Add some Authors: headers based on authors.el.

[gnu-emacs] / src / character.c

diff --git a/src/character.c b/src/character.c
index 37963d1878c3d180237758ee77dd8ba2f1a630c9..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)
@@ -1067,10 +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);