Eliminate trailing zeros from scale tick values. This is done by

calculating a separate format string for these values, rather than
using the same format string as for the current value. The basis is
the same, i.e., enough digits are displayed to distinguish any pair of
adjacent tick values, and all tick values have the same number of
decimals so some may still have trailing zeros.

Code for laying out vertical scales has been adjusted to take account
of the fact that the tick column may now be narrower than the value
column.

1 files changed, 95 insertions, 53 deletions

diff --git a/generic/tkScale.c b/generic/tkScale.c
index 29f8722..455c656 100644
--- a/generic/tkScale.c
+++ b/generic/tkScale.c
@@ -157,7 +157,7 @@ enum command {
  * Forward declarations for procedures defined later in this file:
  */
 
-static void		ComputeFormat(TkScale *scalePtr);
+static void		ComputeFormat(TkScale *scalePtr, int forTicks);
 static void		ComputeScaleGeometry(TkScale *scalePtr);
 static int		ConfigureScale(Tcl_Interp *interp, TkScale *scalePtr,
 			    int objc, Tcl_Obj *const objv[]);
@@ -430,7 +430,7 @@ ScaleWidgetObjCmd(
 	    }
 	    value = TkScalePixelToValue(scalePtr, x, y);
 	}
-	Tcl_SetObjResult(interp, Tcl_ObjPrintf(scalePtr->format, value));
+	Tcl_SetObjResult(interp, Tcl_ObjPrintf(scalePtr->valueFormat, value));
 	break;
     }
     case COMMAND_IDENTIFY: {
@@ -636,7 +636,8 @@ ConfigureScale(
 	    scalePtr->tickInterval = -scalePtr->tickInterval;
 	}
 
-	ComputeFormat(scalePtr);
+	ComputeFormat(scalePtr, 0);
+	ComputeFormat(scalePtr, 1);
 
 	scalePtr->labelLength = scalePtr->label ? (int)strlen(scalePtr->label) : 0;
 
@@ -764,22 +765,22 @@ ScaleWorldChanged(
  *
  * ComputeFormat --
  *
- *	This procedure is invoked to recompute the "format" field of a scale's
- *	widget record, which determines how the value of the scale is
- *	converted to a string.
+ *	This procedure is invoked to recompute the "valueFormat" or
+ *	"tickFormat" field of a scale's widget record, which determines how
+ *	the value of the scale or one of its ticks is converted to a string.
  *
  * Results:
  *	None.
  *
- * Side effects:
- *	The format field of scalePtr is modified.
+ * Side effects: The valueFormat or tickFormat field of scalePtr is modified.
  *
  *----------------------------------------------------------------------
  */
 
 static void
 ComputeFormat(
-    TkScale *scalePtr)		/* Information about scale widget. */
+    TkScale *scalePtr,		/* Information about scale widget. */
+    int forTicks)               /* Do for ticks rather than value */
 {
     double maxValue, x;
     int mostSigDigit, numDigits, leastSigDigit, afterDecimal;
@@ -800,46 +801,60 @@ ComputeFormat(
     }
     mostSigDigit = (int) floor(log10(maxValue));
 
-    /*
-     * If the number of significant digits wasn't specified explicitly,
-     * compute it. It's the difference between the most significant digit
-     * needed to represent any number on the scale and the most significant
-     * digit of the smallest difference between numbers on the scale. In other
-     * words, display enough digits so that at least one digit will be
-     * different between any two adjacent positions of the scale.
-     */
-
-    numDigits = scalePtr->digits;
-    if (numDigits > TCL_MAX_PREC) {
-	numDigits = 0;
-    }
-    if (numDigits <= 0) {
-	if (scalePtr->resolution > 0) {
-	    /*
-	     * A resolution was specified for the scale, so just use it.
-	     */
+    if (forTicks) {
+	/*
+	 * Display only enough digits to ensure adjacent ticks have different
+	 * values
+	 */
 
-	    leastSigDigit = (int) floor(log10(scalePtr->resolution));
+	if (scalePtr->tickInterval > 0) {
+	    numDigits = 1 + mostSigDigit - (int) floor(log10(scalePtr->tickInterval));
 	} else {
-	    /*
-	     * No resolution was specified, so compute the difference in value
-	     * between adjacent pixels and use it for the least significant
-	     * digit.
-	     */
+	    numDigits = 1;
+	}
+    } else {
+	/*
+	 * If the number of significant digits wasn't specified explicitly,
+	 * compute it. It's the difference between the most significant digit
+	 * needed to represent any number on the scale and the most
+	 * significant digit of the smallest difference between numbers on the
+	 * scale. In other words, display enough digits so that at least one
+	 * digit will be different between any two adjacent positions of the
+	 * scale.
+	 */
 
-	    x = fabs(scalePtr->fromValue - scalePtr->toValue);
-	    if (scalePtr->length > 0) {
-		x /= scalePtr->length;
-	    }
-	    if (x > 0){
-		leastSigDigit = (int) floor(log10(x));
+	numDigits = scalePtr->digits;
+	if (numDigits > TCL_MAX_PREC) {
+	    numDigits = 0;
+	}
+	if (numDigits <= 0) {
+	    if (scalePtr->resolution > 0) {
+		/*
+		 * A resolution was specified for the scale, so just use it.
+		 */
+		
+		leastSigDigit = (int) floor(log10(scalePtr->resolution));
 	    } else {
-		leastSigDigit = 0;
+		/*
+		 * No resolution was specified, so compute the difference in
+		 * value between adjacent pixels and use it for the least
+		 * significant digit.
+		 */
+		
+		x = fabs(scalePtr->fromValue - scalePtr->toValue);
+		if (scalePtr->length > 0) {
+		    x /= scalePtr->length;
+		}
+		if (x > 0){
+		    leastSigDigit = (int) floor(log10(x));
+		} else {
+		    leastSigDigit = 0;
+		}
+	    }
+	    numDigits = mostSigDigit - leastSigDigit + 1;
+	    if (numDigits < 1) {
+		numDigits = 1;
 	    }
-	}
-	numDigits = mostSigDigit - leastSigDigit + 1;
-	if (numDigits < 1) {
-	    numDigits = 1;
 	}
     }
 
@@ -863,10 +878,18 @@ ComputeFormat(
     if (mostSigDigit < 0) {
 	fDigits++;			/* Zero to left of decimal point. */
     }
-    if (fDigits <= eDigits) {
-	sprintf(scalePtr->format, "%%.%df", afterDecimal);
+    if (forTicks) {
+	if (fDigits <= eDigits) {
+	    sprintf(scalePtr->tickFormat, "%%.%df", afterDecimal);
+	} else {
+	    sprintf(scalePtr->tickFormat, "%%.%de", numDigits-1);
+	}
     } else {
-	sprintf(scalePtr->format, "%%.%de", numDigits-1);
+	if (fDigits <= eDigits) {
+	    sprintf(scalePtr->valueFormat, "%%.%df", afterDecimal);
+	} else {
+	    sprintf(scalePtr->valueFormat, "%%.%de", numDigits-1);
+	}
     }
 }
 
@@ -894,7 +917,7 @@ ComputeScaleGeometry(
     register TkScale *scalePtr)	/* Information about widget. */
 {
     char valueString[TCL_DOUBLE_SPACE];
-    int tmp, valuePixels, x, y, extraSpace;
+    int tmp, valuePixels, tickPixels, x, y, extraSpace;
     Tk_FontMetrics fm;
 
     Tk_GetFontMetrics(scalePtr->tkfont, &fm);
@@ -940,13 +963,13 @@ ComputeScaleGeometry(
      * whichever length is longer.
      */
 
-    if (snprintf(valueString, TCL_DOUBLE_SPACE, scalePtr->format,
+    if (snprintf(valueString, TCL_DOUBLE_SPACE, scalePtr->valueFormat,
             scalePtr->fromValue) < 0) {
         valueString[TCL_DOUBLE_SPACE - 1] = '\0';
     }
     valuePixels = Tk_TextWidth(scalePtr->tkfont, valueString, -1);
 
-    if (snprintf(valueString, TCL_DOUBLE_SPACE, scalePtr->format,
+    if (snprintf(valueString, TCL_DOUBLE_SPACE, scalePtr->valueFormat,
             scalePtr->toValue) < 0) {
         valueString[TCL_DOUBLE_SPACE - 1] = '\0';
     }
@@ -956,18 +979,37 @@ ComputeScaleGeometry(
     }
 
     /*
+     * Now do the same thing for the tick values
+     */
+
+    if (snprintf(valueString, TCL_DOUBLE_SPACE, scalePtr->tickFormat,
+            scalePtr->fromValue) < 0) {
+        valueString[TCL_DOUBLE_SPACE - 1] = '\0';
+    }
+    tickPixels = Tk_TextWidth(scalePtr->tkfont, valueString, -1);
+
+    if (snprintf(valueString, TCL_DOUBLE_SPACE, scalePtr->tickFormat,
+            scalePtr->toValue) < 0) {
+        valueString[TCL_DOUBLE_SPACE - 1] = '\0';
+    }
+    tmp = Tk_TextWidth(scalePtr->tkfont, valueString, -1);
+    if (tickPixels < tmp) {
+	tickPixels = tmp;
+    }
+
+    /*
      * Assign x-locations to the elements of the scale, working from left to
      * right.
      */
 
     x = scalePtr->inset;
     if ((scalePtr->tickInterval != 0) && (scalePtr->showValue)) {
-	scalePtr->vertTickRightX = x + SPACING + valuePixels;
+	scalePtr->vertTickRightX = x + SPACING + tickPixels;
 	scalePtr->vertValueRightX = scalePtr->vertTickRightX + valuePixels
 		+ fm.ascent/2;
 	x = scalePtr->vertValueRightX + SPACING;
     } else if (scalePtr->tickInterval != 0) {
-	scalePtr->vertTickRightX = x + SPACING + valuePixels;
+	scalePtr->vertTickRightX = x + SPACING + tickPixels;
 	scalePtr->vertValueRightX = scalePtr->vertTickRightX;
 	x = scalePtr->vertTickRightX + SPACING;
     } else if (scalePtr->showValue) {
@@ -1350,7 +1392,7 @@ ScaleSetVariable(
     if (scalePtr->varNamePtr != NULL) {
 	char string[TCL_DOUBLE_SPACE];
 
-        if (snprintf(string, TCL_DOUBLE_SPACE, scalePtr->format,
+        if (snprintf(string, TCL_DOUBLE_SPACE, scalePtr->valueFormat,
                 scalePtr->value) < 0) {
             string[TCL_DOUBLE_SPACE - 1] = '\0';
         }