fix(opencc): CVE-2025-15536

debian/changelog

@@ -1,3 +1,9 @@
+opencc (1.1.9+ds1-1deepin1) unstable; urgency=medium
+
+  * Fix CVE-2025-15536: heap buffer overflow in UTF-8 processing
+
+ -- deepin-ci-robot <[email protected]>  Wed, 29 Apr 2026 19:47:21 +0800
+
 opencc (1.1.9+ds1-1deepin0) unstable; urgency=medium
 
   * No source change upload against GCC 12.

debian/patches/0006-CVE-2025-15536.patch

@@ -0,0 +1,85 @@
+Description: Fix heap buffer overflow in UTF-8 processing
+ Two out-of-bounds read issues were identified in OpenCC's UTF-8
+ processing logic when handling malformed or truncated UTF-8 sequences.
+ .
+ 1) MaxMatchSegmentation:
+    NextCharLength() could return a value larger than the remaining input size.
+    The previous logic subtracted this value from a size_t length counter,
+    potentially causing underflow and subsequent out-of-bounds reads.
+ .
+ 2) Conversion:
+    Similar length handling could allow reads past the end of the input buffer
+    during dictionary matching.
+ .
+ This patch fixes both issues by:
+ - Explicitly tracking the end of the input buffer
+ - Recomputing remaining length on each iteration
+ - Clamping matched character and key lengths to the remaining buffer size
+ - Preventing reads past the null terminator
+Origin: upstream, https://github.com/BYVoid/OpenCC/commit/345c9a50ab07018f1b4439776bad78a0d40778ec
+Bug: https://github.com/BYVoid/OpenCC/issues/997
+Bug-Debian: https://bugs.debian.org/1126286
+Forwarded: not-needed
+Last-Update: 2026-04-29
+
+--- opencc-1.1.9+ds1.orig/src/Conversion.cpp
++++ opencc-1.1.9+ds1/src/Conversion.cpp
+@@ -25,14 +25,30 @@ using namespace opencc;
+
+ std::string Conversion::Convert(const char* phrase) const {
+   std::ostringstream buffer;
++  // Calculate string end to prevent reading beyond null terminator
++  const char* phraseEnd = phrase;
++  while (*phraseEnd != '\0') {
++    phraseEnd++;
++  }
++
+   for (const char* pstr = phrase; *pstr != '\0';) {
+-    Optional<const DictEntry*> matched = dict->MatchPrefix(pstr);
++    size_t remainingLength = phraseEnd - pstr;
++    Optional<const DictEntry*> matched = dict->MatchPrefix(pstr, remainingLength);
+     size_t matchedLength;
+     if (matched.IsNull()) {
+       matchedLength = UTF8Util::NextCharLength(pstr);
++      // Ensure we don't read beyond the null terminator
++      if (matchedLength > remainingLength) {
++        matchedLength = remainingLength;
++      }
+       buffer << UTF8Util::FromSubstr(pstr, matchedLength);
+     } else {
+       matchedLength = matched.Get()->KeyLength();
++      // Defensive: ensure dictionary key length does not exceed remaining input
++      // (MatchPrefix should already guarantee this, but defense in depth)
++      if (matchedLength > remainingLength) {
++        matchedLength = remainingLength;
++      }
+       buffer << matched.Get()->GetDefault();
+     }
+     pstr += matchedLength;
+--- opencc-1.1.9+ds1.orig/src/MaxMatchSegmentation.cpp
++++ opencc-1.1.9+ds1/src/MaxMatchSegmentation.cpp
+@@ -26,12 +26,16 @@ SegmentsPtr MaxMatchSegmentation::Segmen
+   };
+   size_t length = text.length();
+   for (const char* pstr = text.c_str(); *pstr != '\0';) {
++    // Recompute remaining length each iteration to avoid underflow
++    size_t remainingLength = text.c_str() + text.length() - pstr;
+     const Optional<const DictEntry*>& matched = dict->MatchPrefix(pstr, length);
+     size_t matchedLength;
+     if (matched.IsNull()) {
+       matchedLength = UTF8Util::NextCharLength(pstr);
++      // Clamp matchedLength to remaining buffer size
++      if (matchedLength > remainingLength) matchedLength = remainingLength;
+       segLength += matchedLength;
+     } else {
+       clearBuffer();
+       matchedLength = matched.Get()->KeyLength();
+       segments->AddSegment(matched.Get()->Key());
+       segStart = pstr + matchedLength;
+     }
+     pstr += matchedLength;
+-    length -= matchedLength;
++    // No need to track length, we recompute remainingLength each iteration
+   }
+   clearBuffer();
+   return segments;

debian/patches/series

@@ -2,3 +2,4 @@
 0003-no-remote-images-when-reading-docs-on-disk.patch
 0004-Use-system-googletest.patch
 0005-Disable-build-in-setup.py.patch
+0006-CVE-2025-15536.patch