Merge branch 'odin-lang:master' into master

core/encoding/json/parser.odin

@@ -40,7 +40,7 @@ parse_string :: proc(data: string, spec := DEFAULT_SPECIFICATION, parse_integers
 		return parse_object(&p)
 	case .JSON5:
 		return parse_value(&p)
-	case .MJSON:
+	case .SJSON:
 		#partial switch p.curr_token.kind {
 		case .Ident, .String:
 			return parse_object_body(&p, .EOF)

core/encoding/json/types.odin

@@ -33,8 +33,9 @@ package json
 Specification :: enum {
 	JSON,
 	JSON5, // https://json5.org/
-	MJSON, // https://bitsquid.blogspot.com/2009/10/simplified-json-notation.html
-	Bitsquid = MJSON,
+	SJSON, // https://bitsquid.blogspot.com/2009/10/simplified-json-notation.html
+	Bitsquid = SJSON,
+	MJSON = SJSON,
 }
 
 

core/strings/strings.odin

@@ -225,14 +225,23 @@ equal_fold :: proc(u, v: string) -> bool {
 */
 prefix_length :: proc(a, b: string) -> (n: int) {
 	_len := min(len(a), len(b))
-	idx  := 0
 
-	#no_bounds_check for idx < _len && a[idx] == b[idx] {
-		idx += 1
+	// Scan for matches including partial codepoints.
+	#no_bounds_check for n < _len && a[n] == b[n] {
+		n += 1
+	}
 
-		if a[idx] & 128 != 128 {
-			// new codepoint or end of multi-byte codepoint, update match length
-			n = idx
+	// Now scan to ignore partial codepoints.
+	if n > 0 {
+		s := a[:n]
+		n = 0
+		for {
+			r0, w := utf8.decode_rune(s[n:])
+			if r0 != utf8.RUNE_ERROR {
+				n += w
+			} else {
+				break
+			}
 		}
 	}
 	return

core/unicode/utf8/utf8.odin

@@ -90,10 +90,15 @@ encode_rune :: proc(c: rune) -> ([4]u8, int) {
 	return buf, 4
 }
 
+
+decode_rune :: proc{
+	decode_rune_in_string,
+	decode_rune_in_bytes,
+}
 decode_rune_in_string :: #force_inline proc(s: string) -> (rune, int) {
-	return decode_rune(transmute([]u8)s)
+	return decode_rune_in_bytes(transmute([]u8)s)
 }
-decode_rune :: proc(s: []u8) -> (rune, int) {
+decode_rune_in_bytes :: proc(s: []u8) -> (rune, int) {
 	n := len(s)
 	if n < 1 {
 		return RUNE_ERROR, 0
@@ -161,10 +166,15 @@ runes_to_string :: proc(runes: []rune, allocator := context.allocator) -> string
 }
 
 
+decode_last_rune :: proc{
+	decode_last_rune_in_string,
+	decode_last_rune_in_bytes,
+}
+
 decode_last_rune_in_string :: #force_inline proc(s: string) -> (rune, int) {
-	return decode_last_rune(transmute([]u8)s)
+	return decode_last_rune_in_bytes(transmute([]u8)s)
 }
-decode_last_rune :: proc(s: []u8) -> (rune, int) {
+decode_last_rune_in_bytes :: proc(s: []u8) -> (rune, int) {
 	r: rune
 	size: int
 	start, end, limit: int
@@ -297,10 +307,15 @@ rune_start :: #force_inline proc(b: u8) -> bool {
 	return b&0xc0 != 0x80
 }
 
+rune_count :: proc{
+	rune_count_in_string,
+	rune_count_in_bytes,
+}
+
 rune_count_in_string :: #force_inline proc(s: string) -> int {
-	return rune_count(transmute([]u8)s)
+	return rune_count_in_bytes(transmute([]u8)s)
 }
-rune_count :: proc(s: []u8) -> int {
+rune_count_in_bytes :: proc(s: []u8) -> int {
 	count := 0
 	n := len(s)
 
@@ -353,7 +368,14 @@ rune_size :: proc(r: rune) -> int {
 
 // full_rune reports if the bytes in b begin with a full utf-8 encoding of a rune or not
 // An invalid encoding is considered a full rune since it will convert as an error rune of width 1 (RUNE_ERROR)
-full_rune :: proc(b: []byte) -> bool {
+full_rune :: proc{
+	full_rune_in_bytes,
+	full_rune_in_string,
+}
+
+// full_rune_in_bytes reports if the bytes in b begin with a full utf-8 encoding of a rune or not
+// An invalid encoding is considered a full rune since it will convert as an error rune of width 1 (RUNE_ERROR)
+full_rune_in_bytes :: proc(b: []byte) -> bool {
 	n := len(b)
 	if n == 0 {
 		return false
@@ -374,7 +396,7 @@ full_rune :: proc(b: []byte) -> bool {
 // full_rune_in_string reports if the bytes in s begin with a full utf-8 encoding of a rune or not
 // An invalid encoding is considered a full rune since it will convert as an error rune of width 1 (RUNE_ERROR)
 full_rune_in_string :: proc(s: string) -> bool {
-	return full_rune(transmute([]byte)s)
+	return full_rune_in_bytes(transmute([]byte)s)
 }