Attachment "1122671-tcl85.patch" to
ticket [1122671fff]
added by
hobbs
2006-10-06 04:21:14.
Index: generic/tcl.h
===================================================================
RCS file: /cvsroot/tcl/tcl/generic/tcl.h,v
retrieving revision 1.216
diff -u -r1.216 tcl.h
--- generic/tcl.h 26 Sep 2006 14:08:36 -0000 1.216
+++ generic/tcl.h 5 Oct 2006 21:20:17 -0000
@@ -2302,7 +2302,11 @@
/*
* unsigned int isn't 100% accurate as it should be a strict 4-byte value
* (perhaps wchar_t). 64-bit systems may have troubles. The size of this
- * value must be reflected correctly in regcustom.h.
+ * value must be reflected correctly in regcustom.h and
+ * in tclEncoding.c.
+ * XXX: Tcl is currently UCS-2 and planning UTF-16 for the Unicode
+ * XXX: string rep that Tcl_UniChar represents. Changing the size
+ * XXX: of Tcl_UniChar is /not/ supported.
*/
typedef unsigned int Tcl_UniChar;
#else
Index: generic/tclEncoding.c
===================================================================
RCS file: /cvsroot/tcl/tcl/generic/tclEncoding.c,v
retrieving revision 1.44
diff -u -r1.44 tclEncoding.c
--- generic/tclEncoding.c 26 Sep 2006 23:01:11 -0000 1.44
+++ generic/tclEncoding.c 5 Oct 2006 21:20:17 -0000
@@ -2288,9 +2288,10 @@
* correspond to the bytes stored in the
* output buffer. */
{
- CONST Tcl_UniChar *wSrc, *wSrcStart, *wSrcEnd;
+ CONST char *srcStart, *srcEnd;
char *dstEnd, *dstStart;
int result, numChars;
+ Tcl_UniChar ch;
result = TCL_OK;
if ((srcLen % sizeof(Tcl_UniChar)) != 0) {
@@ -2299,33 +2300,31 @@
srcLen *= sizeof(Tcl_UniChar);
}
- wSrc = (Tcl_UniChar *) src;
-
- wSrcStart = (Tcl_UniChar *) src;
- wSrcEnd = (Tcl_UniChar *) (src + srcLen);
+ srcStart = src;
+ srcEnd = src + srcLen;
dstStart = dst;
dstEnd = dst + dstLen - TCL_UTF_MAX;
- for (numChars = 0; wSrc < wSrcEnd; numChars++) {
+ for (numChars = 0; src < srcEnd; numChars++) {
if (dst > dstEnd) {
result = TCL_CONVERT_NOSPACE;
break;
}
-
/*
- * Special case for 1-byte utf chars for speed.
+ * Special case for 1-byte utf chars for speed. Make sure we
+ * work with Tcl_UniChar-size data.
*/
-
- if (*wSrc && *wSrc < 0x80) {
- *dst++ = (char) *wSrc;
+ ch = *(Tcl_UniChar *)src;
+ if (ch && ch < 0x80) {
+ *dst++ = *src;
} else {
- dst += Tcl_UniCharToUtf(*wSrc, dst);
+ dst += Tcl_UniCharToUtf(ch, dst);
}
- wSrc++;
+ src += sizeof(Tcl_UniChar);
}
- *srcReadPtr = (char *) wSrc - (char *) wSrcStart;
+ *srcReadPtr = src - srcStart;
*dstWrotePtr = dst - dstStart;
*dstCharsPtr = numChars;
return result;
@@ -2375,9 +2374,9 @@
* correspond to the bytes stored in the
* output buffer. */
{
- CONST char *srcStart, *srcEnd, *srcClose;
- Tcl_UniChar *wDst, *wDstStart, *wDstEnd;
+ CONST char *srcStart, *srcEnd, *srcClose, *dstStart, *dstEnd;
int result, numChars;
+ Tcl_UniChar ch;
srcStart = src;
srcEnd = src + srcLen;
@@ -2386,9 +2385,8 @@
srcClose -= TCL_UTF_MAX;
}
- wDst = (Tcl_UniChar *) dst;
- wDstStart = (Tcl_UniChar *) dst;
- wDstEnd = (Tcl_UniChar *) (dst + dstLen - sizeof(Tcl_UniChar));
+ dstStart = dst;
+ dstEnd = dst + dstLen - sizeof(Tcl_UniChar);
result = TCL_OK;
for (numChars = 0; src < srcEnd; numChars++) {
@@ -2401,16 +2399,26 @@
result = TCL_CONVERT_MULTIBYTE;
break;
}
- if (wDst > wDstEnd) {
+ if (dst > dstEnd) {
result = TCL_CONVERT_NOSPACE;
break;
- }
- src += TclUtfToUniChar(src, wDst);
- wDst++;
+ }
+ src += TclUtfToUniChar(src, &ch);
+ /*
+ * Need to handle this in a way that won't cause misalignment
+ * by casting dst to a Tcl_UniChar. [Bug 1122671]
+ * XXX: This hard-codes the assumed size of Tcl_UniChar as 2.
+ */
+#ifdef WORDS_BIGENDIAN
+ *dst++ = (ch >> 8);
+ *dst++ = (ch & 0xFF);
+#else
+ *dst++ = (ch & 0xFF);
+ *dst++ = (ch >> 8);
+#endif
}
-
*srcReadPtr = src - srcStart;
- *dstWrotePtr = (char *) wDst - (char *) wDstStart;
+ *dstWrotePtr = dst - dstStart;
*dstCharsPtr = numChars;
return result;
}
Index: tests/encoding.test
===================================================================
RCS file: /cvsroot/tcl/tcl/tests/encoding.test,v
retrieving revision 1.24
diff -u -r1.24 encoding.test
--- tests/encoding.test 8 Feb 2006 21:41:28 -0000 1.24
+++ tests/encoding.test 5 Oct 2006 21:20:17 -0000
@@ -307,8 +307,9 @@
} {1 2 c080}
test encoding-16.1 {UnicodeToUtfProc} {
- encoding convertfrom unicode NN
-} "\u4e4e"
+ set val [encoding convertfrom unicode NN]
+ list $val [format %x [scan $val %c]]
+} "\u4e4e 4e4e"
test encoding-17.1 {UtfToUnicodeProc} {
} {}
