std/conv: fix FmtFloat returns old-style formats for the NaN, Inf, an…

…d -Inf values

std/conv/atof.jule

@@ -62,7 +62,7 @@ fn commonPrefixLenIgnoreCase(s: str, prefix: str): int {
 }
 
 // Returns the floating-point value for the special,
-// possibly signed floating-point representations inf, infinity,
+// possibly signed floating-point representations Inf, infinity,
 // and NaN. The result is ok if a prefix of s contains one
 // of these representations and n is the length of that prefix.
 // The character case is ignored.
@@ -412,7 +412,7 @@ impl decimal {
 		goto out
 
 	overflow:
-		// ±inf
+		// ±Inf
 		mant = 0
 		exp = int(uint(1)<<flt.expbits - 1 + uint(flt.bias))
 		overflow = true
@@ -549,7 +549,7 @@ fn atof32exact(mantissa: u64, mut exp: int, mut neg: bool): (f: f32, ok: bool) {
 }
 
 // Converts the hex floating-point string s
-// to a rounded f32 or f64 value (depending on flt==&f32_info or flt==&f64_info)
+// to a rounded f32 or f64 value (depending on flt==&f32info or flt==&f64info)
 // and returns it as a f64.
 // The string s has already been parsed into a mantissa, exponent, and sign (neg==true for negative).
 // If trunc is true, trailing non-zero bits have been omitted from the mantissa.
@@ -765,7 +765,7 @@ fn parseFloatPrefix(&s: str, bitSize: int): (f64, int, int) {
 // away from the largest floating point number of the given size,
 // Exceptional = Error.OutOfRange.
 //
-// Recognizes the string "NaN", and the (possibly signed) strings "inf" and "infinity"
+// Recognizes the string "NaN", and the (possibly signed) strings "Inf" and "Infinity"
 // as their respective special floating point values. It ignores case when matching.
 fn ParseFloat(s: str, bitSize: int)!: f64 {
 	f, n, err := parseFloatPrefix(s, bitSize)

std/conv/eisel_lemire.jule

@@ -93,7 +93,7 @@ fn eiselLemire64(mut man: u64, exp10: int, neg: bool): (f: f64, ok: bool) {
 		retExp2 += 1
 	}
 	// ret_exp2 is a u64. Zero or underflow means that we're in subnormal
-	// f64 space. 0x7FF or above means that we're in inf/NaN f64 space.
+	// f64 space. 0x7FF or above means that we're in Inf/NaN f64 space.
 	//
 	// The if block is equivalent to (but has fewer branches than):
 	//   if retExp2 <= 0 || retExp2 >= 0x7FF { etc }
@@ -168,7 +168,7 @@ fn eiselLemire32(mut man: u64, exp10: int, neg: bool): (f: f32, ok: bool) {
 		retExp2 += 1
 	}
 	// retExp2 is a u64. Zero or underflow means that we're in subnormal
-	// f32 space. 0xFF or above means that we're in inf/NaN f32 space.
+	// f32 space. 0xFF or above means that we're in Inf/NaN f32 space.
 	//
 	// The if block is equivalent to (but has fewer branches than):
 	//   if retExp2 <= 0 || retExp2 >= 0xFF { etc }

std/conv/ftoa.jule

@@ -104,14 +104,14 @@ fn genericFtoa(mut dst: []byte, val: f64, fmt: byte, mut prec: int, bitSize: int
 
 	match exp {
 	| int(1<<flt.expbits - 1):
-		// +inf, NaN
+		// +Inf, NaN
 		match {
 		| mant != 0:
-			ret append(dst, "nan"...)
+			ret append(dst, "NaN"...)
 		| neg:
-			ret append(dst, "-inf"...)
+			ret append(dst, "-Inf"...)
 		|:
-			ret append(dst, "+inf"...)
+			ret append(dst, "+Inf"...)
 		}
 	| 0:
 		// denormalized