authorgravatar for quae@daurnimator.comdaurnimator <quae@daurnimator.com> 2019-12-28 14:45:52+11:00
committergravatar for quae@daurnimator.comdaurnimator <quae@daurnimator.com> 2019-12-28 16:49:59+11:00
logedb5deb39cc923d6fe2d3de507757d8ba9d74d68
tree4de2d4b045f3168182cff64f0150ac635476543b
parentab6065407d74fc8d63d398c60f9fe653374d9d6d
signature Commit is signed but in an unrecognized format.

std: unicode codepoints are 21 bits

1 files changed, 22 insertions(+), 22 deletions(-)

lib/std/unicode.zig+22-22
......@@ -6,7 +6,7 @@ const mem = std.mem;
66
77/// Returns how many bytes the UTF-8 representation would require
88/// for the given codepoint.
9pub fn utf8CodepointSequenceLength(c: u32) !u3 {
9pub fn utf8CodepointSequenceLength(c: u21) !u3 {
1010 if (c < 0x80) return @as(u3, 1);
1111 if (c < 0x800) return @as(u3, 2);
1212 if (c < 0x10000) return @as(u3, 3);
......@@ -32,7 +32,7 @@ pub fn utf8ByteSequenceLength(first_byte: u8) !u3 {
3232/// out: the out buffer to write to. Must have a len >= utf8CodepointSequenceLength(c).
3333/// Errors: if c cannot be encoded in UTF-8.
3434/// Returns: the number of bytes written to out.
35pub fn utf8Encode(c: u32, out: []u8) !u3 {
35pub fn utf8Encode(c: u21, out: []u8) !u3 {
3636 const length = try utf8CodepointSequenceLength(c);
3737 assert(out.len >= length);
3838 switch (length) {
......@@ -68,9 +68,9 @@ const Utf8DecodeError = Utf8Decode2Error || Utf8Decode3Error || Utf8Decode4Error
6868/// bytes.len must be equal to utf8ByteSequenceLength(bytes[0]) catch unreachable.
6969/// If you already know the length at comptime, you can call one of
7070/// utf8Decode2,utf8Decode3,utf8Decode4 directly instead of this function.
71pub fn utf8Decode(bytes: []const u8) Utf8DecodeError!u32 {
71pub fn utf8Decode(bytes: []const u8) Utf8DecodeError!u21 {
7272 return switch (bytes.len) {
73 1 => @as(u32, bytes[0]),
73 1 => @as(u21, bytes[0]),
7474 2 => utf8Decode2(bytes),
7575 3 => utf8Decode3(bytes),
7676 4 => utf8Decode4(bytes),
......@@ -82,10 +82,10 @@ const Utf8Decode2Error = error{
8282 Utf8ExpectedContinuation,
8383 Utf8OverlongEncoding,
8484};
85pub fn utf8Decode2(bytes: []const u8) Utf8Decode2Error!u32 {
85pub fn utf8Decode2(bytes: []const u8) Utf8Decode2Error!u21 {
8686 assert(bytes.len == 2);
8787 assert(bytes[0] & 0b11100000 == 0b11000000);
88 var value: u32 = bytes[0] & 0b00011111;
88 var value: u21 = bytes[0] & 0b00011111;
8989
9090 if (bytes[1] & 0b11000000 != 0b10000000) return error.Utf8ExpectedContinuation;
9191 value <<= 6;
......@@ -101,10 +101,10 @@ const Utf8Decode3Error = error{
101101 Utf8OverlongEncoding,
102102 Utf8EncodesSurrogateHalf,
103103};
104pub fn utf8Decode3(bytes: []const u8) Utf8Decode3Error!u32 {
104pub fn utf8Decode3(bytes: []const u8) Utf8Decode3Error!u21 {
105105 assert(bytes.len == 3);
106106 assert(bytes[0] & 0b11110000 == 0b11100000);
107 var value: u32 = bytes[0] & 0b00001111;
107 var value: u21 = bytes[0] & 0b00001111;
108108
109109 if (bytes[1] & 0b11000000 != 0b10000000) return error.Utf8ExpectedContinuation;
110110 value <<= 6;
......@@ -125,10 +125,10 @@ const Utf8Decode4Error = error{
125125 Utf8OverlongEncoding,
126126 Utf8CodepointTooLarge,
127127};
128pub fn utf8Decode4(bytes: []const u8) Utf8Decode4Error!u32 {
128pub fn utf8Decode4(bytes: []const u8) Utf8Decode4Error!u21 {
129129 assert(bytes.len == 4);
130130 assert(bytes[0] & 0b11111000 == 0b11110000);
131 var value: u32 = bytes[0] & 0b00000111;
131 var value: u21 = bytes[0] & 0b00000111;
132132
133133 if (bytes[1] & 0b11000000 != 0b10000000) return error.Utf8ExpectedContinuation;
134134 value <<= 6;
......@@ -224,11 +224,11 @@ pub const Utf8Iterator = struct {
224224 return it.bytes[it.i - cp_len .. it.i];
225225 }
226226
227 pub fn nextCodepoint(it: *Utf8Iterator) ?u32 {
227 pub fn nextCodepoint(it: *Utf8Iterator) ?u21 {
228228 const slice = it.nextCodepointSlice() orelse return null;
229229
230230 switch (slice.len) {
231 1 => return @as(u32, slice[0]),
231 1 => return @as(u21, slice[0]),
232232 2 => return utf8Decode2(slice) catch unreachable,
233233 3 => return utf8Decode3(slice) catch unreachable,
234234 4 => return utf8Decode4(slice) catch unreachable,
......@@ -248,19 +248,19 @@ pub const Utf16LeIterator = struct {
248248 };
249249 }
250250
251 pub fn nextCodepoint(it: *Utf16LeIterator) !?u32 {
251 pub fn nextCodepoint(it: *Utf16LeIterator) !?u21 {
252252 assert(it.i <= it.bytes.len);
253253 if (it.i == it.bytes.len) return null;
254 const c0: u32 = mem.readIntSliceLittle(u16, it.bytes[it.i .. it.i + 2]);
255 if (c0 & ~@as(u32, 0x03ff) == 0xd800) {
254 const c0: u21 = mem.readIntSliceLittle(u16, it.bytes[it.i .. it.i + 2]);
255 if (c0 & ~@as(u21, 0x03ff) == 0xd800) {
256256 // surrogate pair
257257 it.i += 2;
258258 if (it.i >= it.bytes.len) return error.DanglingSurrogateHalf;
259 const c1: u32 = mem.readIntSliceLittle(u16, it.bytes[it.i .. it.i + 2]);
260 if (c1 & ~@as(u32, 0x03ff) != 0xdc00) return error.ExpectedSecondSurrogateHalf;
259 const c1: u21 = mem.readIntSliceLittle(u16, it.bytes[it.i .. it.i + 2]);
260 if (c1 & ~@as(u21, 0x03ff) != 0xdc00) return error.ExpectedSecondSurrogateHalf;
261261 it.i += 2;
262262 return 0x10000 + (((c0 & 0x03ff) << 10) | (c1 & 0x03ff));
263 } else if (c0 & ~@as(u32, 0x03ff) == 0xdc00) {
263 } else if (c0 & ~@as(u21, 0x03ff) == 0xdc00) {
264264 return error.UnexpectedSecondSurrogateHalf;
265265 } else {
266266 it.i += 2;
......@@ -304,10 +304,10 @@ fn testUtf8EncodeError() void {
304304 testErrorEncode(0xd800, array[0..], error.Utf8CannotEncodeSurrogateHalf);
305305 testErrorEncode(0xdfff, array[0..], error.Utf8CannotEncodeSurrogateHalf);
306306 testErrorEncode(0x110000, array[0..], error.CodepointTooLarge);
307 testErrorEncode(0xffffffff, array[0..], error.CodepointTooLarge);
307 testErrorEncode(0x1fffff, array[0..], error.CodepointTooLarge);
308308}
309309
310fn testErrorEncode(codePoint: u32, array: []u8, expectedErr: anyerror) void {
310fn testErrorEncode(codePoint: u21, array: []u8, expectedErr: anyerror) void {
311311 testing.expectError(expectedErr, utf8Encode(codePoint, array));
312312}
313313
......@@ -455,11 +455,11 @@ fn testError(bytes: []const u8, expected_err: anyerror) void {
455455 testing.expectError(expected_err, testDecode(bytes));
456456}
457457
458fn testValid(bytes: []const u8, expected_codepoint: u32) void {
458fn testValid(bytes: []const u8, expected_codepoint: u21) void {
459459 testing.expect((testDecode(bytes) catch unreachable) == expected_codepoint);
460460}
461461
462fn testDecode(bytes: []const u8) !u32 {
462fn testDecode(bytes: []const u8) !u21 {
463463 const length = try utf8ByteSequenceLength(bytes[0]);
464464 if (bytes.len < length) return error.UnexpectedEof;
465465 testing.expect(bytes.len == length);