summaryrefslogtreecommitdiff
path: root/apps/plugins/lua/lstrlib.c
diff options
context:
space:
mode:
Diffstat (limited to 'apps/plugins/lua/lstrlib.c')
-rw-r--r--apps/plugins/lua/lstrlib.c580
1 files changed, 365 insertions, 215 deletions
diff --git a/apps/plugins/lua/lstrlib.c b/apps/plugins/lua/lstrlib.c
index 3d6103692f..04cc2f60d6 100644
--- a/apps/plugins/lua/lstrlib.c
+++ b/apps/plugins/lua/lstrlib.c
@@ -1,5 +1,5 @@
1/* 1/*
2** $Id: lstrlib.c,v 1.132.1.4 2008/07/11 17:27:21 roberto Exp $ 2** $Id: lstrlib.c,v 1.178.1.1 2013/04/12 18:48:47 roberto Exp $
3** Standard library for string operations and pattern-matching 3** Standard library for string operations and pattern-matching
4** See Copyright Notice in lua.h 4** See Copyright Notice in lua.h
5*/ 5*/
@@ -20,47 +20,58 @@
20#include "lualib.h" 20#include "lualib.h"
21 21
22 22
23/*
24** maximum number of captures that a pattern can do during
25** pattern-matching. This limit is arbitrary.
26*/
27#if !defined(LUA_MAXCAPTURES)
28#define LUA_MAXCAPTURES 32
29#endif
30
31
23/* macro to `unsign' a character */ 32/* macro to `unsign' a character */
24#define uchar(c) ((unsigned char)(c)) 33#define uchar(c) ((unsigned char)(c))
25 34
26 35
27 36
28static int str_len (lua_State *L) { 37static int str_len (lua_State *L) {
29 size_t l; 38 size_t l;
30 luaL_checklstring(L, 1, &l); 39 luaL_checklstring(L, 1, &l);
31 lua_pushinteger(L, l); 40 lua_pushinteger(L, (lua_Integer)l);
32 return 1; 41 return 1;
33} 42}
34 43
35 44
36static ptrdiff_t posrelat (ptrdiff_t pos, size_t len) { 45/* translate a relative string position: negative means back from end */
37 /* relative string position: negative means back from end */ 46static size_t posrelat (ptrdiff_t pos, size_t len) {
38 if (pos < 0) pos += (ptrdiff_t)len + 1; 47 if (pos >= 0) return (size_t)pos;
39 return (pos >= 0) ? pos : 0; 48 else if (0u - (size_t)pos > len) return 0;
49 else return len - ((size_t)-pos) + 1;
40} 50}
41 51
42 52
43static int str_sub (lua_State *L) { 53static int str_sub (lua_State *L) {
44 size_t l; 54 size_t l;
45 const char *s = luaL_checklstring(L, 1, &l); 55 const char *s = luaL_checklstring(L, 1, &l);
46 ptrdiff_t start = posrelat(luaL_checkinteger(L, 2), l); 56 size_t start = posrelat(luaL_checkinteger(L, 2), l);
47 ptrdiff_t end = posrelat(luaL_optinteger(L, 3, -1), l); 57 size_t end = posrelat(luaL_optinteger(L, 3, -1), l);
48 if (start < 1) start = 1; 58 if (start < 1) start = 1;
49 if (end > (ptrdiff_t)l) end = (ptrdiff_t)l; 59 if (end > l) end = l;
50 if (start <= end) 60 if (start <= end)
51 lua_pushlstring(L, s+start-1, end-start+1); 61 lua_pushlstring(L, s + start - 1, end - start + 1);
52 else lua_pushliteral(L, ""); 62 else lua_pushliteral(L, "");
53 return 1; 63 return 1;
54} 64}
55 65
56 66
57static int str_reverse (lua_State *L) { 67static int str_reverse (lua_State *L) {
58 size_t l; 68 size_t l, i;
59 luaL_Buffer b; 69 luaL_Buffer b;
60 const char *s = luaL_checklstring(L, 1, &l); 70 const char *s = luaL_checklstring(L, 1, &l);
61 luaL_buffinit(L, &b); 71 char *p = luaL_buffinitsize(L, &b, l);
62 while (l--) luaL_addchar(&b, s[l]); 72 for (i = 0; i < l; i++)
63 luaL_pushresult(&b); 73 p[i] = s[l - i - 1];
74 luaL_pushresultsize(&b, l);
64 return 1; 75 return 1;
65} 76}
66 77
@@ -70,10 +81,10 @@ static int str_lower (lua_State *L) {
70 size_t i; 81 size_t i;
71 luaL_Buffer b; 82 luaL_Buffer b;
72 const char *s = luaL_checklstring(L, 1, &l); 83 const char *s = luaL_checklstring(L, 1, &l);
73 luaL_buffinit(L, &b); 84 char *p = luaL_buffinitsize(L, &b, l);
74 for (i=0; i<l; i++) 85 for (i=0; i<l; i++)
75 luaL_addchar(&b, tolower(uchar(s[i]))); 86 p[i] = tolower(uchar(s[i]));
76 luaL_pushresult(&b); 87 luaL_pushresultsize(&b, l);
77 return 1; 88 return 1;
78} 89}
79 90
@@ -83,22 +94,38 @@ static int str_upper (lua_State *L) {
83 size_t i; 94 size_t i;
84 luaL_Buffer b; 95 luaL_Buffer b;
85 const char *s = luaL_checklstring(L, 1, &l); 96 const char *s = luaL_checklstring(L, 1, &l);
86 luaL_buffinit(L, &b); 97 char *p = luaL_buffinitsize(L, &b, l);
87 for (i=0; i<l; i++) 98 for (i=0; i<l; i++)
88 luaL_addchar(&b, toupper(uchar(s[i]))); 99 p[i] = toupper(uchar(s[i]));
89 luaL_pushresult(&b); 100 luaL_pushresultsize(&b, l);
90 return 1; 101 return 1;
91} 102}
92 103
104
105/* reasonable limit to avoid arithmetic overflow */
106#define MAXSIZE ((~(size_t)0) >> 1)
107
93static int str_rep (lua_State *L) { 108static int str_rep (lua_State *L) {
94 size_t l; 109 size_t l, lsep;
95 luaL_Buffer b;
96 const char *s = luaL_checklstring(L, 1, &l); 110 const char *s = luaL_checklstring(L, 1, &l);
97 int n = luaL_checkint(L, 2); 111 int n = luaL_checkint(L, 2);
98 luaL_buffinit(L, &b); 112 const char *sep = luaL_optlstring(L, 3, "", &lsep);
99 while (n-- > 0) 113 if (n <= 0) lua_pushliteral(L, "");
100 luaL_addlstring(&b, s, l); 114 else if (l + lsep < l || l + lsep >= MAXSIZE / n) /* may overflow? */
101 luaL_pushresult(&b); 115 return luaL_error(L, "resulting string too large");
116 else {
117 size_t totallen = n * l + (n - 1) * lsep;
118 luaL_Buffer b;
119 char *p = luaL_buffinitsize(L, &b, totallen);
120 while (n-- > 1) { /* first n-1 copies (followed by separator) */
121 memcpy(p, s, l * sizeof(char)); p += l;
122 if (lsep > 0) { /* avoid empty 'memcpy' (may be expensive) */
123 memcpy(p, sep, lsep * sizeof(char)); p += lsep;
124 }
125 }
126 memcpy(p, s, l * sizeof(char)); /* last copy (not followed by separator) */
127 luaL_pushresultsize(&b, totallen);
128 }
102 return 1; 129 return 1;
103} 130}
104 131
@@ -106,15 +133,15 @@ static int str_rep (lua_State *L) {
106static int str_byte (lua_State *L) { 133static int str_byte (lua_State *L) {
107 size_t l; 134 size_t l;
108 const char *s = luaL_checklstring(L, 1, &l); 135 const char *s = luaL_checklstring(L, 1, &l);
109 ptrdiff_t posi = posrelat(luaL_optinteger(L, 2, 1), l); 136 size_t posi = posrelat(luaL_optinteger(L, 2, 1), l);
110 ptrdiff_t pose = posrelat(luaL_optinteger(L, 3, posi), l); 137 size_t pose = posrelat(luaL_optinteger(L, 3, posi), l);
111 int n, i; 138 int n, i;
112 if (posi <= 0) posi = 1; 139 if (posi < 1) posi = 1;
113 if ((size_t)pose > l) pose = l; 140 if (pose > l) pose = l;
114 if (posi > pose) return 0; /* empty interval; return no values */ 141 if (posi > pose) return 0; /* empty interval; return no values */
115 n = (int)(pose - posi + 1); 142 n = (int)(pose - posi + 1);
116 if (posi + n <= pose) /* overflow? */ 143 if (posi + n <= pose) /* (size_t -> int) overflow? */
117 luaL_error(L, "string slice too long"); 144 return luaL_error(L, "string slice too long");
118 luaL_checkstack(L, n, "string slice too long"); 145 luaL_checkstack(L, n, "string slice too long");
119 for (i=0; i<n; i++) 146 for (i=0; i<n; i++)
120 lua_pushinteger(L, uchar(s[posi+i-1])); 147 lua_pushinteger(L, uchar(s[posi+i-1]));
@@ -126,13 +153,13 @@ static int str_char (lua_State *L) {
126 int n = lua_gettop(L); /* number of arguments */ 153 int n = lua_gettop(L); /* number of arguments */
127 int i; 154 int i;
128 luaL_Buffer b; 155 luaL_Buffer b;
129 luaL_buffinit(L, &b); 156 char *p = luaL_buffinitsize(L, &b, n);
130 for (i=1; i<=n; i++) { 157 for (i=1; i<=n; i++) {
131 int c = luaL_checkint(L, i); 158 int c = luaL_checkint(L, i);
132 luaL_argcheck(L, uchar(c) == c, i, "invalid value"); 159 luaL_argcheck(L, uchar(c) == c, i, "value out of range");
133 luaL_addchar(&b, uchar(c)); 160 p[i - 1] = uchar(c);
134 } 161 }
135 luaL_pushresult(&b); 162 luaL_pushresultsize(&b, n);
136 return 1; 163 return 1;
137} 164}
138 165
@@ -150,7 +177,7 @@ static int str_dump (lua_State *L) {
150 lua_settop(L, 1); 177 lua_settop(L, 1);
151 luaL_buffinit(L,&b); 178 luaL_buffinit(L,&b);
152 if (lua_dump(L, writer, &b) != 0) 179 if (lua_dump(L, writer, &b) != 0)
153 luaL_error(L, "unable to dump given function"); 180 return luaL_error(L, "unable to dump given function");
154 luaL_pushresult(&b); 181 luaL_pushresult(&b);
155 return 1; 182 return 1;
156} 183}
@@ -167,9 +194,12 @@ static int str_dump (lua_State *L) {
167#define CAP_UNFINISHED (-1) 194#define CAP_UNFINISHED (-1)
168#define CAP_POSITION (-2) 195#define CAP_POSITION (-2)
169 196
197
170typedef struct MatchState { 198typedef struct MatchState {
199 int matchdepth; /* control for recursive depth (to avoid C stack overflow) */
171 const char *src_init; /* init of source string */ 200 const char *src_init; /* init of source string */
172 const char *src_end; /* end (`\0') of source string */ 201 const char *src_end; /* end ('\0') of source string */
202 const char *p_end; /* end ('\0') of pattern */
173 lua_State *L; 203 lua_State *L;
174 int level; /* total number of captures (finished or unfinished) */ 204 int level; /* total number of captures (finished or unfinished) */
175 struct { 205 struct {
@@ -179,6 +209,16 @@ typedef struct MatchState {
179} MatchState; 209} MatchState;
180 210
181 211
212/* recursive function */
213static const char *match (MatchState *ms, const char *s, const char *p);
214
215
216/* maximum recursion depth for 'match' */
217#if !defined(MAXCCALLS)
218#define MAXCCALLS 200
219#endif
220
221
182#define L_ESC '%' 222#define L_ESC '%'
183#define SPECIALS "^$*+?.([%-" 223#define SPECIALS "^$*+?.([%-"
184 224
@@ -186,7 +226,7 @@ typedef struct MatchState {
186static int check_capture (MatchState *ms, int l) { 226static int check_capture (MatchState *ms, int l) {
187 l -= '1'; 227 l -= '1';
188 if (l < 0 || l >= ms->level || ms->capture[l].len == CAP_UNFINISHED) 228 if (l < 0 || l >= ms->level || ms->capture[l].len == CAP_UNFINISHED)
189 return luaL_error(ms->L, "invalid capture index"); 229 return luaL_error(ms->L, "invalid capture index %%%d", l + 1);
190 return l; 230 return l;
191} 231}
192 232
@@ -202,16 +242,16 @@ static int capture_to_close (MatchState *ms) {
202static const char *classend (MatchState *ms, const char *p) { 242static const char *classend (MatchState *ms, const char *p) {
203 switch (*p++) { 243 switch (*p++) {
204 case L_ESC: { 244 case L_ESC: {
205 if (*p == '\0') 245 if (p == ms->p_end)
206 luaL_error(ms->L, "malformed pattern (ends with " LUA_QL("%%") ")"); 246 luaL_error(ms->L, "malformed pattern (ends with " LUA_QL("%%") ")");
207 return p+1; 247 return p+1;
208 } 248 }
209 case '[': { 249 case '[': {
210 if (*p == '^') p++; 250 if (*p == '^') p++;
211 do { /* look for a `]' */ 251 do { /* look for a `]' */
212 if (*p == '\0') 252 if (p == ms->p_end)
213 luaL_error(ms->L, "malformed pattern (missing " LUA_QL("]") ")"); 253 luaL_error(ms->L, "malformed pattern (missing " LUA_QL("]") ")");
214 if (*(p++) == L_ESC && *p != '\0') 254 if (*(p++) == L_ESC && p < ms->p_end)
215 p++; /* skip escapes (e.g. `%]') */ 255 p++; /* skip escapes (e.g. `%]') */
216 } while (*p != ']'); 256 } while (*p != ']');
217 return p+1; 257 return p+1;
@@ -229,13 +269,14 @@ static int match_class (int c, int cl) {
229 case 'a' : res = isalpha(c); break; 269 case 'a' : res = isalpha(c); break;
230 case 'c' : res = iscntrl(c); break; 270 case 'c' : res = iscntrl(c); break;
231 case 'd' : res = isdigit(c); break; 271 case 'd' : res = isdigit(c); break;
272 case 'g' : res = isgraph(c); break;
232 case 'l' : res = islower(c); break; 273 case 'l' : res = islower(c); break;
233 case 'p' : res = ispunct(c); break; 274 case 'p' : res = ispunct(c); break;
234 case 's' : res = isspace(c); break; 275 case 's' : res = isspace(c); break;
235 case 'u' : res = isupper(c); break; 276 case 'u' : res = isupper(c); break;
236 case 'w' : res = isalnum(c); break; 277 case 'w' : res = isalnum(c); break;
237 case 'x' : res = isxdigit(c); break; 278 case 'x' : res = isxdigit(c); break;
238 case 'z' : res = (c == 0); break; 279 case 'z' : res = (c == 0); break; /* deprecated option */
239 default: return (cl == c); 280 default: return (cl == c);
240 } 281 }
241 return (islower(cl) ? res : !res); 282 return (islower(cl) ? res : !res);
@@ -265,23 +306,27 @@ static int matchbracketclass (int c, const char *p, const char *ec) {
265} 306}
266 307
267 308
268static int singlematch (int c, const char *p, const char *ep) { 309static int singlematch (MatchState *ms, const char *s, const char *p,
269 switch (*p) { 310 const char *ep) {
270 case '.': return 1; /* matches any char */ 311 if (s >= ms->src_end)
271 case L_ESC: return match_class(c, uchar(*(p+1))); 312 return 0;
272 case '[': return matchbracketclass(c, p, ep-1); 313 else {
273 default: return (uchar(*p) == c); 314 int c = uchar(*s);
315 switch (*p) {
316 case '.': return 1; /* matches any char */
317 case L_ESC: return match_class(c, uchar(*(p+1)));
318 case '[': return matchbracketclass(c, p, ep-1);
319 default: return (uchar(*p) == c);
320 }
274 } 321 }
275} 322}
276 323
277 324
278static const char *match (MatchState *ms, const char *s, const char *p);
279
280
281static const char *matchbalance (MatchState *ms, const char *s, 325static const char *matchbalance (MatchState *ms, const char *s,
282 const char *p) { 326 const char *p) {
283 if (*p == 0 || *(p+1) == 0) 327 if (p >= ms->p_end - 1)
284 luaL_error(ms->L, "unbalanced pattern"); 328 luaL_error(ms->L, "malformed pattern "
329 "(missing arguments to " LUA_QL("%%b") ")");
285 if (*s != *p) return NULL; 330 if (*s != *p) return NULL;
286 else { 331 else {
287 int b = *p; 332 int b = *p;
@@ -301,7 +346,7 @@ static const char *matchbalance (MatchState *ms, const char *s,
301static const char *max_expand (MatchState *ms, const char *s, 346static const char *max_expand (MatchState *ms, const char *s,
302 const char *p, const char *ep) { 347 const char *p, const char *ep) {
303 ptrdiff_t i = 0; /* counts maximum expand for item */ 348 ptrdiff_t i = 0; /* counts maximum expand for item */
304 while ((s+i)<ms->src_end && singlematch(uchar(*(s+i)), p, ep)) 349 while (singlematch(ms, s + i, p, ep))
305 i++; 350 i++;
306 /* keeps trying to match with the maximum repetitions */ 351 /* keeps trying to match with the maximum repetitions */
307 while (i>=0) { 352 while (i>=0) {
@@ -319,7 +364,7 @@ static const char *min_expand (MatchState *ms, const char *s,
319 const char *res = match(ms, s, ep+1); 364 const char *res = match(ms, s, ep+1);
320 if (res != NULL) 365 if (res != NULL)
321 return res; 366 return res;
322 else if (s<ms->src_end && singlematch(uchar(*s), p, ep)) 367 else if (singlematch(ms, s, p, ep))
323 s++; /* try with one more repetition */ 368 s++; /* try with one more repetition */
324 else return NULL; 369 else return NULL;
325 } 370 }
@@ -363,80 +408,105 @@ static const char *match_capture (MatchState *ms, const char *s, int l) {
363 408
364 409
365static const char *match (MatchState *ms, const char *s, const char *p) { 410static const char *match (MatchState *ms, const char *s, const char *p) {
411 if (ms->matchdepth-- == 0)
412 luaL_error(ms->L, "pattern too complex");
366 init: /* using goto's to optimize tail recursion */ 413 init: /* using goto's to optimize tail recursion */
367 switch (*p) { 414 if (p != ms->p_end) { /* end of pattern? */
368 case '(': { /* start capture */ 415 switch (*p) {
369 if (*(p+1) == ')') /* position capture? */ 416 case '(': { /* start capture */
370 return start_capture(ms, s, p+2, CAP_POSITION); 417 if (*(p + 1) == ')') /* position capture? */
371 else 418 s = start_capture(ms, s, p + 2, CAP_POSITION);
372 return start_capture(ms, s, p+1, CAP_UNFINISHED); 419 else
373 } 420 s = start_capture(ms, s, p + 1, CAP_UNFINISHED);
374 case ')': { /* end capture */ 421 break;
375 return end_capture(ms, s, p+1); 422 }
376 } 423 case ')': { /* end capture */
377 case L_ESC: { 424 s = end_capture(ms, s, p + 1);
378 switch (*(p+1)) { 425 break;
379 case 'b': { /* balanced string? */ 426 }
380 s = matchbalance(ms, s, p+2); 427 case '$': {
381 if (s == NULL) return NULL; 428 if ((p + 1) != ms->p_end) /* is the `$' the last char in pattern? */
382 p+=4; goto init; /* else return match(ms, s, p+4); */ 429 goto dflt; /* no; go to default */
383 } 430 s = (s == ms->src_end) ? s : NULL; /* check end of string */
384 case 'f': { /* frontier? */ 431 break;
385 const char *ep; char previous; 432 }
386 p += 2; 433 case L_ESC: { /* escaped sequences not in the format class[*+?-]? */
387 if (*p != '[') 434 switch (*(p + 1)) {
388 luaL_error(ms->L, "missing " LUA_QL("[") " after " 435 case 'b': { /* balanced string? */
389 LUA_QL("%%f") " in pattern"); 436 s = matchbalance(ms, s, p + 2);
390 ep = classend(ms, p); /* points to what is next */ 437 if (s != NULL) {
391 previous = (s == ms->src_init) ? '\0' : *(s-1); 438 p += 4; goto init; /* return match(ms, s, p + 4); */
392 if (matchbracketclass(uchar(previous), p, ep-1) || 439 } /* else fail (s == NULL) */
393 !matchbracketclass(uchar(*s), p, ep-1)) return NULL; 440 break;
394 p=ep; goto init; /* else return match(ms, s, ep); */ 441 }
395 } 442 case 'f': { /* frontier? */
396 default: { 443 const char *ep; char previous;
397 if (isdigit(uchar(*(p+1)))) { /* capture results (%0-%9)? */ 444 p += 2;
398 s = match_capture(ms, s, uchar(*(p+1))); 445 if (*p != '[')
399 if (s == NULL) return NULL; 446 luaL_error(ms->L, "missing " LUA_QL("[") " after "
400 p+=2; goto init; /* else return match(ms, s, p+2) */ 447 LUA_QL("%%f") " in pattern");
448 ep = classend(ms, p); /* points to what is next */
449 previous = (s == ms->src_init) ? '\0' : *(s - 1);
450 if (!matchbracketclass(uchar(previous), p, ep - 1) &&
451 matchbracketclass(uchar(*s), p, ep - 1)) {
452 p = ep; goto init; /* return match(ms, s, ep); */
453 }
454 s = NULL; /* match failed */
455 break;
401 } 456 }
402 goto dflt; /* case default */ 457 case '0': case '1': case '2': case '3':
458 case '4': case '5': case '6': case '7':
459 case '8': case '9': { /* capture results (%0-%9)? */
460 s = match_capture(ms, s, uchar(*(p + 1)));
461 if (s != NULL) {
462 p += 2; goto init; /* return match(ms, s, p + 2) */
463 }
464 break;
465 }
466 default: goto dflt;
403 } 467 }
468 break;
404 } 469 }
405 } 470 default: dflt: { /* pattern class plus optional suffix */
406 case '\0': { /* end of pattern */ 471 const char *ep = classend(ms, p); /* points to optional suffix */
407 return s; /* match succeeded */ 472 /* does not match at least once? */
408 } 473 if (!singlematch(ms, s, p, ep)) {
409 case '$': { 474 if (*ep == '*' || *ep == '?' || *ep == '-') { /* accept empty? */
410 if (*(p+1) == '\0') /* is the `$' the last char in pattern? */ 475 p = ep + 1; goto init; /* return match(ms, s, ep + 1); */
411 return (s == ms->src_end) ? s : NULL; /* check end of string */ 476 }
412 else goto dflt; 477 else /* '+' or no suffix */
413 } 478 s = NULL; /* fail */
414 default: dflt: { /* it is a pattern item */
415 const char *ep = classend(ms, p); /* points to what is next */
416 int m = s<ms->src_end && singlematch(uchar(*s), p, ep);
417 switch (*ep) {
418 case '?': { /* optional */
419 const char *res;
420 if (m && ((res=match(ms, s+1, ep+1)) != NULL))
421 return res;
422 p=ep+1; goto init; /* else return match(ms, s, ep+1); */
423 }
424 case '*': { /* 0 or more repetitions */
425 return max_expand(ms, s, p, ep);
426 }
427 case '+': { /* 1 or more repetitions */
428 return (m ? max_expand(ms, s+1, p, ep) : NULL);
429 }
430 case '-': { /* 0 or more repetitions (minimum) */
431 return min_expand(ms, s, p, ep);
432 } 479 }
433 default: { 480 else { /* matched once */
434 if (!m) return NULL; 481 switch (*ep) { /* handle optional suffix */
435 s++; p=ep; goto init; /* else return match(ms, s+1, ep); */ 482 case '?': { /* optional */
483 const char *res;
484 if ((res = match(ms, s + 1, ep + 1)) != NULL)
485 s = res;
486 else {
487 p = ep + 1; goto init; /* else return match(ms, s, ep + 1); */
488 }
489 break;
490 }
491 case '+': /* 1 or more repetitions */
492 s++; /* 1 match already done */
493 /* go through */
494 case '*': /* 0 or more repetitions */
495 s = max_expand(ms, s, p, ep);
496 break;
497 case '-': /* 0 or more repetitions (minimum) */
498 s = min_expand(ms, s, p, ep);
499 break;
500 default: /* no suffix */
501 s++; p = ep; goto init; /* return match(ms, s + 1, ep); */
502 }
436 } 503 }
504 break;
437 } 505 }
438 } 506 }
439 } 507 }
508 ms->matchdepth++;
509 return s;
440} 510}
441 511
442 512
@@ -492,37 +562,58 @@ static int push_captures (MatchState *ms, const char *s, const char *e) {
492} 562}
493 563
494 564
565/* check whether pattern has no special characters */
566static int nospecials (const char *p, size_t l) {
567 size_t upto = 0;
568 do {
569 if (strpbrk(p + upto, SPECIALS))
570 return 0; /* pattern has a special character */
571 upto += strlen(p + upto) + 1; /* may have more after \0 */
572 } while (upto <= l);
573 return 1; /* no special chars found */
574}
575
576
495static int str_find_aux (lua_State *L, int find) { 577static int str_find_aux (lua_State *L, int find) {
496 size_t l1, l2; 578 size_t ls, lp;
497 const char *s = luaL_checklstring(L, 1, &l1); 579 const char *s = luaL_checklstring(L, 1, &ls);
498 const char *p = luaL_checklstring(L, 2, &l2); 580 const char *p = luaL_checklstring(L, 2, &lp);
499 ptrdiff_t init = posrelat(luaL_optinteger(L, 3, 1), l1) - 1; 581 size_t init = posrelat(luaL_optinteger(L, 3, 1), ls);
500 if (init < 0) init = 0; 582 if (init < 1) init = 1;
501 else if ((size_t)(init) > l1) init = (ptrdiff_t)l1; 583 else if (init > ls + 1) { /* start after string's end? */
502 if (find && (lua_toboolean(L, 4) || /* explicit request? */ 584 lua_pushnil(L); /* cannot find anything */
503 strpbrk(p, SPECIALS) == NULL)) { /* or no special characters? */ 585 return 1;
586 }
587 /* explicit request or no special characters? */
588 if (find && (lua_toboolean(L, 4) || nospecials(p, lp))) {
504 /* do a plain search */ 589 /* do a plain search */
505 const char *s2 = lmemfind(s+init, l1-init, p, l2); 590 const char *s2 = lmemfind(s + init - 1, ls - init + 1, p, lp);
506 if (s2) { 591 if (s2) {
507 lua_pushinteger(L, s2-s+1); 592 lua_pushinteger(L, s2 - s + 1);
508 lua_pushinteger(L, s2-s+l2); 593 lua_pushinteger(L, s2 - s + lp);
509 return 2; 594 return 2;
510 } 595 }
511 } 596 }
512 else { 597 else {
513 MatchState ms; 598 MatchState ms;
514 int anchor = (*p == '^') ? (p++, 1) : 0; 599 const char *s1 = s + init - 1;
515 const char *s1=s+init; 600 int anchor = (*p == '^');
601 if (anchor) {
602 p++; lp--; /* skip anchor character */
603 }
516 ms.L = L; 604 ms.L = L;
605 ms.matchdepth = MAXCCALLS;
517 ms.src_init = s; 606 ms.src_init = s;
518 ms.src_end = s+l1; 607 ms.src_end = s + ls;
608 ms.p_end = p + lp;
519 do { 609 do {
520 const char *res; 610 const char *res;
521 ms.level = 0; 611 ms.level = 0;
612 lua_assert(ms.matchdepth == MAXCCALLS);
522 if ((res=match(&ms, s1, p)) != NULL) { 613 if ((res=match(&ms, s1, p)) != NULL) {
523 if (find) { 614 if (find) {
524 lua_pushinteger(L, s1-s+1); /* start */ 615 lua_pushinteger(L, s1 - s + 1); /* start */
525 lua_pushinteger(L, res-s); /* end */ 616 lua_pushinteger(L, res - s); /* end */
526 return push_captures(&ms, NULL, 0) + 2; 617 return push_captures(&ms, NULL, 0) + 2;
527 } 618 }
528 else 619 else
@@ -547,18 +638,21 @@ static int str_match (lua_State *L) {
547 638
548static int gmatch_aux (lua_State *L) { 639static int gmatch_aux (lua_State *L) {
549 MatchState ms; 640 MatchState ms;
550 size_t ls; 641 size_t ls, lp;
551 const char *s = lua_tolstring(L, lua_upvalueindex(1), &ls); 642 const char *s = lua_tolstring(L, lua_upvalueindex(1), &ls);
552 const char *p = lua_tostring(L, lua_upvalueindex(2)); 643 const char *p = lua_tolstring(L, lua_upvalueindex(2), &lp);
553 const char *src; 644 const char *src;
554 ms.L = L; 645 ms.L = L;
646 ms.matchdepth = MAXCCALLS;
555 ms.src_init = s; 647 ms.src_init = s;
556 ms.src_end = s+ls; 648 ms.src_end = s+ls;
649 ms.p_end = p + lp;
557 for (src = s + (size_t)lua_tointeger(L, lua_upvalueindex(3)); 650 for (src = s + (size_t)lua_tointeger(L, lua_upvalueindex(3));
558 src <= ms.src_end; 651 src <= ms.src_end;
559 src++) { 652 src++) {
560 const char *e; 653 const char *e;
561 ms.level = 0; 654 ms.level = 0;
655 lua_assert(ms.matchdepth == MAXCCALLS);
562 if ((e = match(&ms, src, p)) != NULL) { 656 if ((e = match(&ms, src, p)) != NULL) {
563 lua_Integer newstart = e-s; 657 lua_Integer newstart = e-s;
564 if (e == src) newstart++; /* empty match? go at least one position */ 658 if (e == src) newstart++; /* empty match? go at least one position */
@@ -581,12 +675,6 @@ static int gmatch (lua_State *L) {
581} 675}
582 676
583 677
584static int gfind_nodef (lua_State *L) {
585 return luaL_error(L, LUA_QL("string.gfind") " was renamed to "
586 LUA_QL("string.gmatch"));
587}
588
589
590static void add_s (MatchState *ms, luaL_Buffer *b, const char *s, 678static void add_s (MatchState *ms, luaL_Buffer *b, const char *s,
591 const char *e) { 679 const char *e) {
592 size_t l, i; 680 size_t l, i;
@@ -596,8 +684,12 @@ static void add_s (MatchState *ms, luaL_Buffer *b, const char *s,
596 luaL_addchar(b, news[i]); 684 luaL_addchar(b, news[i]);
597 else { 685 else {
598 i++; /* skip ESC */ 686 i++; /* skip ESC */
599 if (!isdigit(uchar(news[i]))) 687 if (!isdigit(uchar(news[i]))) {
688 if (news[i] != L_ESC)
689 luaL_error(ms->L, "invalid use of " LUA_QL("%c")
690 " in replacement string", L_ESC);
600 luaL_addchar(b, news[i]); 691 luaL_addchar(b, news[i]);
692 }
601 else if (news[i] == '0') 693 else if (news[i] == '0')
602 luaL_addlstring(b, s, e - s); 694 luaL_addlstring(b, s, e - s);
603 else { 695 else {
@@ -610,14 +702,9 @@ static void add_s (MatchState *ms, luaL_Buffer *b, const char *s,
610 702
611 703
612static void add_value (MatchState *ms, luaL_Buffer *b, const char *s, 704static void add_value (MatchState *ms, luaL_Buffer *b, const char *s,
613 const char *e) { 705 const char *e, int tr) {
614 lua_State *L = ms->L; 706 lua_State *L = ms->L;
615 switch (lua_type(L, 3)) { 707 switch (tr) {
616 case LUA_TNUMBER:
617 case LUA_TSTRING: {
618 add_s(ms, b, s, e);
619 return;
620 }
621 case LUA_TFUNCTION: { 708 case LUA_TFUNCTION: {
622 int n; 709 int n;
623 lua_pushvalue(L, 3); 710 lua_pushvalue(L, 3);
@@ -630,41 +717,51 @@ static void add_value (MatchState *ms, luaL_Buffer *b, const char *s,
630 lua_gettable(L, 3); 717 lua_gettable(L, 3);
631 break; 718 break;
632 } 719 }
720 default: { /* LUA_TNUMBER or LUA_TSTRING */
721 add_s(ms, b, s, e);
722 return;
723 }
633 } 724 }
634 if (!lua_toboolean(L, -1)) { /* nil or false? */ 725 if (!lua_toboolean(L, -1)) { /* nil or false? */
635 lua_pop(L, 1); 726 lua_pop(L, 1);
636 lua_pushlstring(L, s, e - s); /* keep original text */ 727 lua_pushlstring(L, s, e - s); /* keep original text */
637 } 728 }
638 else if (!lua_isstring(L, -1)) 729 else if (!lua_isstring(L, -1))
639 luaL_error(L, "invalid replacement value (a %s)", luaL_typename(L, -1)); 730 luaL_error(L, "invalid replacement value (a %s)", luaL_typename(L, -1));
640 luaL_addvalue(b); /* add result to accumulator */ 731 luaL_addvalue(b); /* add result to accumulator */
641} 732}
642 733
643 734
644static int str_gsub (lua_State *L) { 735static int str_gsub (lua_State *L) {
645 size_t srcl; 736 size_t srcl, lp;
646 const char *src = luaL_checklstring(L, 1, &srcl); 737 const char *src = luaL_checklstring(L, 1, &srcl);
647 const char *p = luaL_checkstring(L, 2); 738 const char *p = luaL_checklstring(L, 2, &lp);
648 int tr = lua_type(L, 3); 739 int tr = lua_type(L, 3);
649 int max_s = luaL_optint(L, 4, srcl+1); 740 size_t max_s = luaL_optinteger(L, 4, srcl+1);
650 int anchor = (*p == '^') ? (p++, 1) : 0; 741 int anchor = (*p == '^');
651 int n = 0; 742 size_t n = 0;
652 MatchState ms; 743 MatchState ms;
653 luaL_Buffer b; 744 luaL_Buffer b;
654 luaL_argcheck(L, tr == LUA_TNUMBER || tr == LUA_TSTRING || 745 luaL_argcheck(L, tr == LUA_TNUMBER || tr == LUA_TSTRING ||
655 tr == LUA_TFUNCTION || tr == LUA_TTABLE, 3, 746 tr == LUA_TFUNCTION || tr == LUA_TTABLE, 3,
656 "string/function/table expected"); 747 "string/function/table expected");
657 luaL_buffinit(L, &b); 748 luaL_buffinit(L, &b);
749 if (anchor) {
750 p++; lp--; /* skip anchor character */
751 }
658 ms.L = L; 752 ms.L = L;
753 ms.matchdepth = MAXCCALLS;
659 ms.src_init = src; 754 ms.src_init = src;
660 ms.src_end = src+srcl; 755 ms.src_end = src+srcl;
756 ms.p_end = p + lp;
661 while (n < max_s) { 757 while (n < max_s) {
662 const char *e; 758 const char *e;
663 ms.level = 0; 759 ms.level = 0;
760 lua_assert(ms.matchdepth == MAXCCALLS);
664 e = match(&ms, src, p); 761 e = match(&ms, src, p);
665 if (e) { 762 if (e) {
666 n++; 763 n++;
667 add_value(&ms, &b, src, e); 764 add_value(&ms, &b, src, e, tr);
668 } 765 }
669 if (e && e>src) /* non empty match? */ 766 if (e && e>src) /* non empty match? */
670 src = e; /* skip it */ 767 src = e; /* skip it */
@@ -682,6 +779,46 @@ static int str_gsub (lua_State *L) {
682/* }====================================================== */ 779/* }====================================================== */
683 780
684 781
782
783/*
784** {======================================================
785** STRING FORMAT
786** =======================================================
787*/
788
789/*
790** LUA_INTFRMLEN is the length modifier for integer conversions in
791** 'string.format'; LUA_INTFRM_T is the integer type corresponding to
792** the previous length
793*/
794#if !defined(LUA_INTFRMLEN) /* { */
795#if defined(LUA_USE_LONGLONG)
796
797#define LUA_INTFRMLEN "ll"
798#define LUA_INTFRM_T long long
799
800#else
801
802#define LUA_INTFRMLEN "l"
803#define LUA_INTFRM_T long
804
805#endif
806#endif /* } */
807
808
809/*
810** LUA_FLTFRMLEN is the length modifier for float conversions in
811** 'string.format'; LUA_FLTFRM_T is the float type corresponding to
812** the previous length
813*/
814#if !defined(LUA_FLTFRMLEN)
815
816#define LUA_FLTFRMLEN ""
817#define LUA_FLTFRM_T double
818
819#endif
820
821
685/* maximum size of each formatted item (> len(format('%99.99f', -1e308))) */ 822/* maximum size of each formatted item (> len(format('%99.99f', -1e308))) */
686#define MAX_ITEM 512 823#define MAX_ITEM 512
687/* valid flags in a format specification */ 824/* valid flags in a format specification */
@@ -698,25 +835,20 @@ static void addquoted (lua_State *L, luaL_Buffer *b, int arg) {
698 const char *s = luaL_checklstring(L, arg, &l); 835 const char *s = luaL_checklstring(L, arg, &l);
699 luaL_addchar(b, '"'); 836 luaL_addchar(b, '"');
700 while (l--) { 837 while (l--) {
701 switch (*s) { 838 if (*s == '"' || *s == '\\' || *s == '\n') {
702 case '"': case '\\': case '\n': { 839 luaL_addchar(b, '\\');
703 luaL_addchar(b, '\\'); 840 luaL_addchar(b, *s);
704 luaL_addchar(b, *s);
705 break;
706 }
707 case '\r': {
708 luaL_addlstring(b, "\\r", 2);
709 break;
710 }
711 case '\0': {
712 luaL_addlstring(b, "\\000", 4);
713 break;
714 }
715 default: {
716 luaL_addchar(b, *s);
717 break;
718 }
719 } 841 }
842 else if (*s == '\0' || iscntrl(uchar(*s))) {
843 char buff[10];
844 if (!isdigit(uchar(*(s+1))))
845 snprintf(buff, sizeof(buff), "\\%d", (int)uchar(*s));
846 else
847 snprintf(buff, sizeof(buff), "\\%03d", (int)uchar(*s));
848 luaL_addstring(b, buff);
849 }
850 else
851 luaL_addchar(b, *s);
720 s++; 852 s++;
721 } 853 }
722 luaL_addchar(b, '"'); 854 luaL_addchar(b, '"');
@@ -725,7 +857,7 @@ static void addquoted (lua_State *L, luaL_Buffer *b, int arg) {
725static const char *scanformat (lua_State *L, const char *strfrmt, char *form) { 857static const char *scanformat (lua_State *L, const char *strfrmt, char *form) {
726 const char *p = strfrmt; 858 const char *p = strfrmt;
727 while (*p != '\0' && strchr(FLAGS, *p) != NULL) p++; /* skip flags */ 859 while (*p != '\0' && strchr(FLAGS, *p) != NULL) p++; /* skip flags */
728 if ((size_t)(p - strfrmt) >= sizeof(FLAGS)) 860 if ((size_t)(p - strfrmt) >= sizeof(FLAGS)/sizeof(char))
729 luaL_error(L, "invalid format (repeated flags)"); 861 luaL_error(L, "invalid format (repeated flags)");
730 if (isdigit(uchar(*p))) p++; /* skip width */ 862 if (isdigit(uchar(*p))) p++; /* skip width */
731 if (isdigit(uchar(*p))) p++; /* (2 digits at most) */ 863 if (isdigit(uchar(*p))) p++; /* (2 digits at most) */
@@ -737,23 +869,28 @@ static const char *scanformat (lua_State *L, const char *strfrmt, char *form) {
737 if (isdigit(uchar(*p))) 869 if (isdigit(uchar(*p)))
738 luaL_error(L, "invalid format (width or precision too long)"); 870 luaL_error(L, "invalid format (width or precision too long)");
739 *(form++) = '%'; 871 *(form++) = '%';
740 strncpy(form, strfrmt, p - strfrmt + 1); 872 memcpy(form, strfrmt, (p - strfrmt + 1) * sizeof(char));
741 form += p - strfrmt + 1; 873 form += p - strfrmt + 1;
742 *form = '\0'; 874 *form = '\0';
743 return p; 875 return p;
744} 876}
745 877
746 878
747static void addintlen (char *form) { 879/*
880** add length modifier into formats
881*/
882static void addlenmod (char *form, const char *lenmod) {
748 size_t l = strlen(form); 883 size_t l = strlen(form);
884 size_t lm = strlen(lenmod);
749 char spec = form[l - 1]; 885 char spec = form[l - 1];
750 strcpy(form + l - 1, LUA_INTFRMLEN); 886 strcpy(form + l - 1, lenmod);
751 form[l + sizeof(LUA_INTFRMLEN) - 2] = spec; 887 form[l + lm - 1] = spec;
752 form[l + sizeof(LUA_INTFRMLEN) - 1] = '\0'; 888 form[l + lm] = '\0';
753} 889}
754 890
755 891
756static int str_format (lua_State *L) { 892static int str_format (lua_State *L) {
893 int top = lua_gettop(L);
757 int arg = 1; 894 int arg = 1;
758 size_t sfl; 895 size_t sfl;
759 const char *strfrmt = luaL_checklstring(L, arg, &sfl); 896 const char *strfrmt = luaL_checklstring(L, arg, &sfl);
@@ -767,45 +904,61 @@ static int str_format (lua_State *L) {
767 luaL_addchar(&b, *strfrmt++); /* %% */ 904 luaL_addchar(&b, *strfrmt++); /* %% */
768 else { /* format item */ 905 else { /* format item */
769 char form[MAX_FORMAT]; /* to store the format (`%...') */ 906 char form[MAX_FORMAT]; /* to store the format (`%...') */
770 char buff[MAX_ITEM]; /* to store the formatted item */ 907 char *buff = luaL_prepbuffsize(&b, MAX_ITEM); /* to put formatted item */
771 arg++; 908 int nb = 0; /* number of bytes in added item */
909 if (++arg > top)
910 luaL_argerror(L, arg, "no value");
772 strfrmt = scanformat(L, strfrmt, form); 911 strfrmt = scanformat(L, strfrmt, form);
773 switch (*strfrmt++) { 912 switch (*strfrmt++) {
774 case 'c': { 913 case 'c': {
775 snprintf(buff, MAX_ITEM, form, (int)luaL_checknumber(L, arg)); 914 nb = snprintf(buff, MAX_ITEM, form, luaL_checkint(L, arg));
776 break; 915 break;
777 } 916 }
778 case 'd': case 'i': { 917 case 'd': case 'i': {
779 addintlen(form); 918 lua_Number n = luaL_checknumber(L, arg);
780 snprintf(buff, MAX_ITEM, form, (LUA_INTFRM_T)luaL_checknumber(L, arg)); 919 LUA_INTFRM_T ni = (LUA_INTFRM_T)n;
920 lua_Number diff = n - (lua_Number)ni;
921 luaL_argcheck(L, -1 < diff && diff < 1, arg,
922 "not a number in proper range");
923 addlenmod(form, LUA_INTFRMLEN);
924 nb = snprintf(buff, MAX_ITEM, form, ni);
781 break; 925 break;
782 } 926 }
783 case 'o': case 'u': case 'x': case 'X': { 927 case 'o': case 'u': case 'x': case 'X': {
784 addintlen(form); 928 lua_Number n = luaL_checknumber(L, arg);
785 snprintf(buff, MAX_ITEM, form, (unsigned LUA_INTFRM_T)luaL_checknumber(L, arg)); 929 unsigned LUA_INTFRM_T ni = (unsigned LUA_INTFRM_T)n;
930 lua_Number diff = n - (lua_Number)ni;
931 luaL_argcheck(L, -1 < diff && diff < 1, arg,
932 "not a non-negative number in proper range");
933 addlenmod(form, LUA_INTFRMLEN);
934 nb = snprintf(buff, MAX_ITEM, form, ni);
786 break; 935 break;
787 } 936 }
788 case 'e': case 'E': case 'f': 937 case 'e': case 'E': case 'f':
938#if defined(LUA_USE_AFORMAT)
939 case 'a': case 'A':
940#endif
789 case 'g': case 'G': { 941 case 'g': case 'G': {
790 snprintf(buff, MAX_ITEM, form, (double)luaL_checknumber(L, arg)); 942 addlenmod(form, LUA_FLTFRMLEN);
943 nb = snprintf(buff, MAX_ITEM, form, (LUA_FLTFRM_T)luaL_checknumber(L, arg));
791 break; 944 break;
792 } 945 }
793 case 'q': { 946 case 'q': {
794 addquoted(L, &b, arg); 947 addquoted(L, &b, arg);
795 continue; /* skip the 'addsize' at the end */ 948 break;
796 } 949 }
797 case 's': { 950 case 's': {
798 size_t l; 951 size_t l;
799 const char *s = luaL_checklstring(L, arg, &l); 952 const char *s = luaL_tolstring(L, arg, &l);
800 if (!strchr(form, '.') && l >= 100) { 953 if (!strchr(form, '.') && l >= 100) {
801 /* no precision and string is too long to be formatted; 954 /* no precision and string is too long to be formatted;
802 keep original string */ 955 keep original string */
803 lua_pushvalue(L, arg);
804 luaL_addvalue(&b); 956 luaL_addvalue(&b);
805 continue; /* skip the `addsize' at the end */ 957 break;
806 } 958 }
807 else { 959 else {
808 snprintf(buff, MAX_ITEM, form, s); 960 nb = snprintf(buff, MAX_ITEM, form, s);
961 lua_pop(L, 1); /* remove result from 'luaL_tolstring' */
809 break; 962 break;
810 } 963 }
811 } 964 }
@@ -814,13 +967,15 @@ static int str_format (lua_State *L) {
814 LUA_QL("format"), *(strfrmt - 1)); 967 LUA_QL("format"), *(strfrmt - 1));
815 } 968 }
816 } 969 }
817 luaL_addlstring(&b, buff, strlen(buff)); 970 luaL_addsize(&b, nb);
818 } 971 }
819 } 972 }
820 luaL_pushresult(&b); 973 luaL_pushresult(&b);
821 return 1; 974 return 1;
822} 975}
823 976
977/* }====================================================== */
978
824 979
825static const luaL_Reg strlib[] = { 980static const luaL_Reg strlib[] = {
826 {"byte", str_byte}, 981 {"byte", str_byte},
@@ -828,7 +983,6 @@ static const luaL_Reg strlib[] = {
828 {"dump", str_dump}, 983 {"dump", str_dump},
829 {"find", str_find}, 984 {"find", str_find},
830 {"format", str_format}, 985 {"format", str_format},
831 {"gfind", gfind_nodef},
832 {"gmatch", gmatch}, 986 {"gmatch", gmatch},
833 {"gsub", str_gsub}, 987 {"gsub", str_gsub},
834 {"len", str_len}, 988 {"len", str_len},
@@ -843,13 +997,13 @@ static const luaL_Reg strlib[] = {
843 997
844 998
845static void createmetatable (lua_State *L) { 999static void createmetatable (lua_State *L) {
846 lua_createtable(L, 0, 1); /* create metatable for strings */ 1000 lua_createtable(L, 0, 1); /* table to be metatable for strings */
847 lua_pushliteral(L, ""); /* dummy string */ 1001 lua_pushliteral(L, ""); /* dummy string */
848 lua_pushvalue(L, -2); 1002 lua_pushvalue(L, -2); /* copy table */
849 lua_setmetatable(L, -2); /* set string metatable */ 1003 lua_setmetatable(L, -2); /* set table as metatable for strings */
850 lua_pop(L, 1); /* pop dummy string */ 1004 lua_pop(L, 1); /* pop dummy string */
851 lua_pushvalue(L, -2); /* string library... */ 1005 lua_pushvalue(L, -2); /* get string library */
852 lua_setfield(L, -2, "__index"); /* ...is the __index metamethod */ 1006 lua_setfield(L, -2, "__index"); /* metatable.__index = string */
853 lua_pop(L, 1); /* pop metatable */ 1007 lua_pop(L, 1); /* pop metatable */
854} 1008}
855 1009
@@ -857,12 +1011,8 @@ static void createmetatable (lua_State *L) {
857/* 1011/*
858** Open string library 1012** Open string library
859*/ 1013*/
860LUALIB_API int luaopen_string (lua_State *L) { 1014LUAMOD_API int luaopen_string (lua_State *L) {
861 luaL_register(L, LUA_STRLIBNAME, strlib); 1015 luaL_newlib(L, strlib);
862#if defined(LUA_COMPAT_GFIND)
863 lua_getfield(L, -1, "gmatch");
864 lua_setfield(L, -2, "gfind");
865#endif
866 createmetatable(L); 1016 createmetatable(L);
867 return 1; 1017 return 1;
868} 1018}