beta-0.89.2
[luatex.git] / source / libs / luajit / LuaJIT-src / src / lj_strfmt.c
blobd54e796ae2a31dce9e8b1d9546cdfc9d37d52615
1 /*
2 ** String formatting.
3 ** Copyright (C) 2005-2015 Mike Pall. See Copyright Notice in luajit.h
4 */
6 #include <stdio.h>
8 #define lj_strfmt_c
9 #define LUA_CORE
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"
18 /* -- Format parser ------------------------------------------------------- */
20 static const uint8_t strfmt_map[('x'-'A')+1] = {
21 STRFMT_A,0,0,0,STRFMT_E,0,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
28 SFormat LJ_FASTCALL lj_strfmt_parse(FormatState *fs)
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;
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);
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;
65 sf |= ((prec+1) << STRFMT_SH_PREC);
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));
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;
84 fs->p = p;
85 retlit:
86 fs->len = (MSize)(p - (const uint8_t *)fs->str);
87 return fs->len ? STRFMT_LIT : STRFMT_EOF;
90 /* -- Raw conversions ----------------------------------------------------- */
92 /* Write number to bufer. */
93 char * LJ_FASTCALL lj_strfmt_wnum(char *p, cTValue *o)
95 if (LJ_LIKELY((o->u32.hi << 1) < 0xffe00000)) { /* Finite? */
96 #if __BIONIC__
97 if (tvismzero(o)) { *p++ = '-'; *p++ = '0'; return p; }
98 #endif
99 return p + lua_number2str(p, o->n);
100 } else if (((o->u32.hi & 0x000fffff) | o->u32.lo) != 0) {
101 *p++ = 'n'; *p++ = 'a'; *p++ = 'n';
102 } else if ((o->u32.hi & 0x80000000) == 0) {
103 *p++ = 'i'; *p++ = 'n'; *p++ = 'f';
104 } else {
105 *p++ = '-'; *p++ = 'i'; *p++ = 'n'; *p++ = 'f';
107 return p;
110 #define WINT_R(x, sh, sc) \
111 { uint32_t d = (x*(((1<<sh)+sc-1)/sc))>>sh; x -= d*sc; *p++ = (char)('0'+d); }
113 /* Write integer to buffer. */
114 char * LJ_FASTCALL lj_strfmt_wint(char *p, int32_t k)
116 uint32_t u = (uint32_t)k;
117 if (k < 0) { u = (uint32_t)-k; *p++ = '-'; }
118 if (u < 10000) {
119 if (u < 10) goto dig1; if (u < 100) goto dig2; if (u < 1000) goto dig3;
120 } else {
121 uint32_t v = u / 10000; u -= v * 10000;
122 if (v < 10000) {
123 if (v < 10) goto dig5; if (v < 100) goto dig6; if (v < 1000) goto dig7;
124 } else {
125 uint32_t w = v / 10000; v -= w * 10000;
126 if (w >= 10) WINT_R(w, 10, 10)
127 *p++ = (char)('0'+w);
129 WINT_R(v, 23, 1000)
130 dig7: WINT_R(v, 12, 100)
131 dig6: WINT_R(v, 10, 10)
132 dig5: *p++ = (char)('0'+v);
134 WINT_R(u, 23, 1000)
135 dig3: WINT_R(u, 12, 100)
136 dig2: WINT_R(u, 10, 10)
137 dig1: *p++ = (char)('0'+u);
138 return p;
140 #undef WINT_R
142 /* Write pointer to buffer. */
143 char * LJ_FASTCALL lj_strfmt_wptr(char *p, const void *v)
145 ptrdiff_t x = (ptrdiff_t)v;
146 MSize i, n = STRFMT_MAXBUF_PTR;
147 if (x == 0) {
148 *p++ = 'N'; *p++ = 'U'; *p++ = 'L'; *p++ = 'L';
149 return p;
151 #if LJ_64
152 /* Shorten output for 64 bit pointers. */
153 n = 2+2*4+((x >> 32) ? 2+2*(lj_fls((uint32_t)(x >> 32))>>3) : 0);
154 #endif
155 p[0] = '0';
156 p[1] = 'x';
157 for (i = n-1; i >= 2; i--, x >>= 4)
158 p[i] = "0123456789abcdef"[(x & 15)];
159 return p+n;
162 /* Write ULEB128 to buffer. */
163 char * LJ_FASTCALL lj_strfmt_wuleb128(char *p, uint32_t v)
165 for (; v >= 0x80; v >>= 7)
166 *p++ = (char)((v & 0x7f) | 0x80);
167 *p++ = (char)v;
168 return p;
171 /* Return string or write number to buffer and return pointer to start. */
172 const char *lj_strfmt_wstrnum(char *buf, cTValue *o, MSize *lenp)
174 if (tvisstr(o)) {
175 *lenp = strV(o)->len;
176 return strVdata(o);
177 } else if (tvisint(o)) {
178 *lenp = (MSize)(lj_strfmt_wint(buf, intV(o)) - buf);
179 return buf;
180 } else if (tvisnum(o)) {
181 *lenp = (MSize)(lj_strfmt_wnum(buf, o) - buf);
182 return buf;
183 } else {
184 return NULL;
188 /* -- Unformatted conversions to buffer ----------------------------------- */
190 /* Add integer to buffer. */
191 SBuf * LJ_FASTCALL lj_strfmt_putint(SBuf *sb, int32_t k)
193 setsbufP(sb, lj_strfmt_wint(lj_buf_more(sb, STRFMT_MAXBUF_INT), k));
194 return sb;
197 #if LJ_HASJIT
198 /* Add number to buffer. */
199 SBuf * LJ_FASTCALL lj_strfmt_putnum(SBuf *sb, cTValue *o)
201 setsbufP(sb, lj_strfmt_wnum(lj_buf_more(sb, STRFMT_MAXBUF_NUM), o));
202 return sb;
204 #endif
206 SBuf * LJ_FASTCALL lj_strfmt_putptr(SBuf *sb, const void *v)
208 setsbufP(sb, lj_strfmt_wptr(lj_buf_more(sb, STRFMT_MAXBUF_PTR), v));
209 return sb;
212 /* Add quoted string to buffer. */
213 SBuf * LJ_FASTCALL lj_strfmt_putquoted(SBuf *sb, GCstr *str)
215 const char *s = strdata(str);
216 MSize len = str->len;
217 lj_buf_putb(sb, '"');
218 while (len--) {
219 uint32_t c = (uint32_t)(uint8_t)*s++;
220 char *p = lj_buf_more(sb, 4);
221 if (c == '"' || c == '\\' || c == '\n') {
222 *p++ = '\\';
223 } else if (lj_char_iscntrl(c)) { /* This can only be 0-31 or 127. */
224 uint32_t d;
225 *p++ = '\\';
226 if (c >= 100 || lj_char_isdigit((uint8_t)*s)) {
227 *p++ = (char)('0'+(c >= 100)); if (c >= 100) c -= 100;
228 goto tens;
229 } else if (c >= 10) {
230 tens:
231 d = (c * 205) >> 11; c -= d * 10; *p++ = (char)('0'+d);
233 c += '0';
235 *p++ = (char)c;
236 setsbufP(sb, p);
238 lj_buf_putb(sb, '"');
239 return sb;
242 /* -- Formatted conversions to buffer ------------------------------------- */
244 /* Add formatted char to buffer. */
245 SBuf *lj_strfmt_putfchar(SBuf *sb, SFormat sf, int32_t c)
247 MSize width = STRFMT_WIDTH(sf);
248 char *p = lj_buf_more(sb, width > 1 ? width : 1);
249 if ((sf & STRFMT_F_LEFT)) *p++ = (char)c;
250 while (width-- > 1) *p++ = ' ';
251 if (!(sf & STRFMT_F_LEFT)) *p++ = (char)c;
252 setsbufP(sb, p);
253 return sb;
256 /* Add formatted string to buffer. */
257 SBuf *lj_strfmt_putfstr(SBuf *sb, SFormat sf, GCstr *str)
259 MSize len = str->len <= STRFMT_PREC(sf) ? str->len : STRFMT_PREC(sf);
260 MSize width = STRFMT_WIDTH(sf);
261 char *p = lj_buf_more(sb, width > len ? width : len);
262 if ((sf & STRFMT_F_LEFT)) p = lj_buf_wmem(p, strdata(str), len);
263 while (width-- > len) *p++ = ' ';
264 if (!(sf & STRFMT_F_LEFT)) p = lj_buf_wmem(p, strdata(str), len);
265 setsbufP(sb, p);
266 return sb;
269 /* Add formatted signed/unsigned integer to buffer. */
270 SBuf *lj_strfmt_putfxint(SBuf *sb, SFormat sf, uint64_t k)
272 char buf[STRFMT_MAXBUF_XINT], *q = buf + sizeof(buf), *p;
273 #ifdef LUA_USE_ASSERT
274 char *ps;
275 #endif
276 MSize prefix = 0, len, prec, pprec, width, need;
278 /* Figure out signed prefixes. */
279 if (STRFMT_TYPE(sf) == STRFMT_INT) {
280 if ((int64_t)k < 0) {
281 k = (uint64_t)-(int64_t)k;
282 prefix = 256 + '-';
283 } else if ((sf & STRFMT_F_PLUS)) {
284 prefix = 256 + '+';
285 } else if ((sf & STRFMT_F_SPACE)) {
286 prefix = 256 + ' ';
290 /* Convert number and store to fixed-size buffer in reverse order. */
291 prec = STRFMT_PREC(sf);
292 if ((int32_t)prec >= 0) sf &= ~STRFMT_F_ZERO;
293 if (k == 0) { /* Special-case zero argument. */
294 if (prec != 0 ||
295 (sf & (STRFMT_T_OCT|STRFMT_F_ALT)) == (STRFMT_T_OCT|STRFMT_F_ALT))
296 *--q = '0';
297 } else if (!(sf & (STRFMT_T_HEX|STRFMT_T_OCT))) { /* Decimal. */
298 uint32_t k2;
299 while ((k >> 32)) { *--q = (char)('0' + k % 10); k /= 10; }
300 k2 = (uint32_t)k;
301 do { *--q = (char)('0' + k2 % 10); k2 /= 10; } while (k2);
302 } else if ((sf & STRFMT_T_HEX)) { /* Hex. */
303 const char *hexdig = (sf & STRFMT_F_UPPER) ? "0123456789ABCDEF" :
304 "0123456789abcdef";
305 do { *--q = hexdig[(k & 15)]; k >>= 4; } while (k);
306 if ((sf & STRFMT_F_ALT)) prefix = 512 + ((sf & STRFMT_F_UPPER) ? 'X' : 'x');
307 } else { /* Octal. */
308 do { *--q = (char)('0' + (uint32_t)(k & 7)); k >>= 3; } while (k);
309 if ((sf & STRFMT_F_ALT)) *--q = '0';
312 /* Calculate sizes. */
313 len = (MSize)(buf + sizeof(buf) - q);
314 if ((int32_t)len >= (int32_t)prec) prec = len;
315 width = STRFMT_WIDTH(sf);
316 pprec = prec + (prefix >> 8);
317 need = width > pprec ? width : pprec;
318 p = lj_buf_more(sb, need);
319 #ifdef LUA_USE_ASSERT
320 ps = p;
321 #endif
323 /* Format number with leading/trailing whitespace and zeros. */
324 if ((sf & (STRFMT_F_LEFT|STRFMT_F_ZERO)) == 0)
325 while (width-- > pprec) *p++ = ' ';
326 if (prefix) {
327 if ((char)prefix >= 'X') *p++ = '0';
328 *p++ = (char)prefix;
330 if ((sf & (STRFMT_F_LEFT|STRFMT_F_ZERO)) == STRFMT_F_ZERO)
331 while (width-- > pprec) *p++ = '0';
332 while (prec-- > len) *p++ = '0';
333 while (q < buf + sizeof(buf)) *p++ = *q++; /* Add number itself. */
334 if ((sf & STRFMT_F_LEFT))
335 while (width-- > pprec) *p++ = ' ';
337 lua_assert(need == (MSize)(p - ps));
338 setsbufP(sb, p);
339 return sb;
342 /* Add number formatted as signed integer to buffer. */
343 SBuf *lj_strfmt_putfnum_int(SBuf *sb, SFormat sf, lua_Number n)
345 int64_t k = (int64_t)n;
346 if (checki32(k) && sf == STRFMT_INT)
347 return lj_strfmt_putint(sb, (int32_t)k); /* Shortcut for plain %d. */
348 else
349 return lj_strfmt_putfxint(sb, sf, (uint64_t)k);
352 /* Add number formatted as unsigned integer to buffer. */
353 SBuf *lj_strfmt_putfnum_uint(SBuf *sb, SFormat sf, lua_Number n)
355 int64_t k;
356 if (n >= 9223372036854775808.0)
357 k = (int64_t)(n - 18446744073709551616.0);
358 else
359 k = (int64_t)n;
360 return lj_strfmt_putfxint(sb, sf, (uint64_t)k);
363 /* Max. sprintf buffer size needed. At least #string.format("%.99f", -1e308). */
364 #define STRFMT_FMTNUMBUF 512
366 /* Add formatted floating-point number to buffer. */
367 SBuf *lj_strfmt_putfnum(SBuf *sb, SFormat sf, lua_Number n)
369 TValue tv;
370 tv.n = n;
371 if (LJ_UNLIKELY((tv.u32.hi << 1) >= 0xffe00000)) {
372 /* Canonicalize output of non-finite values. */
373 MSize width = STRFMT_WIDTH(sf), len = 3;
374 int prefix = 0, ch = (sf & STRFMT_F_UPPER) ? 0x202020 : 0;
375 char *p;
376 if (((tv.u32.hi & 0x000fffff) | tv.u32.lo) != 0) {
377 ch ^= ('n' << 16) | ('a' << 8) | 'n';
378 if ((sf & STRFMT_F_SPACE)) prefix = ' ';
379 } else {
380 ch ^= ('i' << 16) | ('n' << 8) | 'f';
381 if ((tv.u32.hi & 0x80000000)) prefix = '-';
382 else if ((sf & STRFMT_F_PLUS)) prefix = '+';
383 else if ((sf & STRFMT_F_SPACE)) prefix = ' ';
385 if (prefix) len = 4;
386 p = lj_buf_more(sb, width > len ? width : len);
387 if (!(sf & STRFMT_F_LEFT)) while (width-- > len) *p++ = ' ';
388 if (prefix) *p++ = prefix;
389 *p++ = (char)(ch >> 16); *p++ = (char)(ch >> 8); *p++ = (char)ch;
390 if ((sf & STRFMT_F_LEFT)) while (width-- > len) *p++ = ' ';
391 setsbufP(sb, p);
392 } else { /* Delegate to sprintf() for now. */
393 uint8_t width = (uint8_t)STRFMT_WIDTH(sf), prec = (uint8_t)STRFMT_PREC(sf);
394 char fmt[1+5+2+3+1+1], *p = fmt;
395 *p++ = '%';
396 if ((sf & STRFMT_F_LEFT)) *p++ = '-';
397 if ((sf & STRFMT_F_PLUS)) *p++ = '+';
398 if ((sf & STRFMT_F_ZERO)) *p++ = '0';
399 if ((sf & STRFMT_F_SPACE)) *p++ = ' ';
400 if ((sf & STRFMT_F_ALT)) *p++ = '#';
401 if (width) {
402 uint8_t x = width / 10, y = width % 10;
403 if (x) *p++ = '0' + x;
404 *p++ = '0' + y;
406 if (prec != 255) {
407 uint8_t x = prec / 10, y = prec % 10;
408 *p++ = '.';
409 if (x) *p++ = '0' + x;
410 *p++ = '0' + y;
412 *p++ = (0x67666561 >> (STRFMT_FP(sf)<<3)) ^ ((sf & STRFMT_F_UPPER)?0x20:0);
413 *p = '\0';
414 p = lj_buf_more(sb, STRFMT_FMTNUMBUF);
415 setsbufP(sb, p + sprintf(p, fmt, n));
417 return sb;
420 /* -- Conversions to strings ---------------------------------------------- */
422 /* Convert integer to string. */
423 GCstr * LJ_FASTCALL lj_strfmt_int(lua_State *L, int32_t k)
425 char buf[STRFMT_MAXBUF_INT];
426 MSize len = (MSize)(lj_strfmt_wint(buf, k) - buf);
427 return lj_str_new(L, buf, len);
430 /* Convert number to string. */
431 GCstr * LJ_FASTCALL lj_strfmt_num(lua_State *L, cTValue *o)
433 char buf[STRFMT_MAXBUF_NUM];
434 MSize len = (MSize)(lj_strfmt_wnum(buf, o) - buf);
435 return lj_str_new(L, buf, len);
438 /* Convert integer or number to string. */
439 GCstr * LJ_FASTCALL lj_strfmt_number(lua_State *L, cTValue *o)
441 return tvisint(o) ? lj_strfmt_int(L, intV(o)) : lj_strfmt_num(L, o);
444 #if LJ_HASJIT
445 /* Convert char value to string. */
446 GCstr * LJ_FASTCALL lj_strfmt_char(lua_State *L, int c)
448 char buf[1];
449 buf[0] = c;
450 return lj_str_new(L, buf, 1);
452 #endif
454 /* Raw conversion of object to string. */
455 GCstr * LJ_FASTCALL lj_strfmt_obj(lua_State *L, cTValue *o)
457 if (tvisstr(o)) {
458 return strV(o);
459 } else if (tvisnumber(o)) {
460 return lj_strfmt_number(L, o);
461 } else if (tvisnil(o)) {
462 return lj_str_newlit(L, "nil");
463 } else if (tvisfalse(o)) {
464 return lj_str_newlit(L, "false");
465 } else if (tvistrue(o)) {
466 return lj_str_newlit(L, "true");
467 } else {
468 char buf[8+2+2+16], *p = buf;
469 p = lj_buf_wmem(p, lj_typename(o), (MSize)strlen(lj_typename(o)));
470 *p++ = ':'; *p++ = ' ';
471 if (tvisfunc(o) && isffunc(funcV(o))) {
472 p = lj_buf_wmem(p, "builtin#", 8);
473 p = lj_strfmt_wint(p, funcV(o)->c.ffid);
474 } else {
475 p = lj_strfmt_wptr(p, lj_obj_ptr(o));
477 return lj_str_new(L, buf, (size_t)(p - buf));
481 /* -- Internal string formatting ------------------------------------------ */
484 ** These functions are only used for lua_pushfstring(), lua_pushvfstring()
485 ** and for internal string formatting (e.g. error messages). Caveat: unlike
486 ** string.format(), only a limited subset of formats and flags are supported!
488 ** LuaJIT has support for a couple more formats than Lua 5.1/5.2:
489 ** - %d %u %o %x with full formatting, 32 bit integers only.
490 ** - %f and other FP formats are really %.14g.
491 ** - %s %c %p without formatting.
494 /* Push formatted message as a string object to Lua stack. va_list variant. */
495 const char *lj_strfmt_pushvf(lua_State *L, const char *fmt, va_list argp)
497 SBuf *sb = lj_buf_tmp_(L);
498 FormatState fs;
499 SFormat sf;
500 GCstr *str;
501 lj_strfmt_init(&fs, fmt, (MSize)strlen(fmt));
502 while ((sf = lj_strfmt_parse(&fs)) != STRFMT_EOF) {
503 switch (STRFMT_TYPE(sf)) {
504 case STRFMT_LIT:
505 lj_buf_putmem(sb, fs.str, fs.len);
506 break;
507 case STRFMT_INT:
508 lj_strfmt_putfxint(sb, sf, va_arg(argp, int32_t));
509 break;
510 case STRFMT_UINT:
511 lj_strfmt_putfxint(sb, sf, va_arg(argp, uint32_t));
512 break;
513 case STRFMT_NUM: {
514 TValue tv;
515 tv.n = va_arg(argp, lua_Number);
516 setsbufP(sb, lj_strfmt_wnum(lj_buf_more(sb, STRFMT_MAXBUF_NUM), &tv));
517 break;
519 case STRFMT_STR: {
520 const char *s = va_arg(argp, char *);
521 if (s == NULL) s = "(null)";
522 lj_buf_putmem(sb, s, (MSize)strlen(s));
523 break;
525 case STRFMT_CHAR:
526 lj_buf_putb(sb, va_arg(argp, int));
527 break;
528 case STRFMT_PTR:
529 lj_strfmt_putptr(sb, va_arg(argp, void *));
530 break;
531 case STRFMT_ERR:
532 default:
533 lj_buf_putb(sb, '?');
534 lua_assert(0);
535 break;
538 str = lj_buf_str(L, sb);
539 setstrV(L, L->top, str);
540 incr_top(L);
541 return strdata(str);
544 /* Push formatted message as a string object to Lua stack. Vararg variant. */
545 const char *lj_strfmt_pushf(lua_State *L, const char *fmt, ...)
547 const char *msg;
548 va_list argp;
549 va_start(argp, fmt);
550 msg = lj_strfmt_pushvf(L, fmt, argp);
551 va_end(argp);
552 return msg;