From 7f9dac9015a1391ad4c92c31d981cc26ab367c06 Mon Sep 17 00:00:00 2001
From: Zhao Dengfeng <2370187+zhaodengfeng@users.noreply.github.com>
Date: Thu, 29 Oct 2026 20:25:42 +0800
Subject: [PATCH] Don't fail getTextContent when the first text in a stream has
 a zero font size (issue 22381)

The text state is cloned before any `Tf` is applied, so its font is `null` and its font size is `0`. When the first text shown uses a zero font size, the font size comparison in `buildTextContentItem` is equal, and the font name comparison then reads `name` from that `null` font. With the default `ignoreErrors`, all of the text on the page was dropped.
---
 src/core/evaluator.js    |   4 ++--
 test/pdfs/.gitignore     |   1 +
 test/pdfs/issue22381.pdf | Bin 0 -> 613 bytes
 test/test_manifest.json  |   7 +++++++
 test/unit/api_spec.js    |  15 +++++++++++++++
 5 files changed, 25 insertions(+), 2 deletions(-)
 create mode 100644 test/pdfs/issue22381.pdf

diff --git a/src/core/evaluator.js b/src/core/evaluator.js
index d974b75a593c0..b5e2dd96cdbef 100644
--- a/src/core/evaluator.js
+++ b/src/core/evaluator.js
@@ -2982,8 +2982,8 @@ class PartialEvaluator {
         currentTextState !== textState &&
         (currentTextState.fontSize !== textState.fontSize ||
           (currentTextState.fontName !== textState.fontName &&
-            (currentTextState.font.name !== textState.font.name ||
-              currentTextState.font.vertical !== textState.font.vertical)))
+            (currentTextState.font?.name !== textState.font?.name ||
+              currentTextState.font?.vertical !== textState.font?.vertical)))
       ) {
         flushTextContentItem();
         currentTextState = textState.clone();
diff --git a/test/pdfs/.gitignore b/test/pdfs/.gitignore
index 6db7cfe8b5073..3313b4d784e02 100644
--- a/test/pdfs/.gitignore
+++ b/test/pdfs/.gitignore
@@ -189,6 +189,7 @@
 !bug1337429.pdf
 !bug1606566.pdf
 !bug2086012.pdf
+!issue22381.pdf
 !issue5564_reduced.pdf
 !canvas.pdf
 !bug1132849.pdf
diff --git a/test/pdfs/issue22381.pdf b/test/pdfs/issue22381.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..b395bb90d435a4acb4c65bcb5988564a822aa099
GIT binary patch
literal 613
zcmZWnO-}+b5WV+T%!LFG+AaGLNC*cMVxmR@8!yB|VFxH#w`N-m`s<z94?t=5uzmAp
z-s@{Svxn)ex{HK@0PJcnhC`6)_a}n9*Iws#3v#Bn=%7c0IgCaEi;Ri9|JzXW<dex9
zEJLQh(-eCsJgAFCu3R^=Oo>FTb>s_XMvv_gmI3j@AOICbG;fa19?CTtVy9C}9o3J5
zm=#HfKawpxO>+=6XltAIt@<+-Tfc*3!0wzcQGbYWN>e$EbOYU;$ucZ%Al<_xZA-m~
zYheNjkuvDM8D~~Gg2$t8sgzMOuStu|wX$`*m36r|tbCO=g9_wWJFF{@n17)+Ykd*j
zvBZrCAQ8c7K@!Jd3>!FO1E{J{oUyong(*dzcwO*%5uN-W7Sb_Km@l=-v8-On!u%j4
ZbR*XC1vjt5d0qO-XA}ma)0w<Z#V-n%pwR#T

literal 0
HcmV?d00001

diff --git a/test/test_manifest.json b/test/test_manifest.json
index b52f3a181c098..185896af37fb7 100644
--- a/test/test_manifest.json
+++ b/test/test_manifest.json
@@ -14647,6 +14647,13 @@
     "rounds": 1,
     "type": "eq"
   },
+  {
+    "id": "issue22381",
+    "file": "pdfs/issue22381.pdf",
+    "md5": "3f5fd4d65e246557510fefa4bd6cebcb",
+    "rounds": 1,
+    "type": "text"
+  },
   {
     "id": "devicen-none-zero-domain",
     "file": "pdfs/devicen-none-zero-domain.pdf",
diff --git a/test/unit/api_spec.js b/test/unit/api_spec.js
index a9b4cd9a28cf1..c9660b3a6cd6a 100644
--- a/test/unit/api_spec.js
+++ b/test/unit/api_spec.js
@@ -5600,6 +5600,21 @@ have written that much by now. So, here’s to squashing bugs.`);
       await loadingTask.destroy();
     });
 
+    it("gets text content when the first text has a zero font size (issue 22381)", async function () {
+      const loadingTask = getDocument(
+        buildGetDocumentParams("issue22381.pdf", { stopAtErrors: true })
+      );
+      const pdfDoc = await loadingTask.promise;
+      const pdfPage = await pdfDoc.getPage(1);
+      const { items } = await pdfPage.getTextContent({
+        disableNormalization: true,
+      });
+      const text = mergeText(items);
+      expect(text).toEqual("hiddenVisible text");
+
+      await loadingTask.destroy();
+    });
+
     it("gets empty structure tree", async function () {
       const tree = await page.getStructTree();