From e00ef49560bb5bd4349d017887c7cd5a2d0ba38e Mon Sep 17 00:00:00 2001
From: "jan.nijtmans" <nijtmans@users.sourceforge.net>
Date: Fri, 15 Nov 2019 09:41:03 +0000
Subject: Protect additional Tcl_UtfToUniChar() call, for the case when not
 enough bytes are available in the buffer any more. Add additional test-cases
 for those situations (upper surrogate followed by somthing other than lower
 surrogate)

---
 generic/tclEncoding.c |  2 +-
 tests/encoding.test   | 24 +++++++++++++++++++++---
 2 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/generic/tclEncoding.c b/generic/tclEncoding.c
index 00b97f5..9e1d262 100644
--- a/generic/tclEncoding.c
+++ b/generic/tclEncoding.c
@@ -2365,7 +2365,7 @@ UtfToUtfProc(
 	    if ((*chPtr & 0xFC00) == 0xD800) {
 		/* A high surrogate character is detected, handle especially */
 		Tcl_UniChar low = *chPtr;
-		size_t len = Tcl_UtfToUniChar(src, &low);
+		size_t len = (src <= srcEnd-3) ? Tcl_UtfToUniChar(src, &low) : 0;
 		if ((low & 0xFC00) != 0xDC00) {
 			*dst++ = (char) (((*chPtr >> 12) | 0xE0) & 0xEF);
 			*dst++ = (char) (((*chPtr >> 6) | 0x80) & 0xBF);
diff --git a/tests/encoding.test b/tests/encoding.test
index cf27190..36fcff6 100644
--- a/tests/encoding.test
+++ b/tests/encoding.test
@@ -339,11 +339,29 @@ test encoding-15.5 {UtfToUtfProc emoji character input} {
     list [string length $x] [string length $y] $y
 } "4 2 \uD83D\uDE02"
 test encoding-15.6 {UtfToUtfProc emoji character output} {
-    set x \uD83D\uDE02
-    set y [encoding convertto utf-8 \uD83D\uDE02]
+    set x \uDE02\uD83D\uDE02\uD83D
+    set y [encoding convertto utf-8 \uDE02\uD83D\uDE02\uD83D]
     binary scan $y H* z
     list [string length $x] [string length $y] $z
-} {2 4 f09f9882}
+} {4 10 edb882f09f9882eda0bd}
+test encoding-15.7 {UtfToUtfProc emoji character output} {
+    set x \uDE02\uD83D\uD83D
+    set y [encoding convertto utf-8 \uDE02\uD83D\uD83D]
+    binary scan $y H* z
+    list [string length $x] [string length $y] $z
+} {3 9 edb882eda0bdeda0bd}
+test encoding-15.8 {UtfToUtfProc emoji character output} {
+    set x \uDE02\uD83D\xE9
+    set y [encoding convertto utf-8 \uDE02\uD83D\xE9]
+    binary scan $y H* z
+    list [string length $x] [string length $y] $z
+} {3 8 edb882eda0bdc3a9}
+test encoding-15.9 {UtfToUtfProc emoji character output} {
+    set x \uDE02\uD83DX
+    set y [encoding convertto utf-8 \uDE02\uD83DX]
+    binary scan $y H* z
+    list [string length $x] [string length $y] $z
+} {3 7 edb882eda0bd58}
 
 test encoding-16.1 {UnicodeToUtfProc} {
     set val [encoding convertfrom unicode NN]
-- 
cgit v0.12