| 1 |
#!/usr/bin/env python3 |
| 2 |
|
| 3 |
import argparse |
| 4 |
|
| 5 |
|
| 6 |
class Codepoint: |
| 7 |
def __init__(self, start: int, end: None | int = None): |
| 8 |
self.start = start |
| 9 |
self.end = start if end is None else end |
| 10 |
self.vs15 = False |
| 11 |
self.vs16 = False |
| 12 |
|
| 13 |
def __repr__(self) -> str: |
| 14 |
return f'{self.start:x}-{self.end:x}, vs15={self.vs15}, vs16={self.vs16}' |
| 15 |
|
| 16 |
|
| 17 |
def main() -> None: |
| 18 |
parser = argparse.ArgumentParser() |
| 19 |
parser.add_argument('input', type=argparse.FileType('r')) |
| 20 |
parser.add_argument('output', type=argparse.FileType('w')) |
| 21 |
opts = parser.parse_args() |
| 22 |
|
| 23 |
codepoints: dict[int, Codepoint] = {} |
| 24 |
|
| 25 |
for line in opts.input: |
| 26 |
line = line.rstrip() |
| 27 |
if not line: |
| 28 |
continue |
| 29 |
if line[0] == '#': |
| 30 |
continue |
| 31 |
|
| 32 |
# Example: "0023 FE0E ; text style; # (1.1) NUMBER SIGN" |
| 33 |
cps, _ = line.split(';', maxsplit=1) # cps = "0023 FE0F " |
| 34 |
cps = cps.strip().split(' ') # cps = ["0023", "FE0F"] |
| 35 |
|
| 36 |
if len(cps) != 2: |
| 37 |
raise NotImplementedError(f'emoji variation sequences with more than one base codepoint: {cps}') |
| 38 |
|
| 39 |
cp, vs = cps # cp = "0023", vs = "FE0F" |
| 40 |
cp = int(cp, 16) # cp = 0x23 |
| 41 |
vs = int(vs, 16) # vs = 0xfe0f |
| 42 |
|
| 43 |
assert vs in [0xfe0e, 0xfe0f] |
| 44 |
|
| 45 |
if cp not in codepoints: |
| 46 |
codepoints[cp] = Codepoint(cp) |
| 47 |
|
| 48 |
assert codepoints[cp].start == cp |
| 49 |
|
| 50 |
if vs == 0xfe0e: |
| 51 |
codepoints[cp].vs15 = True |
| 52 |
else: |
| 53 |
codepoints[cp].vs16 = True |
| 54 |
|
| 55 |
sorted_list = sorted(codepoints.values(), key=lambda cp: cp.start) |
| 56 |
|
| 57 |
compacted: list[Codepoint] = [] |
| 58 |
for i, cp in enumerate(sorted_list): |
| 59 |
assert cp.end == cp.start |
| 60 |
|
| 61 |
if i == 0: |
| 62 |
compacted.append(cp) |
| 63 |
continue |
| 64 |
|
| 65 |
last_cp = compacted[-1] |
| 66 |
if last_cp.end == cp.start - 1 and last_cp.vs15 == cp.vs15 and last_cp.vs16 == cp.vs16: |
| 67 |
compacted[-1].end = cp.start |
| 68 |
else: |
| 69 |
compacted.append(cp) |
| 70 |
|
| 71 |
opts.output.write('#pragma once\n') |
| 72 |
opts.output.write('#include <stdint.h>\n') |
| 73 |
opts.output.write('#include <stdbool.h>\n') |
| 74 |
opts.output.write('\n') |
| 75 |
opts.output.write('struct emoji_vs {\n') |
| 76 |
opts.output.write(' uint32_t start:21;\n') |
| 77 |
opts.output.write(' uint32_t end:21;\n') |
| 78 |
opts.output.write(' bool vs15:1;\n') |
| 79 |
opts.output.write(' bool vs16:1;\n') |
| 80 |
opts.output.write('} __attribute__((packed));\n') |
| 81 |
opts.output.write('_Static_assert(sizeof(struct emoji_vs) == 6, "unexpected struct size");\n') |
| 82 |
opts.output.write('\n') |
| 83 |
opts.output.write('#if defined(FOOT_GRAPHEME_CLUSTERING)\n') |
| 84 |
opts.output.write('\n') |
| 85 |
|
| 86 |
opts.output.write(f'static const struct emoji_vs emoji_vs[{len(compacted)}] = {{\n') |
| 87 |
|
| 88 |
for cp in compacted: |
| 89 |
opts.output.write(' {\n') |
| 90 |
opts.output.write(f' .start = 0x{cp.start:X},\n') |
| 91 |
opts.output.write(f' .end = 0x{cp.end:x},\n') |
| 92 |
opts.output.write(f' .vs15 = {"true" if cp.vs15 else "false"},\n') |
| 93 |
opts.output.write(f' .vs16 = {"true" if cp.vs16 else "false"},\n') |
| 94 |
opts.output.write(' },\n') |
| 95 |
|
| 96 |
opts.output.write('};\n') |
| 97 |
opts.output.write('\n') |
| 98 |
opts.output.write('#endif /* FOOT_GRAPHEME_CLUSTERING */\n') |
| 99 |
|
| 100 |
|
| 101 |
if __name__ == '__main__': |
| 102 |
main() |