wuxu
c7ef3c0047
Extend the bitmap font pipeline with the Chinese Joker glyph subset and encode supported UTF-8 text for Tonc TTE. Move the affine map clear of the expanded font, add a compile-time VRAM overlap guard, and cover every custom glyph with host tests.
114 lines
2.6 KiB
C
114 lines
2.6 KiB
C
#include "chinese_text.h"
|
|
|
|
#include <stdint.h>
|
|
|
|
#define CHINESE_GLYPH_FIRST 128
|
|
|
|
static const uint32_t s_chinese_codepoints[] = {
|
|
#define CHINESE_GLYPH(codepoint) codepoint,
|
|
#include "def_chinese_glyph_table.h"
|
|
#undef CHINESE_GLYPH
|
|
};
|
|
|
|
_Static_assert(
|
|
sizeof(s_chinese_codepoints) / sizeof(s_chinese_codepoints[0]) <= 256 - CHINESE_GLYPH_FIRST,
|
|
"Chinese glyph table exceeds the single-byte encoding range"
|
|
);
|
|
|
|
static uint32_t decode_utf8_codepoint(const unsigned char** cursor)
|
|
{
|
|
const unsigned char* text = *cursor;
|
|
|
|
if (text[0] < 0x80)
|
|
{
|
|
*cursor += 1;
|
|
return text[0];
|
|
}
|
|
|
|
if ((text[0] & 0xE0) == 0xC0 && (text[1] & 0xC0) == 0x80)
|
|
{
|
|
*cursor += 2;
|
|
return ((uint32_t)(text[0] & 0x1F) << 6) | (text[1] & 0x3F);
|
|
}
|
|
|
|
if ((text[0] & 0xF0) == 0xE0 && (text[1] & 0xC0) == 0x80 && (text[2] & 0xC0) == 0x80)
|
|
{
|
|
*cursor += 3;
|
|
return ((uint32_t)(text[0] & 0x0F) << 12) | ((uint32_t)(text[1] & 0x3F) << 6) |
|
|
(text[2] & 0x3F);
|
|
}
|
|
|
|
*cursor += 1;
|
|
return '?';
|
|
}
|
|
|
|
static uint32_t codepoint_to_glyph(uint32_t codepoint)
|
|
{
|
|
for (size_t i = 0; i < sizeof(s_chinese_codepoints) / sizeof(s_chinese_codepoints[0]); i++)
|
|
{
|
|
if (s_chinese_codepoints[i] == codepoint)
|
|
{
|
|
return CHINESE_GLYPH_FIRST + i;
|
|
}
|
|
}
|
|
|
|
return '?';
|
|
}
|
|
|
|
static size_t encode_tte_codepoint(uint32_t codepoint, char* output, size_t output_size)
|
|
{
|
|
if (codepoint < 0x80)
|
|
{
|
|
if (output_size < 1)
|
|
{
|
|
return 0;
|
|
}
|
|
|
|
output[0] = (char)codepoint;
|
|
return 1;
|
|
}
|
|
|
|
if (codepoint <= 0x7FF)
|
|
{
|
|
if (output_size < 2)
|
|
{
|
|
return 0;
|
|
}
|
|
|
|
output[0] = (char)(0xC0 | (codepoint >> 6));
|
|
output[1] = (char)(0x80 | (codepoint & 0x3F));
|
|
return 2;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
size_t chinese_text_encode(const char* utf8, char* output, size_t output_size)
|
|
{
|
|
if (utf8 == NULL || output == NULL || output_size == 0)
|
|
{
|
|
return 0;
|
|
}
|
|
|
|
const unsigned char* cursor = (const unsigned char*)utf8;
|
|
size_t output_len = 0;
|
|
|
|
while (*cursor != '\0')
|
|
{
|
|
uint32_t codepoint = decode_utf8_codepoint(&cursor);
|
|
uint32_t tte_codepoint = codepoint < 0x80 ? codepoint : codepoint_to_glyph(codepoint);
|
|
size_t encoded_size =
|
|
encode_tte_codepoint(tte_codepoint, output + output_len, output_size - output_len - 1);
|
|
|
|
if (encoded_size == 0)
|
|
{
|
|
break;
|
|
}
|
|
|
|
output_len += encoded_size;
|
|
}
|
|
|
|
output[output_len] = '\0';
|
|
return output_len;
|
|
}
|