kordoc 4.21.19 → 4.21.21

package/dist/{parser-JVUZ7EEV.js → parser-KFCY3NIN.js}

@@ -45,19 +45,21 @@ import {
   removeHeaderFooterBlocks,
   removeSideTabs,
   restoreImageBullets,
+  runningLinePatterns,
   splitPageBreakWraps,
   startsNewItem,
+  stripAbsorbedRunningLines,
   trimTrailingEmptyTableCols,
   wrapJoiner
-} from "./chunk-Y5OSPKBU.js";
+} from "./chunk-KBV4WFZ6.js";
 import {
   ocrModelsCached
-} from "./chunk-OWLDTF5Q.js";
-import "./chunk-4VM6BT4W.js";
-import "./chunk-SM42XSI3.js";
+} from "./chunk-WGMYVRSA.js";
+import "./chunk-3TCF5HPP.js";
+import "./chunk-CY4QM2KQ.js";
 import {
   blocksToPages
-} from "./chunk-TC6DQLYN.js";
+} from "./chunk-2ZGP7A5T.js";
 import {
   encodePng
 } from "./chunk-7FCNOKOR.js";
@@ -69,7 +71,7 @@ import {
   escapeLiteralTags,
   stripScriptTags,
   unframeLayoutTables
-} from "./chunk-X6WXU4LA.js";
+} from "./chunk-PTF4NCFM.js";
 import {
   hasRequestedPagesAfter,
   parsePageRange
@@ -79,11 +81,11 @@ import {
   assignTableIds,
   blocksToMarkdown,
   escapeLiteralDollar
-} from "./chunk-RYRUQ73F.js";
+} from "./chunk-RBRHO7WL.js";
 import {
   KordocError,
   sanitizeHref
-} from "./chunk-Z2OOYQ4I.js";
+} from "./chunk-FZMGF6XU.js";
 
 // src/pdf/image-extract.ts
 import { createHash } from "crypto";
@@ -3220,7 +3222,7 @@ function mergeLinkRuns(markdown) {
 
 // src/pdf/formula-ocr.ts
 async function applyFormulaOcr(buffer, blocks, pageFilter, effectivePageCount, warnings, _onProgress) {
-  const formulaMod = await import("./formula-Z5UPXBQM.js");
+  const formulaMod = await import("./formula-P7M7IH2L.js");
   const { FormulaPipeline, ensureFormulaModels } = formulaMod;
   await ensureFormulaModels((p) => {
     if (p.phase === "download" && p.total) {
@@ -3643,7 +3645,7 @@ async function parsePdfDocument(buffer, options) {
       }
       if (targets.size > 0) {
         try {
-          const { runPdfOcr } = await import("./pdf-ocr-4MZELYCR.js");
+          const { runPdfOcr } = await import("./pdf-ocr-ZZ6AJL6Z.js");
           const mode = typeof options?.ocr === "function" ? options.ocr : "builtin";
           const regionPages = new Map([...ocrRegions].filter(([p]) => options?.ocr !== "force" && !isImageBased && !pageQuality.find((q) => q.page === p)?.needsOcr));
           const ocrPageBlocks = await runPdfOcr(ocrBuffer, targets, mode, warnings, options?.onProgress, options?.tables !== false, vectorPageOps, regionPages);
@@ -3712,9 +3714,15 @@ async function parsePdfDocument(buffer, options) {
     }
     if (options?.removeHeaderFooter !== false && (parsedPageCount >= 3 || headerContext.length)) {
       const removed = removeHeaderFooterBlocks(blocks, pageHeights, warnings, noteMarks, false, headerContext);
+      const running = runningLinePatterns(removed.map((ri) => blocks[ri]), pageHeights);
       for (let ri = removed.length - 1; ri >= 0; ri--) {
         blocks.splice(removed[ri], 1);
       }
+      const stripped = stripAbsorbedRunningLines(blocks, pageHeights, running);
+      if (stripped !== blocks) {
+        blocks.length = 0;
+        blocks.push(...stripped);
+      }
       const kept = removeSideTabs(blocks, pageWidths);
       if (kept !== blocks) {
         blocks.length = 0;
@@ -3883,4 +3891,4 @@ export {
   parsePdfDocument,
   removeHeaderFooterBlocks
 };
-//# sourceMappingURL=parser-JVUZ7EEV.js.map
+//# sourceMappingURL=parser-KFCY3NIN.js.map