lua/lstrlib.c

894 lines
24 KiB
C
Raw Normal View History

1993-07-28 21:18:00 +08:00
/*
** $Id: lstrlib.c,v 1.147 2009/12/17 12:50:20 roberto Exp roberto $
1999-12-28 01:33:22 +08:00
** Standard library for string operations and pattern-matching
** See Copyright Notice in lua.h
1993-07-28 21:18:00 +08:00
*/
1993-12-18 02:41:19 +08:00
#include <ctype.h>
2000-05-24 21:54:49 +08:00
#include <stddef.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
1993-07-28 21:18:00 +08:00
2002-12-05 01:38:31 +08:00
#define lstrlib_c
#define LUA_LIB
2002-12-05 01:38:31 +08:00
1993-07-28 21:18:00 +08:00
#include "lua.h"
#include "lauxlib.h"
#include "lualib.h"
1993-07-28 21:18:00 +08:00
/*
** maximum number of captures that a pattern can do during
** pattern-matching. This limit is arbitrary.
*/
#if !defined(LUA_MAXCAPTURES)
#define LUA_MAXCAPTURES 32
#endif
/* macro to `unsign' a character */
#define uchar(c) ((unsigned char)(c))
2000-08-29 01:57:04 +08:00
static int str_len (lua_State *L) {
2000-05-24 21:54:49 +08:00
size_t l;
2002-11-14 23:41:38 +08:00
luaL_checklstring(L, 1, &l);
2007-10-26 03:30:36 +08:00
lua_pushinteger(L, (lua_Integer)l);
2000-08-29 01:57:04 +08:00
return 1;
}
2007-10-26 03:30:36 +08:00
/* translate a relative string position: negative means back from end */
static size_t posrelat (ptrdiff_t pos, size_t len) {
if (pos >= 0) return (size_t)pos;
else if (pos == -pos || (size_t)-pos > len) return 0;
2007-10-26 03:30:36 +08:00
else return len - ((size_t)-pos) + 1;
}
2000-08-29 01:57:04 +08:00
static int str_sub (lua_State *L) {
2000-05-24 21:54:49 +08:00
size_t l;
2002-11-14 23:41:38 +08:00
const char *s = luaL_checklstring(L, 1, &l);
2007-10-26 03:30:36 +08:00
size_t start = posrelat(luaL_checkinteger(L, 2), l);
size_t end = posrelat(luaL_optinteger(L, 3, -1), l);
if (start < 1) start = 1;
2007-10-26 03:30:36 +08:00
if (end > l) end = l;
if (start <= end)
2007-10-26 03:30:36 +08:00
lua_pushlstring(L, s + start - 1, end - start + 1);
else lua_pushliteral(L, "");
2000-08-29 01:57:04 +08:00
return 1;
1993-07-28 21:18:00 +08:00
}
2003-05-14 22:35:54 +08:00
static int str_reverse (lua_State *L) {
size_t l;
luaL_Buffer b;
const char *s = luaL_checklstring(L, 1, &l);
luaL_buffinit(L, &b);
while (l--) luaL_addchar(&b, s[l]);
2003-05-14 22:35:54 +08:00
luaL_pushresult(&b);
return 1;
}
2000-08-29 01:57:04 +08:00
static int str_lower (lua_State *L) {
2000-05-24 21:54:49 +08:00
size_t l;
size_t i;
2000-09-12 01:38:42 +08:00
luaL_Buffer b;
2002-11-14 23:41:38 +08:00
const char *s = luaL_checklstring(L, 1, &l);
2000-09-12 01:38:42 +08:00
luaL_buffinit(L, &b);
1998-03-07 00:54:42 +08:00
for (i=0; i<l; i++)
luaL_addchar(&b, tolower(uchar(s[i])));
2000-09-12 01:38:42 +08:00
luaL_pushresult(&b);
2000-08-29 01:57:04 +08:00
return 1;
}
2000-08-29 01:57:04 +08:00
static int str_upper (lua_State *L) {
2000-05-24 21:54:49 +08:00
size_t l;
size_t i;
2000-09-12 01:38:42 +08:00
luaL_Buffer b;
2002-11-14 23:41:38 +08:00
const char *s = luaL_checklstring(L, 1, &l);
2000-09-12 01:38:42 +08:00
luaL_buffinit(L, &b);
1998-03-07 00:54:42 +08:00
for (i=0; i<l; i++)
luaL_addchar(&b, toupper(uchar(s[i])));
2000-09-12 01:38:42 +08:00
luaL_pushresult(&b);
2000-08-29 01:57:04 +08:00
return 1;
}
1993-07-28 21:18:00 +08:00
2000-08-29 01:57:04 +08:00
static int str_rep (lua_State *L) {
2000-05-24 21:54:49 +08:00
size_t l;
2000-09-12 01:38:42 +08:00
luaL_Buffer b;
2002-11-14 23:41:38 +08:00
const char *s = luaL_checklstring(L, 1, &l);
int n = luaL_checkint(L, 2);
2000-09-12 01:38:42 +08:00
luaL_buffinit(L, &b);
1996-09-21 02:20:44 +08:00
while (n-- > 0)
2000-09-12 01:38:42 +08:00
luaL_addlstring(&b, s, l);
luaL_pushresult(&b);
2000-08-29 01:57:04 +08:00
return 1;
1996-09-21 02:20:44 +08:00
}
2000-08-29 01:57:04 +08:00
static int str_byte (lua_State *L) {
2000-05-24 21:54:49 +08:00
size_t l;
2002-11-14 23:41:38 +08:00
const char *s = luaL_checklstring(L, 1, &l);
2007-10-26 03:30:36 +08:00
size_t posi = posrelat(luaL_optinteger(L, 2, 1), l);
size_t pose = posrelat(luaL_optinteger(L, 3, posi), l);
2004-08-07 01:35:38 +08:00
int n, i;
2007-10-26 03:30:36 +08:00
if (posi < 1) posi = 1;
if (pose > l) pose = l;
if (posi > pose) return 0; /* empty interval; return no values */
2005-10-19 21:05:11 +08:00
n = (int)(pose - posi + 1);
if (posi + n <= pose) /* overflow? */
luaL_error(L, "string slice too long");
2004-08-07 01:35:38 +08:00
luaL_checkstack(L, n, "string slice too long");
for (i=0; i<n; i++)
lua_pushinteger(L, uchar(s[posi+i-1]));
return n;
}
2000-08-29 01:57:04 +08:00
static int str_char (lua_State *L) {
int n = lua_gettop(L); /* number of arguments */
int i;
2000-09-12 01:38:42 +08:00
luaL_Buffer b;
luaL_buffinit(L, &b);
2000-08-29 01:57:04 +08:00
for (i=1; i<=n; i++) {
2002-11-14 23:41:38 +08:00
int c = luaL_checkint(L, i);
luaL_argcheck(L, uchar(c) == c, i, "invalid value");
luaL_addchar(&b, uchar(c));
}
2000-09-12 01:38:42 +08:00
luaL_pushresult(&b);
2000-08-29 01:57:04 +08:00
return 1;
1998-03-25 04:14:25 +08:00
}
2002-11-26 01:33:33 +08:00
static int writer (lua_State *L, const void* b, size_t size, void* B) {
(void)L;
luaL_addlstring((luaL_Buffer*) B, (const char *)b, size);
return 0;
2002-11-26 01:33:33 +08:00
}
static int str_dump (lua_State *L) {
luaL_Buffer b;
luaL_checktype(L, 1, LUA_TFUNCTION);
2004-06-08 22:31:15 +08:00
lua_settop(L, 1);
2002-11-26 01:33:33 +08:00
luaL_buffinit(L,&b);
if (lua_dump(L, writer, &b) != 0)
2002-11-26 01:33:33 +08:00
luaL_error(L, "unable to dump given function");
luaL_pushresult(&b);
return 1;
}
/*
** {======================================================
** PATTERN MATCHING
** =======================================================
*/
#define CAP_UNFINISHED (-1)
#define CAP_POSITION (-2)
2000-11-23 21:49:35 +08:00
typedef struct MatchState {
const char *src_init; /* init of source string */
const char *src_end; /* end (`\0') of source string */
2002-02-09 06:39:36 +08:00
lua_State *L;
int level; /* total number of captures (finished or unfinished) */
struct {
const char *init;
2004-12-01 23:46:06 +08:00
ptrdiff_t len;
} capture[LUA_MAXCAPTURES];
2000-11-23 21:49:35 +08:00
} MatchState;
#define L_ESC '%'
#define SPECIALS "^$*+?.([%-"
2000-11-23 21:49:35 +08:00
static int check_capture (MatchState *ms, int l) {
l -= '1';
if (l < 0 || l >= ms->level || ms->capture[l].len == CAP_UNFINISHED)
2006-06-23 00:12:59 +08:00
return luaL_error(ms->L, "invalid capture index %%%d", l + 1);
return l;
}
2000-11-23 21:49:35 +08:00
static int capture_to_close (MatchState *ms) {
int level = ms->level;
for (level--; level>=0; level--)
if (ms->capture[level].len == CAP_UNFINISHED) return level;
2002-06-18 23:16:18 +08:00
return luaL_error(ms->L, "invalid pattern capture");
}
2004-11-20 00:59:08 +08:00
static const char *classend (MatchState *ms, const char *p) {
switch (*p++) {
case L_ESC: {
if (*p == '\0')
2005-05-18 03:49:15 +08:00
luaL_error(ms->L, "malformed pattern (ends with " LUA_QL("%%") ")");
return p+1;
2002-08-22 03:39:31 +08:00
}
case '[': {
if (*p == '^') p++;
2001-02-23 02:59:59 +08:00
do { /* look for a `]' */
if (*p == '\0')
2005-05-18 03:49:15 +08:00
luaL_error(ms->L, "malformed pattern (missing " LUA_QL("]") ")");
if (*(p++) == L_ESC && *p != '\0')
2001-03-03 01:40:08 +08:00
p++; /* skip escapes (e.g. `%]') */
} while (*p != ']');
return p+1;
2002-08-22 03:39:31 +08:00
}
default: {
return p;
2002-08-22 03:39:31 +08:00
}
}
}
static int match_class (int c, int cl) {
int res;
switch (tolower(cl)) {
case 'a' : res = isalpha(c); break;
case 'c' : res = iscntrl(c); break;
case 'd' : res = isdigit(c); break;
case 'l' : res = islower(c); break;
case 'p' : res = ispunct(c); break;
case 's' : res = isspace(c); break;
case 'u' : res = isupper(c); break;
case 'w' : res = isalnum(c); break;
case 'x' : res = isxdigit(c); break;
case 'z' : res = (c == 0); break;
default: return (cl == c);
}
return (islower(cl) ? res : !res);
}
static int matchbracketclass (int c, const char *p, const char *ec) {
int sig = 1;
if (*(p+1) == '^') {
sig = 0;
2001-02-23 02:59:59 +08:00
p++; /* skip the `^' */
}
2001-03-03 01:40:08 +08:00
while (++p < ec) {
if (*p == L_ESC) {
p++;
if (match_class(c, uchar(*p)))
return sig;
}
else if ((*(p+1) == '-') && (p+2 < ec)) {
p+=2;
if (uchar(*(p-2)) <= c && c <= uchar(*p))
return sig;
}
else if (uchar(*p) == c) return sig;
}
return !sig;
}
2004-11-20 00:59:08 +08:00
static int singlematch (int c, const char *p, const char *ep) {
switch (*p) {
2002-08-22 03:39:31 +08:00
case '.': return 1; /* matches any char */
case L_ESC: return match_class(c, uchar(*(p+1)));
2002-08-22 03:39:31 +08:00
case '[': return matchbracketclass(c, p, ep-1);
default: return (uchar(*p) == c);
}
}
static const char *match (MatchState *ms, const char *s, const char *p);
static const char *matchbalance (MatchState *ms, const char *s,
const char *p) {
if (*p == 0 || *(p+1) == 0)
2002-06-18 23:16:18 +08:00
luaL_error(ms->L, "unbalanced pattern");
if (*s != *p) return NULL;
else {
int b = *p;
int e = *(p+1);
int cont = 1;
2000-11-23 21:49:35 +08:00
while (++s < ms->src_end) {
if (*s == e) {
if (--cont == 0) return s+1;
}
else if (*s == b) cont++;
}
}
return NULL; /* string ends out of balance */
}
static const char *max_expand (MatchState *ms, const char *s,
const char *p, const char *ep) {
2004-12-01 23:46:06 +08:00
ptrdiff_t i = 0; /* counts maximum expand for item */
2004-11-20 00:59:08 +08:00
while ((s+i)<ms->src_end && singlematch(uchar(*(s+i)), p, ep))
i++;
2000-08-31 21:30:22 +08:00
/* keeps trying to match with the maximum repetitions */
while (i>=0) {
const char *res = match(ms, (s+i), ep+1);
if (res) return res;
i--; /* else didn't match; reduce 1 repetition to try again */
}
return NULL;
}
static const char *min_expand (MatchState *ms, const char *s,
const char *p, const char *ep) {
for (;;) {
const char *res = match(ms, s, ep+1);
if (res != NULL)
return res;
2004-11-20 00:59:08 +08:00
else if (s<ms->src_end && singlematch(uchar(*s), p, ep))
s++; /* try with one more repetition */
else return NULL;
}
}
static const char *start_capture (MatchState *ms, const char *s,
const char *p, int what) {
const char *res;
2000-11-23 21:49:35 +08:00
int level = ms->level;
if (level >= LUA_MAXCAPTURES) luaL_error(ms->L, "too many captures");
2000-11-23 21:49:35 +08:00
ms->capture[level].init = s;
ms->capture[level].len = what;
2000-11-23 21:49:35 +08:00
ms->level = level+1;
if ((res=match(ms, s, p)) == NULL) /* match failed? */
2000-11-23 21:49:35 +08:00
ms->level--; /* undo capture */
return res;
}
static const char *end_capture (MatchState *ms, const char *s,
const char *p) {
2000-11-23 21:49:35 +08:00
int l = capture_to_close(ms);
const char *res;
2000-11-23 21:49:35 +08:00
ms->capture[l].len = s - ms->capture[l].init; /* close capture */
if ((res = match(ms, s, p)) == NULL) /* match failed? */
ms->capture[l].len = CAP_UNFINISHED; /* undo capture */
return res;
}
static const char *match_capture (MatchState *ms, const char *s, int l) {
2001-03-03 01:40:08 +08:00
size_t len;
l = check_capture(ms, l);
len = ms->capture[l].len;
2000-11-23 21:49:35 +08:00
if ((size_t)(ms->src_end-s) >= len &&
memcmp(ms->capture[l].init, s, len) == 0)
return s+len;
else return NULL;
}
static const char *match (MatchState *ms, const char *s, const char *p) {
init: /* using goto's to optimize tail recursion */
switch (*p) {
2002-08-22 03:39:31 +08:00
case '(': { /* start capture */
if (*(p+1) == ')') /* position capture? */
return start_capture(ms, s, p+2, CAP_POSITION);
else
return start_capture(ms, s, p+1, CAP_UNFINISHED);
2002-08-22 03:39:31 +08:00
}
case ')': { /* end capture */
return end_capture(ms, s, p+1);
2002-08-22 03:39:31 +08:00
}
case '\0': { /* end of pattern */
return s; /* match succeeded */
}
case '$': {
if (*(p+1) == '\0') /* is the `$' the last char in pattern? */
return (s == ms->src_end) ? s : NULL; /* check end of string */
else goto dflt;
}
case L_ESC: { /* escaped sequences not in the format class[*+?-]? */
2002-08-24 03:45:24 +08:00
switch (*(p+1)) {
case 'b': { /* balanced string? */
s = matchbalance(ms, s, p+2);
if (s == NULL) return NULL;
p+=4; goto init; /* else return match(ms, s, p+4); */
}
case 'f': { /* frontier? */
const char *ep; char previous;
p += 2;
if (*p != '[')
2005-05-18 03:49:15 +08:00
luaL_error(ms->L, "missing " LUA_QL("[") " after "
LUA_QL("%%f") " in pattern");
2004-11-20 00:59:08 +08:00
ep = classend(ms, p); /* points to what is next */
2002-08-24 03:45:24 +08:00
previous = (s == ms->src_init) ? '\0' : *(s-1);
if (matchbracketclass(uchar(previous), p, ep-1) ||
!matchbracketclass(uchar(*s), p, ep-1)) return NULL;
p=ep; goto init; /* else return match(ms, s, ep); */
}
case '0': case '1': case '2': case '3':
case '4': case '5': case '6': case '7':
case '8': case '9': { /* capture results (%0-%9)? */
s = match_capture(ms, s, uchar(*(p+1)));
if (s == NULL) return NULL;
p+=2; goto init; /* else return match(ms, s, p+2) */
2002-08-24 03:45:24 +08:00
}
default: break; /* go through to 'dflt' */
}
2002-08-22 03:39:31 +08:00
}
default: dflt: { /* pattern class plus optional sufix */
2004-11-20 00:59:08 +08:00
const char *ep = classend(ms, p); /* points to what is next */
int m = s<ms->src_end && singlematch(uchar(*s), p, ep);
switch (*ep) {
case '?': { /* optional */
const char *res;
2000-11-23 21:49:35 +08:00
if (m && ((res=match(ms, s+1, ep+1)) != NULL))
1998-01-09 22:44:55 +08:00
return res;
2000-11-23 21:49:35 +08:00
p=ep+1; goto init; /* else return match(ms, s, ep+1); */
1998-01-09 22:44:55 +08:00
}
2002-08-22 03:39:31 +08:00
case '*': { /* 0 or more repetitions */
2000-11-23 21:49:35 +08:00
return max_expand(ms, s, p, ep);
2002-08-22 03:39:31 +08:00
}
case '+': { /* 1 or more repetitions */
2000-11-23 21:49:35 +08:00
return (m ? max_expand(ms, s+1, p, ep) : NULL);
2002-08-22 03:39:31 +08:00
}
case '-': { /* 0 or more repetitions (minimum) */
2000-11-23 21:49:35 +08:00
return min_expand(ms, s, p, ep);
2002-08-22 03:39:31 +08:00
}
default: {
if (!m) return NULL;
2000-11-23 21:49:35 +08:00
s++; p=ep; goto init; /* else return match(ms, s+1, ep); */
2002-08-22 03:39:31 +08:00
}
}
}
}
}
static const char *lmemfind (const char *s1, size_t l1,
const char *s2, size_t l2) {
2000-06-12 22:37:18 +08:00
if (l2 == 0) return s1; /* empty strings are everywhere */
else if (l2 > l1) return NULL; /* avoids a negative `l1' */
else {
const char *init; /* to search for a `*s2' inside `s1' */
l2--; /* 1st char will be checked by `memchr' */
l1 = l1-l2; /* `s2' cannot be found after that */
while (l1 > 0 && (init = (const char *)memchr(s1, *s2, l1)) != NULL) {
init++; /* 1st char is already checked */
if (memcmp(init, s2+1, l2) == 0)
return init-1;
else { /* correct `l1' and `s1' to try again */
l1 -= init-s1;
s1 = init;
}
}
return NULL; /* not found */
}
}
static void push_onecapture (MatchState *ms, int i, const char *s,
const char *e) {
if (i >= ms->level) {
if (i == 0) /* ms->level == 0, too */
lua_pushlstring(ms->L, s, e - s); /* add whole match */
else
luaL_error(ms->L, "invalid capture index");
}
else {
ptrdiff_t l = ms->capture[i].len;
if (l == CAP_UNFINISHED) luaL_error(ms->L, "unfinished capture");
if (l == CAP_POSITION)
lua_pushinteger(ms->L, ms->capture[i].init - ms->src_init + 1);
else
lua_pushlstring(ms->L, ms->capture[i].init, l);
}
}
2002-03-11 21:29:40 +08:00
static int push_captures (MatchState *ms, const char *s, const char *e) {
2000-08-29 01:57:04 +08:00
int i;
int nlevels = (ms->level == 0 && s) ? 1 : ms->level;
luaL_checkstack(ms->L, nlevels, "too many captures");
for (i = 0; i < nlevels; i++)
push_onecapture(ms, i, s, e);
return nlevels; /* number of strings pushed */
2000-08-29 01:57:04 +08:00
}
static int str_find_aux (lua_State *L, int find) {
2000-05-24 21:54:49 +08:00
size_t l1, l2;
2002-11-14 23:41:38 +08:00
const char *s = luaL_checklstring(L, 1, &l1);
const char *p = luaL_checklstring(L, 2, &l2);
2007-10-26 03:30:36 +08:00
size_t init = posrelat(luaL_optinteger(L, 3, 1), l1);
if (init < 1) init = 1;
else if (init > l1 + 1) { /* start after string's end? */
lua_pushnil(L); /* cannot find anything */
return 1;
}
if (find && (lua_toboolean(L, 4) || /* explicit request? */
strpbrk(p, SPECIALS) == NULL)) { /* or no special characters? */
2002-01-04 01:27:43 +08:00
/* do a plain search */
2007-10-26 03:30:36 +08:00
const char *s2 = lmemfind(s + init - 1, l1 - init + 1, p, l2);
if (s2) {
2007-10-26 03:30:36 +08:00
lua_pushinteger(L, s2 - s + 1);
lua_pushinteger(L, s2 - s + l2);
2000-08-29 01:57:04 +08:00
return 2;
}
}
else {
2000-11-23 21:49:35 +08:00
MatchState ms;
int anchor = (*p == '^') ? (p++, 1) : 0;
2007-10-26 03:30:36 +08:00
const char *s1 = s + init - 1;
2000-11-23 21:49:35 +08:00
ms.L = L;
ms.src_init = s;
2007-10-26 03:30:36 +08:00
ms.src_end = s + l1;
do {
const char *res;
2000-11-23 21:49:35 +08:00
ms.level = 0;
if ((res=match(&ms, s1, p)) != NULL) {
if (find) {
2007-10-26 03:30:36 +08:00
lua_pushinteger(L, s1 - s + 1); /* start */
lua_pushinteger(L, res - s); /* end */
return push_captures(&ms, NULL, 0) + 2;
}
else
return push_captures(&ms, s1, res);
}
} while (s1++ < ms.src_end && !anchor);
}
lua_pushnil(L); /* not found */
2000-08-29 01:57:04 +08:00
return 1;
}
static int str_find (lua_State *L) {
return str_find_aux(L, 1);
}
static int str_match (lua_State *L) {
return str_find_aux(L, 0);
}
static int gmatch_aux (lua_State *L) {
2002-03-11 21:29:40 +08:00
MatchState ms;
2005-05-17 03:21:11 +08:00
size_t ls;
const char *s = lua_tolstring(L, lua_upvalueindex(1), &ls);
2002-03-11 21:29:40 +08:00
const char *p = lua_tostring(L, lua_upvalueindex(2));
const char *src;
ms.L = L;
ms.src_init = s;
ms.src_end = s+ls;
for (src = s + (size_t)lua_tointeger(L, lua_upvalueindex(3));
2002-03-11 21:29:40 +08:00
src <= ms.src_end;
src++) {
const char *e;
ms.level = 0;
if ((e = match(&ms, src, p)) != NULL) {
2005-05-31 22:25:18 +08:00
lua_Integer newstart = e-s;
2002-03-11 21:29:40 +08:00
if (e == src) newstart++; /* empty match? go at least one position */
lua_pushinteger(L, newstart);
2002-03-11 21:29:40 +08:00
lua_replace(L, lua_upvalueindex(3));
return push_captures(&ms, src, e);
}
}
return 0; /* not found */
}
static int gmatch (lua_State *L) {
2002-11-14 23:41:38 +08:00
luaL_checkstring(L, 1);
luaL_checkstring(L, 2);
2002-03-11 21:29:40 +08:00
lua_settop(L, 2);
lua_pushinteger(L, 0);
lua_pushcclosure(L, gmatch_aux, 3);
2002-03-11 21:29:40 +08:00
return 1;
}
static int gfind_nodef (lua_State *L) {
return luaL_error(L, LUA_QL("string.gfind") " was renamed to "
LUA_QL("string.gmatch"));
}
static void add_s (MatchState *ms, luaL_Buffer *b, const char *s,
const char *e) {
size_t l, i;
const char *news = lua_tolstring(ms->L, 3, &l);
for (i = 0; i < l; i++) {
if (news[i] != L_ESC)
luaL_addchar(b, news[i]);
else {
i++; /* skip ESC */
if (!isdigit(uchar(news[i]))) {
if (news[i] != L_ESC)
luaL_error(ms->L, "invalid use of " LUA_QL("%c")
" in replacement string", L_ESC);
luaL_addchar(b, news[i]);
}
else if (news[i] == '0')
luaL_addlstring(b, s, e - s);
else {
push_onecapture(ms, news[i] - '1', s, e);
luaL_addvalue(b); /* add capture to accumulated result */
}
}
}
}
static void add_value (MatchState *ms, luaL_Buffer *b, const char *s,
const char *e, int tr) {
lua_State *L = ms->L;
switch (tr) {
case LUA_TFUNCTION: {
int n;
lua_pushvalue(L, 3);
n = push_captures(ms, s, e);
lua_call(L, n, 1);
break;
}
case LUA_TTABLE: {
push_onecapture(ms, 0, s, e);
lua_gettable(L, 3);
break;
}
default: { /* LUA_TNUMBER or LUA_TSTRING */
add_s(ms, b, s, e);
return;
}
}
if (!lua_toboolean(L, -1)) { /* nil or false? */
lua_pop(L, 1);
lua_pushlstring(L, s, e - s); /* keep original text */
}
else if (!lua_isstring(L, -1))
2006-09-11 22:07:24 +08:00
luaL_error(L, "invalid replacement value (a %s)", luaL_typename(L, -1));
luaL_addvalue(b); /* add result to accumulator */
}
2000-08-29 01:57:04 +08:00
static int str_gsub (lua_State *L) {
2000-05-24 21:54:49 +08:00
size_t srcl;
2002-11-14 23:41:38 +08:00
const char *src = luaL_checklstring(L, 1, &srcl);
const char *p = luaL_checkstring(L, 2);
int tr = lua_type(L, 3);
2007-10-26 03:30:36 +08:00
size_t max_s = luaL_optinteger(L, 4, srcl+1);
int anchor = (*p == '^') ? (p++, 1) : 0;
2007-10-26 03:30:36 +08:00
size_t n = 0;
2000-11-23 21:49:35 +08:00
MatchState ms;
2000-09-12 01:38:42 +08:00
luaL_Buffer b;
luaL_argcheck(L, tr == LUA_TNUMBER || tr == LUA_TSTRING ||
tr == LUA_TFUNCTION || tr == LUA_TTABLE, 3,
"string/function/table expected");
2000-09-12 01:38:42 +08:00
luaL_buffinit(L, &b);
2000-11-23 21:49:35 +08:00
ms.L = L;
ms.src_init = src;
2000-11-23 21:49:35 +08:00
ms.src_end = src+srcl;
while (n < max_s) {
const char *e;
2000-11-23 21:49:35 +08:00
ms.level = 0;
e = match(&ms, src, p);
if (e) {
n++;
add_value(&ms, &b, src, e, tr);
}
if (e && e>src) /* non empty match? */
src = e; /* skip it */
2000-11-23 21:49:35 +08:00
else if (src < ms.src_end)
luaL_addchar(&b, *src++);
else break;
if (anchor) break;
}
2000-11-23 21:49:35 +08:00
luaL_addlstring(&b, src, ms.src_end-src);
2000-09-12 01:38:42 +08:00
luaL_pushresult(&b);
lua_pushinteger(L, n); /* number of substitutions */
2000-08-29 01:57:04 +08:00
return 2;
}
1993-07-28 21:18:00 +08:00
/* }====================================================== */
/*
** length modifier for integer conversions ** in 'string.format' and
** integer type corresponding to the previous length
*/
#if defined(LUA_USELONGLONG)
#define LUA_INTFRMLEN "ll"
#define LUA_INTFRM_T long long
#else
#define LUA_INTFRMLEN "l"
#define LUA_INTFRM_T long
#endif
2000-11-23 21:49:35 +08:00
/* maximum size of each formatted item (> len(format('%99.99f', -1e308))) */
#define MAX_ITEM 512
/* valid flags in a format specification */
#define FLAGS "-+ #0"
/*
** maximum size of each format specification (such as '%-099.99d')
** (+10 accounts for %99.99x plus margin of error)
*/
#define MAX_FORMAT (sizeof(FLAGS) + sizeof(LUA_INTFRMLEN) + 10)
2000-11-23 21:49:35 +08:00
2004-11-20 00:59:08 +08:00
static void addquoted (lua_State *L, luaL_Buffer *b, int arg) {
2000-05-24 21:54:49 +08:00
size_t l;
2002-11-14 23:41:38 +08:00
const char *s = luaL_checklstring(L, arg, &l);
luaL_addchar(b, '"');
1998-07-13 00:13:45 +08:00
while (l--) {
if (*s == '"' || *s == '\\' || *s == '\n') {
luaL_addchar(b, '\\');
luaL_addchar(b, *s);
1998-07-13 00:13:45 +08:00
}
else if (*s == '\0' || iscntrl(uchar(*s))) {
char buff[10];
if (*s != '\0' && !isdigit(uchar(*(s+1))))
sprintf(buff, "\\%d", uchar(*s));
else
sprintf(buff, "\\%03d", uchar(*s));
luaL_addstring(b, buff);
}
else
luaL_addchar(b, *s);
1998-07-13 00:13:45 +08:00
s++;
1996-03-14 23:52:35 +08:00
}
luaL_addchar(b, '"');
1996-03-14 23:52:35 +08:00
}
static const char *scanformat (lua_State *L, const char *strfrmt, char *form) {
const char *p = strfrmt;
while (*p != '\0' && strchr(FLAGS, *p) != NULL) p++; /* skip flags */
if ((size_t)(p - strfrmt) >= sizeof(FLAGS))
luaL_error(L, "invalid format (repeated flags)");
2001-02-23 01:15:18 +08:00
if (isdigit(uchar(*p))) p++; /* skip width */
if (isdigit(uchar(*p))) p++; /* (2 digits at most) */
if (*p == '.') {
2000-11-23 21:49:35 +08:00
p++;
2001-02-23 01:15:18 +08:00
if (isdigit(uchar(*p))) p++; /* skip precision */
if (isdigit(uchar(*p))) p++; /* (2 digits at most) */
2000-11-23 21:49:35 +08:00
}
2001-02-23 01:15:18 +08:00
if (isdigit(uchar(*p)))
2002-06-18 23:16:18 +08:00
luaL_error(L, "invalid format (width or precision too long)");
*(form++) = '%';
2007-11-12 22:10:09 +08:00
memcpy(form, strfrmt, p - strfrmt + 1);
form += p - strfrmt + 1;
*form = '\0';
2000-11-23 21:49:35 +08:00
return p;
}
static void addintlen (char *form) {
2005-12-21 20:59:43 +08:00
size_t l = strlen(form);
char spec = form[l - 1];
strcpy(form + l - 1, LUA_INTFRMLEN);
form[l + sizeof(LUA_INTFRMLEN) - 2] = spec;
form[l + sizeof(LUA_INTFRMLEN) - 1] = '\0';
}
2000-08-29 01:57:04 +08:00
static int str_format (lua_State *L) {
int arg = 1;
size_t sfl;
2002-11-14 23:41:38 +08:00
const char *strfrmt = luaL_checklstring(L, arg, &sfl);
const char *strfrmt_end = strfrmt+sfl;
2000-09-12 01:38:42 +08:00
luaL_Buffer b;
luaL_buffinit(L, &b);
while (strfrmt < strfrmt_end) {
if (*strfrmt != L_ESC)
luaL_addchar(&b, *strfrmt++);
else if (*++strfrmt == L_ESC)
luaL_addchar(&b, *strfrmt++); /* %% */
else { /* format item */
char form[MAX_FORMAT]; /* to store the format (`%...') */
char buff[MAX_ITEM]; /* to store the formatted item */
arg++;
strfrmt = scanformat(L, strfrmt, form);
switch (*strfrmt++) {
case 'c': {
sprintf(buff, form, luaL_checkint(L, arg));
break;
}
case 'd': case 'i':
2002-08-22 03:39:31 +08:00
case 'o': case 'u': case 'x': case 'X': {
lua_Number n = luaL_checknumber(L, arg);
LUA_INTFRM_T r = (n < 0) ? (LUA_INTFRM_T)n :
(LUA_INTFRM_T)(unsigned LUA_INTFRM_T)n;
addintlen(form);
sprintf(buff, form, r);
break;
2002-08-22 03:39:31 +08:00
}
case 'e': case 'E': case 'f':
2002-08-22 03:39:31 +08:00
case 'g': case 'G': {
sprintf(buff, form, (double)luaL_checknumber(L, arg));
break;
2002-08-22 03:39:31 +08:00
}
case 'q': {
2004-11-20 00:59:08 +08:00
addquoted(L, &b, arg);
continue; /* skip the 'addsize' at the end */
2002-08-22 03:39:31 +08:00
}
case 's': {
2000-05-24 21:54:49 +08:00
size_t l;
2002-11-14 23:41:38 +08:00
const char *s = luaL_checklstring(L, arg, &l);
if (!strchr(form, '.') && l >= 100) {
2000-03-03 22:58:26 +08:00
/* no precision and string is too long to be formatted;
1999-02-05 19:22:58 +08:00
keep original string */
2000-09-12 01:38:42 +08:00
lua_pushvalue(L, arg);
luaL_addvalue(&b);
2001-02-23 02:59:59 +08:00
continue; /* skip the `addsize' at the end */
1999-02-05 19:22:58 +08:00
}
else {
sprintf(buff, form, s);
break;
}
}
2002-08-22 03:39:31 +08:00
default: { /* also treat cases `pnLlh' */
return luaL_error(L, "invalid option " LUA_QL("%%%c") " to "
LUA_QL("format"), *(strfrmt - 1));
2002-08-22 03:39:31 +08:00
}
}
2000-09-12 01:38:42 +08:00
luaL_addlstring(&b, buff, strlen(buff));
}
}
2000-09-12 01:38:42 +08:00
luaL_pushresult(&b);
2000-08-29 01:57:04 +08:00
return 1;
}
1993-07-28 21:18:00 +08:00
2005-08-27 01:36:32 +08:00
static const luaL_Reg strlib[] = {
2002-03-20 20:54:08 +08:00
{"byte", str_byte},
{"char", str_char},
2002-11-26 01:33:33 +08:00
{"dump", str_dump},
2002-03-20 20:54:08 +08:00
{"find", str_find},
{"format", str_format},
{"gfind", gfind_nodef},
{"gmatch", gmatch},
2002-03-20 20:54:08 +08:00
{"gsub", str_gsub},
{"len", str_len},
{"lower", str_lower},
{"match", str_match},
{"rep", str_rep},
{"reverse", str_reverse},
{"sub", str_sub},
{"upper", str_upper},
2002-03-20 20:54:08 +08:00
{NULL, NULL}
};
2005-05-05 23:34:03 +08:00
static void createmetatable (lua_State *L) {
lua_createtable(L, 0, 1); /* create metatable for strings */
2005-05-05 23:34:03 +08:00
lua_pushliteral(L, ""); /* dummy string */
lua_pushvalue(L, -2);
lua_setmetatable(L, -2); /* set string metatable */
lua_pop(L, 1); /* pop dummy string */
lua_pushvalue(L, -2); /* string library... */
lua_setfield(L, -2, "__index"); /* ...is the __index metamethod */
lua_pop(L, 1); /* pop metatable */
}
1993-07-28 21:18:00 +08:00
/*
** Open string library
*/
LUAMOD_API int luaopen_string (lua_State *L) {
luaL_register(L, LUA_STRLIBNAME, strlib);
2005-05-05 23:34:03 +08:00
createmetatable(L);
2002-12-20 18:26:33 +08:00
return 1;
1993-07-28 21:18:00 +08:00
}
2002-03-20 20:54:08 +08:00