Proposed solution for 3613609: lsort -nocase does not sort non-ASCII correctly

4 files changed, 34 insertions, 6 deletions

diff --git a/generic/tclCmdIL.c b/generic/tclCmdIL.c
index 152e61d..98ec8b4 100644
--- a/generic/tclCmdIL.c
+++ b/generic/tclCmdIL.c
@@ -2807,7 +2807,7 @@ Tcl_LsearchObjCmd(
 	    dataType = INTEGER;
 	    break;
 	case LSEARCH_NOCASE:		/* -nocase */
-	    strCmpFn = strcasecmp;
+	    strCmpFn = TclUtfCasecmp;
 	    noCase = 1;
 	    break;
 	case LSEARCH_NOT:		/* -not */
@@ -3209,7 +3209,7 @@ Tcl_LsearchObjCmd(
 			 */
 
 			if (noCase) {
-			    match = (strcasecmp(bytes, patternBytes) == 0);
+			    match = (TclUtfCasecmp(bytes, patternBytes) == 0);
 			} else {
 			    match = (memcmp(bytes, patternBytes,
 				    (size_t) length) == 0);
@@ -3712,7 +3712,7 @@ Tcl_LsortObjCmd(
 		goto done1;
 	    }
 	    elementArray[i].index.intValue = a;
-	} else if (sortInfo.sortMode == SORTMODE_REAL) {
+	} else if (sortMode == SORTMODE_REAL) {
 	    double a;
 	    if (Tcl_GetDoubleFromObj(sortInfo.interp, indexPtr, &a) != TCL_OK) {
 		sortInfo.resultCode = TCL_ERROR;
@@ -3790,7 +3790,7 @@ Tcl_LsortObjCmd(
     ckfree((char *)elementArray);
 
   done:
-    if (sortInfo.sortMode == SORTMODE_COMMAND) {
+    if (sortMode == SORTMODE_COMMAND) {
 	TclDecrRefCount(sortInfo.compareCmdPtr);
 	TclDecrRefCount(listObj);
 	sortInfo.compareCmdPtr = NULL;
@@ -3932,7 +3932,7 @@ SortCompare(
 	order = strcmp(elemPtr1->index.strValuePtr,
 		elemPtr2->index.strValuePtr);
     } else if (infoPtr->sortMode == SORTMODE_ASCII_NC) {
-	order = strcasecmp(elemPtr1->index.strValuePtr,
+	order = TclUtfCasecmp(elemPtr1->index.strValuePtr,
 		elemPtr2->index.strValuePtr);
     } else if (infoPtr->sortMode == SORTMODE_DICTIONARY) {
 	order = DictionaryCompare(elemPtr1->index.strValuePtr,
diff --git a/generic/tclCmdMZ.c b/generic/tclCmdMZ.c
index 0ad77aa..6fd468c 100644
--- a/generic/tclCmdMZ.c
+++ b/generic/tclCmdMZ.c
@@ -3436,7 +3436,7 @@ Tcl_SwitchObjCmd(
 	    i++;
 	    goto finishedOptions;
 	case OPT_NOCASE:
-	    strCmpFn = strcasecmp;
+	    strCmpFn = TclUtfCasecmp;
 	    noCase = 1;
 	    break;
 
diff --git a/generic/tclInt.h b/generic/tclInt.h
index 92251fe..dc28b97 100644
--- a/generic/tclInt.h
+++ b/generic/tclInt.h
@@ -2762,6 +2762,7 @@ MODULE_SCOPE Tcl_WideInt TclpGetWideClicks(void);
 MODULE_SCOPE double	TclpWideClicksToNanoseconds(Tcl_WideInt clicks);
 #endif
 MODULE_SCOPE Tcl_Obj *	TclDisassembleByteCodeObj(Tcl_Obj *objPtr);
+MODULE_SCOPE int TclUtfCasecmp(CONST char *cs, CONST char *ct);
 
 /*
  *----------------------------------------------------------------
diff --git a/generic/tclUtf.c b/generic/tclUtf.c
index 83900e9..9dacb53 100644
--- a/generic/tclUtf.c
+++ b/generic/tclUtf.c
@@ -1101,6 +1101,33 @@ Tcl_UtfNcasecmp(
     }
     return 0;
 }
+
+
+/* Replacement for strcasecmp in Tcl core, in places where UTF-8 should be handled. */
+int
+TclUtfCasecmp(
+    CONST char *cs,		/* UTF string to compare to ct. */
+    CONST char *ct)		/* UTF string cs is compared to. */
+{
+    Tcl_UniChar ch1, ch2;
+    char c;
+
+    do {
+
+	/* If c == '\0', loop should end. */
+   	c = *cs;
+
+	cs += TclUtfToUniChar(cs, &ch1);
+	ct += TclUtfToUniChar(ct, &ch2);
+	if (ch1 != ch2) {
+	    ch1 = Tcl_UniCharToLower(ch1);
+	    ch2 = Tcl_UniCharToLower(ch2);
+	    if (ch1 != ch2) break;
+	}
+    } while (c);
+    return (ch1 - ch2);
+}
+
 
 /*
  *----------------------------------------------------------------------