Merge pull request #17558 from calixteman/bug1669097

Print correctly documents containing chars with an unicode greater than 0xFFFF (bug 1669097)

src/core/annotation.js

@@ -3826,7 +3826,7 @@ class FreeTextAnnotation extends MarkupAnnotation {
           fontColor,
           strokeAlpha
         );
-        this._streams.push(this.appearance, FakeUnicodeFont.toUnicodeStream);
+        this._streams.push(this.appearance);
       } else {
         warn(
           "FreeTextAnnotation: OffscreenCanvas is not supported, annotation may not render correctly."

src/core/core_utils.js

@@ -386,6 +386,17 @@ const XMLEntities = {
   /* ' */ 0x27: "'",
 };
 
+function* codePointIter(str) {
+  for (let i = 0, ii = str.length; i < ii; i++) {
+    const char = str.codePointAt(i);
+    if (char > 0xd7ff && (char < 0xe000 || char > 0xfffd)) {
+      // char is represented by two u16
+      i++;
+    }
+    yield char;
+  }
+}
+
 function encodeToXmlString(str) {
   const buffer = [];
   let start = 0;
@@ -602,6 +613,7 @@ function getRotationMatrix(rotation, width, height) {
 
 export {
   arrayBuffersToBytes,
+  codePointIter,
   collectActions,
   encodeToXmlString,
   escapePDFName,

src/core/default_appearance.js

@@ -13,13 +13,14 @@
  * limitations under the License.
  */
 
-import { Dict, Name } from "./primitives.js";
 import {
+  codePointIter,
   escapePDFName,
   getRotationMatrix,
   numberToString,
   stringToUTF16HexString,
 } from "./core_utils.js";
+import { Dict, Name } from "./primitives.js";
 import {
   LINE_DESCENT_FACTOR,
   LINE_FACTOR,
@@ -251,35 +252,6 @@ class FakeUnicodeFont {
     );
   }
 
-  get toUnicodeRef() {
-    if (!FakeUnicodeFont._toUnicodeRef) {
-      const toUnicode = `/CIDInit /ProcSet findresource begin
-12 dict begin
-begincmap
-/CIDSystemInfo
-<< /Registry (Adobe)
-/Ordering (UCS) /Supplement 0 >> def
-/CMapName /Adobe-Identity-UCS def
-/CMapType 2 def
-1 begincodespacerange
-<0000> <FFFF>
-endcodespacerange
-1 beginbfrange
-<0000> <FFFF> <0000>
-endbfrange
-endcmap CMapName currentdict /CMap defineresource pop end end`;
-      const toUnicodeStream = (FakeUnicodeFont.toUnicodeStream =
-        new StringStream(toUnicode));
-      const toUnicodeDict = new Dict(this.xref);
-      toUnicodeStream.dict = toUnicodeDict;
-      toUnicodeDict.set("Length", toUnicode.length);
-      FakeUnicodeFont._toUnicodeRef =
-        this.xref.getNewPersistentRef(toUnicodeStream);
-    }
-
-    return FakeUnicodeFont._toUnicodeRef;
-  }
-
   get fontDescriptorRef() {
     if (!FakeUnicodeFont._fontDescriptorRef) {
       const fontDescriptor = new Dict(this.xref);
@@ -350,7 +322,7 @@ endcmap CMapName currentdict /CMap defineresource pop end end`;
     baseFont.set("Subtype", Name.get("Type0"));
     baseFont.set("Encoding", Name.get("Identity-H"));
     baseFont.set("DescendantFonts", [this.descendantFontRef]);
-    baseFont.set("ToUnicode", this.toUnicodeRef);
+    baseFont.set("ToUnicode", Name.get("Identity-H"));
 
     return this.xref.getNewPersistentRef(baseFont);
   }
@@ -420,8 +392,8 @@ endcmap CMapName currentdict /CMap defineresource pop end end`;
       // languages, like arabic, it'd be wrong because of ligatures.
       const lineWidth = ctx.measureText(line).width;
       maxWidth = Math.max(maxWidth, lineWidth);
-      for (const char of line.split("")) {
+      for (const code of codePointIter(line)) {
-        const code = char.charCodeAt(0);
+        const char = String.fromCodePoint(code);
         let width = this.widths.get(code);
         if (width === undefined) {
           const metrics = ctx.measureText(char);

test/pdfs/.gitignore

@@ -627,3 +627,4 @@
 !file_pdfjs_form.pdf
 !issue17492.pdf
 !issue17540.pdf
+!bug1669097.pdf

test/test_manifest.json

@@ -9650,5 +9650,30 @@
         "structTreeParentId": null
       }
     }
+  },
+  {
+    "id": "bug1669097",
+    "file": "pdfs/bug1669097.pdf",
+    "md5": "561b3abac1fe49e1c9cd265cbf4a456e",
+    "rounds": 1,
+    "type": "eq",
+    "save": true,
+    "print": true,
+    "annotationStorage": {
+      "24R": {
+        "value": "😇👽🖖"
+      },
+      "pdfjs_internal_editor_0": {
+        "annotationType": 3,
+        "color": [0, 0, 0],
+        "fontSize": 10,
+        "value": "😇😇😇😇😇😇👽👽👽👽👽👽🖖",
+        "pageIndex": 0,
+        "rect": [267, 638, 452, 658],
+        "rotation": 0,
+        "structTreeParentId": null,
+        "id": null
+      }
+    }
   }
 ]