gh-mirror/mikepaul-LuaJIT
1
0
Fork 0
mirror of https://github.com/LuaJIT/LuaJIT.git synced 2025-02-08 07:34:07 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Use string buffer for string.format().

Browse Source
This commit is contained in:
Mike Pall 2013-03-20 22:45:52 +01:00
parent fecde1b222
commit f1dbd65c0e
3 changed files with 117 additions and 104 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

178
src/lib_string.c
View File

@ -702,76 +702,81 @@ LJLIB_CF(string_gsub)
/* ------------------------------------------------------------------------ */ /* ------------------------------------------------------------------------ */
/* maximum size of each formatted item (> len(format('%99.99f', -1e308))) */ /* Max. buffer size needed (at least #string.format("%99.99f", -1e308)). */
#define MAX_FMTITEM 512 #define STRING_FMT_MAXBUF 512
/* valid flags in a format specification */ /* Valid format specifier flags. */
#define FMT_FLAGS "-+ #0" #define STRING_FMT_FLAGS "-+ #0"
/* /* Max. format specifier size. */
** maximum size of each format specification (such as '%-099.99d') #define STRING_FMT_MAXSPEC \
** (+10 accounts for %99.99x plus margin of error) (sizeof(STRING_FMT_FLAGS) + sizeof(LUA_INTFRMLEN) + 10)
*/
#define MAX_FMTSPEC (sizeof(FMT_FLAGS) + sizeof(LUA_INTFRMLEN) + 10)
static void addquoted(lua_State *L, luaL_Buffer *b, int arg) /* Add quoted string to buffer. */
static void string_fmt_quoted(SBuf *sb, GCstr *str)
{ {
GCstr *str = lj_lib_checkstr(L, arg);
int32_t len = (int32_t)str->len;
const char *s = strdata(str); const char *s = strdata(str);
luaL_addchar(b, '"'); MSize len = str->len;
lj_buf_putb(sb, '"');
while (len--) { while (len--) {
uint32_t c = uchar(*s); uint32_t c = (uint32_t)(uint8_t)*s++;
char *p = lj_buf_more(sb, 4);
if (c == '"' || c == '\\' || c == '\n') { if (c == '"' || c == '\\' || c == '\n') {
luaL_addchar(b, '\\'); *p++ = '\\';
} else if (lj_char_iscntrl(c)) { /* This can only be 0-31 or 127. */ } else if (lj_char_iscntrl(c)) { /* This can only be 0-31 or 127. */
uint32_t d; uint32_t d;
luaL_addchar(b, '\\'); *p++ = '\\';
if (c >= 100 || lj_char_isdigit(uchar(s[1]))) { if (c >= 100 || lj_char_isdigit((uint8_t)*s)) {
luaL_addchar(b, '0'+(c >= 100)); if (c >= 100) c -= 100; *p++ = (char)('0'+(c >= 100)); if (c >= 100) c -= 100;
goto tens; goto tens;
} else if (c >= 10) { } else if (c >= 10) {
tens: tens:
d = (c * 205) >> 11; c -= d * 10; luaL_addchar(b, '0'+d); d = (c * 205) >> 11; c -= d * 10; *p++ = (char)('0'+d);
} }
c += '0'; c += '0';
} }
luaL_addchar(b, c); *p++ = (char)c;
s++; setsbufP(sb, p);
} }
luaL_addchar(b, '"'); lj_buf_putb(sb, '"');
} }
static const char *scanformat(lua_State *L, const char *strfrmt, char *form) /* Scan format and generate format specifier. */
static const char *string_fmt_scan(lua_State *L, char *spec, const char *fmt)
{ {
const char *p = strfrmt; const char *p = fmt;
while (*p != '\0' && strchr(FMT_FLAGS, *p) != NULL) p++; /* skip flags */ while (*p && strchr(STRING_FMT_FLAGS, *p) != NULL) p++; /* Skip flags. */
if ((size_t)(p - strfrmt) >= sizeof(FMT_FLAGS)) if ((size_t)(p - fmt) >= sizeof(STRING_FMT_FLAGS))
lj_err_caller(L, LJ_ERR_STRFMTR); lj_err_caller(L, LJ_ERR_STRFMTR);
if (lj_char_isdigit(uchar(*p))) p++; /* skip width */ if (lj_char_isdigit((uint8_t)*p)) p++; /* Skip max. 2 digits for width. */
if (lj_char_isdigit(uchar(*p))) p++; /* (2 digits at most) */ if (lj_char_isdigit((uint8_t)*p)) p++;
if (*p == '.') { if (*p == '.') {
p++; p++;
if (lj_char_isdigit(uchar(*p))) p++; /* skip precision */ if (lj_char_isdigit((uint8_t)*p)) p++; /* Skip max. 2 digits for prec. */
if (lj_char_isdigit(uchar(*p))) p++; /* (2 digits at most) */ if (lj_char_isdigit((uint8_t)*p)) p++;
} }
if (lj_char_isdigit(uchar(*p))) if (lj_char_isdigit((uint8_t)*p))
lj_err_caller(L, LJ_ERR_STRFMTW); lj_err_caller(L, LJ_ERR_STRFMTW);
*(form++) = '%'; *spec++ = '%';
strncpy(form, strfrmt, (size_t)(p - strfrmt + 1)); strncpy(spec, fmt, (size_t)(p - fmt + 1));
form += p - strfrmt + 1; spec += p - fmt + 1;
*form = '\0'; *spec = '\0';
return p; return p;
} }
static void addintlen(char *form) /* Patch LUA_INTRFRMLEN into integer format specifier. */
static void string_fmt_intfmt(char *spec)
{ {
size_t l = strlen(form); char c;
char spec = form[l - 1]; do {
strcpy(form + l - 1, LUA_INTFRMLEN); c = *spec++;
form[l + sizeof(LUA_INTFRMLEN) - 2] = spec; } while (*spec);
form[l + sizeof(LUA_INTFRMLEN) - 1] = '\0'; *--spec = (LUA_INTFRMLEN)[0];
if ((LUA_INTFRMLEN)[1]) *++spec = (LUA_INTFRMLEN)[1];
*++spec = c;
*++spec = '\0';
} }
static unsigned LUA_INTFRM_T num2intfrm(lua_State *L, int arg) /* Derive sprintf argument for integer format. Ugly. */
static LUA_INTFRM_T string_fmt_intarg(lua_State *L, int arg)
{ {
if (sizeof(LUA_INTFRM_T) == 4) { if (sizeof(LUA_INTFRM_T) == 4) {
return (LUA_INTFRM_T)lj_lib_checkbit(L, arg); return (LUA_INTFRM_T)lj_lib_checkbit(L, arg);
@ -786,7 +791,8 @@ static unsigned LUA_INTFRM_T num2intfrm(lua_State *L, int arg)
} }
} }
static unsigned LUA_INTFRM_T num2uintfrm(lua_State *L, int arg) /* Derive sprintf argument for unsigned integer format. Ugly. */
static unsigned LUA_INTFRM_T string_fmt_uintarg(lua_State *L, int arg)
{ {
if (sizeof(LUA_INTFRM_T) == 4) { if (sizeof(LUA_INTFRM_T) == 4) {
return (unsigned LUA_INTFRM_T)lj_lib_checkbit(L, arg); return (unsigned LUA_INTFRM_T)lj_lib_checkbit(L, arg);
@ -803,7 +809,8 @@ static unsigned LUA_INTFRM_T num2uintfrm(lua_State *L, int arg)
} }
} }
static GCstr *meta_tostring(lua_State *L, int arg) /* Emulate tostring() inline. */
static GCstr *string_fmt_tostring(lua_State *L, int arg)
{ {
TValue *o = L->base+arg-1; TValue *o = L->base+arg-1;
cTValue *mo; cTValue *mo;
@ -841,33 +848,33 @@ static GCstr *meta_tostring(lua_State *L, int arg)
LJLIB_CF(string_format) LJLIB_CF(string_format)
{ {
int arg = 1, top = (int)(L->top - L->base); int arg = 1, top = (int)(L->top - L->base);
GCstr *fmt = lj_lib_checkstr(L, arg); GCstr *sfmt = lj_lib_checkstr(L, arg);
const char *strfrmt = strdata(fmt); const char *fmt = strdata(sfmt);
const char *strfrmt_end = strfrmt + fmt->len; const char *efmt = fmt + sfmt->len;
luaL_Buffer b; SBuf *sb = &G(L)->tmpbuf;
luaL_buffinit(L, &b); setmref(sb->L, L);
while (strfrmt < strfrmt_end) { lj_buf_reset(sb);
if (*strfrmt != L_ESC) { while (fmt < efmt) {
luaL_addchar(&b, *strfrmt++); if (*fmt != L_ESC || *++fmt == L_ESC) {
} else if (*++strfrmt == L_ESC) { lj_buf_putb(sb, *fmt++);
luaL_addchar(&b, *strfrmt++); /* %% */ } else {
} else { /* format item */ char buf[STRING_FMT_MAXBUF];
char form[MAX_FMTSPEC]; /* to store the format (`%...') */ char spec[STRING_FMT_MAXSPEC];
char buff[MAX_FMTITEM]; /* to store the formatted item */ MSize len = 0;
if (++arg > top) if (++arg > top)
luaL_argerror(L, arg, lj_obj_typename[0]); luaL_argerror(L, arg, lj_obj_typename[0]);
strfrmt = scanformat(L, strfrmt, form); fmt = string_fmt_scan(L, spec, fmt);
switch (*strfrmt++) { switch (*fmt++) {
case 'c': case 'c':
sprintf(buff, form, lj_lib_checkint(L, arg)); len = (MSize)sprintf(buf, spec, lj_lib_checkint(L, arg));
break; break;
case 'd': case 'i': case 'd': case 'i':
addintlen(form); string_fmt_intfmt(spec);
sprintf(buff, form, num2intfrm(L, arg)); len = (MSize)sprintf(buf, spec, string_fmt_intarg(L, arg));
break; break;
case 'o': case 'u': case 'x': case 'X': case 'o': case 'u': case 'x': case 'X':
addintlen(form); string_fmt_intfmt(spec);
sprintf(buff, form, num2uintfrm(L, arg)); len = (MSize)sprintf(buf, spec, string_fmt_uintarg(L, arg));
break; break;
case 'e': case 'E': case 'f': case 'g': case 'G': case 'a': case 'A': { case 'e': case 'E': case 'f': case 'g': case 'G': case 'a': case 'A': {
TValue tv; TValue tv;
@ -875,48 +882,45 @@ LJLIB_CF(string_format)
if (LJ_UNLIKELY((tv.u32.hi << 1) >= 0xffe00000)) { if (LJ_UNLIKELY((tv.u32.hi << 1) >= 0xffe00000)) {
/* Canonicalize output of non-finite values. */ /* Canonicalize output of non-finite values. */
char *p, nbuf[LJ_STR_NUMBUF]; char *p, nbuf[LJ_STR_NUMBUF];
MSize len = lj_str_bufnum(nbuf, &tv); MSize n = lj_str_bufnum(nbuf, &tv);
if (strfrmt[-1] < 'a') { if (fmt[-1] < 'a') {
nbuf[len-3] = nbuf[len-3] - 0x20; nbuf[n-3] = nbuf[n-3] - 0x20;
nbuf[len-2] = nbuf[len-2] - 0x20; nbuf[n-2] = nbuf[n-2] - 0x20;
nbuf[len-1] = nbuf[len-1] - 0x20; nbuf[n-1] = nbuf[n-1] - 0x20;
} }
nbuf[len] = '\0'; nbuf[n] = '\0';
for (p = form; *p < 'A' && *p != '.'; p++) ; for (p = spec; *p < 'A' && *p != '.'; p++) ;
*p++ = 's'; *p = '\0'; *p++ = 's'; *p = '\0';
sprintf(buff, form, nbuf); len = (MSize)sprintf(buf, spec, nbuf);
break; break;
} }
sprintf(buff, form, (double)tv.n); len = (MSize)sprintf(buf, spec, (double)tv.n);
break; break;
} }
case 'q': case 'q':
addquoted(L, &b, arg); string_fmt_quoted(sb, lj_lib_checkstr(L, arg));
continue; continue;
case 'p': case 'p':
lj_str_pushf(L, "%p", lua_topointer(L, arg)); len = lj_str_bufptr(buf, lua_topointer(L, arg));
luaL_addvalue(&b); break;
continue;
case 's': { case 's': {
GCstr *str = meta_tostring(L, arg); GCstr *str = string_fmt_tostring(L, arg);
if (!strchr(form, '.') && str->len >= 100) { if (!strchr(spec, '.') && str->len >= 100) { /* Format overflow? */
/* no precision and string is too long to be formatted; lj_buf_putmem(sb, strdata(str), str->len); /* Use orig string. */
keep original string */
setstrV(L, L->top++, str);
luaL_addvalue(&b);
continue; continue;
} }
sprintf(buff, form, strdata(str)); len = (MSize)sprintf(buf, spec, strdata(str));
break; break;
} }
default: default:
lj_err_callerv(L, LJ_ERR_STRFMTO, *(strfrmt -1)); lj_err_callerv(L, LJ_ERR_STRFMTO, fmt[-1] ? fmt[-1] : ' ');
break; break;
} }
luaL_addlstring(&b, buff, strlen(buff)); lj_buf_putmem(sb, buf, len);
} }
} }
luaL_pushresult(&b); setstrV(L, L->top-1, lj_buf_str(L, sb));
lj_gc_check(L);
return 1; return 1;
} }

41
src/lj_str.c
View File

@ -186,7 +186,7 @@ MSize LJ_FASTCALL lj_str_bufnum(char *s, cTValue *o)
} }
/* Print integer to buffer. Returns pointer to start (!= buffer start). */ /* Print integer to buffer. Returns pointer to start (!= buffer start). */
static char * str_bufint(char *p, int32_t k) static char *str_bufint(char *p, int32_t k)
{ {
uint32_t u = (uint32_t)(k < 0 ? -k : k); uint32_t u = (uint32_t)(k < 0 ? -k : k);
p += LJ_STR_INTBUF; p += LJ_STR_INTBUF;
@ -195,6 +195,26 @@ static char * str_bufint(char *p, int32_t k)
return p; return p;
} }
/* Print pointer to buffer. */
MSize LJ_FASTCALL lj_str_bufptr(char *p, const void *v)
{
ptrdiff_t x = (ptrdiff_t)v;
MSize i, n = LJ_STR_PTRBUF;
if (x == 0) {
p[0] = 'N'; p[1] = 'U'; p[2] = 'L'; p[3] = 'L';
return 4;
}
#if LJ_64
/* Shorten output for 64 bit pointers. */
n = 2+2*4+((x >> 32) ? 2+2*(lj_fls((uint32_t)(x >> 32))>>3) : 0);
#endif
p[0] = '0';
p[1] = 'x';
for (i = n-1; i >= 2; i--, x >>= 4)
p[i] = "0123456789abcdef"[(x & 15)];
return n;
}
/* Print TValue to buffer (only for numbers) and return pointer to start. */ /* Print TValue to buffer (only for numbers) and return pointer to start. */
const char *lj_str_buftv(char *buf, cTValue *o, MSize *lenp) const char *lj_str_buftv(char *buf, cTValue *o, MSize *lenp)
{ {
@ -275,22 +295,9 @@ const char *lj_str_pushvf(lua_State *L, const char *fmt, va_list argp)
} }
case 'p': { case 'p': {
#define FMTP_CHARS (2*sizeof(ptrdiff_t)) #define FMTP_CHARS (2*sizeof(ptrdiff_t))
char buf[2+FMTP_CHARS]; char buf[LJ_STR_PTRBUF];
ptrdiff_t p = (ptrdiff_t)(va_arg(argp, void *)); MSize len = lj_str_bufptr(buf, va_arg(argp, void *));
ptrdiff_t i, lasti = 2+FMTP_CHARS; lj_buf_putmem(sb, buf, len);
if (p == 0) {
lj_buf_putmem(sb, "NULL", 4);
break;
}
#if LJ_64
/* Shorten output for 64 bit pointers. */
lasti = 2+2*4+((p >> 32) ? 2+2*(lj_fls((uint32_t)(p >> 32))>>3) : 0);
#endif
buf[0] = '0';
buf[1] = 'x';
for (i = lasti-1; i >= 2; i--, p >>= 4)
buf[i] = "0123456789abcdef"[(p & 15)];
lj_buf_putmem(sb, buf, (MSize)lasti);
break; break;
} }
case '%': case '%':

2
src/lj_str.h
View File

@ -21,6 +21,7 @@ LJ_FUNC void LJ_FASTCALL lj_str_free(global_State *g, GCstr *s);
/* Type conversions. */ /* Type conversions. */
LJ_FUNC MSize LJ_FASTCALL lj_str_bufnum(char *s, cTValue *o); LJ_FUNC MSize LJ_FASTCALL lj_str_bufnum(char *s, cTValue *o);
LJ_FUNC MSize LJ_FASTCALL lj_str_bufptr(char *p, const void *v);
LJ_FUNC const char *lj_str_buftv(char *buf, cTValue *o, MSize *lenp); LJ_FUNC const char *lj_str_buftv(char *buf, cTValue *o, MSize *lenp);
LJ_FUNCA GCstr * LJ_FASTCALL lj_str_fromnum(lua_State *L, const lua_Number *np); LJ_FUNCA GCstr * LJ_FASTCALL lj_str_fromnum(lua_State *L, const lua_Number *np);
LJ_FUNC GCstr * LJ_FASTCALL lj_str_fromint(lua_State *L, int32_t k); LJ_FUNC GCstr * LJ_FASTCALL lj_str_fromint(lua_State *L, int32_t k);
@ -29,6 +30,7 @@ LJ_FUNCA GCstr * LJ_FASTCALL lj_str_fromnumber(lua_State *L, cTValue *o);
#define LJ_STR_INTBUF (1+10) #define LJ_STR_INTBUF (1+10)
#define LJ_STR_NUMBUF LUAI_MAXNUMBER2STR #define LJ_STR_NUMBUF LUAI_MAXNUMBER2STR
#define LJ_STR_NUMBERBUF LUAI_MAXNUMBER2STR #define LJ_STR_NUMBERBUF LUAI_MAXNUMBER2STR
#define LJ_STR_PTRBUF (2*sizeof(ptrdiff_t)+2)
/* String formatting. */ /* String formatting. */
LJ_FUNC const char *lj_str_pushvf(lua_State *L, const char *fmt, va_list argp); LJ_FUNC const char *lj_str_pushvf(lua_State *L, const char *fmt, va_list argp);