kordoc 3.12.0 → 3.13.0

package/dist/{parser-5FSEFR3R.js → parser-XUNGOH76.js}

@@ -4,7 +4,7 @@ import {
   HEADING_RATIO_H2,
   HEADING_RATIO_H3,
   blocksToMarkdown
-} from "./chunk-BUDM6ULT.js";
+} from "./chunk-YXOT7P5B.js";
 import {
   parsePageRange
 } from "./chunk-MOL7MDBG.js";
@@ -12,7 +12,7 @@ import {
   KordocError,
   safeMax,
   safeMin
-} from "./chunk-7Z43WEML.js";
+} from "./chunk-Q6UXVJXI.js";
 
 // src/pdf/line-extract.ts
 import { OPS } from "pdfjs-dist/legacy/build/pdf.mjs";
@@ -2721,6 +2721,54 @@ function wrapStrikethroughRuns(items) {
     arr[arr.length - 1].text = arr[arr.length - 1].text + "~~";
   }
 }
+var PROSEBOX_FULLWIDTH_MIN = 0.6;
+var PROSEBOX_LONG_CELL_CHARS = 80;
+var PROSEBOX_LONG_CELL_MIN = 3;
+var PROSEBOX_LONG_CELL_RATIO = 0.4;
+var PROSEBOX_X_TOL = 8;
+function verticalCoverageAt(verticals, x, yMin, yMax) {
+  const tol = PROSEBOX_X_TOL;
+  const spans = [];
+  for (const v of verticals) {
+    if (Math.abs(v.x1 - x) > tol) continue;
+    const lo = Math.max(v.y1, yMin), hi = Math.min(v.y2, yMax);
+    if (hi > lo) spans.push([lo, hi]);
+  }
+  if (spans.length === 0) return 0;
+  spans.sort((a, b) => a[0] - b[0]);
+  let total = 0, s = spans[0][0], e = spans[0][1];
+  for (let i = 1; i < spans.length; i++) {
+    if (spans[i][0] <= e) {
+      if (spans[i][1] > e) e = spans[i][1];
+    } else {
+      total += e - s;
+      s = spans[i][0];
+      e = spans[i][1];
+    }
+  }
+  return total + (e - s);
+}
+function isProseBoxGrid(grid, verticals, table) {
+  const numCols = grid.colXs.length - 1;
+  if (numCols < 2 || grid.rowYs.length < 3) return false;
+  const gyMax = grid.rowYs[0], gyMin = grid.rowYs[grid.rowYs.length - 1];
+  const span = gyMax - gyMin;
+  if (span <= 0) return false;
+  const interior = grid.colXs.slice(1, -1);
+  let fullWidthHeight = 0;
+  for (let r = 0; r < grid.rowYs.length - 1; r++) {
+    const top = grid.rowYs[r], bot = grid.rowYs[r + 1];
+    const h = top - bot;
+    if (h <= 0) continue;
+    const hasDivider = interior.some((cx) => verticalCoverageAt(verticals, cx, bot, top) >= h * 0.5);
+    if (!hasDivider) fullWidthHeight += h;
+  }
+  if (fullWidthHeight < span * PROSEBOX_FULLWIDTH_MIN) return false;
+  const texts = table.cells.flat().map((c) => c.text.trim()).filter(Boolean);
+  const longCells = texts.filter((s) => s.length > PROSEBOX_LONG_CELL_CHARS).length;
+  if (longCells < PROSEBOX_LONG_CELL_MIN || longCells < texts.length * PROSEBOX_LONG_CELL_RATIO) return false;
+  return true;
+}
 function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
   const blocks = [];
   const usedItems = /* @__PURE__ */ new Set();
@@ -2795,6 +2843,10 @@ function extractBlocksWithGrids(items, pageNum, grids, horizontals, verticals) {
     };
     const hasContent = finalGrid.some((row) => row.some((cell) => cell.text.trim() !== ""));
     if (!hasContent) continue;
+    if (isProseBoxGrid(grid, verticals, irTable)) {
+      for (const it of tableItems) usedItems.delete(it);
+      continue;
+    }
     const tableBbox = {
       page: pageNum,
       x: grid.bbox.x1,
@@ -3474,4 +3526,4 @@ export {
   parsePdfDocument,
   removeHeaderFooterBlocks
 };
-//# sourceMappingURL=parser-5FSEFR3R.js.map
+//# sourceMappingURL=parser-XUNGOH76.js.map