PATCH: alternative wcwidth() implementation

[Peter Stephenson <pws@xxxxxxx>]

This attempts to test for a broken wcwidth() and uses Markus Kuhn's if
necessary.  It's a bit desperate, because we need to assume wchar_t is a
Unicode code point, but at this stage we are desperate.

Jun T. will have to decide if this is helping on MacOS, otherwise it's
back to the drawing board.

I note we should supposedly be defining _XOPEN_SOURCE to get wcwidth(),
at least with glibc, but apparently generally don't (though I did for the
configure test).  I'm a bit frightened of changing this, but POSIX does
suggest it's part of the XSI extension, so probably we ought to.
Indeed, we do define _XOPEN_SOURCE_EXTENDED if needed for curses.h;
probably the right thing to do is in the #else case, #define
_XOPEN_SOURCE if MULTIBYTE_SUPPORT is present.

Index: configure.ac
===================================================================
RCS file: /cvsroot/zsh/zsh/configure.ac,v
retrieving revision 1.96
diff -u -r1.96 configure.ac
--- configure.ac	28 Mar 2008 11:56:46 -0000	1.96
+++ configure.ac	22 Apr 2008 10:36:01 -0000
@@ -2266,8 +2266,63 @@
 ])
 AH_TEMPLATE([MULTIBYTE_SUPPORT],
 [Define to 1 if you want support for multibyte character sets.])
+AH_TEMPLATE([BROKEN_WCWIDTH],
+[Define to 1 if the wcwidth() function is present but broken.])
 if test x$zsh_cv_c_unicode_support = xyes; then
   AC_DEFINE(MULTIBYTE_SUPPORT)
+
+  dnl Test for a wcwidth() implementation that gives the wrong width for
+  dnl zero-width combining characters.
+  dnl For the test we use a combining acute accent (\u0301).
+  dnl We input it as UTF-8 since that is the standard we can rely
+  dnl upon most:  we can't rely on a wchar_t being stored as a
+  dnl Unicode code point on all systems.
+  dnl The programme returns 0 only if all the conditions for brokenness
+  dnl are met:
+  dnl - the programme compiled, linked and ran
+  dnl - we successfully set a UTF-8 locale
+  dnl - the locale we set plausibly converted the UTF-8 string
+  dnl   for a zero-width combining character (the only way to be
+  dnl   100% sure would be to output it and ask if it looked right)
+  dnl - the converted wide character gave a non-zero width.
+  dnl locale -a is a fallback; on most systems we should find en_US.UTF-8.
+  [locale_prog='char *my_locales[] = {
+  "en_US.UTF-8", "en_GB.UTF-8", "en.UTF-8", '
+  locale_prog="$locale_prog"`locale -a 2>/dev/null | \
+    sed -e 's/utf8/UTF-8/' | grep UTF-8 | \
+    while read line; do echo " \"$line\","; done;`
+  locale_prog="$locale_prog 0 };
+  #define _XOPEN_SOURCE
+  #include <stdlib.h>
+  #include <locale.h>
+  #include <wchar.h>
+
+  int main() {
+    char **localep;
+    char comb_acute_mb[] = { (char)0xcc, (char)0x81 };
+    wchar_t wc;
+
+    for (localep = my_locales; *localep; localep++)
+      if (setlocale(LC_ALL, *localep) &&
+          mbtowc(&wc, comb_acute_mb, 2) == 2)
+	  break;
+    if (!*localep)
+      return 1;
+    if (wcwidth(wc) == 0)
+      return 1;
+    return 0;
+  }
+  "]
+
+  AC_CACHE_CHECK(if the wcwidth() function is broken,
+  zsh_cv_c_broken_wcwidth,
+  [AC_TRY_RUN([$locale_prog],
+  zsh_cv_c_broken_wcwidth=yes,
+  zsh_cv_c_broken_wcwidth=no,
+  zsh_cv_c_broken_wcwidth=no)])
+  if test x$zsh_cv_c_broken_wcwdith = xyes; then
+    AC_DEFINE(BROKEN_WCWIDTH)
+  fi
 fi
 
 dnl
Index: Src/builtin.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/builtin.c,v
retrieving revision 1.190
diff -u -r1.190 builtin.c
--- Src/builtin.c	3 Apr 2008 15:20:22 -0000	1.190
+++ Src/builtin.c	22 Apr 2008 10:36:06 -0000
@@ -3668,7 +3668,7 @@
 			width += l;
 			break;
 		    }
-		    wcw = wcwidth(wc);
+		    wcw = WCWIDTH(wc);
 		    /* treat unprintable as 0 */
 		    if (wcw > 0)
 			width += wcw;
Index: Src/compat.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/compat.c,v
retrieving revision 1.16
diff -u -r1.16 compat.c
--- Src/compat.c	3 Mar 2008 12:29:56 -0000	1.16
+++ Src/compat.c	22 Apr 2008 10:36:07 -0000
@@ -549,4 +549,327 @@
 		*endptr = any ? s - 1 : nptr;
 	return (acc);
 }
-#endif
+#endif /* HAVE_STRTOUL */
+
+/**/
+#ifdef BROKEN_WCWIDTH
+
+/*
+ * This is an implementation of wcwidth() and wcswidth() (defined in
+ * IEEE Std 1002.1-2001) for Unicode.
+ *
+ * http://www.opengroup.org/onlinepubs/007904975/functions/wcwidth.html
+ * http://www.opengroup.org/onlinepubs/007904975/functions/wcswidth.html
+ *
+ * In fixed-width output devices, Latin characters all occupy a single
+ * "cell" position of equal width, whereas ideographic CJK characters
+ * occupy two such cells. Interoperability between terminal-line
+ * applications and (teletype-style) character terminals using the
+ * UTF-8 encoding requires agreement on which character should advance
+ * the cursor by how many cell positions. No established formal
+ * standards exist at present on which Unicode character shall occupy
+ * how many cell positions on character terminals. These routines are
+ * a first attempt of defining such behavior based on simple rules
+ * applied to data provided by the Unicode Consortium.
+ *
+ * For some graphical characters, the Unicode standard explicitly
+ * defines a character-cell width via the definition of the East Asian
+ * FullWidth (F), Wide (W), Half-width (H), and Narrow (Na) classes.
+ * In all these cases, there is no ambiguity about which width a
+ * terminal shall use. For characters in the East Asian Ambiguous (A)
+ * class, the width choice depends purely on a preference of backward
+ * compatibility with either historic CJK or Western practice.
+ * Choosing single-width for these characters is easy to justify as
+ * the appropriate long-term solution, as the CJK practice of
+ * displaying these characters as double-width comes from historic
+ * implementation simplicity (8-bit encoded characters were displayed
+ * single-width and 16-bit ones double-width, even for Greek,
+ * Cyrillic, etc.) and not any typographic considerations.
+ *
+ * Much less clear is the choice of width for the Not East Asian
+ * (Neutral) class. Existing practice does not dictate a width for any
+ * of these characters. It would nevertheless make sense
+ * typographically to allocate two character cells to characters such
+ * as for instance EM SPACE or VOLUME INTEGRAL, which cannot be
+ * represented adequately with a single-width glyph. The following
+ * routines at present merely assign a single-cell width to all
+ * neutral characters, in the interest of simplicity. This is not
+ * entirely satisfactory and should be reconsidered before
+ * establishing a formal standard in this area. At the moment, the
+ * decision which Not East Asian (Neutral) characters should be
+ * represented by double-width glyphs cannot yet be answered by
+ * applying a simple rule from the Unicode database content. Setting
+ * up a proper standard for the behavior of UTF-8 character terminals
+ * will require a careful analysis not only of each Unicode character,
+ * but also of each presentation form, something the author of these
+ * routines has avoided to do so far.
+ *
+ * http://www.unicode.org/unicode/reports/tr11/
+ *
+ * Markus Kuhn -- 2007-05-26 (Unicode 5.0)
+ *
+ * Permission to use, copy, modify, and distribute this software
+ * for any purpose and without fee is hereby granted. The author
+ * disclaims all warranties with regard to this software.
+ *
+ * Latest version: http://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c
+ */
+
+struct interval {
+  int first;
+  int last;
+};
+
+/* auxiliary function for binary search in interval table */
+static int bisearch(wchar_t ucs, const struct interval *table, int max) {
+  int min = 0;
+  int mid;
+
+  if (ucs < table[0].first || ucs > table[max].last)
+    return 0;
+  while (max >= min) {
+    mid = (min + max) / 2;
+    if (ucs > table[mid].last)
+      min = mid + 1;
+    else if (ucs < table[mid].first)
+      max = mid - 1;
+    else
+      return 1;
+  }
+
+  return 0;
+}
+
+
+/* The following two functions define the column width of an ISO 10646
+ * character as follows:
+ *
+ *    - The null character (U+0000) has a column width of 0.
+ *
+ *    - Other C0/C1 control characters and DEL will lead to a return
+ *      value of -1.
+ *
+ *    - Non-spacing and enclosing combining characters (general
+ *      category code Mn or Me in the Unicode database) have a
+ *      column width of 0.
+ *
+ *    - SOFT HYPHEN (U+00AD) has a column width of 1.
+ *
+ *    - Other format characters (general category code Cf in the Unicode
+ *      database) and ZERO WIDTH SPACE (U+200B) have a column width of 0.
+ *
+ *    - Hangul Jamo medial vowels and final consonants (U+1160-U+11FF)
+ *      have a column width of 0.
+ *
+ *    - Spacing characters in the East Asian Wide (W) or East Asian
+ *      Full-width (F) category as defined in Unicode Technical
+ *      Report #11 have a column width of 2.
+ *
+ *    - All remaining characters (including all printable
+ *      ISO 8859-1 and WGL4 characters, Unicode control characters,
+ *      etc.) have a column width of 1.
+ *
+ * This implementation assumes that wchar_t characters are encoded
+ * in ISO 10646.
+ */
+
+/**/
+int
+mk_wcwidth(wchar_t ucs)
+{
+  /* sorted list of non-overlapping intervals of non-spacing characters */
+  /* generated by "uniset +cat=Me +cat=Mn +cat=Cf -00AD +1160-11FF +200B c" */
+  static const struct interval combining[] = {
+    { 0x0300, 0x036F }, { 0x0483, 0x0486 }, { 0x0488, 0x0489 },
+    { 0x0591, 0x05BD }, { 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 },
+    { 0x05C4, 0x05C5 }, { 0x05C7, 0x05C7 }, { 0x0600, 0x0603 },
+    { 0x0610, 0x0615 }, { 0x064B, 0x065E }, { 0x0670, 0x0670 },
+    { 0x06D6, 0x06E4 }, { 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED },
+    { 0x070F, 0x070F }, { 0x0711, 0x0711 }, { 0x0730, 0x074A },
+    { 0x07A6, 0x07B0 }, { 0x07EB, 0x07F3 }, { 0x0901, 0x0902 },
+    { 0x093C, 0x093C }, { 0x0941, 0x0948 }, { 0x094D, 0x094D },
+    { 0x0951, 0x0954 }, { 0x0962, 0x0963 }, { 0x0981, 0x0981 },
+    { 0x09BC, 0x09BC }, { 0x09C1, 0x09C4 }, { 0x09CD, 0x09CD },
+    { 0x09E2, 0x09E3 }, { 0x0A01, 0x0A02 }, { 0x0A3C, 0x0A3C },
+    { 0x0A41, 0x0A42 }, { 0x0A47, 0x0A48 }, { 0x0A4B, 0x0A4D },
+    { 0x0A70, 0x0A71 }, { 0x0A81, 0x0A82 }, { 0x0ABC, 0x0ABC },
+    { 0x0AC1, 0x0AC5 }, { 0x0AC7, 0x0AC8 }, { 0x0ACD, 0x0ACD },
+    { 0x0AE2, 0x0AE3 }, { 0x0B01, 0x0B01 }, { 0x0B3C, 0x0B3C },
+    { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B43 }, { 0x0B4D, 0x0B4D },
+    { 0x0B56, 0x0B56 }, { 0x0B82, 0x0B82 }, { 0x0BC0, 0x0BC0 },
+    { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 }, { 0x0C46, 0x0C48 },
+    { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 }, { 0x0CBC, 0x0CBC },
+    { 0x0CBF, 0x0CBF }, { 0x0CC6, 0x0CC6 }, { 0x0CCC, 0x0CCD },
+    { 0x0CE2, 0x0CE3 }, { 0x0D41, 0x0D43 }, { 0x0D4D, 0x0D4D },
+    { 0x0DCA, 0x0DCA }, { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 },
+    { 0x0E31, 0x0E31 }, { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E },
+    { 0x0EB1, 0x0EB1 }, { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC },
+    { 0x0EC8, 0x0ECD }, { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 },
+    { 0x0F37, 0x0F37 }, { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E },
+    { 0x0F80, 0x0F84 }, { 0x0F86, 0x0F87 }, { 0x0F90, 0x0F97 },
+    { 0x0F99, 0x0FBC }, { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 },
+    { 0x1032, 0x1032 }, { 0x1036, 0x1037 }, { 0x1039, 0x1039 },
+    { 0x1058, 0x1059 }, { 0x1160, 0x11FF }, { 0x135F, 0x135F },
+    { 0x1712, 0x1714 }, { 0x1732, 0x1734 }, { 0x1752, 0x1753 },
+    { 0x1772, 0x1773 }, { 0x17B4, 0x17B5 }, { 0x17B7, 0x17BD },
+    { 0x17C6, 0x17C6 }, { 0x17C9, 0x17D3 }, { 0x17DD, 0x17DD },
+    { 0x180B, 0x180D }, { 0x18A9, 0x18A9 }, { 0x1920, 0x1922 },
+    { 0x1927, 0x1928 }, { 0x1932, 0x1932 }, { 0x1939, 0x193B },
+    { 0x1A17, 0x1A18 }, { 0x1B00, 0x1B03 }, { 0x1B34, 0x1B34 },
+    { 0x1B36, 0x1B3A }, { 0x1B3C, 0x1B3C }, { 0x1B42, 0x1B42 },
+    { 0x1B6B, 0x1B73 }, { 0x1DC0, 0x1DCA }, { 0x1DFE, 0x1DFF },
+    { 0x200B, 0x200F }, { 0x202A, 0x202E }, { 0x2060, 0x2063 },
+    { 0x206A, 0x206F }, { 0x20D0, 0x20EF }, { 0x302A, 0x302F },
+    { 0x3099, 0x309A }, { 0xA806, 0xA806 }, { 0xA80B, 0xA80B },
+    { 0xA825, 0xA826 }, { 0xFB1E, 0xFB1E }, { 0xFE00, 0xFE0F },
+    { 0xFE20, 0xFE23 }, { 0xFEFF, 0xFEFF }, { 0xFFF9, 0xFFFB },
+    { 0x10A01, 0x10A03 }, { 0x10A05, 0x10A06 }, { 0x10A0C, 0x10A0F },
+    { 0x10A38, 0x10A3A }, { 0x10A3F, 0x10A3F }, { 0x1D167, 0x1D169 },
+    { 0x1D173, 0x1D182 }, { 0x1D185, 0x1D18B }, { 0x1D1AA, 0x1D1AD },
+    { 0x1D242, 0x1D244 }, { 0xE0001, 0xE0001 }, { 0xE0020, 0xE007F },
+    { 0xE0100, 0xE01EF }
+  };
+
+  /* test for 8-bit control characters */
+  if (ucs == 0)
+    return 0;
+  if (ucs < 32 || (ucs >= 0x7f && ucs < 0xa0))
+    return -1;
+
+  /* binary search in table of non-spacing characters */
+  if (bisearch(ucs, combining,
+	       sizeof(combining) / sizeof(struct interval) - 1))
+    return 0;
+
+  /* if we arrive here, ucs is not a combining or C0/C1 control character */
+
+  return 1 + 
+    (ucs >= 0x1100 &&
+     (ucs <= 0x115f ||                    /* Hangul Jamo init. consonants */
+      ucs == 0x2329 || ucs == 0x232a ||
+      (ucs >= 0x2e80 && ucs <= 0xa4cf &&
+       ucs != 0x303f) ||                  /* CJK ... Yi */
+      (ucs >= 0xac00 && ucs <= 0xd7a3) || /* Hangul Syllables */
+      (ucs >= 0xf900 && ucs <= 0xfaff) || /* CJK Compatibility Ideographs */
+      (ucs >= 0xfe10 && ucs <= 0xfe19) || /* Vertical forms */
+      (ucs >= 0xfe30 && ucs <= 0xfe6f) || /* CJK Compatibility Forms */
+      (ucs >= 0xff00 && ucs <= 0xff60) || /* Fullwidth Forms */
+      (ucs >= 0xffe0 && ucs <= 0xffe6) ||
+      (ucs >= 0x20000 && ucs <= 0x2fffd) ||
+      (ucs >= 0x30000 && ucs <= 0x3fffd)));
+}
+
+
+/*
+ * The following functions are part of the original wcwidth.c:
+ * we don't use them but I've kept them in case - pws.
+ */
+#if 0
+int mk_wcswidth(const wchar_t *pwcs, size_t n)
+{
+  int w, width = 0;
+
+  for (;*pwcs && n-- > 0; pwcs++)
+    if ((w = mk_wcwidth(*pwcs)) < 0)
+      return -1;
+    else
+      width += w;
+
+  return width;
+}
+
+
+/*
+ * The following functions are the same as mk_wcwidth() and
+ * mk_wcswidth(), except that spacing characters in the East Asian
+ * Ambiguous (A) category as defined in Unicode Technical Report #11
+ * have a column width of 2. This variant might be useful for users of
+ * CJK legacy encodings who want to migrate to UCS without changing
+ * the traditional terminal character-width behaviour. It is not
+ * otherwise recommended for general use.
+ */
+int mk_wcwidth_cjk(wchar_t ucs)
+{
+  /* sorted list of non-overlapping intervals of East Asian Ambiguous
+   * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */
+  static const struct interval ambiguous[] = {
+    { 0x00A1, 0x00A1 }, { 0x00A4, 0x00A4 }, { 0x00A7, 0x00A8 },
+    { 0x00AA, 0x00AA }, { 0x00AE, 0x00AE }, { 0x00B0, 0x00B4 },
+    { 0x00B6, 0x00BA }, { 0x00BC, 0x00BF }, { 0x00C6, 0x00C6 },
+    { 0x00D0, 0x00D0 }, { 0x00D7, 0x00D8 }, { 0x00DE, 0x00E1 },
+    { 0x00E6, 0x00E6 }, { 0x00E8, 0x00EA }, { 0x00EC, 0x00ED },
+    { 0x00F0, 0x00F0 }, { 0x00F2, 0x00F3 }, { 0x00F7, 0x00FA },
+    { 0x00FC, 0x00FC }, { 0x00FE, 0x00FE }, { 0x0101, 0x0101 },
+    { 0x0111, 0x0111 }, { 0x0113, 0x0113 }, { 0x011B, 0x011B },
+    { 0x0126, 0x0127 }, { 0x012B, 0x012B }, { 0x0131, 0x0133 },
+    { 0x0138, 0x0138 }, { 0x013F, 0x0142 }, { 0x0144, 0x0144 },
+    { 0x0148, 0x014B }, { 0x014D, 0x014D }, { 0x0152, 0x0153 },
+    { 0x0166, 0x0167 }, { 0x016B, 0x016B }, { 0x01CE, 0x01CE },
+    { 0x01D0, 0x01D0 }, { 0x01D2, 0x01D2 }, { 0x01D4, 0x01D4 },
+    { 0x01D6, 0x01D6 }, { 0x01D8, 0x01D8 }, { 0x01DA, 0x01DA },
+    { 0x01DC, 0x01DC }, { 0x0251, 0x0251 }, { 0x0261, 0x0261 },
+    { 0x02C4, 0x02C4 }, { 0x02C7, 0x02C7 }, { 0x02C9, 0x02CB },
+    { 0x02CD, 0x02CD }, { 0x02D0, 0x02D0 }, { 0x02D8, 0x02DB },
+    { 0x02DD, 0x02DD }, { 0x02DF, 0x02DF }, { 0x0391, 0x03A1 },
+    { 0x03A3, 0x03A9 }, { 0x03B1, 0x03C1 }, { 0x03C3, 0x03C9 },
+    { 0x0401, 0x0401 }, { 0x0410, 0x044F }, { 0x0451, 0x0451 },
+    { 0x2010, 0x2010 }, { 0x2013, 0x2016 }, { 0x2018, 0x2019 },
+    { 0x201C, 0x201D }, { 0x2020, 0x2022 }, { 0x2024, 0x2027 },
+    { 0x2030, 0x2030 }, { 0x2032, 0x2033 }, { 0x2035, 0x2035 },
+    { 0x203B, 0x203B }, { 0x203E, 0x203E }, { 0x2074, 0x2074 },
+    { 0x207F, 0x207F }, { 0x2081, 0x2084 }, { 0x20AC, 0x20AC },
+    { 0x2103, 0x2103 }, { 0x2105, 0x2105 }, { 0x2109, 0x2109 },
+    { 0x2113, 0x2113 }, { 0x2116, 0x2116 }, { 0x2121, 0x2122 },
+    { 0x2126, 0x2126 }, { 0x212B, 0x212B }, { 0x2153, 0x2154 },
+    { 0x215B, 0x215E }, { 0x2160, 0x216B }, { 0x2170, 0x2179 },
+    { 0x2190, 0x2199 }, { 0x21B8, 0x21B9 }, { 0x21D2, 0x21D2 },
+    { 0x21D4, 0x21D4 }, { 0x21E7, 0x21E7 }, { 0x2200, 0x2200 },
+    { 0x2202, 0x2203 }, { 0x2207, 0x2208 }, { 0x220B, 0x220B },
+    { 0x220F, 0x220F }, { 0x2211, 0x2211 }, { 0x2215, 0x2215 },
+    { 0x221A, 0x221A }, { 0x221D, 0x2220 }, { 0x2223, 0x2223 },
+    { 0x2225, 0x2225 }, { 0x2227, 0x222C }, { 0x222E, 0x222E },
+    { 0x2234, 0x2237 }, { 0x223C, 0x223D }, { 0x2248, 0x2248 },
+    { 0x224C, 0x224C }, { 0x2252, 0x2252 }, { 0x2260, 0x2261 },
+    { 0x2264, 0x2267 }, { 0x226A, 0x226B }, { 0x226E, 0x226F },
+    { 0x2282, 0x2283 }, { 0x2286, 0x2287 }, { 0x2295, 0x2295 },
+    { 0x2299, 0x2299 }, { 0x22A5, 0x22A5 }, { 0x22BF, 0x22BF },
+    { 0x2312, 0x2312 }, { 0x2460, 0x24E9 }, { 0x24EB, 0x254B },
+    { 0x2550, 0x2573 }, { 0x2580, 0x258F }, { 0x2592, 0x2595 },
+    { 0x25A0, 0x25A1 }, { 0x25A3, 0x25A9 }, { 0x25B2, 0x25B3 },
+    { 0x25B6, 0x25B7 }, { 0x25BC, 0x25BD }, { 0x25C0, 0x25C1 },
+    { 0x25C6, 0x25C8 }, { 0x25CB, 0x25CB }, { 0x25CE, 0x25D1 },
+    { 0x25E2, 0x25E5 }, { 0x25EF, 0x25EF }, { 0x2605, 0x2606 },
+    { 0x2609, 0x2609 }, { 0x260E, 0x260F }, { 0x2614, 0x2615 },
+    { 0x261C, 0x261C }, { 0x261E, 0x261E }, { 0x2640, 0x2640 },
+    { 0x2642, 0x2642 }, { 0x2660, 0x2661 }, { 0x2663, 0x2665 },
+    { 0x2667, 0x266A }, { 0x266C, 0x266D }, { 0x266F, 0x266F },
+    { 0x273D, 0x273D }, { 0x2776, 0x277F }, { 0xE000, 0xF8FF },
+    { 0xFFFD, 0xFFFD }, { 0xF0000, 0xFFFFD }, { 0x100000, 0x10FFFD }
+  };
+
+  /* binary search in table of non-spacing characters */
+  if (bisearch(ucs, ambiguous,
+	       sizeof(ambiguous) / sizeof(struct interval) - 1))
+    return 2;
+
+  return mk_wcwidth(ucs);
+}
+
+
+int mk_wcswidth_cjk(const wchar_t *pwcs, size_t n)
+{
+  int w, width = 0;
+
+  for (;*pwcs && n-- > 0; pwcs++)
+    if ((w = mk_wcwidth_cjk(*pwcs)) < 0)
+      return -1;
+    else
+      width += w;
+
+  return width;
+}
+#endif /* 0 */
+
+/**/
+#endif /* BROKEN_WCWIDTH */
+
Index: Src/prompt.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/prompt.c,v
retrieving revision 1.46
diff -u -r1.46 prompt.c
--- Src/prompt.c	17 Feb 2008 18:15:10 -0000	1.46
+++ Src/prompt.c	22 Apr 2008 10:36:07 -0000
@@ -960,10 +960,10 @@
 		break;
 	    default:
 		/*
-		 * If the character isn't printable, wcwidth() returns
+		 * If the character isn't printable, WCWIDTH() returns
 		 * -1.  We assume width 1.
 		 */
-		wcw = wcwidth(wc);
+		wcw = WCWIDTH(wc);
 		if (wcw >= 0)
 		    w += wcw;
 		else
@@ -1177,7 +1177,7 @@
 				remw--;
 				break;
 			    default:
-				wcw = wcwidth(cc);
+				wcw = WCWIDTH(cc);
 				if (wcw >= 0)
 				    remw -= wcw;
 				else
@@ -1249,7 +1249,7 @@
 				maxwidth--;
 				break;
 			    default:
-				wcw = wcwidth(cc);
+				wcw = WCWIDTH(cc);
 				if (wcw >= 0)
 				    maxwidth -= wcw;
 				else
Index: Src/utils.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/utils.c,v
retrieving revision 1.187
diff -u -r1.187 utils.c
--- Src/utils.c	21 Apr 2008 11:49:57 -0000	1.187
+++ Src/utils.c	22 Apr 2008 10:36:09 -0000
@@ -548,7 +548,7 @@
     }
 
     if (widthp) {
-	int wcw = wcwidth(c);
+	int wcw = WCWIDTH(c);
 	*widthp = (s - buf);
 	if (wcw >= 0)
 	    *widthp += wcw;
@@ -581,7 +581,7 @@
     /* assume a single-byte character if not valid */
     if (wc == WEOF || unset(MULTIBYTE))
 	return 1;
-    wcw = wcwidth(wc);
+    wcw = WCWIDTH(wc);
     /* if not printable, assume width 1 */
     if (wcw < 0)
 	return 1;
@@ -4097,7 +4097,7 @@
 		 * Returns -1 if not a printable character.  We
 		 * turn this into 1 for backward compatibility.
 		 */
-		int wcw = wcwidth(wc);
+		int wcw = WCWIDTH(wc);
 		if (wcw >= 0)
 		    num += wcw;
 		else
Index: Src/zsh.h
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/zsh.h,v
retrieving revision 1.124
diff -u -r1.124 zsh.h
--- Src/zsh.h	21 Apr 2008 11:49:58 -0000	1.124
+++ Src/zsh.h	22 Apr 2008 10:36:11 -0000
@@ -2256,12 +2256,21 @@
 #define MB_METASTRWIDTH(str)	mb_metastrlen(str, 1)
 #define MB_METASTRLEN2(str, widthp)	mb_metastrlen(str, widthp)
 
+#ifdef BROKEN_WCWIDTH
+#define WCWIDTH(wc)	wcwidth_mk(wc)
+#else
+#define WCWIDTH(wc)	wcwidth(wc)
+#endif
 /*
- * Note WCWIDTH() takes wint_t, typically as a convchar_t.
+ * Note WCWIDTH_WINT() takes wint_t, typically as a convchar_t.
  * It's written to use the wint_t from mb_metacharlenconv() without
  * further tests.
+ *
+ * This version has a non-multibyte definition that simply returns
+ * 1.  We never expose WCWIDTH() in the non-multibyte world since
+ * it's just a proxy for wcwidth() itself.
  */
-#define WCWIDTH(wc)	zwcwidth(wc)
+#define WCWIDTH_WINT(wc)	zwcwidth(wc)
 
 #define MB_INCOMPLETE	((size_t)-2)
 #define MB_INVALID	((size_t)-1)
@@ -2286,9 +2295,6 @@
  *
  * wc is assumed to be a wchar_t (i.e. we don't need zwcwidth).
  *
- * This may need to be more careful if we import a wcwidth() for
- * compatibility to try to avoid clashes with the system library.
- *
  * Pedantic note: in Unicode, a combining character need not be
  * zero length.  However, we are concerned here about display;
  * we simply need to know whether the character will be displayed
@@ -2296,7 +2302,15 @@
  * sense throughout the shell.  I am not aware of a way of
  * detecting the Unicode trait in standard libraries.
  */
-#define IS_COMBINING(wc)	(wcwidth(wc) == 0)
+#ifdef BROKEN_WCWIDTH
+/*
+ * We can't be quite sure the wcwidth we've provided is entirely
+ * in agreement with the system's, so be extra safe.
+ */
+#define IS_COMBINING(wc)	(WCWIDTH(wc) == 0 && !iswcntrl(wc))
+#else
+#define IS_COMBINING(wc)	(WCWIDTH(wc) == 0)
+#endif
 /*
  * Test for the base of a combining character.
  *
@@ -2305,7 +2319,7 @@
  * is, as long as it has non-zero width.  We need to avoid all forms of
  * space because the shell will split words on any whitespace.
  */
-#define IS_BASECHAR(wc)		(iswgraph(wc) && wcwidth(wc) > 0)
+#define IS_BASECHAR(wc)		(iswgraph(wc) && WCWIDTH(wc) > 0)
 
 #else /* not MULTIBYTE_SUPPORT */
 
@@ -2317,7 +2331,7 @@
 #define MB_METASTRWIDTH(str)	ztrlen(str)
 #define MB_METASTRLEN2(str, widthp)	ztrlen(str)
 
-#define WCWIDTH(c)	(1)
+#define WCWIDTH_WINT(c)	(1)
 
 /* Leave character or string as is. */
 #define ZWC(c)	c
Index: Src/Zle/complist.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/Zle/complist.c,v
retrieving revision 1.112
diff -u -r1.112 complist.c
--- Src/Zle/complist.c	20 Apr 2008 21:17:29 -0000	1.112
+++ Src/Zle/complist.c	22 Apr 2008 10:36:13 -0000
@@ -1017,7 +1017,7 @@
 	}
 	else
 #endif
-	    width = WCWIDTH(cchar);
+	    width = WCWIDTH_WINT(cchar);
 
 	if (doesc && cchar == ZWC('%')) {
 	    p += len;
Index: Src/Zle/zle_misc.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/Zle/zle_misc.c,v
retrieving revision 1.51
diff -u -r1.51 zle_misc.c
--- Src/Zle/zle_misc.c	21 Apr 2008 17:58:59 -0000	1.51
+++ Src/Zle/zle_misc.c	22 Apr 2008 10:36:17 -0000
@@ -50,7 +50,8 @@
     if (insmode)
 	spaceinline(m * len);
     else {
-	int pos = zlecs, diff, i;
+	int pos = zlecs, diff, i, width;
+	int lastchar_base = 0;
 
 	/*
 	 * Calculate the number of character positions we are
@@ -59,10 +60,30 @@
 	 * (i.e. even if control, or double width, or with combining
 	 * characters) is treated as 1 for the purpose of replacing
 	 * what's there already.
+	 *
+	 * However, to do that we need to know whether combining
+	 * characters are going to be treated as such.  Here
+	 * lastchar_base = 0 is something that isn't a combining
+	 * character and won't be treated by zsh as the base for one
+	 * lastchar_base = 1 is a suitable base character.
+	 * We don't change lastchar_type for combining characters
+	 * since they only depend on the base character.
+	 *
+	 * The difference won't usually matter, but it means,
+	 * for example, that a combining character inserted at
+	 * the start of the line will overwrite rather than
+	 * insert, which would be illogical; similarly at the
+	 * start of a command line argument.
 	 */
 	for (i = 0, count = 0; i < len; i++) {
-	    int width = wcwidth(zstr[i]);
-	    count += (width != 0) ? 1 : 0;
+	    if (IS_COMBINING(zstr[i])) {
+		/* width is zero unless we didn't have a base character */
+		if (!lastchar_base)
+		    count += 1;
+	    } else {
+		count++;
+		lastchar_base = IS_BASECHAR(zstr[i]);
+	    }
 	}
 	/*
 	 * Ensure we replace a complete combining character
Index: Src/Zle/zle_refresh.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/Zle/zle_refresh.c,v
retrieving revision 1.60
diff -u -r1.60 zle_refresh.c
--- Src/Zle/zle_refresh.c	21 Apr 2008 11:49:58 -0000	1.60
+++ Src/Zle/zle_refresh.c	22 Apr 2008 10:36:17 -0000
@@ -1222,7 +1222,7 @@
 	    }
 	}
 #ifdef MULTIBYTE_SUPPORT
-	else if (iswprint(*t) && (width = wcwidth(*t)) > 0) {
+	else if (iswprint(*t) && (width = WCWIDTH(*t)) > 0) {
 	    int ichars;
 	    if (width > rpms.sen - rpms.s) {
 		int started = 0;
@@ -1397,7 +1397,7 @@
 	for (; u < outputline + outll; u++) {
 #ifdef MULTIBYTE_SUPPORT
 	    if (iswprint(*u)) {
-		int width = wcwidth(*u);
+		int width = WCWIDTH(*u);
 		/* Handle wide characters as above */
 		if (width > rpms.sen - rpms.s) {
 		    do {
@@ -2144,7 +2144,7 @@
    characters occupying more than one column.  We could flag that
    this has happened (since it's not that common to have characters
    wider than one column), but for now it's easier not to use the
-   trick if we are using wcwidth() on the prompt.  It's not that
+   trick if we are using WCWIDTH() on the prompt.  It's not that
    common to be editing in the middle of the prompt anyway, I would
    think.
    */
@@ -2264,7 +2264,7 @@
 	if (tmpline[t0] == ZWC('\t'))
 	    vsiz = (vsiz | 7) + 2;
 #ifdef MULTIBYTE_SUPPORT
-	else if (iswprint(tmpline[t0]) && (width = wcwidth(tmpline[t0]) > 0)) {
+	else if (iswprint(tmpline[t0]) && (width = WCWIDTH(tmpline[t0]) > 0)) {
 	    vsiz += width;
 	    if (isset(COMBININGCHARS) && IS_BASECHAR(tmpline[t0])) {
 		while (t0 < tmpll-1 && IS_COMBINING(tmpline[t0+1]))
@@ -2340,7 +2340,7 @@
 	    vp++;
 #ifdef MULTIBYTE_SUPPORT
 	} else if (iswprint(tmpline[t0]) &&
-		   (width = wcwidth(tmpline[t0])) > 0) {
+		   (width = WCWIDTH(tmpline[t0])) > 0) {
 	    int ichars;
 	    if (isset(COMBININGCHARS) && IS_BASECHAR(tmpline[t0])) {
 		/*
Index: Src/Zle/zle_tricky.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/Zle/zle_tricky.c,v
retrieving revision 1.93
diff -u -r1.93 zle_tricky.c
--- Src/Zle/zle_tricky.c	14 Apr 2008 14:57:54 -0000	1.93
+++ Src/Zle/zle_tricky.c	22 Apr 2008 10:36:17 -0000
@@ -2375,7 +2375,7 @@
 		    }
 		} else
 		    p += clen;
-		cc += WCWIDTH(cchar);
+		cc += WCWIDTH_WINT(cchar);
 		if (dopr && !(cc % columns))
 			fputs(" \010", shout);
 	    }


-- 
Peter Stephenson <pws@xxxxxxx>                  Software Engineer
CSR PLC, Churchill House, Cambridge Business Park, Cowley Road
Cambridge, CB4 0WZ, UK                          Tel: +44 (0)1223 692070