSpan,
         rowSpan: cell.rowSpan
       };
@@ -3519,18 +3584,11 @@ function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
     let finalGrid = irGrid;
     let finalRows = numRows;
     let rebuiltUsed = false;
-    if (numRows <= 2 && numCols >= 3) {
+    if (numRows <= 2 && numCols >= 3 && !nestedAttached) {
       const rebuilt = normalizeUndersegmentedTable(irGrid, grid.colXs, textItems);
       if (rebuilt) {
         rebuiltUsed = true;
-        finalGrid = rebuilt.map((row) => row.map((rawText) => {
-          const cleaned = rawText.replace(/^[\s]*[-–—]\s*\d+\s*[-–—][\s]*$/gm, "").trim();
-          return {
-            text: cleaned.split("\n").map((line) => collapseEvenSpacing(line)).join("\n"),
-            colSpan: 1,
-            rowSpan: 1
-          };
-        }));
+        finalGrid = rebuilt.map((row) => row.map((rawText) => ({ text: cleanCellText(rawText), colSpan: 1, rowSpan: 1 })));
         finalRows = finalGrid.length;
       }
     }
@@ -3547,6 +3605,11 @@ function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
     };
     const hasContent = finalGrid.some((row) => row.some((cell) => cell.text.trim() !== ""));
     if (!hasContent) continue;
+    if (grid.clipParent) {
+      const nb = { page: pageNum, x: grid.bbox.x1, y: grid.bbox.y1, width: grid.bbox.x2 - grid.bbox.x1, height: grid.bbox.y2 - grid.bbox.y1 };
+      pendingNested.push({ parent: grid.clipParent, block: { type: "table", table: irTable, pageNumber: pageNum, bbox: nb } });
+      continue;
+    }
     if (!grid.cells && isProseBoxGrid(grid, verticals, irTable)) {
       for (const it of tableItems) usedItems.delete(it);
       continue;
@@ -3568,6 +3631,7 @@ function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
     }
     blocks.push({ type: "table", table: irTable, pageNumber: pageNum, bbox: tableBbox });
   }
+  for (const p of pendingNested) blocks.push(p.block);
   let remaining = items.filter((i) => !usedItems.has(i));
   const groupSizes = [];
   let finalTextBlocks = [];
@@ -3610,7 +3674,7 @@ function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
     if (gutterX !== null) {
       const gx = gutterX;
       const allY = remaining.map((i) => i.y);
-      const pageH = _chunk6MF7JRG3cjs.safeMax.call(void 0, allY) - _chunk6MF7JRG3cjs.safeMin.call(void 0, allY);
+      const pageH = _chunk62IU4NV3cjs.safeMax.call(void 0, allY) - _chunk62IU4NV3cjs.safeMin.call(void 0, allY);
       const gapThreshold = Math.max(15, pageH * 0.03);
       const sides = [
         remaining.filter((i) => i.x + i.w <= gx),
@@ -3630,7 +3694,7 @@ function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
       finalTextBlocks = detectListBlocks(textBlocks);
     } else {
       const allY = remaining.map((i) => i.y);
-      const pageH = _chunk6MF7JRG3cjs.safeMax.call(void 0, allY) - _chunk6MF7JRG3cjs.safeMin.call(void 0, allY);
+      const pageH = _chunk62IU4NV3cjs.safeMax.call(void 0, allY) - _chunk62IU4NV3cjs.safeMin.call(void 0, allY);
       const groups = xyCutOrder(remaining, Math.max(15, pageH * 0.03));
       const textBlocks = [];
       for (const group of groups) {
@@ -3838,7 +3902,7 @@ function extractPageBlocksFallback(items, pageNum, fullPage = false, detectTable
       blocks.push({ type: "paragraph", text: tableText, pageNumber: pageNum, bbox, style: dominantStyle(items) });
     } else {
       const allY = items.map((i) => i.y);
-      const pageHeight = _chunk6MF7JRG3cjs.safeMax.call(void 0, allY) - _chunk6MF7JRG3cjs.safeMin.call(void 0, allY);
+      const pageHeight = _chunk62IU4NV3cjs.safeMax.call(void 0, allY) - _chunk62IU4NV3cjs.safeMin.call(void 0, allY);
       const gapThreshold = Math.max(15, pageHeight * 0.03);
       const orderedGroups = proseCutX !== null ? splitTwoColumnProse(items, proseCutX) : xyCutOrder(items, gapThreshold);
       for (const group of orderedGroups) {
@@ -3877,4 +3941,4 @@ function extractPageBlocksFallback(items, pageNum, fullPage = false, detectTable
 
 
 exports.extractImageRegions = extractImageRegions; exports.filterHiddenText = filterHiddenText; exports.collapseEvenSpacing = collapseEvenSpacing; exports.normalizeItems = normalizeItems; exports.computeMedianFontSizeFromFreq = computeMedianFontSizeFromFreq; exports.detectHeadings = detectHeadings; exports.detectMarkerHeadings = detectMarkerHeadings; exports.detectTableCaptions = detectTableCaptions; exports.detectKoreanListBlocks = detectKoreanListBlocks; exports.removeHeaderFooterBlocks = removeHeaderFooterBlocks; exports.extractPageBlocksWithLines = extractPageBlocksWithLines; exports.mergeCrossPageTables = mergeCrossPageTables;
-//# sourceMappingURL=chunk-MXZXZTKW.cjs.map
+//# sourceMappingURL=chunk-ONNAHSPV.cjs.map