1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
// This file is part of the Luau programming language and is licensed under MIT License; see LICENSE.txt for details
// This code is based on Lua 5.x implementation licensed under MIT License; see lua_LICENSE.txt for details
#include "lualib.h"
#include "lcommon.h"
#define MAXUNICODE 0x10FFFF
#define iscont(p) ((*(p)&0xC0) == 0x80)
/* from strlib */
/* translate a relative string position: negative means back from end */
static int u_posrelat(int pos, size_t len)
{
if (pos >= 0)
return pos;
else if (0u - (size_t)pos > len)
return 0;
else
return (int)len + pos + 1;
}
/*
** Decode one UTF-8 sequence, returning NULL if byte sequence is invalid.
*/
static const char* utf8_decode(const char* o, int* val)
{
static const unsigned int limits[] = {0xFF, 0x7F, 0x7FF, 0xFFFF};
const unsigned char* s = (const unsigned char*)o;
unsigned int c = s[0];
unsigned int res = 0; /* final result */
if (c < 0x80) /* ascii? */
res = c;
else
{
int count = 0; /* to count number of continuation bytes */
while (c & 0x40)
{ /* still have continuation bytes? */
int cc = s[++count]; /* read next byte */
if ((cc & 0xC0) != 0x80) /* not a continuation byte? */
return NULL; /* invalid byte sequence */
res = (res << 6) | (cc & 0x3F); /* add lower 6 bits from cont. byte */
c <<= 1; /* to test next bit */
}
res |= ((c & 0x7F) << (count * 5)); /* add first byte */
if (count > 3 || res > MAXUNICODE || res <= limits[count])
return NULL; /* invalid byte sequence */
s += count; /* skip continuation bytes read */
}
if (val)
*val = res;
return (const char*)s + 1; /* +1 to include first byte */
}
/*
** utf8len(s [, i [, j]]) --> number of characters that start in the
** range [i,j], or nil + current position if 's' is not well formed in
** that interval
*/
static int utflen(lua_State* L)
{
int n = 0;
size_t len;
const char* s = luaL_checklstring(L, 1, &len);
int posi = u_posrelat(luaL_optinteger(L, 2, 1), len);
int posj = u_posrelat(luaL_optinteger(L, 3, -1), len);
luaL_argcheck(L, 1 <= posi && --posi <= (int)len, 2, "initial position out of string");
luaL_argcheck(L, --posj < (int)len, 3, "final position out of string");
while (posi <= posj)
{
const char* s1 = utf8_decode(s + posi, NULL);
if (s1 == NULL)
{ /* conversion error? */
lua_pushnil(L); /* return nil ... */
lua_pushinteger(L, posi + 1); /* ... and current position */
return 2;
}
posi = (int)(s1 - s);
n++;
}
lua_pushinteger(L, n);
return 1;
}
/*
** codepoint(s, [i, [j]]) -> returns codepoints for all characters
** that start in the range [i,j]
*/
static int codepoint(lua_State* L)
{
size_t len;
const char* s = luaL_checklstring(L, 1, &len);
int posi = u_posrelat(luaL_optinteger(L, 2, 1), len);
int pose = u_posrelat(luaL_optinteger(L, 3, posi), len);
int n;
const char* se;
luaL_argcheck(L, posi >= 1, 2, "out of range");
luaL_argcheck(L, pose <= (int)len, 3, "out of range");
if (posi > pose)
return 0; /* empty interval; return no values */
if (pose - posi >= INT_MAX) /* (int -> int) overflow? */
luaL_error(L, "string slice too long");
n = (int)(pose - posi) + 1;
luaL_checkstack(L, n, "string slice too long");
n = 0;
se = s + pose;
for (s += posi - 1; s < se;)
{
int code;
s = utf8_decode(s, &code);
if (s == NULL)
luaL_error(L, "invalid UTF-8 code");
lua_pushinteger(L, code);
n++;
}
return n;
}
// from Lua 5.3 lobject.h
#define UTF8BUFFSZ 8
// from Lua 5.3 lobject.c, copied verbatim + static
static int luaO_utf8esc(char* buff, unsigned long x)
{
int n = 1; /* number of bytes put in buffer (backwards) */
LUAU_ASSERT(x <= 0x10FFFF);
if (x < 0x80) /* ascii? */
buff[UTF8BUFFSZ - 1] = cast_to(char, x);
else
{ /* need continuation bytes */
unsigned int mfb = 0x3f; /* maximum that fits in first byte */
do
{ /* add continuation bytes */
buff[UTF8BUFFSZ - (n++)] = cast_to(char, 0x80 | (x & 0x3f));
x >>= 6; /* remove added bits */
mfb >>= 1; /* now there is one less bit available in first byte */
} while (x > mfb); /* still needs continuation byte? */
buff[UTF8BUFFSZ - n] = cast_to(char, (~mfb << 1) | x); /* add first byte */
}
return n;
}
// lighter replacement for pushutfchar; doesn't push any string onto the stack
static int buffutfchar(lua_State* L, int arg, char* buff, const char** charstr)
{
int code = luaL_checkinteger(L, arg);
luaL_argcheck(L, 0 <= code && code <= MAXUNICODE, arg, "value out of range");
int l = luaO_utf8esc(buff, cast_to(long, code));
*charstr = buff + UTF8BUFFSZ - l;
return l;
}
/*
** utfchar(n1, n2, ...) -> char(n1)..char(n2)...
**
** This version avoids the need to make more invasive upgrades elsewhere (like
** implementing the %U escape in lua_pushfstring) and avoids pushing string
** objects for each codepoint in the multi-argument case. -Jovanni
*/
static int utfchar(lua_State* L)
{
char buff[UTF8BUFFSZ];
const char* charstr;
int n = lua_gettop(L); /* number of arguments */
if (n == 1)
{ /* optimize common case of single char */
int l = buffutfchar(L, 1, buff, &charstr);
lua_pushlstring(L, charstr, l);
}
else
{
luaL_Buffer b;
luaL_buffinit(L, &b);
for (int i = 1; i <= n; i++)
{
int l = buffutfchar(L, i, buff, &charstr);
luaL_addlstring(&b, charstr, l);
}
luaL_pushresult(&b);
}
return 1;
}
/*
** offset(s, n, [i]) -> index where n-th character counting from
** position 'i' starts; 0 means character at 'i'.
*/
static int byteoffset(lua_State* L)
{
size_t len;
const char* s = luaL_checklstring(L, 1, &len);
int n = luaL_checkinteger(L, 2);
int posi = (n >= 0) ? 1 : (int)len + 1;
posi = u_posrelat(luaL_optinteger(L, 3, posi), len);
luaL_argcheck(L, 1 <= posi && --posi <= (int)len, 3, "position out of range");
if (n == 0)
{
/* find beginning of current byte sequence */
while (posi > 0 && iscont(s + posi))
posi--;
}
else
{
if (iscont(s + posi))
luaL_error(L, "initial position is a continuation byte");
if (n < 0)
{
while (n < 0 && posi > 0)
{ /* move back */
do
{ /* find beginning of previous character */
posi--;
} while (posi > 0 && iscont(s + posi));
n++;
}
}
else
{
n--; /* do not move for 1st character */
while (n > 0 && posi < (int)len)
{
do
{ /* find beginning of next character */
posi++;
} while (iscont(s + posi)); /* (cannot pass final '\0') */
n--;
}
}
}
if (n == 0) /* did it find given character? */
lua_pushinteger(L, posi + 1);
else /* no such character */
lua_pushnil(L);
return 1;
}
static int iter_aux(lua_State* L)
{
size_t len;
const char* s = luaL_checklstring(L, 1, &len);
int n = lua_tointeger(L, 2) - 1;
if (n < 0) /* first iteration? */
n = 0; /* start from here */
else if (n < (int)len)
{
n++; /* skip current byte */
while (iscont(s + n))
n++; /* and its continuations */
}
if (n >= (int)len)
return 0; /* no more codepoints */
else
{
int code;
const char* next = utf8_decode(s + n, &code);
if (next == NULL || iscont(next))
luaL_error(L, "invalid UTF-8 code");
lua_pushinteger(L, n + 1);
lua_pushinteger(L, code);
return 2;
}
}
static int iter_codes(lua_State* L)
{
luaL_checkstring(L, 1);
lua_pushcfunction(L, iter_aux, NULL);
lua_pushvalue(L, 1);
lua_pushinteger(L, 0);
return 3;
}
/* pattern to match a single UTF-8 character */
#define UTF8PATT "[\0-\x7F\xC2-\xF4][\x80-\xBF]*"
static const luaL_Reg funcs[] = {
{"offset", byteoffset},
{"codepoint", codepoint},
{"char", utfchar},
{"len", utflen},
{"codes", iter_codes},
{NULL, NULL},
};
int luaopen_utf8(lua_State* L)
{
luaL_register(L, LUA_UTF8LIBNAME, funcs);
lua_pushlstring(L, UTF8PATT, sizeof(UTF8PATT) / sizeof(char) - 1);
lua_setfield(L, -2, "charpattern");
return 1;
}