1 | /* |
2 | ** $Id: lutf8lib.c,v 1.13 2014/11/02 19:19:04 roberto Exp $ |
3 | ** Standard library for UTF-8 manipulation |
4 | ** Modified by the LOVE Development Team to work with Lua 5.1's API |
5 | */ |
6 | |
7 | /****************************************************************************** |
8 | * Copyright (C) 1994-2015 Lua.org, PUC-Rio, 2015 LOVE Development Team. |
9 | * |
10 | * Permission is hereby granted, free of charge, to any person obtaining |
11 | * a copy of this software and associated documentation files (the |
12 | * "Software"), to deal in the Software without restriction, including |
13 | * without limitation the rights to use, copy, modify, merge, publish, |
14 | * distribute, sublicense, and/or sell copies of the Software, and to |
15 | * permit persons to whom the Software is furnished to do so, subject to |
16 | * the following conditions: |
17 | * |
18 | * The above copyright notice and this permission notice shall be |
19 | * included in all copies or substantial portions of the Software. |
20 | * |
21 | * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, |
22 | * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF |
23 | * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. |
24 | * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY |
25 | * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, |
26 | * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE |
27 | * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. |
28 | ******************************************************************************/ |
29 | |
30 | #define lutf8lib_c |
31 | |
32 | #include "lprefix.h" |
33 | |
34 | |
35 | #include <assert.h> |
36 | #include <stdlib.h> |
37 | #include <string.h> |
38 | |
39 | #include "lutf8lib.h" |
40 | |
41 | #include "lauxlib.h" |
42 | #include "lualib.h" |
43 | |
44 | #define MAXUNICODE 0x10FFFF |
45 | |
46 | /* size of buffer for 'utf8esc' function (taken from lobject.h) */ |
47 | #define UTF8BUFFSZ 8 |
48 | |
49 | #define iscont(p) ((*(p) & 0xC0) == 0x80) |
50 | |
51 | |
52 | /* from strlib */ |
53 | /* translate a relative string position: negative means back from end */ |
54 | static lua_Integer u_posrelat (lua_Integer pos, size_t len) { |
55 | if (pos >= 0) return pos; |
56 | else if (0u - (size_t)pos > len) return 0; |
57 | else return (lua_Integer)len + pos + 1; |
58 | } |
59 | |
60 | |
61 | /* |
62 | ** Decode one UTF-8 sequence, returning NULL if byte sequence is invalid. |
63 | */ |
64 | static const char *utf8_decode (const char *o, int *val) { |
65 | static unsigned int limits[] = {0xFF, 0x7F, 0x7FF, 0xFFFF}; |
66 | const unsigned char *s = (const unsigned char *)o; |
67 | unsigned int c = s[0]; |
68 | unsigned int res = 0; /* final result */ |
69 | if (c < 0x80) /* ascii? */ |
70 | res = c; |
71 | else { |
72 | int count = 0; /* to count number of continuation bytes */ |
73 | while (c & 0x40) { /* still have continuation bytes? */ |
74 | int cc = s[++count]; /* read next byte */ |
75 | if ((cc & 0xC0) != 0x80) /* not a continuation byte? */ |
76 | return NULL; /* invalid byte sequence */ |
77 | res = (res << 6) | (cc & 0x3F); /* add lower 6 bits from cont. byte */ |
78 | c <<= 1; /* to test next bit */ |
79 | } |
80 | res |= ((c & 0x7F) << (count * 5)); /* add first byte */ |
81 | if (count > 3 || res > MAXUNICODE || res <= limits[count]) |
82 | return NULL; /* invalid byte sequence */ |
83 | s += count; /* skip continuation bytes read */ |
84 | } |
85 | if (val) *val = res; |
86 | return (const char *)s + 1; /* +1 to include first byte */ |
87 | } |
88 | |
89 | |
90 | /* |
91 | ** utf8len(s [, i [, j]]) --> number of characters that start in the |
92 | ** range [i,j], or nil + current position if 's' is not well formed in |
93 | ** that interval |
94 | */ |
95 | static int utflen (lua_State *L) { |
96 | int n = 0; |
97 | size_t len; |
98 | const char *s = luaL_checklstring(L, 1, &len); |
99 | lua_Integer posi = u_posrelat(luaL_optinteger(L, 2, 1), len); |
100 | lua_Integer posj = u_posrelat(luaL_optinteger(L, 3, -1), len); |
101 | luaL_argcheck(L, 1 <= posi && --posi <= (lua_Integer)len, 2, |
102 | "initial position out of string" ); |
103 | luaL_argcheck(L, --posj < (lua_Integer)len, 3, |
104 | "final position out of string" ); |
105 | while (posi <= posj) { |
106 | const char *s1 = utf8_decode(s + posi, NULL); |
107 | if (s1 == NULL) { /* conversion error? */ |
108 | lua_pushnil(L); /* return nil ... */ |
109 | lua_pushinteger(L, posi + 1); /* ... and current position */ |
110 | return 2; |
111 | } |
112 | posi = s1 - s; |
113 | n++; |
114 | } |
115 | lua_pushinteger(L, n); |
116 | return 1; |
117 | } |
118 | |
119 | |
120 | /* |
121 | ** codepoint(s, [i, [j]]) -> returns codepoints for all characters |
122 | ** that start in the range [i,j] |
123 | */ |
124 | static int codepoint (lua_State *L) { |
125 | size_t len; |
126 | const char *s = luaL_checklstring(L, 1, &len); |
127 | lua_Integer posi = u_posrelat(luaL_optinteger(L, 2, 1), len); |
128 | lua_Integer pose = u_posrelat(luaL_optinteger(L, 3, posi), len); |
129 | int n; |
130 | const char *se; |
131 | luaL_argcheck(L, posi >= 1, 2, "out of range" ); |
132 | luaL_argcheck(L, pose <= (lua_Integer)len, 3, "out of range" ); |
133 | if (posi > pose) return 0; /* empty interval; return no values */ |
134 | n = (int)(pose - posi + 1); |
135 | if (posi + n <= pose) /* (lua_Integer -> int) overflow? */ |
136 | return luaL_error(L, "string slice too long" ); |
137 | luaL_checkstack(L, n, "string slice too long" ); |
138 | n = 0; |
139 | se = s + pose; |
140 | for (s += posi - 1; s < se;) { |
141 | int code; |
142 | s = utf8_decode(s, &code); |
143 | if (s == NULL) |
144 | return luaL_error(L, "invalid UTF-8 code" ); |
145 | lua_pushinteger(L, code); |
146 | n++; |
147 | } |
148 | return n; |
149 | } |
150 | |
151 | |
152 | /* taken from lobject.c */ |
153 | static int utf8esc (char *buff, unsigned long x) { |
154 | int n = 1; /* number of bytes put in buffer (backwards) */ |
155 | lua_assert(x <= 0x10FFFF); |
156 | if (x < 0x80) /* ascii? */ |
157 | buff[UTF8BUFFSZ - 1] = (char) x; |
158 | else { /* need continuation bytes */ |
159 | unsigned int mfb = 0x3f; /* maximum that fits in first byte */ |
160 | do { /* add continuation bytes */ |
161 | buff[UTF8BUFFSZ - (n++)] = (char) (0x80 | (x & 0x3f)); |
162 | x >>= 6; /* remove added bits */ |
163 | mfb >>= 1; /* now there is one less bit available in first byte */ |
164 | } while (x > mfb); /* still needs continuation byte? */ |
165 | buff[UTF8BUFFSZ - n] = (char) ((~mfb << 1) | x); /* add first byte */ |
166 | } |
167 | return n; |
168 | } |
169 | |
170 | static void pushutfchar (lua_State *L, int arg) { |
171 | lua_Integer code = luaL_checkinteger(L, arg); |
172 | luaL_argcheck(L, 0 <= code && code <= MAXUNICODE, arg, "value out of range" ); |
173 | |
174 | /* the %U string format does not exist in lua 5.1 or 5.2, so we emulate it */ |
175 | /* (code from luaO_pushvfstring in lobject.c) */ |
176 | char buff[UTF8BUFFSZ]; |
177 | int l = utf8esc(buff, (long) code); |
178 | lua_pushlstring(L, buff + UTF8BUFFSZ - l, l); |
179 | } |
180 | |
181 | |
182 | /* |
183 | ** utfchar(n1, n2, ...) -> char(n1)..char(n2)... |
184 | */ |
185 | static int utfchar (lua_State *L) { |
186 | int n = lua_gettop(L); /* number of arguments */ |
187 | if (n == 1) /* optimize common case of single char */ |
188 | pushutfchar(L, 1); |
189 | else { |
190 | int i; |
191 | luaL_Buffer b; |
192 | luaL_buffinit(L, &b); |
193 | for (i = 1; i <= n; i++) { |
194 | pushutfchar(L, i); |
195 | luaL_addvalue(&b); |
196 | } |
197 | luaL_pushresult(&b); |
198 | } |
199 | return 1; |
200 | } |
201 | |
202 | |
203 | /* |
204 | ** offset(s, n, [i]) -> index where n-th character counting from |
205 | ** position 'i' starts; 0 means character at 'i'. |
206 | */ |
207 | static int byteoffset (lua_State *L) { |
208 | size_t len; |
209 | const char *s = luaL_checklstring(L, 1, &len); |
210 | lua_Integer n = luaL_checkinteger(L, 2); |
211 | lua_Integer posi = (n >= 0) ? 1 : len + 1; |
212 | posi = u_posrelat(luaL_optinteger(L, 3, posi), len); |
213 | luaL_argcheck(L, 1 <= posi && --posi <= (lua_Integer)len, 3, |
214 | "position out of range" ); |
215 | if (n == 0) { |
216 | /* find beginning of current byte sequence */ |
217 | while (posi > 0 && iscont(s + posi)) posi--; |
218 | } |
219 | else { |
220 | if (iscont(s + posi)) |
221 | luaL_error(L, "initial position is a continuation byte" ); |
222 | if (n < 0) { |
223 | while (n < 0 && posi > 0) { /* move back */ |
224 | do { /* find beginning of previous character */ |
225 | posi--; |
226 | } while (posi > 0 && iscont(s + posi)); |
227 | n++; |
228 | } |
229 | } |
230 | else { |
231 | n--; /* do not move for 1st character */ |
232 | while (n > 0 && posi < (lua_Integer)len) { |
233 | do { /* find beginning of next character */ |
234 | posi++; |
235 | } while (iscont(s + posi)); /* (cannot pass final '\0') */ |
236 | n--; |
237 | } |
238 | } |
239 | } |
240 | if (n == 0) /* did it find given character? */ |
241 | lua_pushinteger(L, posi + 1); |
242 | else /* no such character */ |
243 | lua_pushnil(L); |
244 | return 1; |
245 | } |
246 | |
247 | |
248 | static int iter_aux (lua_State *L) { |
249 | size_t len; |
250 | const char *s = luaL_checklstring(L, 1, &len); |
251 | lua_Integer n = lua_tointeger(L, 2) - 1; |
252 | if (n < 0) /* first iteration? */ |
253 | n = 0; /* start from here */ |
254 | else if (n < (lua_Integer)len) { |
255 | n++; /* skip current byte */ |
256 | while (iscont(s + n)) n++; /* and its continuations */ |
257 | } |
258 | if (n >= (lua_Integer)len) |
259 | return 0; /* no more codepoints */ |
260 | else { |
261 | int code; |
262 | const char *next = utf8_decode(s + n, &code); |
263 | if (next == NULL || iscont(next)) |
264 | return luaL_error(L, "invalid UTF-8 code" ); |
265 | lua_pushinteger(L, n + 1); |
266 | lua_pushinteger(L, code); |
267 | return 2; |
268 | } |
269 | } |
270 | |
271 | |
272 | static int iter_codes (lua_State *L) { |
273 | luaL_checkstring(L, 1); |
274 | lua_pushcfunction(L, iter_aux); |
275 | lua_pushvalue(L, 1); |
276 | lua_pushinteger(L, 0); |
277 | return 3; |
278 | } |
279 | |
280 | |
281 | /* pattern to match a single UTF-8 character */ |
282 | #if LUA_VERSION_NUM >= 502 |
283 | #define UTF8PATT "[\0-\x7F\xC2-\xF4][\x80-\xBF]*" |
284 | #else |
285 | /* lua 5.1 doesn't support literal null bytes in patterns */ |
286 | #define UTF8PATT "[%z\x01-\x7F\xC2-\xF4][\x80-\xBF]*" |
287 | #endif |
288 | |
289 | |
290 | static struct luaL_Reg funcs[] = { |
291 | {"offset" , byteoffset}, |
292 | {"codepoint" , codepoint}, |
293 | {"char" , utfchar}, |
294 | {"len" , utflen}, |
295 | {"codes" , iter_codes}, |
296 | /* placeholders */ |
297 | {"charpattern" , NULL}, |
298 | {NULL, NULL} |
299 | }; |
300 | |
301 | |
302 | /* modified version of luaopen_utf8, designed to work with lua 5.1-5.3 */ |
303 | int luaopen_luautf8 (lua_State *L) { |
304 | luaL_Reg *l; |
305 | lua_createtable(L, 0, (int) (sizeof(funcs) / sizeof(luaL_Reg)) - 1); |
306 | for (l = funcs; l->name != NULL; l++) { |
307 | if (l->func != NULL) { |
308 | lua_pushcfunction(L, l->func); |
309 | lua_setfield(L, -2, l->name); |
310 | } |
311 | } |
312 | lua_pushlstring(L, UTF8PATT, sizeof(UTF8PATT) / sizeof(char) - 1); |
313 | lua_setfield(L, -2, "charpattern" ); |
314 | return 1; |
315 | } |
316 | |