Merge 8.7

2 files changed, 27 insertions, 15 deletions

diff --git a/generic/tclEncoding.c b/generic/tclEncoding.c
index 9d896e2..07acb13 100644
--- a/generic/tclEncoding.c
+++ b/generic/tclEncoding.c
@@ -570,7 +570,7 @@ TclInitEncodingSubsystem(void)
     type.nullSize	= 1;
     type.clientData	= INT2PTR(ENCODING_UTF);
     Tcl_CreateEncoding(&type);
-    type.clientData	= INT2PTR(TCL_ENCODING_NOCOMPLAIN);
+    type.clientData	= INT2PTR(0);
     type.encodingName	= "cesu-8";
     Tcl_CreateEncoding(&type);
 
@@ -2325,13 +2325,13 @@ UtfToUtfProc(
 
 	    *dst++ = *src++;
 	} else if ((UCHAR(*src) == 0xC0) && (src + 1 < srcEnd)
-		&& (UCHAR(src[1]) == 0x80) && (flags & ENCODING_UTF) && (!(flags & ENCODING_INPUT)
+		&& (UCHAR(src[1]) == 0x80) && !(flags & TCL_ENCODING_MODIFIED) && (!(flags & ENCODING_INPUT)
 			|| ((flags & TCL_ENCODING_STRICT) == TCL_ENCODING_STRICT)
 			|| (flags & ENCODING_FAILINDEX))) {
 	    /*
 	     * If in input mode, and -strict or -failindex is specified: This is an error.
 	     */
-	    if (flags & ENCODING_INPUT) {
+	    if ((STOPONERROR) && (flags & ENCODING_INPUT)) {
 		result = TCL_CONVERT_SYNTAX;
 		break;
 	    }
@@ -2366,15 +2366,21 @@ UtfToUtfProc(
 	    }
 	    dst += Tcl_UniCharToUtf(ch, dst);
 	} else {
-	    const char *saveSrc = src;
 	    size_t len = TclUtfToUCS4(src, &ch);
-	    if ((len < 2) && (ch != 0) && (flags & ENCODING_INPUT)
-		    && (((flags & TCL_ENCODING_STRICT) == TCL_ENCODING_STRICT))) {
-		result = TCL_CONVERT_SYNTAX;
-		break;
+	    if (flags & ENCODING_INPUT) {
+		if ((len < 2) && (ch != 0)
+			&& (((flags & TCL_ENCODING_STRICT) == TCL_ENCODING_STRICT) || (flags & ENCODING_FAILINDEX))) {
+		    result = TCL_CONVERT_SYNTAX;
+		    break;
+		} else if ((ch > 0xFFFF) && !(flags & ENCODING_UTF)
+			&& (((flags & TCL_ENCODING_STRICT) == TCL_ENCODING_STRICT) || (flags & ENCODING_FAILINDEX))) {
+		    result = TCL_CONVERT_SYNTAX;
+		    break;
+		}
 	    }
+	    const char *saveSrc = src;
 	    src += len;
-	    if (!(flags & ENCODING_UTF) && (ch > 0x3FF)) {
+	    if (!(flags & ENCODING_UTF) && !(flags & ENCODING_INPUT) && (ch > 0x3FF)) {
 		if (ch > 0xFFFF) {
 		    /* CESU-8 6-byte sequence for chars > U+FFFF */
 		    ch -= 0x10000;
diff --git a/tests/encoding.test b/tests/encoding.test
index 4cf2e4c..ed41937 100644
--- a/tests/encoding.test
+++ b/tests/encoding.test
@@ -460,17 +460,20 @@ test encoding-15.26 {UtfToUtfProc CESU-8} {
     encoding convertfrom cesu-8 \xC0\x80
 } \x00
 test encoding-15.27 {UtfToUtfProc -strict CESU-8} {
-    encoding convertfrom -strict cesu-8 \xC0\x80
+    encoding convertfrom -strict cesu-8 \x00
 } \x00
-test encoding-15.28 {UtfToUtfProc -strict CESU-8} {
+test encoding-15.28 {UtfToUtfProc -strict CESU-8} -body {
     encoding convertfrom -strict cesu-8 \xC0\x80
-} \x00
+} -returnCodes 1 -result {unexpected byte sequence starting at index 0: '\xC0'}
 test encoding-15.29 {UtfToUtfProc CESU-8} {
     encoding convertto cesu-8 \x00
-} \xC0\x80
+} \x00
 test encoding-15.30 {UtfToUtfProc -strict CESU-8} {
     encoding convertto -strict cesu-8 \x00
-} \xC0\x80
+} \x00
+test encoding-15.31 {UtfToUtfProc -strict CESU-8 (bytes F0-F4 are invalid)} -body {
+    encoding convertfrom -strict cesu-8 \xF1\x86\x83\x9C
+} -returnCodes 1 -result {unexpected byte sequence starting at index 0: '\xF1'}
 
 test encoding-16.1 {Utf16ToUtfProc} -body {
     set val [encoding convertfrom utf-16 NN]
@@ -618,9 +621,12 @@ test encoding-19.3 {TableFromUtfProc} -body {
 test encoding-19.4 {TableFromUtfProc} -body {
     list [encoding convertfrom -failindex idx ascii AÁ] [set idx]
 } -result {A 1}
-test encoding-19.4 {TableFromUtfProc} -body {
+test encoding-19.5 {TableFromUtfProc} -body {
     list [encoding convertfrom -failindex idx -strict ascii AÁ] [set idx]
 } -result {A 1}
+test encoding-19.6 {TableFromUtfProc} -body {
+    list [encoding convertfrom -failindex idx -strict ascii AÁB] [set idx]
+} -result {A 1}
 
 test encoding-20.1 {TableFreefProc} {
 } {}