(print_string): Don't ignore garbage bytes following a

[gnu-emacs] / src / casefiddle.c

diff --git a/src/casefiddle.c b/src/casefiddle.c
index 2ca61ab51ec6b38840eaa74e1af4a2537543ecfb..b6d8b21f606674cf31ca72bc2d9f0d856ccbf7fc 100644 (file)
--- a/src/casefiddle.c
+++ b/src/casefiddle.c
@@ -1,5 +1,5 @@
 /* GNU Emacs case conversion functions.
-   Copyright (C) 1985, 1994 Free Software Foundation, Inc.
+   Copyright (C) 1985, 1994, 1997 Free Software Foundation, Inc.
 
 This file is part of GNU Emacs.
 
@@ -57,15 +57,19 @@ casify_object (flag, obj)
            }
          return obj;
        }
+
       if (STRINGP (obj))
        {
-         int multibyte = !NILP (current_buffer->enable_multibyte_characters);
+         int multibyte = STRING_MULTIBYTE (obj);
 
          obj = Fcopy_sequence (obj);
-         len = XSTRING (obj)->size;
-         for (i = 0; i < len; i++)
+         len = STRING_BYTES (XSTRING (obj));
+
+         /* Scan all single-byte characters from start of string.  */
+         for (i = 0; i < len;)
            {
              c = XSTRING (obj)->data[i];
+
              if (multibyte && c >= 0x80)
                /* A multibyte character can't be handled in this
                    simple loop.  */
@@ -75,15 +79,25 @@ casify_object (flag, obj)
              else if (!UPPERCASEP (c)
                       && (!inword || flag != CASE_CAPITALIZE_UP))
                c = UPCASE1 (c);
+             /* If this char won't fit in a single-byte string.
+                fall out to the multibyte case.  */
+             if (multibyte ? ! ASCII_BYTE_P (c)
+                 : ! SINGLE_BYTE_CHAR_P (c))
+               break;
+
              XSTRING (obj)->data[i] = c;
              if ((int) flag >= (int) CASE_CAPITALIZE)
                inword = SYNTAX (c) == Sword;
+             i++;
            }
+
+         /* If we didn't do the whole string as single-byte,
+            scan the rest in a more complex way.  */
          if (i < len)
            {
              /* The work is not yet finished because of a multibyte
                 character just encountered.  */
-             int fromlen, tolen, j = i;
+             int fromlen, tolen, j = i, j_byte = i;
              char *buf
                = (char *) alloca ((len - i) * MAX_LENGTH_OF_MULTI_BYTE_FORM
                                   + i);
@@ -92,6 +106,7 @@ casify_object (flag, obj)
              /* Copy data already handled.  */
              bcopy (XSTRING (obj)->data, buf, i);
 
+             /* From now on, I counts bytes.  */
              while (i < len)
                {
                  c = STRING_CHAR_AND_LENGTH (XSTRING (obj)->data + i,
@@ -102,13 +117,15 @@ casify_object (flag, obj)
                           && (!inword || flag != CASE_CAPITALIZE_UP))
                    c = UPCASE1 (c);
                  tolen = CHAR_STRING (c, workbuf, str);
-                 bcopy (str, buf + j, tolen);
+                 bcopy (str, buf + j_byte, tolen);
                  i += fromlen;
-                 j += tolen;
+                 j++;
+                 j_byte += tolen;
                  if ((int) flag >= (int) CASE_CAPITALIZE)
                    inword = SYNTAX (c) == Sword;
                }
-             obj = make_string (buf, j);
+             obj = make_specified_string (buf, j, j_byte,
+                                          STRING_MULTIBYTE (obj));
            }
          return obj;
        }
@@ -165,6 +182,7 @@ The argument object is not altered--the value is a copy.")
 /* flag is CASE_UP, CASE_DOWN or CASE_CAPITALIZE or CASE_CAPITALIZE_UP.
    b and e specify range of buffer to operate on. */
 
+void
 casify_region (flag, b, e)
      enum case_action flag;
      Lisp_Object b, e;
@@ -387,6 +405,7 @@ With negative argument, capitalize previous words but do not move.")
   return Qnil;
 }
 \f
+void
 syms_of_casefiddle ()
 {
   Qidentity = intern ("identity");
@@ -404,6 +423,7 @@ syms_of_casefiddle ()
   defsubr (&Scapitalize_word);
 }
 
+void
 keys_of_casefiddle ()
 {
   initial_define_key (control_x_map, Ctl('U'), "upcase-region");