Add some Authors: headers based on authors.el.

[gnu-emacs] / src / character.c

diff --git a/src/character.c b/src/character.c
index b2acf36ec158148002cb42d5ab4a41d5c74d1c34..12a95203527a803b82e7462d8a92bcb388d1c1a2 100644 (file)
--- a/src/character.c
+++ b/src/character.c
@@ -1,6 +1,6 @@
 /* Basic character support.
 
-Copyright (C) 2001-2012  Free Software Foundation, Inc.
+Copyright (C) 2001-2014 Free Software Foundation, Inc.
 Copyright (C) 1995, 1997, 1998, 2001 Electrotechnical Laboratory, JAPAN.
   Licensed to the Free Software Foundation.
 Copyright (C) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011
@@ -29,14 +29,11 @@ along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.  */
 #include <config.h>
 #endif
 
-#define CHARACTER_INLINE EXTERN_INLINE
-
 #include <stdio.h>
 
 #ifdef emacs
 
 #include <sys/types.h>
-#include <setjmp.h>
 #include <intprops.h>
 #include "lisp.h"
 #include "character.h"
@@ -127,8 +124,6 @@ char_string (unsigned int c, unsigned char *p)
       c &= ~CHAR_MODIFIER_MASK;
     }
 
-  MAYBE_UNIFY_CHAR (c);
-
   if (c <= MAX_3_BYTE_CHAR)
     {
       bytes = CHAR_STRING (c, p);
@@ -177,11 +172,14 @@ string_char (const unsigned char *p, const unsigned char **advanced, int *len)
 
   if (*p < 0x80 || ! (*p & 0x20) || ! (*p & 0x10))
     {
+      /* 1-, 2-, and 3-byte sequences can be handled by the macro.  */
       c = STRING_CHAR_ADVANCE (p);
     }
   else if (! (*p & 0x08))
     {
-      c = ((((p)[0] & 0xF) << 18)
+      /* A 4-byte sequence of this form:
+        11110xxx 10xxxxxx 10xxxxxx 10xxxxxx  */
+      c = ((((p)[0] & 0x7) << 18)
           | (((p)[1] & 0x3F) << 12)
           | (((p)[2] & 0x3F) << 6)
           | ((p)[3] & 0x3F));
@@ -189,15 +187,20 @@ string_char (const unsigned char *p, const unsigned char **advanced, int *len)
     }
   else
     {
-      c = ((((p)[1] & 0x3F) << 18)
+      /* A 5-byte sequence of this form:
+
+        111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
+
+        Note that the top 4 `x's are always 0, so shifting p[1] can
+        never exceed the maximum valid character codepoint. */
+      c = (/* (((p)[0] & 0x3) << 24) ... always 0, so no need to shift. */
+          (((p)[1] & 0x3F) << 18)
           | (((p)[2] & 0x3F) << 12)
           | (((p)[3] & 0x3F) << 6)
           | ((p)[4] & 0x3F));
       p += 5;
     }
 
-  MAYBE_UNIFY_CHAR (c);
-
   if (len)
     *len = p - saved_p;
   if (advanced)
@@ -427,7 +430,7 @@ lisp_string_width (Lisp_Object string, ptrdiff_t precision,
   /* This set multibyte to 0 even if STRING is multibyte when it
      contains only ascii and eight-bit-graphic, but that's
      intentional.  */
-  int multibyte = len < SBYTES (string);
+  bool multibyte = len < SBYTES (string);
   unsigned char *str = SDATA (string);
   ptrdiff_t i = 0, i_byte = 0;
   ptrdiff_t width = 0;
@@ -541,7 +544,7 @@ multibyte_chars_in_text (const unsigned char *ptr, ptrdiff_t nbytes)
       int len = MULTIBYTE_LENGTH (ptr, endp);
 
       if (len == 0)
-       abort ();
+       emacs_abort ();
       ptr += len;
       chars++;
     }
@@ -765,13 +768,10 @@ str_as_unibyte (unsigned char *str, ptrdiff_t bytes)
    corresponding byte and store in DST.  CHARS is the number of
    characters in SRC.  The value is the number of bytes stored in DST.
    Usually, the value is the same as CHARS, but is less than it if SRC
-   contains a non-ASCII, non-eight-bit character.  If ACCEPT_LATIN_1
-   is nonzero, a Latin-1 character is accepted and converted to a byte
-   of that character code.
-   Note: Currently the arg ACCEPT_LATIN_1 is not used.  */
+   contains a non-ASCII, non-eight-bit character.  */
 
 ptrdiff_t
-str_to_unibyte (const unsigned char *src, unsigned char *dst, ptrdiff_t chars, int accept_latin_1)
+str_to_unibyte (const unsigned char *src, unsigned char *dst, ptrdiff_t chars)
 {
   ptrdiff_t i;
 
@@ -781,8 +781,7 @@ str_to_unibyte (const unsigned char *src, unsigned char *dst, ptrdiff_t chars, i
 
       if (CHAR_BYTE8_P (c))
        c = CHAR_TO_BYTE8 (c);
-      else if (! ASCII_CHAR_P (c)
-              && (! accept_latin_1 || c >= 0x100))
+      else if (! ASCII_CHAR_P (c))
        return i;
       *dst++ = c;
     }
@@ -793,7 +792,7 @@ str_to_unibyte (const unsigned char *src, unsigned char *dst, ptrdiff_t chars, i
 static ptrdiff_t
 string_count_byte8 (Lisp_Object string)
 {
-  int multibyte = STRING_MULTIBYTE (string);
+  bool multibyte = STRING_MULTIBYTE (string);
   ptrdiff_t nbytes = SBYTES (string);
   unsigned char *p = SDATA (string);
   unsigned char *pend = p + nbytes;
@@ -825,7 +824,7 @@ string_escape_byte8 (Lisp_Object string)
 {
   ptrdiff_t nchars = SCHARS (string);
   ptrdiff_t nbytes = SBYTES (string);
-  int multibyte = STRING_MULTIBYTE (string);
+  bool multibyte = STRING_MULTIBYTE (string);
   ptrdiff_t byte8_count;
   const unsigned char *src, *src_end;
   unsigned char *dst;
@@ -1071,10 +1070,6 @@ A char-table for width (columns) of each character.  */);
               doc: /* Char table of script symbols.
 It has one extra slot whose value is a list of script symbols.  */);
 
-  /* Intern this now in case it isn't already done.
-     Setting this variable twice is harmless.
-     But don't staticpro it here--that is done in alloc.c.  */
-  Qchar_table_extra_slots = intern_c_string ("char-table-extra-slots");
   DEFSYM (Qchar_script_table, "char-script-table");
   Fput (Qchar_script_table, Qchar_table_extra_slots, make_number (1));
   Vchar_script_table = Fmake_char_table (Qchar_script_table, Qnil);