pike.git/ src/ stralloc.c

Branch: Tag:

2017-10-08

2017-10-08 11:37:20 by Arne Goedeke <el@laramies.com>

utf8_to_string: implement 7bit fast path

Implement a fast path for the length calculation for 7 bit strings.

2831:    return ((x * 0xCD) >> 8) >> 2;   }    + ATTRIBUTE((noinline)) + static size_t pike_string_utf8_decode_length_slowpath(size_t len, const unsigned char *in, +  const unsigned char *end, INT32 args, +  int extended, int *_shift) { +  int shift = 0; +  +  for(; in < end; in++) { +  unsigned int c = *in; +  len++; +  if (LIKELY(!(c & 0x80))) continue; +  int cont = 0; +  +  /* From table 3-6 in the Unicode standard 4.0: Well-Formed UTF-8 +  * Byte Sequences +  * +  * Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte +  * 000000-00007f 00-7f +  * 000080-0007ff c2-df 80-bf +  * 000800-000fff e0 a0-bf 80-bf +  * 001000-00cfff e1-ec 80-bf 80-bf +  * 00d000-00d7ff ed 80-9f 80-bf +  * 00e000-00ffff ee-ef 80-bf 80-bf +  * 010000-03ffff f0 90-bf 80-bf 80-bf +  * 040000-0fffff f1-f3 80-bf 80-bf 80-bf +  * 100000-10ffff f4 80-8f 80-bf 80-bf +  */ +  +  if ((c & 0xc0) == 0x80) { +  bad_arg_error ("utf8_to_string", Pike_sp - args, args, 1, +  NULL, Pike_sp - args, +  "Invalid continuation character 0x%02x.\n", c); +  } +  + #define GET_CHAR(in, c) do { \ +  in++; \ +  if (in >= end) \ +  bad_arg_error ("utf8_to_string", Pike_sp - args, args, 1, \ +  NULL, Pike_sp - args, \ +  "Truncated UTF-8 sequence at end of string.\n"); \ +  c = *in; \ +  } while(0) + #define GET_CONT_CHAR(in, c) do { \ +  GET_CHAR(in, c); \ +  if ((c & 0xc0) != 0x80) \ +  bad_arg_error ("utf8_to_string", Pike_sp - args, args, 1, \ +  NULL, Pike_sp - args, \ +  "Expected continuation character, " \ +  "got 0x%02x.\n", \ +  c); \ +  } while (0) +  + #define UTF8_SEQ_ERROR(prefix, c, problem) do { \ +  bad_arg_error ("utf8_to_string", Pike_sp - args, args, 1, \ +  NULL, Pike_sp - args, \ +  "UTF-8 sequence beginning with %s0x%02x " \ +  " %s.\n", \ +  prefix, c, problem); \ +  } while (0) +  +  if ((c & 0xe0) == 0xc0) { +  /* 11bit */ +  if (!(c & 0x1e)) +  UTF8_SEQ_ERROR ("", c, "is a non-shortest form"); +  cont = 1; +  if (c & 0x1c) { +  if (shift < 1) { +  shift = 1; +  } +  } +  } +  +  else if ((c & 0xf0) == 0xe0) { +  /* 16bit */ +  if (c == 0xe0) { +  GET_CONT_CHAR (in, c); +  if (!(c & 0x20)) +  UTF8_SEQ_ERROR ("0xe0 ", c, "is a non-shortest form"); +  cont = 1; +  } +  else if (!(extended & 1) && c == 0xed) { +  GET_CONT_CHAR (in, c); +  if (c & 0x20) { +  /* Surrogate. */ +  if (!(extended & 2)) { +  UTF8_SEQ_ERROR ("0xed ", c, "would decode to " +  "a UTF-16 surrogate character"); +  } +  if (c & 0x10) { +  UTF8_SEQ_ERROR ("0xed ", c, "would decode to " +  "a UTF-16 low surrogate character"); +  } +  GET_CONT_CHAR(in, c); +  +  GET_CHAR (in, c); +  if (c != 0xed) { +  UTF8_SEQ_ERROR ("", c, "UTF-16 low surrogate " +  "character required"); +  } +  GET_CONT_CHAR (in, c); +  if ((c & 0xf0) != 0xb0) { +  UTF8_SEQ_ERROR ("0xed ", c, "UTF-16 low surrogate " +  "character required"); +  } +  shift = 2; +  } +  cont = 1; +  } +  else +  cont = 2; +  if (shift < 1) { +  shift = 1; +  } +  } +  +  else { +  if ((c & 0xf8) == 0xf0) { +  /* 21bit */ +  if (c == 0xf0) { +  GET_CONT_CHAR (in, c); +  if (!(c & 0x30)) +  UTF8_SEQ_ERROR ("0xf0 ", c, "is a non-shortest form"); +  cont = 2; +  } +  else if (!(extended & 1)) { +  if (c > 0xf4) +  UTF8_SEQ_ERROR ("", c, "would decode to " +  "a character outside the valid UTF-8 range"); +  else if (c == 0xf4) { +  GET_CONT_CHAR (in, c); +  if (c > 0x8f) +  UTF8_SEQ_ERROR ("0xf4 ", c, "would decode to " +  "a character outside the valid UTF-8 range"); +  cont = 2; +  } +  else +  cont = 3; +  } +  else +  cont = 3; +  } +  +  else if (c == 0xff) +  bad_arg_error ("utf8_to_string", Pike_sp - args, args, 1, +  NULL, Pike_sp - args, +  "Invalid character 0xff"); +  +  else if (!(extended & 1)) +  UTF8_SEQ_ERROR ("", c, "would decode to " +  "a character outside the valid UTF-8 range"); +  +  else { +  if ((c & 0xfc) == 0xf8) { +  /* 26bit */ +  if (c == 0xf8) { +  GET_CONT_CHAR (in, c); +  if (!(c & 0x38)) +  UTF8_SEQ_ERROR ("0xf8 ", c, "is a non-shortest form"); +  cont = 3; +  } +  else +  cont = 4; +  } else if ((c & 0xfe) == 0xfc) { +  /* 31bit */ +  if (c == 0xfc) { +  GET_CONT_CHAR (in, c); +  if (!(c & 0x3c)) +  UTF8_SEQ_ERROR ("0xfc ", c, "is a non-shortest form"); +  cont = 4; +  } +  else +  cont = 5; +  } else if (c == 0xfe) { +  /* 36bit */ +  GET_CONT_CHAR (in, c); +  if (!(c & 0x3e)) +  UTF8_SEQ_ERROR ("0xfe ", c, "is a non-shortest form"); +  else if (c & 0x3c) +  UTF8_SEQ_ERROR ("0xfe ", c, "would decode to " +  "a too large character value"); +  cont = 5; +  } +  } +  +  shift = 2; +  } +  +  while(cont--) +  GET_CONT_CHAR (in, c); +  + #undef GET_CHAR + #undef GET_CONT_CHAR + #undef UTF8_SEQ_ERROR +  } +  +  *_shift = shift; +  return len; + } +  + PMOD_EXPORT size_t pike_string_utf8_decode_length(const struct pike_string *s, INT32 args, int extended, +  int *_shift) { +  static volatile poptype foo = (poptype)0x8080808080808080ULL; +  const unsigned char *in; +  size_t len = 0; +  size_t elen; +  const poptype mask = foo; +  const poptype *in8 = (poptype*)STR0(s); +  const poptype *end8 = in8 + (s->len / sizeof(poptype)); +  +  if (in8 < end8) { +  const size_t tail = (size_t)(end8 - in8) % 4; +  elen = tail*sizeof(poptype); +  poptype a = 0, b = 0, c = 0, d = 0; +  +  in8 += tail; +  switch (tail) { +  do { +  case 0: +  in8 += 4; +  elen = sizeof(poptype)*4; +  a = in8[-4]; +  case 3: +  b = in8[-3]; +  case 2: +  c = in8[-2]; +  case 1: +  d = in8[-1]; +  +  a &= mask; +  b &= mask; +  c &= mask; +  d &= mask; +  +  if (UNLIKELY(a | b | c | d)) { +  /* we have to begin from the beginning of the last chunk */ +  in = (const unsigned char*)(in8) - elen; +  goto not_7bit; +  } +  len += elen; +  } while (in8 < end8); +  break; +  default: UNREACHABLE(break); +  } +  } +  +  /* process the single byte tail */ +  +  elen = (size_t)s->len % sizeof(poptype); +  +  if (elen) { +  poptype a = 0; +  +  in = (const unsigned char*)in8; +  +  switch (7-elen) { +  case 0: a |= in[0] & 0x80; +  case 1: a |= in[1] & 0x80; +  case 2: a |= in[2] & 0x80; +  case 3: a |= in[3] & 0x80; +  case 4: a |= in[4] & 0x80; +  case 5: a |= in[5] & 0x80; +  case 6: a |= in[6] & 0x80; break; +  default: UNREACHABLE(break); +  } +  +  if (UNLIKELY(a)) { +  goto not_7bit; +  } +  len += elen; +  } +  +  *_shift = 0; +  return len; + not_7bit: +  +  return pike_string_utf8_decode_length_slowpath(len, in, (const unsigned char*)STR0(s) + s->len, args, +  extended, _shift); + } +    PMOD_EXPORT size_t pike_string_utf8_length(const struct pike_string *s, INT32 args, int extended) {    size_t len = s->len;    size_t elen = s->len;