libgo/runtime/go-rune.c

   1 /* go-rune.c -- rune functions for Go.
   2
   3    Copyright 2009, 2010 The Go Authors. All rights reserved.
   4    Use of this source code is governed by a BSD-style
   5    license that can be found in the LICENSE file.  */
   6
   7 #include <stddef.h>
   8
   9 #include "runtime.h"
  10 #include "go-string.h"
  11
  12 /* Get a character from the UTF-8 string STR, of length LEN.  Store
  13    the Unicode character, if any, in *RUNE.  Return the number of
  14    characters used from STR.  */
  15
  16 int
  17 __go_get_rune (const unsigned char *str, size_t len, int32 *rune)
  18 {
  19   int c, c1, c2, c3, l;
  20
  21   /* Default to the "replacement character".  */
  22   *rune = 0xfffd;
  23
  24   if (len <= 0)
  25     return 1;
  26
  27   c = *str;
  28   if (c <= 0x7f)
  29     {
  30       *rune = c;
  31       return 1;
  32     }
  33
  34   if (len <= 1)
  35     return 1;
  36
  37   c1 = str[1];
  38   if ((c & 0xe0) == 0xc0
  39       && (c1 & 0xc0) == 0x80)
  40     {
  41       l = (((c & 0x1f) << 6) + (c1 & 0x3f));
  42       if (l <= 0x7f)
  43         return 1;
  44       *rune = l;
  45       return 2;
  46     }
  47
  48   if (len <= 2)
  49     return 1;
  50
  51   c2 = str[2];
  52   if ((c & 0xf0) == 0xe0
  53       && (c1 & 0xc0) == 0x80
  54       && (c2 & 0xc0) == 0x80)
  55     {
  56       l = (((c & 0xf) << 12)
  57            + ((c1 & 0x3f) << 6)
  58            + (c2 & 0x3f));
  59
  60       if (l <= 0x7ff)
  61         return 1;
  62
  63       if (l >= 0xd800 && l < 0xe000)
  64         {
  65           /* Invalid surrogate half; return replace character.  */
  66           return 1;
  67         }
  68
  69       *rune = l;
  70
  71       return 3;
  72     }
  73
  74   if (len <= 3)
  75     return 1;
  76
  77   c3 = str[3];
  78   if ((c & 0xf8) == 0xf0
  79       && (c1 & 0xc0) == 0x80
  80       && (c2 & 0xc0) == 0x80
  81       && (c3 & 0xc0) == 0x80)
  82     {
  83       l = (((c & 0x7) << 18)
  84            + ((c1 & 0x3f) << 12)
  85            + ((c2 & 0x3f) << 6)
  86            + (c3 & 0x3f));
  87
  88       if (l <= 0xffff || l > 0x10ffff)
  89         return 1;
  90
  91       *rune = l;
  92       return 4;
  93     }
  94
  95   /* Invalid encoding.  Return 1 so that we advance.  */
  96   return 1;
  97 }