diff options
| -rw-r--r-- | README.md | 28 | ||||
| -rw-r--r-- | build.zig | 27 | ||||
| -rw-r--r-- | codegen/emoji.zig | 146 | ||||
| -rw-r--r-- | src/Emoji.zig | 132 |
4 files changed, 333 insertions, 0 deletions
| @@ -520,6 +520,34 @@ test "Scripts" { | |||
| 520 | } | 520 | } |
| 521 | ``` | 521 | ``` |
| 522 | 522 | ||
| 523 | ## Emoji | ||
| 524 | |||
| 525 | To get information about emoji and emoji-like characters, use the `Emoji` module. | ||
| 526 | |||
| 527 | In your `build.zig`: | ||
| 528 | |||
| 529 | ```zig | ||
| 530 | exe.root_module.addImport("Emoji", zg.module("Emoji")); | ||
| 531 | ``` | ||
| 532 | |||
| 533 | In your code: | ||
| 534 | |||
| 535 | ```zig | ||
| 536 | const Emoji = @import("Emoji"); | ||
| 537 | |||
| 538 | test "Emoji" { | ||
| 539 | const emoji = try Emoji.init(allocator); | ||
| 540 | defer emoji.deinit(allocator); | ||
| 541 | |||
| 542 | try expect(emoji.isEmoji(0x1F415)); // π | ||
| 543 | try expect(emoji.isEmojiPresentation(0x1F408)); // π | ||
| 544 | try expect(emoji.isEmojiModifier(0x1F3FF)); // πΏ | ||
| 545 | try expect(emoji.isEmojiModifierBase(0x1F977)); // π₯· | ||
| 546 | try expect(emoji.isEmojiComponent(0x1F9B0)); // π¦° | ||
| 547 | try expect(emoji.isExtendedPictographic(0x1F005)); // π | ||
| 548 | } | ||
| 549 | ``` | ||
| 550 | |||
| 523 | ## Relation to Ziglyph | 551 | ## Relation to Ziglyph |
| 524 | 552 | ||
| 525 | zg is a total re-write of some of the components of Ziglyph. The idea was to | 553 | zg is a total re-write of some of the components of Ziglyph. The idea was to |
| @@ -22,6 +22,16 @@ pub fn build(b: *std.Build) void { | |||
| 22 | const run_gbp_gen_exe = b.addRunArtifact(gbp_gen_exe); | 22 | const run_gbp_gen_exe = b.addRunArtifact(gbp_gen_exe); |
| 23 | const gbp_gen_out = run_gbp_gen_exe.addOutputFileArg("gbp.bin.z"); | 23 | const gbp_gen_out = run_gbp_gen_exe.addOutputFileArg("gbp.bin.z"); |
| 24 | 24 | ||
| 25 | // Emoji | ||
| 26 | const emoji_gen_exe = b.addExecutable(.{ | ||
| 27 | .name = "emoji", | ||
| 28 | .root_source_file = b.path("codegen/emoji.zig"), | ||
| 29 | .target = b.graph.host, | ||
| 30 | .optimize = .Debug, | ||
| 31 | }); | ||
| 32 | const run_emoji_gen_exe = b.addRunArtifact(emoji_gen_exe); | ||
| 33 | const emoji_gen_out = run_emoji_gen_exe.addOutputFileArg("emoji.bin.z"); | ||
| 34 | |||
| 25 | // Display width | 35 | // Display width |
| 26 | const cjk = b.option(bool, "cjk", "Ambiguous code points are wide (display width: 2).") orelse false; | 36 | const cjk = b.option(bool, "cjk", "Ambiguous code points are wide (display width: 2).") orelse false; |
| 27 | const options = b.addOptions(); | 37 | const options = b.addOptions(); |
| @@ -215,6 +225,22 @@ pub fn build(b: *std.Build) void { | |||
| 215 | }); | 225 | }); |
| 216 | const grapheme_tr = b.addRunArtifact(grapheme_t); | 226 | const grapheme_tr = b.addRunArtifact(grapheme_t); |
| 217 | 227 | ||
| 228 | // Emoji module | ||
| 229 | const emoji = b.addModule("Emoji", .{ | ||
| 230 | .root_source_file = b.path("src/Emoji.zig"), | ||
| 231 | .target = target, | ||
| 232 | .optimize = optimize, | ||
| 233 | }); | ||
| 234 | emoji.addAnonymousImport("emoji", .{ .root_source_file = emoji_gen_out }); | ||
| 235 | |||
| 236 | const emoji_t = b.addTest(.{ | ||
| 237 | .name = "Emoji", | ||
| 238 | .root_module = emoji, | ||
| 239 | .target = target, | ||
| 240 | .optimize = optimize, | ||
| 241 | }); | ||
| 242 | const emoji_tr = b.addRunArtifact(emoji_t); | ||
| 243 | |||
| 218 | // ASCII utilities | 244 | // ASCII utilities |
| 219 | const ascii = b.addModule("ascii", .{ | 245 | const ascii = b.addModule("ascii", .{ |
| 220 | .root_source_file = b.path("src/ascii.zig"), | 246 | .root_source_file = b.path("src/ascii.zig"), |
| @@ -452,6 +478,7 @@ pub fn build(b: *std.Build) void { | |||
| 452 | test_step.dependOn(&code_point_tr.step); | 478 | test_step.dependOn(&code_point_tr.step); |
| 453 | test_step.dependOn(&display_width_tr.step); | 479 | test_step.dependOn(&display_width_tr.step); |
| 454 | test_step.dependOn(&grapheme_tr.step); | 480 | test_step.dependOn(&grapheme_tr.step); |
| 481 | test_step.dependOn(&emoji_tr.step); | ||
| 455 | test_step.dependOn(&ascii_tr.step); | 482 | test_step.dependOn(&ascii_tr.step); |
| 456 | test_step.dependOn(&ccc_data_tr.step); | 483 | test_step.dependOn(&ccc_data_tr.step); |
| 457 | test_step.dependOn(&canon_data_tr.step); | 484 | test_step.dependOn(&canon_data_tr.step); |
diff --git a/codegen/emoji.zig b/codegen/emoji.zig new file mode 100644 index 0000000..0a4dbe6 --- /dev/null +++ b/codegen/emoji.zig | |||
| @@ -0,0 +1,146 @@ | |||
| 1 | const std = @import("std"); | ||
| 2 | const builtin = @import("builtin"); | ||
| 3 | |||
| 4 | pub const Emoji = packed struct { | ||
| 5 | Emoji: bool = false, | ||
| 6 | Emoji_Presentation: bool = false, | ||
| 7 | Emoji_Modifier: bool = false, | ||
| 8 | Emoji_Modifier_Base: bool = false, | ||
| 9 | Emoji_Component: bool = false, | ||
| 10 | Extended_Pictographic: bool = false, | ||
| 11 | }; | ||
| 12 | |||
| 13 | const block_size = 256; | ||
| 14 | const Block = [block_size]u6; | ||
| 15 | |||
| 16 | comptime { | ||
| 17 | if (@bitSizeOf(u6) != @bitSizeOf(Emoji)) { | ||
| 18 | @compileError("Emoji doesn't have expected bit size."); | ||
| 19 | } | ||
| 20 | } | ||
| 21 | |||
| 22 | const BlockMap = std.HashMap( | ||
| 23 | Block, | ||
| 24 | u16, | ||
| 25 | struct { | ||
| 26 | pub fn hash(_: @This(), k: Block) u64 { | ||
| 27 | var hasher = std.hash.Wyhash.init(0); | ||
| 28 | std.hash.autoHashStrat(&hasher, k, .DeepRecursive); | ||
| 29 | return hasher.final(); | ||
| 30 | } | ||
| 31 | |||
| 32 | pub fn eql(_: @This(), a: Block, b: Block) bool { | ||
| 33 | return std.mem.eql(u6, &a, &b); | ||
| 34 | } | ||
| 35 | }, | ||
| 36 | std.hash_map.default_max_load_percentage, | ||
| 37 | ); | ||
| 38 | |||
| 39 | pub fn main() !void { | ||
| 40 | var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator); | ||
| 41 | defer arena.deinit(); | ||
| 42 | const allocator = arena.allocator(); | ||
| 43 | |||
| 44 | var emoji_map = std.AutoHashMap(u21, Emoji).init(allocator); | ||
| 45 | defer emoji_map.deinit(); | ||
| 46 | |||
| 47 | var line_buf: [4096]u8 = undefined; | ||
| 48 | |||
| 49 | // Process Emoji | ||
| 50 | var in_file = try std.fs.cwd().openFile("data/unicode/emoji/emoji-data.txt", .{}); | ||
| 51 | defer in_file.close(); | ||
| 52 | var in_buf = std.io.bufferedReader(in_file.reader()); | ||
| 53 | const in_reader = in_buf.reader(); | ||
| 54 | |||
| 55 | while (try in_reader.readUntilDelimiterOrEof(&line_buf, '\n')) |line| { | ||
| 56 | if (line.len == 0 or line[0] == '#') continue; | ||
| 57 | const no_comment = if (std.mem.indexOfScalar(u8, line, '#')) |octo| line[0..octo] else line; | ||
| 58 | |||
| 59 | var field_iter = std.mem.tokenizeAny(u8, no_comment, "; "); | ||
| 60 | var current_code: [2]u21 = undefined; | ||
| 61 | |||
| 62 | var i: usize = 0; | ||
| 63 | while (field_iter.next()) |field| : (i += 1) { | ||
| 64 | switch (i) { | ||
| 65 | 0 => { | ||
| 66 | // Code point(s) | ||
| 67 | if (std.mem.indexOf(u8, field, "..")) |dots| { | ||
| 68 | current_code = .{ | ||
| 69 | try std.fmt.parseInt(u21, field[0..dots], 16), | ||
| 70 | try std.fmt.parseInt(u21, field[dots + 2 ..], 16), | ||
| 71 | }; | ||
| 72 | } else { | ||
| 73 | const code = try std.fmt.parseInt(u21, field, 16); | ||
| 74 | current_code = .{ code, code }; | ||
| 75 | } | ||
| 76 | }, | ||
| 77 | 1 => { | ||
| 78 | // Prop | ||
| 79 | const prop = std.meta.stringToEnum(std.meta.FieldEnum(Emoji), field) orelse return error.InvalidProp; | ||
| 80 | for (current_code[0]..current_code[1] + 1) |code| { | ||
| 81 | const cp: u21 = @intCast(code); | ||
| 82 | const gop = try emoji_map.getOrPut(cp); | ||
| 83 | if (!gop.found_existing) gop.value_ptr.* = .{}; | ||
| 84 | switch (prop) { | ||
| 85 | inline else => |tag| { | ||
| 86 | @field(gop.value_ptr.*, @tagName(tag)) = true; | ||
| 87 | }, | ||
| 88 | } | ||
| 89 | } | ||
| 90 | }, | ||
| 91 | else => {}, | ||
| 92 | } | ||
| 93 | } | ||
| 94 | } | ||
| 95 | |||
| 96 | var blocks_map = BlockMap.init(allocator); | ||
| 97 | defer blocks_map.deinit(); | ||
| 98 | |||
| 99 | var stage1 = std.ArrayList(u16).init(allocator); | ||
| 100 | defer stage1.deinit(); | ||
| 101 | |||
| 102 | var stage2 = std.ArrayList(u6).init(allocator); | ||
| 103 | defer stage2.deinit(); | ||
| 104 | |||
| 105 | var block: Block = [_]u6{0} ** block_size; | ||
| 106 | var block_len: u16 = 0; | ||
| 107 | |||
| 108 | for (0..0x110000) |i| { | ||
| 109 | const cp: u21 = @intCast(i); | ||
| 110 | const emoji = emoji_map.get(cp) orelse Emoji{}; | ||
| 111 | |||
| 112 | block[block_len] = @bitCast(emoji); | ||
| 113 | block_len += 1; | ||
| 114 | |||
| 115 | if (block_len < block_size and cp != 0x10ffff) continue; | ||
| 116 | |||
| 117 | const gop = try blocks_map.getOrPut(block); | ||
| 118 | if (!gop.found_existing) { | ||
| 119 | gop.value_ptr.* = @intCast(stage2.items.len); | ||
| 120 | try stage2.appendSlice(&block); | ||
| 121 | } | ||
| 122 | |||
| 123 | try stage1.append(gop.value_ptr.*); | ||
| 124 | block_len = 0; | ||
| 125 | } | ||
| 126 | |||
| 127 | var args_iter = try std.process.argsWithAllocator(allocator); | ||
| 128 | defer args_iter.deinit(); | ||
| 129 | _ = args_iter.skip(); | ||
| 130 | const output_path = args_iter.next() orelse @panic("No output file arg!"); | ||
| 131 | |||
| 132 | const compressor = std.compress.flate.deflate.compressor; | ||
| 133 | var out_file = try std.fs.cwd().createFile(output_path, .{}); | ||
| 134 | defer out_file.close(); | ||
| 135 | var out_comp = try compressor(.raw, out_file.writer(), .{ .level = .best }); | ||
| 136 | const writer = out_comp.writer(); | ||
| 137 | |||
| 138 | const endian = builtin.cpu.arch.endian(); | ||
| 139 | try writer.writeInt(u16, @intCast(stage1.items.len), endian); | ||
| 140 | for (stage1.items) |i| try writer.writeInt(u16, i, endian); | ||
| 141 | |||
| 142 | try writer.writeInt(u16, @intCast(stage2.items.len), endian); | ||
| 143 | for (stage2.items) |i| try writer.writeInt(u8, i, endian); | ||
| 144 | |||
| 145 | try out_comp.flush(); | ||
| 146 | } | ||
diff --git a/src/Emoji.zig b/src/Emoji.zig new file mode 100644 index 0000000..bf7014d --- /dev/null +++ b/src/Emoji.zig | |||
| @@ -0,0 +1,132 @@ | |||
| 1 | const std = @import("std"); | ||
| 2 | const builtin = @import("builtin"); | ||
| 3 | const mem = std.mem; | ||
| 4 | const Allocator = mem.Allocator; | ||
| 5 | const compress = std.compress; | ||
| 6 | const unicode = std.unicode; | ||
| 7 | |||
| 8 | const CodePoint = @import("code_point").CodePoint; | ||
| 9 | const CodePointIterator = @import("code_point").Iterator; | ||
| 10 | |||
| 11 | s1: []u16 = undefined, | ||
| 12 | s2: []u6 = undefined, | ||
| 13 | |||
| 14 | const Emoji = @This(); | ||
| 15 | |||
| 16 | // This must be an exact match of `Emoji` from `codegen/emoji.zig`. | ||
| 17 | pub const Properties = packed struct { | ||
| 18 | Emoji: bool = false, | ||
| 19 | Emoji_Presentation: bool = false, | ||
| 20 | Emoji_Modifier: bool = false, | ||
| 21 | Emoji_Modifier_Base: bool = false, | ||
| 22 | Emoji_Component: bool = false, | ||
| 23 | Extended_Pictographic: bool = false, | ||
| 24 | }; | ||
| 25 | |||
| 26 | pub fn init(allocator: Allocator) Allocator.Error!Emoji { | ||
| 27 | var emoji = Emoji{}; | ||
| 28 | try emoji.setup(allocator); | ||
| 29 | return emoji; | ||
| 30 | } | ||
| 31 | |||
| 32 | pub fn setup(emoji: *Emoji, allocator: Allocator) Allocator.Error!void { | ||
| 33 | const decompressor = compress.flate.inflate.decompressor; | ||
| 34 | const in_bytes = @embedFile("emoji"); | ||
| 35 | var in_fbs = std.io.fixedBufferStream(in_bytes); | ||
| 36 | var in_decomp = decompressor(.raw, in_fbs.reader()); | ||
| 37 | var reader = in_decomp.reader(); | ||
| 38 | |||
| 39 | const endian = builtin.cpu.arch.endian(); | ||
| 40 | |||
| 41 | const s1_len: u16 = reader.readInt(u16, endian) catch unreachable; | ||
| 42 | emoji.s1 = try allocator.alloc(u16, s1_len); | ||
| 43 | errdefer allocator.free(emoji.s1); | ||
| 44 | for (0..s1_len) |i| emoji.s1[i] = reader.readInt(u16, endian) catch unreachable; | ||
| 45 | |||
| 46 | const s2_len: u16 = reader.readInt(u16, endian) catch unreachable; | ||
| 47 | emoji.s2 = try allocator.alloc(u6, s2_len); | ||
| 48 | errdefer allocator.free(emoji.s2); | ||
| 49 | for (0..s2_len) |i| emoji.s2[i] = @intCast(reader.readInt(u8, endian) catch unreachable); | ||
| 50 | } | ||
| 51 | |||
| 52 | pub fn deinit(emoji: *const Emoji, allocator: Allocator) void { | ||
| 53 | allocator.free(emoji.s1); | ||
| 54 | allocator.free(emoji.s2); | ||
| 55 | } | ||
| 56 | |||
| 57 | /// Lookup the emoji properties for a code point. | ||
| 58 | fn properties(emoji: Emoji, cp: u21) Properties { | ||
| 59 | return @bitCast(emoji.s2[emoji.s1[cp >> 8] + (cp & 0xff)]); | ||
| 60 | } | ||
| 61 | |||
| 62 | pub fn isEmoji(emoji: Emoji, cp: u21) bool { | ||
| 63 | return properties(emoji, cp).Emoji; | ||
| 64 | } | ||
| 65 | |||
| 66 | pub fn isEmojiPresentation(emoji: Emoji, cp: u21) bool { | ||
| 67 | return properties(emoji, cp).Emoji_Presentation; | ||
| 68 | } | ||
| 69 | |||
| 70 | pub fn isEmojiModifier(emoji: Emoji, cp: u21) bool { | ||
| 71 | return properties(emoji, cp).Emoji_Modifier; | ||
| 72 | } | ||
| 73 | |||
| 74 | pub fn isEmojiModifierBase(emoji: Emoji, cp: u21) bool { | ||
| 75 | return properties(emoji, cp).Emoji_Modifier_Base; | ||
| 76 | } | ||
| 77 | |||
| 78 | pub fn isEmojiComponent(emoji: Emoji, cp: u21) bool { | ||
| 79 | return properties(emoji, cp).Emoji_Component; | ||
| 80 | } | ||
| 81 | |||
| 82 | pub fn isExtendedPictographic(emoji: Emoji, cp: u21) bool { | ||
| 83 | return properties(emoji, cp).Extended_Pictographic; | ||
| 84 | } | ||
| 85 | |||
| 86 | test "isEmoji" { | ||
| 87 | const emoji = try Emoji.init(std.testing.allocator); | ||
| 88 | defer emoji.deinit(std.testing.allocator); | ||
| 89 | |||
| 90 | try std.testing.expect(emoji.isEmoji(0x1F415)); // π | ||
| 91 | try std.testing.expect(!emoji.isEmoji(0x3042)); // γ | ||
| 92 | } | ||
| 93 | |||
| 94 | test "isEmojiPresentation" { | ||
| 95 | const emoji = try Emoji.init(std.testing.allocator); | ||
| 96 | defer emoji.deinit(std.testing.allocator); | ||
| 97 | |||
| 98 | try std.testing.expect(emoji.isEmojiPresentation(0x1F408)); // π | ||
| 99 | try std.testing.expect(!emoji.isEmojiPresentation(0x267E)); // βΎοΈ | ||
| 100 | } | ||
| 101 | |||
| 102 | test "isEmojiModifier" { | ||
| 103 | const emoji = try Emoji.init(std.testing.allocator); | ||
| 104 | defer emoji.deinit(std.testing.allocator); | ||
| 105 | |||
| 106 | try std.testing.expect(emoji.isEmojiModifier(0x1F3FF)); // πΏ | ||
| 107 | try std.testing.expect(!emoji.isEmojiModifier(0x1F385)); // π | ||
| 108 | } | ||
| 109 | |||
| 110 | test "isEmojiModifierBase" { | ||
| 111 | const emoji = try Emoji.init(std.testing.allocator); | ||
| 112 | defer emoji.deinit(std.testing.allocator); | ||
| 113 | |||
| 114 | try std.testing.expect(emoji.isEmojiModifierBase(0x1F977)); // π₯· | ||
| 115 | try std.testing.expect(!emoji.isEmojiModifierBase(0x1F4F8)); // πΈ | ||
| 116 | } | ||
| 117 | |||
| 118 | test "isEmojiComponent" { | ||
| 119 | const emoji = try Emoji.init(std.testing.allocator); | ||
| 120 | defer emoji.deinit(std.testing.allocator); | ||
| 121 | |||
| 122 | try std.testing.expect(emoji.isEmojiComponent(0x1F9B0)); // π¦° | ||
| 123 | try std.testing.expect(!emoji.isEmojiComponent(0x1F9B5)); // 𦡠| ||
| 124 | } | ||
| 125 | |||
| 126 | test "isExtendedPictographic" { | ||
| 127 | const emoji = try Emoji.init(std.testing.allocator); | ||
| 128 | defer emoji.deinit(std.testing.allocator); | ||
| 129 | |||
| 130 | try std.testing.expect(emoji.isExtendedPictographic(0x1F005)); // π | ||
| 131 | try std.testing.expect(!emoji.isExtendedPictographic(0x2A)); // * | ||
| 132 | } | ||