pike.git/ src/ builtin_functions.c

2005-04-02

2005-04-02 12:24:43 by Martin Stjernholm <mast@lysator.liu.se>

• 57dd67d149e8ecb0476b0a78db421351dc27f44c (133 lines) (+90/-43) ^{[ Show | Annotate ]}
  Branch: 7.9

Optimized utf8_to_string a bit now when non-shortest forms aren't
permitted.

Rev: src/builtin_functions.c:1.595

2:   || This file is part of Pike. For copyright information see COPYRIGHT.   || Pike is distributed under GPL, LGPL and MPL. See the file COPYING   || for more information. - || $Id: builtin_functions.c,v 1.594 2005/04/02 11:43:03 mast Exp $ + || $Id: builtin_functions.c,v 1.595 2005/04/02 12:24:43 mast Exp $   */      #include "global.h"
1930:    get_all_args("utf8_to_string", args, "%S.%i", &in, &extended);       for(i=0; i < in->len; i++) { -  unsigned int c = ((unsigned char *)in->str)[i]; +  unsigned int c = STR0(in)[i];    len++;    if (c & 0x80) {    int cont = 0;
1964:    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 = ((unsigned char *)(in->str))[i]; \ +  c = STR0 (in)[i]; \    if ((c & 0xc0) != 0x80) \    bad_arg_error ("utf8_to_string", Pike_sp - args, args, 1, \    NULL, Pike_sp - args, \
2016:    }       else { -  if (shift < 2) -  shift = 2; -  +     if ((c & 0xf8) == 0xf0) {    /* 21bit */    if (c == 0xf0) {
2087:    cont = 5;    }    } +  +  if (shift < 2) +  shift = 2;    }       while(cont--)    GET_CONT_CHAR (in, i, c); -  +  + #undef GET_CONT_CHAR + #undef UTF8_SEQ_ERROR    }    }    if (len == in->len) {
2101:       out = begin_wide_shared_string(len, shift);    -  for(j=i=0; i < in->len; i++) { -  unsigned int c = ((unsigned char *)in->str)[i]; +  switch (shift) { +  case 0: { +  p_wchar0 *out_str = (p_wchar0 *) out->str; +  for(j=i=0; i < in->len;) { +  unsigned int c = STR0(in)[i++]; +  /* NOTE: No tests here since we've already tested the string above. */ +  if (c & 0x80) { +  /* 11bit */ +  unsigned int c2 = STR0(in)[i++] & 0x3f; +  c &= 0x1f; +  c = (c << 6) | c2; +  } +  out_str[j++] = c; +  } +  break; +  }    -  +  case 1: { +  p_wchar1 *out_str = (p_wchar1 *) out->str; +  for(j=i=0; i < in->len;) { +  unsigned int c = STR0(in)[i++]; +  /* NOTE: No tests here since we've already tested the string above. */    if (c & 0x80) { -  int cont = 0; +  if ((c & 0xe0) == 0xc0) { +  /* 11bit */ +  unsigned int c2 = STR0(in)[i++] & 0x3f; +  c &= 0x1f; +  c = (c << 6) | c2; +  } else { +  /* 16bit */ +  unsigned int c2 = STR0(in)[i++] & 0x3f; +  unsigned int c3 = STR0(in)[i++] & 0x3f; +  c &= 0x0f; +  c = (c << 12) | (c2 << 6) | c3; +  } +  } +  out_str[j++] = c; +  } +  break; +  }    -  /* NOTE: No tests here since we've already tested the string -  * above. -  */ +  case 2: { +  p_wchar2 *out_str = (p_wchar2 *) out->str; +  for(j=i=0; i < in->len;) { +  unsigned int c = STR0(in)[i++]; +  /* NOTE: No tests here since we've already tested the string above. */ +  if (c & 0x80) { +  int cont = 0;    if ((c & 0xe0) == 0xc0) {    /* 11bit */    cont = 1;
2136:    c = 0;    }    while(cont--) { -  unsigned INT32 c2 = ((unsigned char *)(in->str))[++i] & 0x3f; +  unsigned int c2 = STR0(in)[i++] & 0x3f;    c = (c << 6) | c2;    }    } -  low_set_index(out, j++, c); +  out_str[j++] = c;    } -  +  break; +  } +  } +    #ifdef PIKE_DEBUG    if (j != len) {    Pike_fatal("utf8_to_string(): Calculated and actual lengths differ: %d != %d\n",    len, j);    }   #endif /* PIKE_DEBUG */ -  out = end_shared_string(out); +  out = low_end_shared_string(out);    pop_n_elems(args);    push_string(out);   }