src/3rdparty: numeric conversion functions now return the number of characters added

src/3rdparty/modp_numtoa.c

@@ -34,7 +34,8 @@ static void strreverse(char* begin, char* end)
 }
 
 // Expects 'str' to have been made using "%e" scientific notation format string
-static void sn_strip_trailing_zeros(char* str)
+// Returns the number of characters removed
+static size_t sn_strip_trailing_zeros(char* str)
 	{
 	char* frac = 0;
 
@@ -53,7 +54,7 @@ static void sn_strip_trailing_zeros(char* str)
 		}
 
 	if ( ! frac )
-		return;
+		return 0;
 
 	char* start_dec = frac;
 	char* exp = 0;
@@ -84,8 +85,11 @@ static void sn_strip_trailing_zeros(char* str)
 	if ( trailing_zeros == start_dec )
 		--trailing_zeros;
 
-	if ( trailing_zeros && exp )
+	if ( ! trailing_zeros || ! exp )
-		{
+		return 0;
+
+	char* start_exp = exp;
+
 	for ( ; ; )
 		{
 		*trailing_zeros = *exp;
@@ -96,10 +100,11 @@ static void sn_strip_trailing_zeros(char* str)
 		++trailing_zeros;
 		++exp;
 		}
-		}
+
+	return exp - start_exp;
 	}
 
-void modp_itoa10(int32_t value, char* str)
+size_t modp_itoa10(int32_t value, char* str)
 {
     char* wstr=str;
     // Take care of sign
@@ -111,9 +116,10 @@ void modp_itoa10(int32_t value, char* str)
 
     // Reverse string
     strreverse(str,wstr-1);
+    return wstr - str - 1;
 }
 
-void modp_uitoa10(uint32_t value, char* str)
+size_t modp_uitoa10(uint32_t value, char* str)
 {
     char* wstr=str;
     // Conversion. Number is reversed.
@@ -121,9 +127,10 @@ void modp_uitoa10(uint32_t value, char* str)
     *wstr='\0';
     // Reverse string
     strreverse(str, wstr-1);
+    return wstr - str - 1;
 }
 
-void modp_litoa10(int64_t value, char* str)
+size_t modp_litoa10(int64_t value, char* str)
 {
     char* wstr=str;
     uint64_t uvalue = (value < 0) ? (value == INT64_MIN ? (uint64_t)(INT64_MAX) + 1 : -value) : value;
@@ -135,9 +142,10 @@ void modp_litoa10(int64_t value, char* str)
 
     // Reverse string
     strreverse(str,wstr-1);
+    return wstr - str - 1;
 }
 
-void modp_ulitoa10(uint64_t value, char* str)
+size_t modp_ulitoa10(uint64_t value, char* str)
 {
     char* wstr=str;
     // Conversion. Number is reversed.
@@ -145,9 +153,10 @@ void modp_ulitoa10(uint64_t value, char* str)
     *wstr='\0';
     // Reverse string
     strreverse(str, wstr-1);
+    return wstr - str - 1;
 }
 
-void modp_dtoa(double value, char* str, int prec)
+size_t modp_dtoa(double value, char* str, int prec)
 {
     /* Hacky test for NaN
      * under -fast-math this won't work, but then you also won't
@@ -156,7 +165,7 @@ void modp_dtoa(double value, char* str, int prec)
      */
     if (! (value == value)) {
         str[0] = 'n'; str[1] = 'a'; str[2] = 'n'; str[3] = '\0';
-        return;
+        return 3;
     }
 
     /* we'll work in positive values and deal with the
@@ -177,9 +186,9 @@ void modp_dtoa(double value, char* str, int prec)
       which can be 100s of characters overflowing your buffers == bad
     */
     if (value >= thres_max) {
-        sprintf(str, "%.*e", DBL_DECIMAL_DIG - 1, neg ? -value : value);
+        int n = sprintf(str, "%.*e", DBL_DECIMAL_DIG - 1, neg ? -value : value);
-        sn_strip_trailing_zeros(str);
+        n -= sn_strip_trailing_zeros(str);
-        return;
+        return n;
     }
 
     double diff = 0.0;
@@ -242,12 +251,13 @@ void modp_dtoa(double value, char* str, int prec)
     }
     *wstr='\0';
     strreverse(str, wstr-1);
+    return wstr - str - 1;
 }
 
 
 // This is near identical to modp_dtoa above
 //   The differnce is noted below
-void modp_dtoa2(double value, char* str, int prec)
+size_t modp_dtoa2(double value, char* str, int prec)
 {
     /* Hacky test for NaN
      * under -fast-math this won't work, but then you also won't
@@ -256,7 +266,7 @@ void modp_dtoa2(double value, char* str, int prec)
      */
     if (! (value == value)) {
         str[0] = 'n'; str[1] = 'a'; str[2] = 'n'; str[3] = '\0';
-        return;
+        return 3;
     }
 
     /* we'll work in positive values and deal with the
@@ -277,9 +287,9 @@ void modp_dtoa2(double value, char* str, int prec)
       which can be 100s of characters overflowing your buffers == bad
     */
     if (value >= thres_max) {
-        sprintf(str, "%.*e", DBL_DECIMAL_DIG - 1, neg ? -value : value);
+        int n = sprintf(str, "%.*e", DBL_DECIMAL_DIG - 1, neg ? -value : value);
-        sn_strip_trailing_zeros(str);
+        n -= sn_strip_trailing_zeros(str);
-        return;
+        return n;
     }
 
     int count;
@@ -296,9 +306,9 @@ void modp_dtoa2(double value, char* str, int prec)
     double smallest = _pow10r[prec];
 
     if (value != 0.0 && value < smallest) {
-        sprintf(str, "%.*e", DBL_DECIMAL_DIG - 1, neg ? -value : value);
+        int n = sprintf(str, "%.*e", DBL_DECIMAL_DIG - 1, neg ? -value : value);
-        sn_strip_trailing_zeros(str);
+        n -= sn_strip_trailing_zeros(str);
-        return;
+        return n;
     }
 
     int whole = (int) value;
@@ -362,11 +372,12 @@ void modp_dtoa2(double value, char* str, int prec)
     }
     *wstr='\0';
     strreverse(str, wstr-1);
+    return wstr - str - 1;
 }
 
 // This is near identical to modp_dtoa2 above, excep that it never uses
 // exponential notation and requires a buffer length.
-void modp_dtoa3(double value, char* str, int n, int prec)
+size_t modp_dtoa3(double value, char* str, int n, int prec)
 {
     /* Hacky test for NaN
      * under -fast-math this won't work, but then you also won't
@@ -375,7 +386,7 @@ void modp_dtoa3(double value, char* str, int n, int prec)
      */
     if (! (value == value)) {
         str[0] = 'n'; str[1] = 'a'; str[2] = 'n'; str[3] = '\0';
-        return;
+        return 3;
     }
 
     /* we'll work in positive values and deal with the
@@ -409,7 +420,7 @@ void modp_dtoa3(double value, char* str, int n, int prec)
         if ( i < 0 || i >= n ) {
         // Error or truncated output.
             snprintf(str, n, "NAN");
-            return;
+            return 3;
             }
 
         /* Remove trailing zeros. */
@@ -421,7 +432,7 @@ void modp_dtoa3(double value, char* str, int n, int prec)
             --p;
 
         *++p = '\0';
-        return;
+        return p - str - 1;
 
         /* ---- End of modified part.. */
     }
@@ -491,4 +502,5 @@ void modp_dtoa3(double value, char* str, int n, int prec)
     }
     *wstr='\0';
     strreverse(str, wstr-1);
+    return wstr - str - 1;
 }

src/3rdparty/modp_numtoa.h

@@ -34,37 +34,38 @@
 BEGIN_C
 
 #include <stdint.h>
+#include <stddef.h>
 
-/** \brief convert an signed integer to char buffer
+/** \brief convert an signed integer to char buffer, return # characters added
  *
  * \param[in] value
  * \param[out] buf the output buffer.  Should be 16 chars or more.
  */
-void modp_itoa10(int32_t value, char* buf);
+size_t modp_itoa10(int32_t value, char* buf);
 
-/** \brief convert an unsigned integer to char buffer
+/** \brief convert an unsigned integer to char buffer, return # characters added
  *
  * \param[in] value
  * \param[out] buf The output buffer, should be 16 chars or more.
  */
-void modp_uitoa10(uint32_t value, char* buf);
+size_t modp_uitoa10(uint32_t value, char* buf);
 
-/** \brief convert an signed long integer to char buffer
+/** \brief convert an signed long integer to char buffer, return # characters added
  *
  * \param[in] value
  * \param[out] buf the output buffer.  Should be 24 chars or more.
  */
-void modp_litoa10(int64_t value, char* buf);
+size_t modp_litoa10(int64_t value, char* buf);
 
-/** \brief convert an unsigned long integer to char buffer
+/** \brief convert an unsigned long integer to char buffer, return # characters added
  *
  * \param[in] value
  * \param[out] buf The output buffer, should be 24 chars or more.
  */
-void modp_ulitoa10(uint64_t value, char* buf);
+size_t modp_ulitoa10(uint64_t value, char* buf);
 
 /** \brief convert a floating point number to char buffer with
- *         fixed-precision format
+ *         fixed-precision format, return # characters added
  *
  * This is similar to "%.[0-9]f" in the printf style.  It will include
  * trailing zeros
@@ -78,10 +79,11 @@ void modp_ulitoa10(uint64_t value, char* buf);
  * \param[in] precision  Number of digits to the right of the decimal point.
  *    Can only be 0-9.
  */
-void modp_dtoa(double value, char* buf, int precision);
+size_t modp_dtoa(double value, char* buf, int precision);
 
 /** \brief convert a floating point number to char buffer with a
- *         variable-precision format, and no trailing zeros
+ *         variable-precision format, and no trailing zero, return
+ *         number of characters added
  *
  * This is similar to "%.[0-9]f" in the printf style, except it will
  * NOT include trailing zeros after the decimal point.  This type
@@ -100,15 +102,15 @@ void modp_dtoa(double value, char* buf, int precision);
  * \param[in] precision  Number of digits to the right of the decimal point.
  *    Can only be 0-9.
  */
-void modp_dtoa2(double value, char* buf, int precision);
+size_t modp_dtoa2(double value, char* buf, int precision);
 
 /** \brief convert a floating point number to char buffer with a
  *         variable-precision format, no trailing zeros, and no
- *   	scientific notation.
+ *   	scientific notation, return number of characters added
  *
  *  Other than avoiding scientific notation, this is the same as mop_dtoa2. It does however
  *  require the max buffer length. The buffer will always be null-terminated.
  */
-void modp_dtoa3(double value, char* buf, int n, int precision);
+size_t modp_dtoa3(double value, char* buf, int n, int precision);
 
 END_C