From f9c734a1854a7699cc2827ee6b061a14e49baba2 Mon Sep 17 00:00:00 2001
From: Jay Hemnani <193022578+jayhemnani9910@users.noreply.github.com>
Date: Sun, 8 Nov 2026 06:02:04 +0530
Subject: [PATCH] Look up the space glyph in ToUnicode by string in
 Font.spaceWidth

ToUnicodeMap.charCodeOf expects a string, but spaceWidth passed the
code point as a number, so the lookup never matched and the width came
from widths[32] or defaultWidth instead of the font's space glyph. With
a wrong space width the word-space threshold is too small, and
letter-spaced text such as LibreOffice headings comes out as "T R A C K".
---
 src/core/fonts.js                   |   7 +++++--
 test/pdfs/.gitignore                |   1 +
 test/pdfs/tounicode_space_width.pdf | Bin 0 -> 1261 bytes
 test/unit/api_spec.js               |  15 +++++++++++++++
 4 files changed, 21 insertions(+), 2 deletions(-)
 create mode 100644 test/pdfs/tounicode_space_width.pdf

diff --git a/src/core/fonts.js b/src/core/fonts.js
index 81f06c9b09f15..4b70c775601f4 100644
--- a/src/core/fonts.js
+++ b/src/core/fonts.js
@@ -3602,8 +3602,11 @@ class Font {
         }
       }
       // ... via toUnicode map
-      if (!charcode && this.toUnicode) {
-        charcode = this.toUnicode.charCodeOf(glyphUnicode);
+      if (!charcode && this.toUnicode && glyphUnicode !== undefined) {
+        charcode =
+          this.toUnicode instanceof IdentityToUnicodeMap
+            ? this.toUnicode.charCodeOf(glyphUnicode)
+            : this.toUnicode.charCodeOf(String.fromCodePoint(glyphUnicode));
       }
       // setting it to unicode if negative or undefined
       if (charcode <= 0) {
diff --git a/test/pdfs/.gitignore b/test/pdfs/.gitignore
index 65b515800a631..135c13c042493 100644
--- a/test/pdfs/.gitignore
+++ b/test/pdfs/.gitignore
@@ -1062,6 +1062,7 @@
 !issue18768_text_matrix_spacing.pdf
 !issue18768_type3_wide_glyph.pdf
 !issue18768_word_spaces.pdf
+!tounicode_space_width.pdf
 !issue19954_reduced.pdf
 !issue1936_reduced.pdf
 !issue22462.pdf
diff --git a/test/pdfs/tounicode_space_width.pdf b/test/pdfs/tounicode_space_width.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..3564054f7eb452f3b1075f22b23403bb0d5262c8
GIT binary patch
literal 1261
zcmZWpOK;*Z5WeSE%%xKIz}gNZDH1{|v}oD3+X_<c0rg-Khk6B0<d{|2U%xYsp${U^
z*yDM8^Uc^7PdBrk_&$gPaH!XRqR|LgerOc1sdRE%Z-70?jcOrAg9VJokuoJ2(*CnU
zXUFbz*}|$%<}Vnf-WvzvgCm`sJ5m~}(6r0L9#pC2q<(`HNBdO7kPTyuw@_`pw*>+s
zq_Y}5{6auN88Hw2Fs*SIt%X6*9%0k|nEhUvje7;bQ-aqztK=@4<QP@}Cd?~HIAokd
zUczd`#keQZWW?hj`#}zZOoB{<Jldbzz2raPHYZ3SY2cJfID+e%;0t8S{o47SkWv7f
z$X0bHKh*ZGa=MTRTHDr5U!_I1*}V+w6xeTFx>syTDg<P43~8F3WSm_gskf%6OKmou
z$}aIq#cp&}DQpdKzQWMR&jQ#rat!PaSlh@#y?D~|`q^lVq#*smI_-~z1=jmRKNNug
zHl5$hjdnN&wl0>+L8VP;L$hJ6Hrh~6N?o|H^SWXun~0p3ht{dx+*GwMVu21A`T&>L
zIIZd%ut!@eOEq3TPnWn~?we+-b~JLF(3h&hKo7F{C3iUe-mNz;@m#w@4}Hi2i$)92
z5Mn_O(Sf0&JrP!#je-&9V;Ie5W5C9KK8w=MV_g;0Dw--XPX;97ZtX?iuLu6vAM$}$
zC7pgkBK7MSlVP4NA`KOh0;f`ny`_ugne*@oxHj0*?j}J~H{?(M^O1bBSS8+qh@U=4
zlcb+O1@F2c_){?Ny5#CiCj<^SeIM-N0lqw6bp7GEZ~D>KS*f?m`p0Ui|0&3DChMBs
Wnea%pPFm-&3!Y@r#l`L8Ecy>3>|Kih

literal 0
HcmV?d00001

diff --git a/test/unit/api_spec.js b/test/unit/api_spec.js
index 2c8bdbf6491fe..744310124a0a8 100644
--- a/test/unit/api_spec.js
+++ b/test/unit/api_spec.js
@@ -5370,6 +5370,21 @@ ABCD E`);
       await loadingTask.destroy();
     });
 
+    it("gets text content, using the ToUnicode space width for letter-spaced text", async function () {
+      const loadingTask = getDocument(
+        buildGetDocumentParams("tounicode_space_width.pdf")
+      );
+      const pdfDoc = await loadingTask.promise;
+      const pdfPage = await pdfDoc.getPage(1);
+      const { items } = await pdfPage.getTextContent({
+        disableNormalization: true,
+      });
+
+      expect(mergeText(items)).toEqual("TRACK TACK");
+
+      await loadingTask.destroy();
+    });
+
     it("gets text content, preserving various word spaces", async function () {
       const loadingTask = getDocument(
         buildGetDocumentParams("issue18768_word_spaces.pdf")