Fix [6811a00819]: lsearch performance degradation on Tcl 8.6.11 release (thanks, Sergey!)

11 files changed, 48 insertions, 49 deletions

diff --git a/generic/tclBinary.c b/generic/tclBinary.c
index 8be4b75..570f4d5 100644
--- a/generic/tclBinary.c
+++ b/generic/tclBinary.c
@@ -436,7 +436,7 @@ Tcl_GetBytesFromObj(
 		irPtr = TclFetchInternalRep(objPtr, &tclByteArrayType);
 		baPtr = GET_BYTEARRAY(irPtr);
 		nonbyte = TclUtfAtIndex(Tcl_GetString(objPtr), baPtr->bad);
-		Tcl_UtfToUniChar(nonbyte, &ucs4);
+		TclUtfToUniChar(nonbyte, &ucs4);
 
 		Tcl_SetObjResult(interp, Tcl_ObjPrintf(
 			"expected byte sequence but character %d "
@@ -2654,7 +2654,7 @@ BinaryDecodeHex(
     if (pure) {
 	ucs4 = c;
     } else {
-	Tcl_UtfToUniChar((const char *)(data - 1), &ucs4);
+	TclUtfToUniChar((const char *)(data - 1), &ucs4);
     }
     TclDecrRefCount(resultObj);
     Tcl_SetObjResult(interp, Tcl_ObjPrintf(
@@ -3113,7 +3113,7 @@ BinaryDecodeUu(
     if (pure) {
 	ucs4 = c;
     } else {
-	Tcl_UtfToUniChar((const char *)(data - 1), &ucs4);
+	TclUtfToUniChar((const char *)(data - 1), &ucs4);
     }
     Tcl_SetObjResult(interp, Tcl_ObjPrintf(
 	    "invalid uuencode character \"%c\" (U+%06X) at position %d",
@@ -3288,7 +3288,7 @@ BinaryDecode64(
 	 * of a multi-byte character. */
 
 	/* Safe because we know data is NUL-terminated */
-	Tcl_UtfToUniChar((const char *)(data - 1), &ucs4);
+	TclUtfToUniChar((const char *)(data - 1), &ucs4);
     }
 
     Tcl_SetObjResult(interp, Tcl_ObjPrintf(
diff --git a/generic/tclCmdIL.c b/generic/tclCmdIL.c
index 116dbbf..de96267 100644
--- a/generic/tclCmdIL.c
+++ b/generic/tclCmdIL.c
@@ -5431,8 +5431,8 @@ DictionaryCompare(
 	 */
 
 	if ((*left != '\0') && (*right != '\0')) {
-	    left += Tcl_UtfToUniChar(left, &uniLeft);
-	    right += Tcl_UtfToUniChar(right, &uniRight);
+	    left += TclUtfToUniChar(left, &uniLeft);
+	    right += TclUtfToUniChar(right, &uniRight);
 
 	    /*
 	     * Convert both chars to lower for the comparison, because
diff --git a/generic/tclCmdMZ.c b/generic/tclCmdMZ.c
index c6176a6..1df9dd1 100644
--- a/generic/tclCmdMZ.c
+++ b/generic/tclCmdMZ.c
@@ -1222,7 +1222,7 @@ Tcl_SplitObjCmd(
 	Tcl_InitHashTable(&charReuseTable, TCL_ONE_WORD_KEYS);
 
 	for ( ; stringPtr < end; stringPtr += len) {
-	    len = Tcl_UtfToUniChar(stringPtr, &ch);
+	    len = TclUtfToUniChar(stringPtr, &ch);
 	    hPtr = Tcl_CreateHashEntry(&charReuseTable, INT2PTR(ch), &isNew);
 	    if (isNew) {
 		TclNewStringObj(objPtr, stringPtr, len);
@@ -1268,9 +1268,9 @@ Tcl_SplitObjCmd(
 	splitEnd = splitChars + splitCharLen;
 
 	for (element = stringPtr; stringPtr < end; stringPtr += len) {
-	    len = Tcl_UtfToUniChar(stringPtr, &ch);
+	    len = TclUtfToUniChar(stringPtr, &ch);
 	    for (p = splitChars; p < splitEnd; p += splitLen) {
-		splitLen = Tcl_UtfToUniChar(p, &splitChar);
+		splitLen = TclUtfToUniChar(p, &splitChar);
 		if (ch == splitChar) {
 		    TclNewStringObj(objPtr, element, stringPtr - element);
 		    Tcl_ListObjAppendElement(NULL, listPtr, objPtr);
@@ -1898,7 +1898,7 @@ StringIsCmd(
 	for (; string1 < end; string1 += length2, failat++) {
 	    int ucs4;
 
-	    length2 = Tcl_UtfToUniChar(string1, &ucs4);
+	    length2 = TclUtfToUniChar(string1, &ucs4);
 	    if (!chcomp(ucs4)) {
 		result = 0;
 		break;
diff --git a/generic/tclCompExpr.c b/generic/tclCompExpr.c
index 187c25c..7761ddd 100644
--- a/generic/tclCompExpr.c
+++ b/generic/tclCompExpr.c
@@ -2146,13 +2146,13 @@ ParseLexeme(
 
     if (!TclIsBareword(*start) || *start == '_') {
 	if (Tcl_UtfCharComplete(start, numBytes)) {
-	    scanned = Tcl_UtfToUniChar(start, &ch);
+	    scanned = TclUtfToUniChar(start, &ch);
 	} else {
 	    char utfBytes[8];
 
 	    memcpy(utfBytes, start, numBytes);
 	    utfBytes[numBytes] = '\0';
-	    scanned = Tcl_UtfToUniChar(utfBytes, &ch);
+	    scanned = TclUtfToUniChar(utfBytes, &ch);
 	}
 	*lexemePtr = INVALID;
 	Tcl_DecrRefCount(literal);
diff --git a/generic/tclDisassemble.c b/generic/tclDisassemble.c
index 81bfa08..a096a85 100644
--- a/generic/tclDisassemble.c
+++ b/generic/tclDisassemble.c
@@ -875,7 +875,7 @@ PrintSourceToObj(
     for (;  (*p != '\0') && (i < maxChars);  p+=len) {
 	int ucs4;
 
-	len = Tcl_UtfToUniChar(p, &ucs4);
+	len = TclUtfToUniChar(p, &ucs4);
 	switch (ucs4) {
 	case '"':
 	    Tcl_AppendToObj(appendObj, "\\\"", -1);
diff --git a/generic/tclEncoding.c b/generic/tclEncoding.c
index 34d51e1..39c6ee3 100644
--- a/generic/tclEncoding.c
+++ b/generic/tclEncoding.c
@@ -1639,7 +1639,7 @@ Tcl_UtfToExternalDStringEx(
 		    Tcl_Size pos = TclNumUtfChars(srcStart, nBytesProcessed);
 		    int ucs4;
 		    char buf[TCL_INTEGER_SPACE];
-		    Tcl_UtfToUniChar(&srcStart[nBytesProcessed], &ucs4);
+		    TclUtfToUniChar(&srcStart[nBytesProcessed], &ucs4);
 		    snprintf(buf, sizeof(buf), "%" TCL_SIZE_MODIFIER "d", nBytesProcessed);
 		    Tcl_SetObjResult(
 			interp,
@@ -2604,12 +2604,12 @@ UtfToUtfProc(
 		/* TCL_ENCODING_PROFILE_TCL8 */
 		char chbuf[2];
 		chbuf[0] = UCHAR(*src++); chbuf[1] = 0;
-		Tcl_UtfToUniChar(chbuf, &ch);
+		TclUtfToUniChar(chbuf, &ch);
 	    }
 	    dst += Tcl_UniCharToUtf(ch, dst);
 	} else {
 	    int low;
-	    size_t len = Tcl_UtfToUniChar(src, &ch);
+	    size_t len = TclUtfToUniChar(src, &ch);
 	    if (flags & ENCODING_INPUT) {
 		if (((len < 2) && (ch != 0)) || ((ch > 0xFFFF) && !(flags & ENCODING_UTF))) {
 		    if (PROFILE_STRICT(profile)) {
@@ -2646,7 +2646,7 @@ UtfToUtfProc(
 		    ch = UNICODE_REPLACE_CHAR;
 		} else {
 		    low = ch;
-		    len = (src <= srcEnd - 3) ? Tcl_UtfToUniChar(src, &low) : 0;
+		    len = (src <= srcEnd - 3) ? TclUtfToUniChar(src, &low) : 0;
 
 		    if ((!LOW_SURROGATE(low)) || (ch & 0x400)) {
 
@@ -2906,7 +2906,7 @@ UtfToUtf32Proc(
 	    result = TCL_CONVERT_NOSPACE;
 	    break;
 	}
-	len = Tcl_UtfToUniChar(src, &ch);
+	len = TclUtfToUniChar(src, &ch);
 	if (SURROGATE(ch)) {
 	    if (PROFILE_STRICT(flags)) {
 		result = TCL_CONVERT_UNKNOWN;
@@ -3184,7 +3184,7 @@ UtfToUtf16Proc(
 	    result = TCL_CONVERT_NOSPACE;
 	    break;
 	}
-	len = Tcl_UtfToUniChar(src, &ch);
+	len = TclUtfToUniChar(src, &ch);
 	if (SURROGATE(ch)) {
 	    if (PROFILE_STRICT(flags)) {
 		result = TCL_CONVERT_UNKNOWN;
@@ -3431,7 +3431,7 @@ TableToUtfProc(
 	    } else {
 		char chbuf[2];
 		chbuf[0] = byte; chbuf[1] = 0;
-		Tcl_UtfToUniChar(chbuf, &ch);
+		TclUtfToUniChar(chbuf, &ch);
 	    }
 	}
 
diff --git a/generic/tclParse.c b/generic/tclParse.c
index aab69f3..281eee5 100644
--- a/generic/tclParse.c
+++ b/generic/tclParse.c
@@ -936,13 +936,13 @@ TclParseBackslash(
 	 */
 
 	if (Tcl_UtfCharComplete(p, numBytes - 1)) {
-	    count = Tcl_UtfToUniChar(p, &unichar) + 1;	/* +1 for '\' */
+	    count = TclUtfToUniChar(p, &unichar) + 1;	/* +1 for '\' */
 	} else {
 	    char utfBytes[8];
 
 	    memcpy(utfBytes, p, numBytes - 1);
 	    utfBytes[numBytes - 1] = '\0';
-	    count = Tcl_UtfToUniChar(utfBytes, &unichar) + 1;
+	    count = TclUtfToUniChar(utfBytes, &unichar) + 1;
 	}
 	result = unichar;
 	break;
diff --git a/generic/tclScan.c b/generic/tclScan.c
index 4c141ab..d2dcc70 100644
--- a/generic/tclScan.c
+++ b/generic/tclScan.c
@@ -879,7 +879,7 @@ Tcl_ScanObjCmd(
 	     * Scan a single Unicode character.
 	     */
 
-	    offset = Tcl_UtfToUniChar(string, &i);
+	    offset = TclUtfToUniChar(string, &i);
 	    string += offset;
 	    if (!(flags & SCAN_SUPPRESS)) {
 		TclNewIntObj(objPtr, i);
diff --git a/generic/tclStringObj.c b/generic/tclStringObj.c
index bcab176..5ec026f 100644
--- a/generic/tclStringObj.c
+++ b/generic/tclStringObj.c
@@ -4383,7 +4383,7 @@ TclStringReverse(
 		 * skip calling Tcl_UtfCharComplete() here.
 		 */
 
-		int bytesInChar = Tcl_UtfToUniChar(from, &chw);
+		int bytesInChar = TclUtfToUniChar(from, &chw);
 
 		ReverseBytes((unsigned char *)to, (unsigned char *)from,
 			bytesInChar);
diff --git a/generic/tclUtf.c b/generic/tclUtf.c
index c685733..2fa0e80 100644
--- a/generic/tclUtf.c
+++ b/generic/tclUtf.c
@@ -443,7 +443,6 @@ static const unsigned short cp1252[32] = {
    0x2DC, 0x2122, 0x0161, 0x203A, 0x0153,   0x9D, 0x017E, 0x0178
 };
 
-#undef Tcl_UtfToUniChar
 Tcl_Size
 Tcl_UtfToUniChar(
     const char *src,	/* The UTF-8 string. */
@@ -685,11 +684,11 @@ Tcl_UtfToUniCharDString(
     endPtr = src + length;
     optPtr = endPtr - 4;
     while (p <= optPtr) {
-	p += Tcl_UtfToUniChar(p, &ch);
+	p += TclUtfToUniChar(p, &ch);
 	*w++ = ch;
     }
     while ((p < endPtr) && Tcl_UtfCharComplete(p, endPtr-p)) {
-	p += Tcl_UtfToUniChar(p, &ch);
+	p += TclUtfToUniChar(p, &ch);
 	*w++ = ch;
     }
     while (p < endPtr) {
@@ -937,7 +936,7 @@ Tcl_UtfFindFirst(
     int ch)			/* The Unicode character to search for. */
 {
     while (1) {
-	int find, len = Tcl_UtfToUniChar(src, &find);
+	int find, len = TclUtfToUniChar(src, &find);
 
 	if (find == ch) {
 	    return src;
@@ -976,7 +975,7 @@ Tcl_UtfFindLast(
     const char *last = NULL;
 
     while (1) {
-	int find, len = Tcl_UtfToUniChar(src, &find);
+	int find, len = TclUtfToUniChar(src, &find);
 
 	if (find == ch) {
 	    last = src;
@@ -1203,7 +1202,7 @@ Tcl_UniCharAtIndex(
 	/* Index points at character following high Surrogate */
 	return -1;
     }
-    Tcl_UtfToUniChar(src, &i);
+    TclUtfToUniChar(src, &i);
     return i;
 }
 
@@ -1344,7 +1343,7 @@ Tcl_UtfToUpper(
 
     src = dst = str;
     while (*src) {
-	len = Tcl_UtfToUniChar(src, &ch);
+	len = TclUtfToUniChar(src, &ch);
 	upChar = Tcl_UniCharToUpper(ch);
 
 	/*
@@ -1397,7 +1396,7 @@ Tcl_UtfToLower(
 
     src = dst = str;
     while (*src) {
-	len = Tcl_UtfToUniChar(src, &ch);
+	len = TclUtfToUniChar(src, &ch);
 	lowChar = Tcl_UniCharToLower(ch);
 
 	/*
@@ -1453,7 +1452,7 @@ Tcl_UtfToTitle(
     src = dst = str;
 
     if (*src) {
-	len = Tcl_UtfToUniChar(src, &ch);
+	len = TclUtfToUniChar(src, &ch);
 	titleChar = Tcl_UniCharToTitle(ch);
 
 	if ((len < TclUtfCount(titleChar)) || ((titleChar & ~0x7FF) == 0xD800)) {
@@ -1465,7 +1464,7 @@ Tcl_UtfToTitle(
 	src += len;
     }
     while (*src) {
-	len = Tcl_UtfToUniChar(src, &ch);
+	len = TclUtfToUniChar(src, &ch);
 	lowChar = ch;
 	/* Special exception for Georgian Asomtavruli chars, no titlecase. */
 	if ((unsigned)(lowChar - 0x1C90) >= 0x30) {
diff --git a/generic/tclUtil.c b/generic/tclUtil.c
index 25f84e4..2219c84 100644
--- a/generic/tclUtil.c
+++ b/generic/tclUtil.c
@@ -1717,7 +1717,7 @@ TclTrimRight(
 	pp = Tcl_UtfPrev(p, bytes);
 	do {
 	    pp += pInc;
- 	    pInc = Tcl_UtfToUniChar(pp, &ch1);
+ 	    pInc = TclUtfToUniChar(pp, &ch1);
 	} while (pp + pInc < p);
 
 	/*
@@ -1725,7 +1725,7 @@ TclTrimRight(
 	 */
 
 	do {
-	    pInc = Tcl_UtfToUniChar(q, &ch2);
+	    pInc = TclUtfToUniChar(q, &ch2);
 
 	    if (ch1 == ch2) {
 		break;
@@ -1790,7 +1790,7 @@ TclTrimLeft(
      */
 
     do {
-	Tcl_Size pInc = Tcl_UtfToUniChar(p, &ch1);
+	Tcl_Size pInc = TclUtfToUniChar(p, &ch1);
 	const char *q = trim;
 	Tcl_Size bytesLeft = numTrim;
 
@@ -1799,7 +1799,7 @@ TclTrimLeft(
 	 */
 
 	do {
-	    Tcl_Size qInc = Tcl_UtfToUniChar(q, &ch2);
+	    Tcl_Size qInc = TclUtfToUniChar(q, &ch2);
 
 	    if (ch1 == ch2) {
 		break;
@@ -1866,7 +1866,7 @@ TclTrim(
 	if (numBytes > 0) {
 	    int ch;
 	    const char *first = bytes + trimLeft;
-	    bytes += Tcl_UtfToUniChar(first, &ch);
+	    bytes += TclUtfToUniChar(first, &ch);
 	    numBytes -= (bytes - first);
 
 	    if (numBytes > 0) {
@@ -2221,7 +2221,7 @@ Tcl_StringCaseMatch(
 		ch2 = (int)
 			(nocase ? tolower(UCHAR(*pattern)) : UCHAR(*pattern));
 	    } else {
-		Tcl_UtfToUniChar(pattern, &ch2);
+		TclUtfToUniChar(pattern, &ch2);
 		if (nocase) {
 		    ch2 = Tcl_UniCharToLower(ch2);
 		}
@@ -2237,7 +2237,7 @@ Tcl_StringCaseMatch(
 		if ((p != '[') && (p != '?') && (p != '\\')) {
 		    if (nocase) {
 			while (*str) {
-			    charLen = Tcl_UtfToUniChar(str, &ch1);
+			    charLen = TclUtfToUniChar(str, &ch1);
 			    if (ch2==ch1 || ch2==Tcl_UniCharToLower(ch1)) {
 				break;
 			    }
@@ -2251,7 +2251,7 @@ Tcl_StringCaseMatch(
 			 */
 
 			while (*str) {
-			    charLen = Tcl_UtfToUniChar(str, &ch1);
+			    charLen = TclUtfToUniChar(str, &ch1);
 			    if (ch2 == ch1) {
 				break;
 			    }
@@ -2265,7 +2265,7 @@ Tcl_StringCaseMatch(
 		if (*str == '\0') {
 		    return 0;
 		}
-		str += Tcl_UtfToUniChar(str, &ch1);
+		str += TclUtfToUniChar(str, &ch1);
 	    }
 	}
 
@@ -2276,7 +2276,7 @@ Tcl_StringCaseMatch(
 
 	if (p == '?') {
 	    pattern++;
-	    str += Tcl_UtfToUniChar(str, &ch1);
+	    str += TclUtfToUniChar(str, &ch1);
 	    continue;
 	}
 
@@ -2295,7 +2295,7 @@ Tcl_StringCaseMatch(
 			(nocase ? tolower(UCHAR(*str)) : UCHAR(*str));
 		str++;
 	    } else {
-		str += Tcl_UtfToUniChar(str, &ch1);
+		str += TclUtfToUniChar(str, &ch1);
 		if (nocase) {
 		    ch1 = Tcl_UniCharToLower(ch1);
 		}
@@ -2309,7 +2309,7 @@ Tcl_StringCaseMatch(
 			    ? tolower(UCHAR(*pattern)) : UCHAR(*pattern));
 		    pattern++;
 		} else {
-		    pattern += Tcl_UtfToUniChar(pattern, &startChar);
+		    pattern += TclUtfToUniChar(pattern, &startChar);
 		    if (nocase) {
 			startChar = Tcl_UniCharToLower(startChar);
 		    }
@@ -2324,7 +2324,7 @@ Tcl_StringCaseMatch(
 				? tolower(UCHAR(*pattern)) : UCHAR(*pattern));
 			pattern++;
 		    } else {
-			pattern += Tcl_UtfToUniChar(pattern, &endChar);
+			pattern += TclUtfToUniChar(pattern, &endChar);
 			if (nocase) {
 			    endChar = Tcl_UniCharToLower(endChar);
 			}
@@ -2372,8 +2372,8 @@ Tcl_StringCaseMatch(
 	 * each string match.
 	 */
 
-	str += Tcl_UtfToUniChar(str, &ch1);
-	pattern += Tcl_UtfToUniChar(pattern, &ch2);
+	str += TclUtfToUniChar(str, &ch1);
+	pattern += TclUtfToUniChar(pattern, &ch2);
 	if (nocase) {
 	    if (Tcl_UniCharToLower(ch1) != Tcl_UniCharToLower(ch2)) {
 		return 0;