Better UTF-8 surrogate handling, only functional when TCL_UTF_MAX>3

11 files changed, 75 insertions, 45 deletions

diff --git a/generic/tclBinary.c b/generic/tclBinary.c
index d0d9d5e..bb918f2 100644
--- a/generic/tclBinary.c
+++ b/generic/tclBinary.c
@@ -454,7 +454,7 @@ SetByteArrayFromAny(
     const char *src, *srcEnd;
     unsigned char *dst;
     ByteArray *byteArrayPtr;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
 
     if (objPtr->typePtr != &tclByteArrayType) {
 	src = TclGetStringFromObj(objPtr, &length);
@@ -1210,7 +1210,7 @@ BinaryFormatCmd(
 
  badField:
     {
-	Tcl_UniChar ch;
+	Tcl_UniChar ch = 0;
 	char buf[TCL_UTF_MAX + 1];
 
 	TclUtfToUniChar(errorString, &ch);
@@ -1580,7 +1580,7 @@ BinaryScanCmd(
 
  badField:
     {
-	Tcl_UniChar ch;
+	Tcl_UniChar ch = 0;
 	char buf[TCL_UTF_MAX + 1];
 
 	TclUtfToUniChar(errorString, &ch);
diff --git a/generic/tclCmdIL.c b/generic/tclCmdIL.c
index ba9e1cf..e3c5f10 100644
--- a/generic/tclCmdIL.c
+++ b/generic/tclCmdIL.c
@@ -4345,7 +4345,7 @@ static int
 DictionaryCompare(
     const char *left, const char *right)	/* The strings to compare. */
 {
-    Tcl_UniChar uniLeft, uniRight, uniLeftLower, uniRightLower;
+    Tcl_UniChar uniLeft = 0, uniRight = 0, uniLeftLower, uniRightLower;
     int diff, zeros;
     int secondaryDiff = 0;
 
diff --git a/generic/tclCmdMZ.c b/generic/tclCmdMZ.c
index 3f79ca4..7010495 100644
--- a/generic/tclCmdMZ.c
+++ b/generic/tclCmdMZ.c
@@ -1037,7 +1037,7 @@ Tcl_SplitObjCmd(
     int objc,			/* Number of arguments. */
     Tcl_Obj *const objv[])	/* Argument objects. */
 {
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     int len;
     const char *splitChars;
     const char *stringPtr;
@@ -1122,7 +1122,7 @@ Tcl_SplitObjCmd(
     } else {
 	const char *element, *p, *splitEnd;
 	int splitLen;
-	Tcl_UniChar splitChar;
+	Tcl_UniChar splitChar = 0;
 
 	/*
 	 * Normal case: split on any of a given set of characters. Discard
@@ -1451,7 +1451,7 @@ StringIsCmd(
     Tcl_Obj *const objv[])	/* Argument objects. */
 {
     const char *string1, *end, *stop;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     int (*chcomp)(int) = NULL;	/* The UniChar comparison function. */
     int i, failat = 0, result = 1, strict = 0, index, length1, length2;
     Tcl_Obj *objPtr, *failVarObj = NULL;
@@ -2436,7 +2436,7 @@ StringStartCmd(
     int objc,			/* Number of arguments. */
     Tcl_Obj *const objv[])	/* Argument objects. */
 {
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     const char *p, *string;
     int cur, index, length, numChars;
 
@@ -2497,7 +2497,7 @@ StringEndCmd(
     int objc,			/* Number of arguments. */
     Tcl_Obj *const objv[])	/* Argument objects. */
 {
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     const char *p, *end, *string;
     int cur, index, length, numChars;
 
diff --git a/generic/tclCompExpr.c b/generic/tclCompExpr.c
index 59eecf9..9c7ab8d 100644
--- a/generic/tclCompExpr.c
+++ b/generic/tclCompExpr.c
@@ -1885,7 +1885,7 @@ ParseLexeme(
 {
     const char *end;
     int scanned;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     Tcl_Obj *literal = NULL;
     unsigned char byte;
 
diff --git a/generic/tclEncoding.c b/generic/tclEncoding.c
index b4acb5f..8450128 100644
--- a/generic/tclEncoding.c
+++ b/generic/tclEncoding.c
@@ -2296,7 +2296,7 @@ UtfToUtfProc(
     const char *srcStart, *srcEnd, *srcClose;
     const char *dstStart, *dstEnd;
     int result, numChars, charLimit = INT_MAX;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
 
     result = TCL_OK;
 
@@ -2345,8 +2345,8 @@ UtfToUtfProc(
 	} else if (!Tcl_UtfCharComplete(src, srcEnd - src)) {
 	    /*
 	     * Always check before using TclUtfToUniChar. Not doing can so
-	     * cause it run beyond the endof the buffer! If we happen such an
-	     * incomplete char its byts are made to represent themselves.
+	     * cause it run beyond the end of the buffer! If we happen such an
+	     * incomplete char its bytes are made to represent themselves.
 	     */
 
 	    ch = (unsigned char) *src;
@@ -2410,7 +2410,7 @@ UnicodeToUtfProc(
     const char *srcStart, *srcEnd;
     const char *dstEnd, *dstStart;
     int result, numChars, charLimit = INT_MAX;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
 
     if (flags & TCL_ENCODING_CHAR_LIMIT) {
 	charLimit = *dstCharsPtr;
@@ -2500,7 +2500,7 @@ UtfToUnicodeProc(
 {
     const char *srcStart, *srcEnd, *srcClose, *dstStart, *dstEnd;
     int result, numChars;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
 
     srcStart = src;
     srcEnd = src + srcLen;
@@ -2610,7 +2610,7 @@ TableToUtfProc(
     const char *srcStart, *srcEnd;
     const char *dstEnd, *dstStart, *prefixBytes;
     int result, byte, numChars, charLimit = INT_MAX;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     const unsigned short *const *toUnicode;
     const unsigned short *pageZero;
     TableEncodingData *dataPtr = clientData;
@@ -2722,7 +2722,7 @@ TableFromUtfProc(
 {
     const char *srcStart, *srcEnd, *srcClose;
     const char *dstStart, *dstEnd, *prefixBytes;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     int result, len, word, numChars;
     TableEncodingData *dataPtr = clientData;
     const unsigned short *const *fromUnicode;
@@ -2856,7 +2856,7 @@ Iso88591ToUtfProc(
 
     result = TCL_OK;
     for (numChars = 0; src < srcEnd && numChars <= charLimit; numChars++) {
-	Tcl_UniChar ch;
+	Tcl_UniChar ch = 0;
 
 	if (dst > dstEnd) {
 	    result = TCL_CONVERT_NOSPACE;
@@ -2942,7 +2942,7 @@ Iso88591FromUtfProc(
     dstEnd = dst + dstLen - 1;
 
     for (numChars = 0; src < srcEnd; numChars++) {
-	Tcl_UniChar ch;
+	Tcl_UniChar ch = 0;
 	int len;
 
 	if ((src > srcClose) && (!Tcl_UtfCharComplete(src, srcEnd - src))) {
@@ -3329,7 +3329,7 @@ EscapeFromUtfProc(
     for (numChars = 0; src < srcEnd; numChars++) {
 	unsigned len;
 	int word;
-	Tcl_UniChar ch;
+	Tcl_UniChar ch = 0;
 
 	if ((src > srcClose) && (!Tcl_UtfCharComplete(src, srcEnd - src))) {
 	    /*
diff --git a/generic/tclLoad.c b/generic/tclLoad.c
index 942e6b4..f1bd248 100644
--- a/generic/tclLoad.c
+++ b/generic/tclLoad.c
@@ -130,7 +130,7 @@ Tcl_LoadObjCmd(
     Tcl_PackageInitProc *initProc;
     const char *p, *fullFileName, *packageName;
     Tcl_LoadHandle loadHandle;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     unsigned len;
     int index, flags = 0;
     Tcl_Obj *const *savedobjv = objv;
diff --git a/generic/tclParse.c b/generic/tclParse.c
index 3ecf4a5..f2cf322 100644
--- a/generic/tclParse.c
+++ b/generic/tclParse.c
@@ -841,7 +841,7 @@ TclParseBackslash(
 				 * written there. */
 {
     register const char *p = src+1;
-    Tcl_UniChar unichar;
+    Tcl_UniChar unichar = 0;
     int result;
     int count;
     char buf[TCL_UTF_MAX];
diff --git a/generic/tclScan.c b/generic/tclScan.c
index 17069eb..7a6a8a2 100644
--- a/generic/tclScan.c
+++ b/generic/tclScan.c
@@ -257,7 +257,7 @@ ValidateFormat(
 {
     int gotXpg, gotSequential, value, i, flags;
     char *end;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
     int objIndex, xpgSize, nspace = numVars;
     int *nassign = TclStackAlloc(interp, nspace * sizeof(int));
     char buf[TCL_UTF_MAX+1];
diff --git a/generic/tclStringObj.c b/generic/tclStringObj.c
index 4a3b6f1..0cafffb 100644
--- a/generic/tclStringObj.c
+++ b/generic/tclStringObj.c
@@ -1709,7 +1709,7 @@ Tcl_AppendFormatToObj(
 #endif
 	int newXpg, numChars, allocSegment = 0, segmentLimit, segmentNumBytes;
 	Tcl_Obj *segment;
-	Tcl_UniChar ch;
+	Tcl_UniChar ch = 0;
 	int step = TclUtfToUniChar(format, &ch);
 
 	format += step;
@@ -2693,7 +2693,7 @@ TclStringObjReverse(
     Tcl_Obj *objPtr)
 {
     String *stringPtr;
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
 
     if (TclIsPureByteArray(objPtr)) {
 	int numBytes;
diff --git a/generic/tclUtf.c b/generic/tclUtf.c
index 52b4291..db941e2 100644
--- a/generic/tclUtf.c
+++ b/generic/tclUtf.c
@@ -134,7 +134,7 @@ UtfCount(
  *---------------------------------------------------------------------------
  */
 
-INLINE int
+int
 Tcl_UniCharToUtf(
     int ch,			/* The Tcl_UniChar to be stored in the
 				 * buffer. */
@@ -259,6 +259,15 @@ Tcl_UniCharToUtfDString(
  *	Tcl_UtfCharComplete() before calling this routine to ensure that
  *	enough bytes remain in the string.
  *
+ *	If TCL_UTF_MAX == 4, special handling of Surrogate pairs is done:
+ *	For any UTF-8 string containing a character outside of the BMP, the
+ *	first call to this function will fill *chPtr with the high surrogate
+ *	and generate a return value of 0. Calling Tcl_UtfToUniChar again
+ *	will produce the low surrogate and a return value of 4. Because *chPtr
+ *	is used to remember whether the high surrogate is already produced, it
+ *	is recommended to initialize the variable it points to as 0 before
+ *	the first call to Tcl_UtfToUniChar is done.
+ *
  * Results:
  *	*chPtr is filled with the Tcl_UniChar, and the return value is the
  *	number of bytes from the UTF-8 string that were consumed.
@@ -278,7 +287,7 @@ Tcl_UtfToUniChar(
     register int byte;
 
     /*
-     * Unroll 1 to 3 byte UTF-8 sequences, use loop to handle longer ones.
+     * Unroll 1 to 3 (or 4) byte UTF-8 sequences.
      */
 
     byte = *((unsigned char *) src);
@@ -331,12 +340,30 @@ Tcl_UtfToUniChar(
 	    /*
 	     * Four-byte-character lead byte followed by three trail bytes.
 	     */
-
+#if TCL_UTF_MAX == 4
+	    Tcl_UniChar surrogate;
+
+	    byte = (((byte & 0x07) << 18) | ((src[1] & 0x3F) << 12)
+		    | ((src[2] & 0x3F) << 6) | (src[3] & 0x3F)) - 0x10000;
+	    surrogate = (Tcl_UniChar) (0xD800 + (byte >> 10));
+	    if (byte & 0x100000) {
+		/* out of range, < 0x10000 or > 0x10ffff */
+	    } else if (*chPtr != surrogate) {
+		/* produce high surrogate, but don't advance source pointer */
+		*chPtr = surrogate;
+		return 0;
+	    } else {
+		/* produce low surrogate, and advance source pointer */
+		*chPtr = (Tcl_UniChar) (0xDC00 | (byte & 0x3FF));
+		return 4;
+	    }
+#else
 	    *chPtr = (Tcl_UniChar) (((byte & 0x07) << 18) | ((src[1] & 0x3F) << 12)
 		    | ((src[2] & 0x3F) << 6) | (src[3] & 0x3F));
 	    if ((unsigned)(*chPtr - 0x10000) <= 0xFFFFF) {
 		return 4;
 	    }
+#endif
 	}
 
 	/*
@@ -377,7 +404,7 @@ Tcl_UtfToUniCharDString(
 				 * appended to this previously initialized
 				 * DString. */
 {
-    Tcl_UniChar *w, *wString;
+    Tcl_UniChar ch, *w, *wString;
     const char *p, *end;
     int oldLength;
 
@@ -399,8 +426,8 @@ Tcl_UtfToUniCharDString(
     w = wString;
     end = src + length;
     for (p = src; p < end; ) {
-	p += TclUtfToUniChar(p, w);
-	w++;
+	p += TclUtfToUniChar(p, &ch);
+	*w++ = ch;
     }
     *w = '\0';
     Tcl_DStringSetLength(dsPtr,
@@ -434,9 +461,8 @@ Tcl_UtfCharComplete(
 				 * a complete UTF-8 character. */
     int length)			/* Length of above string in bytes. */
 {
-    int ch;
+    int ch = *((unsigned char *) src);
 
-    ch = *((unsigned char *) src);
     return length >= totalBytes[ch];
 }
 
@@ -464,8 +490,7 @@ Tcl_NumUtfChars(
     int length)			/* The length of the string in bytes, or -1
 				 * for strlen(string). */
 {
-    Tcl_UniChar ch;
-    register Tcl_UniChar *chPtr = &ch;
+    Tcl_UniChar ch = 0;
     register int i;
 
     /*
@@ -478,7 +503,7 @@ Tcl_NumUtfChars(
     i = 0;
     if (length < 0) {
 	while (*src != '\0') {
-	    src += TclUtfToUniChar(src, chPtr);
+	    src += TclUtfToUniChar(src, &ch);
 	    i++;
 	}
     } else {
@@ -489,7 +514,7 @@ Tcl_NumUtfChars(
 		length--;
 		src++;
 	    } else {
-		n = Tcl_UtfToUniChar(src, chPtr);
+		n = Tcl_UtfToUniChar(src, &ch);
 		length -= n;
 		src += n;
 	    }
@@ -524,7 +549,7 @@ Tcl_UtfFindFirst(
     int ch)			/* The Tcl_UniChar to search for. */
 {
     int len;
-    Tcl_UniChar find;
+    Tcl_UniChar find = 0;
 
     while (1) {
 	len = TclUtfToUniChar(src, &find);
@@ -563,7 +588,7 @@ Tcl_UtfFindLast(
     int ch)			/* The Tcl_UniChar to search for. */
 {
     int len;
-    Tcl_UniChar find;
+    Tcl_UniChar find = 0;
     const char *last;
 
     last = NULL;
@@ -603,9 +628,15 @@ const char *
 Tcl_UtfNext(
     const char *src)		/* The current location in the string. */
 {
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
+    int len = TclUtfToUniChar(src, &ch);
 
-    return src + TclUtfToUniChar(src, &ch);
+#if TCL_UTF_MAX == 4
+    if (len == 0) {
+      len = TclUtfToUniChar(src, &ch);
+    }
+#endif
+    return src + len;
 }
 
 /*
@@ -638,8 +669,7 @@ Tcl_UtfPrev(
     const char *look;
     int i, byte;
 
-    src--;
-    look = src;
+    look = --src;
     for (i = 0; i < TCL_UTF_MAX; i++) {
 	if (look < start) {
 	    if (src < start) {
@@ -712,7 +742,7 @@ Tcl_UtfAtIndex(
     register const char *src,	/* The UTF-8 string. */
     register int index)		/* The position of the desired character. */
 {
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
 
     while (index > 0) {
 	index--;
diff --git a/generic/tclUtil.c b/generic/tclUtil.c
index 553593c..0eddb00 100644
--- a/generic/tclUtil.c
+++ b/generic/tclUtil.c
@@ -1646,7 +1646,7 @@ Tcl_Backslash(
 				 * src, unless NULL. */
 {
     char buf[TCL_UTF_MAX];
-    Tcl_UniChar ch;
+    Tcl_UniChar ch = 0;
 
     Tcl_UtfBackslash(src, readPtr, buf);
     TclUtfToUniChar(buf, &ch);