source/libs/luajit/LuaJIT-src/src/lj_strfmt.c

   1 /*
   2 ** String formatting.
   3 ** Copyright (C) 2005-2016 Mike Pall. See Copyright Notice in luajit.h
   4 */
   5
   6 #include <stdio.h>
   7
   8 #define lj_strfmt_c
   9 #define LUA_CORE
  10
  11 #include "lj_obj.h"
  12 #include "lj_buf.h"
  13 #include "lj_str.h"
  14 #include "lj_state.h"
  15 #include "lj_char.h"
  16 #include "lj_strfmt.h"
  17
  18 /* -- Format parser ------------------------------------------------------- */
  19
  20 static const uint8_t strfmt_map[('x'-'A')+1] = {
  21   STRFMT_A,0,0,0,STRFMT_E,STRFMT_F,STRFMT_G,0,0,0,0,0,0,
  22   0,0,0,0,0,0,0,0,0,0,STRFMT_X,0,0,
  23   0,0,0,0,0,0,
  24   STRFMT_A,0,STRFMT_C,STRFMT_D,STRFMT_E,STRFMT_F,STRFMT_G,0,STRFMT_I,0,0,0,0,
  25   0,STRFMT_O,STRFMT_P,STRFMT_Q,0,STRFMT_S,0,STRFMT_U,0,0,STRFMT_X
  26 };
  27
  28 SFormat LJ_FASTCALL lj_strfmt_parse(FormatState *fs)
  29 {
  30   const uint8_t *p = fs->p, *e = fs->e;
  31   fs->str = (const char *)p;
  32   for (; p < e; p++) {
  33     if (*p == '%') {  /* Escape char? */
  34       if (p[1] == '%') {  /* '%%'? */
  35         fs->p = ++p+1;
  36         goto retlit;
  37       } else {
  38         SFormat sf = 0;
  39         uint32_t c;
  40         if (p != (const uint8_t *)fs->str)
  41           break;
  42         for (p++; (uint32_t)*p - ' ' <= (uint32_t)('0' - ' '); p++) {
  43           /* Parse flags. */
  44           if (*p == '-') sf |= STRFMT_F_LEFT;
  45           else if (*p == '+') sf |= STRFMT_F_PLUS;
  46           else if (*p == '0') sf |= STRFMT_F_ZERO;
  47           else if (*p == ' ') sf |= STRFMT_F_SPACE;
  48           else if (*p == '#') sf |= STRFMT_F_ALT;
  49           else break;
  50         }
  51         if ((uint32_t)*p - '0' < 10) {  /* Parse width. */
  52           uint32_t width = (uint32_t)*p++ - '0';
  53           if ((uint32_t)*p - '0' < 10)
  54             width = (uint32_t)*p++ - '0' + width*10;
  55           sf |= (width << STRFMT_SH_WIDTH);
  56         }
  57         if (*p == '.') {  /* Parse precision. */
  58           uint32_t prec = 0;
  59           p++;
  60           if ((uint32_t)*p - '0' < 10) {
  61             prec = (uint32_t)*p++ - '0';
  62             if ((uint32_t)*p - '0' < 10)
  63               prec = (uint32_t)*p++ - '0' + prec*10;
  64           }
  65           sf |= ((prec+1) << STRFMT_SH_PREC);
  66         }
  67         /* Parse conversion. */
  68         c = (uint32_t)*p - 'A';
  69         if (LJ_LIKELY(c <= (uint32_t)('x' - 'A'))) {
  70           uint32_t sx = strfmt_map[c];
  71           if (sx) {
  72             fs->p = p+1;
  73             return (sf | sx | ((c & 0x20) ? 0 : STRFMT_F_UPPER));
  74           }
  75         }
  76         /* Return error location. */
  77         if (*p >= 32) p++;
  78         fs->len = (MSize)(p - (const uint8_t *)fs->str);
  79         fs->p = fs->e;
  80         return STRFMT_ERR;
  81       }
  82     }
  83   }
  84   fs->p = p;
  85 retlit:
  86   fs->len = (MSize)(p - (const uint8_t *)fs->str);
  87   return fs->len ? STRFMT_LIT : STRFMT_EOF;
  88 }
  89
  90 /* -- Raw conversions ----------------------------------------------------- */
  91
  92 #define WINT_R(x, sh, sc) \
  93   { uint32_t d = (x*(((1<<sh)+sc-1)/sc))>>sh; x -= d*sc; *p++ = (char)('0'+d); }
  94
  95 /* Write integer to buffer. */
  96 char * LJ_FASTCALL lj_strfmt_wint(char *p, int32_t k)
  97 {
  98   uint32_t u = (uint32_t)k;
  99   if (k < 0) { u = (uint32_t)-k; *p++ = '-'; }
 100   if (u < 10000) {
 101     if (u < 10) goto dig1; if (u < 100) goto dig2; if (u < 1000) goto dig3;
 102   } else {
 103     uint32_t v = u / 10000; u -= v * 10000;
 104     if (v < 10000) {
 105       if (v < 10) goto dig5; if (v < 100) goto dig6; if (v < 1000) goto dig7;
 106     } else {
 107       uint32_t w = v / 10000; v -= w * 10000;
 108       if (w >= 10) WINT_R(w, 10, 10)
 109       *p++ = (char)('0'+w);
 110     }
 111     WINT_R(v, 23, 1000)
 112     dig7: WINT_R(v, 12, 100)
 113     dig6: WINT_R(v, 10, 10)
 114     dig5: *p++ = (char)('0'+v);
 115   }
 116   WINT_R(u, 23, 1000)
 117   dig3: WINT_R(u, 12, 100)
 118   dig2: WINT_R(u, 10, 10)
 119   dig1: *p++ = (char)('0'+u);
 120   return p;
 121 }
 122 #undef WINT_R
 123
 124 /* Write pointer to buffer. */
 125 char * LJ_FASTCALL lj_strfmt_wptr(char *p, const void *v)
 126 {
 127   ptrdiff_t x = (ptrdiff_t)v;
 128   MSize i, n = STRFMT_MAXBUF_PTR;
 129   if (x == 0) {
 130     *p++ = 'N'; *p++ = 'U'; *p++ = 'L'; *p++ = 'L';
 131     return p;
 132   }
 133 #if LJ_64
 134   /* Shorten output for 64 bit pointers. */
 135   n = 2+2*4+((x >> 32) ? 2+2*(lj_fls((uint32_t)(x >> 32))>>3) : 0);
 136 #endif
 137   p[0] = '0';
 138   p[1] = 'x';
 139   for (i = n-1; i >= 2; i--, x >>= 4)
 140     p[i] = "0123456789abcdef"[(x & 15)];
 141   return p+n;
 142 }
 143
 144 /* Write ULEB128 to buffer. */
 145 char * LJ_FASTCALL lj_strfmt_wuleb128(char *p, uint32_t v)
 146 {
 147   for (; v >= 0x80; v >>= 7)
 148     *p++ = (char)((v & 0x7f) | 0x80);
 149   *p++ = (char)v;
 150   return p;
 151 }
 152
 153 /* Return string or write number to tmp buffer and return pointer to start. */
 154 const char *lj_strfmt_wstrnum(lua_State *L, cTValue *o, MSize *lenp)
 155 {
 156   SBuf *sb;
 157   if (tvisstr(o)) {
 158     *lenp = strV(o)->len;
 159     return strVdata(o);
 160   } else if (tvisint(o)) {
 161     sb = lj_strfmt_putint(lj_buf_tmp_(L), intV(o));
 162   } else if (tvisnum(o)) {
 163     sb = lj_strfmt_putfnum(lj_buf_tmp_(L), STRFMT_G14, o->n);
 164   } else {
 165     return NULL;
 166   }
 167   *lenp = sbuflen(sb);
 168   return sbufB(sb);
 169 }
 170
 171 /* -- Unformatted conversions to buffer ----------------------------------- */
 172
 173 /* Add integer to buffer. */
 174 SBuf * LJ_FASTCALL lj_strfmt_putint(SBuf *sb, int32_t k)
 175 {
 176   setsbufP(sb, lj_strfmt_wint(lj_buf_more(sb, STRFMT_MAXBUF_INT), k));
 177   return sb;
 178 }
 179
 180 #if LJ_HASJIT
 181 /* Add number to buffer. */
 182 SBuf * LJ_FASTCALL lj_strfmt_putnum(SBuf *sb, cTValue *o)
 183 {
 184   return lj_strfmt_putfnum(sb, STRFMT_G14, o->n);
 185 }
 186 #endif
 187
 188 SBuf * LJ_FASTCALL lj_strfmt_putptr(SBuf *sb, const void *v)
 189 {
 190   setsbufP(sb, lj_strfmt_wptr(lj_buf_more(sb, STRFMT_MAXBUF_PTR), v));
 191   return sb;
 192 }
 193
 194 /* Add quoted string to buffer. */
 195 SBuf * LJ_FASTCALL lj_strfmt_putquoted(SBuf *sb, GCstr *str)
 196 {
 197   const char *s = strdata(str);
 198   MSize len = str->len;
 199   lj_buf_putb(sb, '"');
 200   while (len--) {
 201     uint32_t c = (uint32_t)(uint8_t)*s++;
 202     char *p = lj_buf_more(sb, 4);
 203     if (c == '"' || c == '\\' || c == '\n') {
 204       *p++ = '\\';
 205     } else if (lj_char_iscntrl(c)) {  /* This can only be 0-31 or 127. */
 206       uint32_t d;
 207       *p++ = '\\';
 208       if (c >= 100 || lj_char_isdigit((uint8_t)*s)) {
 209         *p++ = (char)('0'+(c >= 100)); if (c >= 100) c -= 100;
 210         goto tens;
 211       } else if (c >= 10) {
 212       tens:
 213         d = (c * 205) >> 11; c -= d * 10; *p++ = (char)('0'+d);
 214       }
 215       c += '0';
 216     }
 217     *p++ = (char)c;
 218     setsbufP(sb, p);
 219   }
 220   lj_buf_putb(sb, '"');
 221   return sb;
 222 }
 223
 224 /* -- Formatted conversions to buffer ------------------------------------- */
 225
 226 /* Add formatted char to buffer. */
 227 SBuf *lj_strfmt_putfchar(SBuf *sb, SFormat sf, int32_t c)
 228 {
 229   MSize width = STRFMT_WIDTH(sf);
 230   char *p = lj_buf_more(sb, width > 1 ? width : 1);
 231   if ((sf & STRFMT_F_LEFT)) *p++ = (char)c;
 232   while (width-- > 1) *p++ = ' ';
 233   if (!(sf & STRFMT_F_LEFT)) *p++ = (char)c;
 234   setsbufP(sb, p);
 235   return sb;
 236 }
 237
 238 /* Add formatted string to buffer. */
 239 SBuf *lj_strfmt_putfstr(SBuf *sb, SFormat sf, GCstr *str)
 240 {
 241   MSize len = str->len <= STRFMT_PREC(sf) ? str->len : STRFMT_PREC(sf);
 242   MSize width = STRFMT_WIDTH(sf);
 243   char *p = lj_buf_more(sb, width > len ? width : len);
 244   if ((sf & STRFMT_F_LEFT)) p = lj_buf_wmem(p, strdata(str), len);
 245   while (width-- > len) *p++ = ' ';
 246   if (!(sf & STRFMT_F_LEFT)) p = lj_buf_wmem(p, strdata(str), len);
 247   setsbufP(sb, p);
 248   return sb;
 249 }
 250
 251 /* Add formatted signed/unsigned integer to buffer. */
 252 SBuf *lj_strfmt_putfxint(SBuf *sb, SFormat sf, uint64_t k)
 253 {
 254   char buf[STRFMT_MAXBUF_XINT], *q = buf + sizeof(buf), *p;
 255 #ifdef LUA_USE_ASSERT
 256   char *ps;
 257 #endif
 258   MSize prefix = 0, len, prec, pprec, width, need;
 259
 260   /* Figure out signed prefixes. */
 261   if (STRFMT_TYPE(sf) == STRFMT_INT) {
 262     if ((int64_t)k < 0) {
 263       k = (uint64_t)-(int64_t)k;
 264       prefix = 256 + '-';
 265     } else if ((sf & STRFMT_F_PLUS)) {
 266       prefix = 256 + '+';
 267     } else if ((sf & STRFMT_F_SPACE)) {
 268       prefix = 256 + ' ';
 269     }
 270   }
 271
 272   /* Convert number and store to fixed-size buffer in reverse order. */
 273   prec = STRFMT_PREC(sf);
 274   if ((int32_t)prec >= 0) sf &= ~STRFMT_F_ZERO;
 275   if (k == 0) {  /* Special-case zero argument. */
 276     if (prec != 0 ||
 277         (sf & (STRFMT_T_OCT|STRFMT_F_ALT)) == (STRFMT_T_OCT|STRFMT_F_ALT))
 278       *--q = '0';
 279   } else if (!(sf & (STRFMT_T_HEX|STRFMT_T_OCT))) {  /* Decimal. */
 280     uint32_t k2;
 281     while ((k >> 32)) { *--q = (char)('0' + k % 10); k /= 10; }
 282     k2 = (uint32_t)k;
 283     do { *--q = (char)('0' + k2 % 10); k2 /= 10; } while (k2);
 284   } else if ((sf & STRFMT_T_HEX)) {  /* Hex. */
 285     const char *hexdig = (sf & STRFMT_F_UPPER) ? "0123456789ABCDEF" :
 286                                                  "0123456789abcdef";
 287     do { *--q = hexdig[(k & 15)]; k >>= 4; } while (k);
 288     if ((sf & STRFMT_F_ALT)) prefix = 512 + ((sf & STRFMT_F_UPPER) ? 'X' : 'x');
 289   } else {  /* Octal. */
 290     do { *--q = (char)('0' + (uint32_t)(k & 7)); k >>= 3; } while (k);
 291     if ((sf & STRFMT_F_ALT)) *--q = '0';
 292   }
 293
 294   /* Calculate sizes. */
 295   len = (MSize)(buf + sizeof(buf) - q);
 296   if ((int32_t)len >= (int32_t)prec) prec = len;
 297   width = STRFMT_WIDTH(sf);
 298   pprec = prec + (prefix >> 8);
 299   need = width > pprec ? width : pprec;
 300   p = lj_buf_more(sb, need);
 301 #ifdef LUA_USE_ASSERT
 302   ps = p;
 303 #endif
 304
 305   /* Format number with leading/trailing whitespace and zeros. */
 306   if ((sf & (STRFMT_F_LEFT|STRFMT_F_ZERO)) == 0)
 307     while (width-- > pprec) *p++ = ' ';
 308   if (prefix) {
 309     if ((char)prefix >= 'X') *p++ = '0';
 310     *p++ = (char)prefix;
 311   }
 312   if ((sf & (STRFMT_F_LEFT|STRFMT_F_ZERO)) == STRFMT_F_ZERO)
 313     while (width-- > pprec) *p++ = '0';
 314   while (prec-- > len) *p++ = '0';
 315   while (q < buf + sizeof(buf)) *p++ = *q++;  /* Add number itself. */
 316   if ((sf & STRFMT_F_LEFT))
 317     while (width-- > pprec) *p++ = ' ';
 318
 319   lua_assert(need == (MSize)(p - ps));
 320   setsbufP(sb, p);
 321   return sb;
 322 }
 323
 324 /* Add number formatted as signed integer to buffer. */
 325 SBuf *lj_strfmt_putfnum_int(SBuf *sb, SFormat sf, lua_Number n)
 326 {
 327   int64_t k = (int64_t)n;
 328   if (checki32(k) && sf == STRFMT_INT)
 329     return lj_strfmt_putint(sb, (int32_t)k);  /* Shortcut for plain %d. */
 330   else
 331     return lj_strfmt_putfxint(sb, sf, (uint64_t)k);
 332 }
 333
 334 /* Add number formatted as unsigned integer to buffer. */
 335 SBuf *lj_strfmt_putfnum_uint(SBuf *sb, SFormat sf, lua_Number n)
 336 {
 337   int64_t k;
 338   if (n >= 9223372036854775808.0)
 339     k = (int64_t)(n - 18446744073709551616.0);
 340   else
 341     k = (int64_t)n;
 342   return lj_strfmt_putfxint(sb, sf, (uint64_t)k);
 343 }
 344
 345 /* -- Conversions to strings ---------------------------------------------- */
 346
 347 /* Convert integer to string. */
 348 GCstr * LJ_FASTCALL lj_strfmt_int(lua_State *L, int32_t k)
 349 {
 350   char buf[STRFMT_MAXBUF_INT];
 351   MSize len = (MSize)(lj_strfmt_wint(buf, k) - buf);
 352   return lj_str_new(L, buf, len);
 353 }
 354
 355 /* Convert integer or number to string. */
 356 GCstr * LJ_FASTCALL lj_strfmt_number(lua_State *L, cTValue *o)
 357 {
 358   return tvisint(o) ? lj_strfmt_int(L, intV(o)) : lj_strfmt_num(L, o);
 359 }
 360
 361 #if LJ_HASJIT
 362 /* Convert char value to string. */
 363 GCstr * LJ_FASTCALL lj_strfmt_char(lua_State *L, int c)
 364 {
 365   char buf[1];
 366   buf[0] = c;
 367   return lj_str_new(L, buf, 1);
 368 }
 369 #endif
 370
 371 /* Raw conversion of object to string. */
 372 GCstr * LJ_FASTCALL lj_strfmt_obj(lua_State *L, cTValue *o)
 373 {
 374   if (tvisstr(o)) {
 375     return strV(o);
 376   } else if (tvisnumber(o)) {
 377     return lj_strfmt_number(L, o);
 378   } else if (tvisnil(o)) {
 379     return lj_str_newlit(L, "nil");
 380   } else if (tvisfalse(o)) {
 381     return lj_str_newlit(L, "false");
 382   } else if (tvistrue(o)) {
 383     return lj_str_newlit(L, "true");
 384   } else {
 385     char buf[8+2+2+16], *p = buf;
 386     p = lj_buf_wmem(p, lj_typename(o), (MSize)strlen(lj_typename(o)));
 387     *p++ = ':'; *p++ = ' ';
 388     if (tvisfunc(o) && isffunc(funcV(o))) {
 389       p = lj_buf_wmem(p, "builtin#", 8);
 390       p = lj_strfmt_wint(p, funcV(o)->c.ffid);
 391     } else {
 392       p = lj_strfmt_wptr(p, lj_obj_ptr(o));
 393     }
 394     return lj_str_new(L, buf, (size_t)(p - buf));
 395   }
 396 }
 397
 398 /* -- Internal string formatting ------------------------------------------ */
 399
 400 /*
 401 ** These functions are only used for lua_pushfstring(), lua_pushvfstring()
 402 ** and for internal string formatting (e.g. error messages). Caveat: unlike
 403 ** string.format(), only a limited subset of formats and flags are supported!
 404 **
 405 ** LuaJIT has support for a couple more formats than Lua 5.1/5.2:
 406 ** - %d %u %o %x with full formatting, 32 bit integers only.
 407 ** - %f and other FP formats are really %.14g.
 408 ** - %s %c %p without formatting.
 409 */
 410
 411 /* Push formatted message as a string object to Lua stack. va_list variant. */
 412 const char *lj_strfmt_pushvf(lua_State *L, const char *fmt, va_list argp)
 413 {
 414   SBuf *sb = lj_buf_tmp_(L);
 415   FormatState fs;
 416   SFormat sf;
 417   GCstr *str;
 418   lj_strfmt_init(&fs, fmt, (MSize)strlen(fmt));
 419   while ((sf = lj_strfmt_parse(&fs)) != STRFMT_EOF) {
 420     switch (STRFMT_TYPE(sf)) {
 421     case STRFMT_LIT:
 422       lj_buf_putmem(sb, fs.str, fs.len);
 423       break;
 424     case STRFMT_INT:
 425       lj_strfmt_putfxint(sb, sf, va_arg(argp, int32_t));
 426       break;
 427     case STRFMT_UINT:
 428       lj_strfmt_putfxint(sb, sf, va_arg(argp, uint32_t));
 429       break;
 430     case STRFMT_NUM:
 431       lj_strfmt_putfnum(sb, STRFMT_G14, va_arg(argp, lua_Number));
 432       break;
 433     case STRFMT_STR: {
 434       const char *s = va_arg(argp, char *);
 435       if (s == NULL) s = "(null)";
 436       lj_buf_putmem(sb, s, (MSize)strlen(s));
 437       break;
 438       }
 439     case STRFMT_CHAR:
 440       lj_buf_putb(sb, va_arg(argp, int));
 441       break;
 442     case STRFMT_PTR:
 443       lj_strfmt_putptr(sb, va_arg(argp, void *));
 444       break;
 445     case STRFMT_ERR:
 446     default:
 447       lj_buf_putb(sb, '?');
 448       lua_assert(0);
 449       break;
 450     }
 451   }
 452   str = lj_buf_str(L, sb);
 453   setstrV(L, L->top, str);
 454   incr_top(L);
 455   return strdata(str);
 456 }
 457
 458 /* Push formatted message as a string object to Lua stack. Vararg variant. */
 459 const char *lj_strfmt_pushf(lua_State *L, const char *fmt, ...)
 460 {
 461   const char *msg;
 462   va_list argp;
 463   va_start(argp, fmt);
 464   msg = lj_strfmt_pushvf(L, fmt, argp);
 465   va_end(argp);
 466   return msg;
 467 }
 468