Merge pull request #16186 from Snuffleupagus/issue-16176

Support multi-byte ToUnicode entries, when using predefined CMaps (issue 16176)

src/core/evaluator.js

@@ -3632,7 +3632,8 @@ class PartialEvaluator {
         fetchBuiltInCMap: this._fetchBuiltInCMapBound,
         useCMap: null,
       });
-      const toUnicode = [];
+      const toUnicode = [],
+        buf = [];
       properties.cMap.forEach(function (charcode, cid) {
         if (cid > 0xffff) {
           throw new FormatError("Max size of CID is 65,535");
@@ -3641,9 +3642,12 @@ class PartialEvaluator {
         // obtained in step (d), producing a Unicode value.
         const ucs2 = ucs2CMap.lookup(cid);
         if (ucs2) {
-          toUnicode[charcode] = String.fromCharCode(
-            (ucs2.charCodeAt(0) << 8) + ucs2.charCodeAt(1)
-          );
+          buf.length = 0;
+          // Support multi-byte entries (fixes issue16176.pdf).
+          for (let i = 0, ii = ucs2.length; i < ii; i += 2) {
+            buf.push((ucs2.charCodeAt(i) << 8) + ucs2.charCodeAt(i + 1));
+          }
+          toUnicode[charcode] = String.fromCharCode(...buf);
         }
       });
       return new ToUnicodeMap(toUnicode);

test/pdfs/.gitignore

@@ -118,6 +118,7 @@
 !issue6127.pdf
 !issue7891_bc0.pdf
 !issue11242_reduced.pdf
+!issue16176.pdf
 !issue11279.pdf
 !issue11362.pdf
 !issue13325_reduced.pdf

test/unit/api_spec.js

@@ -29,6 +29,7 @@ import {
 } from "../../src/shared/util.js";
 import {
   buildGetDocumentParams,
+  CMAP_URL,
   DefaultFileReaderFactory,
   TEST_PDFS_PATH,
 } from "./test_utils.js";
@@ -2623,6 +2624,23 @@ Caron Broadcasting, Inc., an Ohio corporation (“Lessee”).`)
       await loadingTask.destroy();
     });
 
+    it("gets text content with multi-byte entries, using predefined CMaps (issue 16176)", async function () {
+      const loadingTask = getDocument(
+        buildGetDocumentParams("issue16176.pdf", {
+          cMapUrl: CMAP_URL,
+          useWorkerFetch: false,
+        })
+      );
+      const pdfDoc = await loadingTask.promise;
+      const pdfPage = await pdfDoc.getPage(1);
+      const { items } = await pdfPage.getTextContent();
+      const text = mergeText(items);
+
+      expect(text).toEqual("𠮷");
+
+      await loadingTask.destroy();
+    });
+
     it("gets empty structure tree", async function () {
       const tree = await page.getStructTree();