michael@0: /*
michael@0:  * Copyright © 2011,2012  Google, Inc.
michael@0:  *
michael@0:  *  This is part of HarfBuzz, a text shaping library.
michael@0:  *
michael@0:  * Permission is hereby granted, without written agreement and without
michael@0:  * license or royalty fees, to use, copy, modify, and distribute this
michael@0:  * software and its documentation for any purpose, provided that the
michael@0:  * above copyright notice and the following two paragraphs appear in
michael@0:  * all copies of this software.
michael@0:  *
michael@0:  * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
michael@0:  * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
michael@0:  * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
michael@0:  * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
michael@0:  * DAMAGE.
michael@0:  *
michael@0:  * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
michael@0:  * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
michael@0:  * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
michael@0:  * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
michael@0:  * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
michael@0:  *
michael@0:  * Google Author(s): Behdad Esfahbod
michael@0:  */
michael@0: 
michael@0: #ifndef HB_UTF_PRIVATE_HH
michael@0: #define HB_UTF_PRIVATE_HH
michael@0: 
michael@0: #include "hb-private.hh"
michael@0: 
michael@0: 
michael@0: /* UTF-8 */
michael@0: 
michael@0: #define HB_UTF8_COMPUTE(Char, Mask, Len) \
michael@0:   if (Char < 128) { Len = 1; Mask = 0x7f; } \
michael@0:   else if ((Char & 0xe0) == 0xc0) { Len = 2; Mask = 0x1f; } \
michael@0:   else if ((Char & 0xf0) == 0xe0) { Len = 3; Mask = 0x0f; } \
michael@0:   else if ((Char & 0xf8) == 0xf0) { Len = 4; Mask = 0x07; } \
michael@0:   else Len = 0;
michael@0: 
michael@0: static inline const uint8_t *
michael@0: hb_utf_next (const uint8_t *text,
michael@0: 	     const uint8_t *end,
michael@0: 	     hb_codepoint_t *unicode)
michael@0: {
michael@0:   hb_codepoint_t c = *text, mask;
michael@0:   unsigned int len;
michael@0: 
michael@0:   /* TODO check for overlong sequences? */
michael@0: 
michael@0:   HB_UTF8_COMPUTE (c, mask, len);
michael@0:   if (unlikely (!len || (unsigned int) (end - text) < len)) {
michael@0:     *unicode = -1;
michael@0:     return text + 1;
michael@0:   } else {
michael@0:     hb_codepoint_t result;
michael@0:     unsigned int i;
michael@0:     result = c & mask;
michael@0:     for (i = 1; i < len; i++)
michael@0:       {
michael@0: 	if (unlikely ((text[i] & 0xc0) != 0x80))
michael@0: 	  {
michael@0: 	    *unicode = -1;
michael@0: 	    return text + 1;
michael@0: 	  }
michael@0: 	result <<= 6;
michael@0: 	result |= (text[i] & 0x3f);
michael@0:       }
michael@0:     *unicode = result;
michael@0:     return text + len;
michael@0:   }
michael@0: }
michael@0: 
michael@0: static inline const uint8_t *
michael@0: hb_utf_prev (const uint8_t *text,
michael@0: 	     const uint8_t *start,
michael@0: 	     hb_codepoint_t *unicode)
michael@0: {
michael@0:   const uint8_t *end = text--;
michael@0:   while (start < text && (*text & 0xc0) == 0x80 && end - text < 4)
michael@0:     text--;
michael@0: 
michael@0:   hb_codepoint_t c = *text, mask;
michael@0:   unsigned int len;
michael@0: 
michael@0:   /* TODO check for overlong sequences? */
michael@0: 
michael@0:   HB_UTF8_COMPUTE (c, mask, len);
michael@0:   if (unlikely (!len || (unsigned int) (end - text) != len)) {
michael@0:     *unicode = -1;
michael@0:     return end - 1;
michael@0:   } else {
michael@0:     hb_codepoint_t result;
michael@0:     unsigned int i;
michael@0:     result = c & mask;
michael@0:     for (i = 1; i < len; i++)
michael@0:       {
michael@0: 	result <<= 6;
michael@0: 	result |= (text[i] & 0x3f);
michael@0:       }
michael@0:     *unicode = result;
michael@0:     return text;
michael@0:   }
michael@0: }
michael@0: 
michael@0: 
michael@0: static inline unsigned int
michael@0: hb_utf_strlen (const uint8_t *text)
michael@0: {
michael@0:   return strlen ((const char *) text);
michael@0: }
michael@0: 
michael@0: 
michael@0: /* UTF-16 */
michael@0: 
michael@0: static inline const uint16_t *
michael@0: hb_utf_next (const uint16_t *text,
michael@0: 	     const uint16_t *end,
michael@0: 	     hb_codepoint_t *unicode)
michael@0: {
michael@0:   hb_codepoint_t c = *text++;
michael@0: 
michael@0:   if (unlikely (hb_in_range<hb_codepoint_t> (c, 0xd800, 0xdbff)))
michael@0:   {
michael@0:     /* high surrogate */
michael@0:     hb_codepoint_t l;
michael@0:     if (text < end && ((l = *text), likely (hb_in_range<hb_codepoint_t> (l, 0xdc00, 0xdfff))))
michael@0:     {
michael@0:       /* low surrogate */
michael@0:       *unicode = (c << 10) + l - ((0xd800 << 10) - 0x10000 + 0xdc00);
michael@0:        text++;
michael@0:     } else
michael@0:       *unicode = -1;
michael@0:   } else
michael@0:     *unicode = c;
michael@0: 
michael@0:   return text;
michael@0: }
michael@0: 
michael@0: static inline const uint16_t *
michael@0: hb_utf_prev (const uint16_t *text,
michael@0: 	     const uint16_t *start,
michael@0: 	     hb_codepoint_t *unicode)
michael@0: {
michael@0:   hb_codepoint_t c = *--text;
michael@0: 
michael@0:   if (unlikely (hb_in_range<hb_codepoint_t> (c, 0xdc00, 0xdfff)))
michael@0:   {
michael@0:     /* low surrogate */
michael@0:     hb_codepoint_t h;
michael@0:     if (start < text && ((h = *(text - 1)), likely (hb_in_range<hb_codepoint_t> (h, 0xd800, 0xdbff))))
michael@0:     {
michael@0:       /* high surrogate */
michael@0:       *unicode = (h << 10) + c - ((0xd800 << 10) - 0x10000 + 0xdc00);
michael@0:        text--;
michael@0:     } else
michael@0:       *unicode = -1;
michael@0:   } else
michael@0:     *unicode = c;
michael@0: 
michael@0:   return text;
michael@0: }
michael@0: 
michael@0: 
michael@0: static inline unsigned int
michael@0: hb_utf_strlen (const uint16_t *text)
michael@0: {
michael@0:   unsigned int l = 0;
michael@0:   while (*text++) l++;
michael@0:   return l;
michael@0: }
michael@0: 
michael@0: 
michael@0: /* UTF-32 */
michael@0: 
michael@0: static inline const uint32_t *
michael@0: hb_utf_next (const uint32_t *text,
michael@0: 	     const uint32_t *end HB_UNUSED,
michael@0: 	     hb_codepoint_t *unicode)
michael@0: {
michael@0:   *unicode = *text++;
michael@0:   return text;
michael@0: }
michael@0: 
michael@0: static inline const uint32_t *
michael@0: hb_utf_prev (const uint32_t *text,
michael@0: 	     const uint32_t *start HB_UNUSED,
michael@0: 	     hb_codepoint_t *unicode)
michael@0: {
michael@0:   *unicode = *--text;
michael@0:   return text;
michael@0: }
michael@0: 
michael@0: static inline unsigned int
michael@0: hb_utf_strlen (const uint32_t *text)
michael@0: {
michael@0:   unsigned int l = 0;
michael@0:   while (*text++) l++;
michael@0:   return l;
michael@0: }
michael@0: 
michael@0: 
michael@0: #endif /* HB_UTF_PRIVATE_HH */