Add zsh+euc, Zsh with EUC encoding support.

This is just an experimental hack and cannot happily be merged into
the upstream.  Zsh's line editor apparently needs a rewrite in order
to support multi-byte encodings because it strongly relies on the
single-byte character scheme.

These patches are mostly based on the work by ono@ono.org (Thanks!):

	http://www.ono.org/software/zsh-euc/

What I did over this is disable the hack for non-EUC locales.  Maybe
the patches can be moved to shells/zsh in the future, but it's
premature for the moment.

Notes:
- forward-char, backward-char and backward-delete-char with no numeric
  argument should work properly with this hack.
- Completion and redisplay should work fine.
- There can be some trivial side-effects.
- JIS X0201-Roman and JIS X0208-Kanji are supported.
- JIS X0201-Katakana and JIS X0212 Kanji are NOT supported.
- Only tested with the EUC-JP (ja_JP.eucJP) locale.  I'm not sure if
  it works for GB 2312/CNS 11643-1/KS X 1001.  Any feedbacks is
  welcome, especially a patch if it does not work. :)

10 files changed, 401 insertions, 0 deletions

diff --git a/shells/Makefile b/shells/Makefile
index e2d663fd4547..c147ebb71150 100644
--- a/shells/Makefile
+++ b/shells/Makefile
@@ -21,6 +21,7 @@
     SUBDIR += vshnu
     SUBDIR += wapsh
     SUBDIR += zsh
+    SUBDIR += zsh+euc_hack
     SUBDIR += zsh-devel
 
 .include <bsd.port.subdir.mk>
diff --git a/shells/zsh+euc_hack/Makefile b/shells/zsh+euc_hack/Makefile
new file mode 100644
index 000000000000..1be9d36dde4a
--- /dev/null
+++ b/shells/zsh+euc_hack/Makefile
@@ -0,0 +1,18 @@
+# New ports collection makefile for:	zsh with EUC encoding support
+# Date created:		12 May 2002
+# Whom:			Akinori MUSHA aka knu <knu@idaemons.org>
+#
+# $FreeBSD$
+#
+
+MASTERDIR=	${.CURDIR}/../zsh
+
+PKGNAMESUFFIX=	+euc_hack
+
+MAINTAINER=	knu@FreeBSD.org
+
+EXTRA_PATCHES=	${.CURDIR}/files/patch-*
+
+COMMENT=	${.CURDIR}/pkg-comment
+
+.include "${MASTERDIR}/Makefile"
diff --git a/shells/zsh+euc_hack/files/patch-Src::Zle::compmatch.c b/shells/zsh+euc_hack/files/patch-Src::Zle::compmatch.c
new file mode 100644
index 000000000000..3b5fb085e82f
--- /dev/null
+++ b/shells/zsh+euc_hack/files/patch-Src::Zle::compmatch.c
@@ -0,0 +1,57 @@
+--- Src/Zle/compmatch.c.orig	Tue Apr  3 20:25:13 2001
++++ Src/Zle/compmatch.c	Thu May  9 07:59:05 2002
+@@ -1309,6 +1309,9 @@
+     while (la && lb) {
+ 	if (*sa != *sb) {
+ 	    /* Different characters, try the matchers. */
++#ifdef ZSH_EUC
++ono:
++#endif
+ 	    for (t = 0, ms = bmatchers; ms && !t; ms = ms->next) {
+ 		mp = ms->matcher;
+ 		if (mp && !mp->flags && mp->wlen > 0 && mp->llen > 0 &&
+@@ -1354,6 +1357,13 @@
+ 	    }
+ 	    if (!t)
+ 		break;
++#ifdef ZSH_EUC
++	} else if (locale_is_euc &&
++	  (_mbmap_euc[*(unsigned char*)sa] & _MB1) &&
++	  (_mbmap_euc[*((unsigned char*)sa+1)] & _MB2) &&
++	  *(sa+1) != *(sb+1)) {
++	    goto ono;
++#endif
+ 	} else {
+ 	    /* Same character, just take it. */
+ 	    if (rr <= 1) {
+@@ -1580,6 +1590,30 @@
+ 	if (check_cmdata(md, sfx))
+ 	    return ret;
+ 
++#ifdef ZSH_EUC
++	if (locale_is_euc) {
++	    if (sfx)
++		for (l = 0, p = str, q = md->str;
++		     l < len && l < md->len && p[ind] == q[ind];
++		     l++, p += add, q += add);
++	    else
++		for (l = 0, p = str, q = md->str;;) {
++		    if (!(l < len && l < md->len))
++			break;
++
++		    if ((_mbmap_euc[*(unsigned char*)p] & _MB1) && (_mbmap_euc[*((unsigned char*)p+1)] & _MB2)) {
++			if (*p == *q && *(p+1) == *(q+1)) {
++			    l+=2, p+=2, q+=2;
++			} else
++			    break;
++		    } else {
++			if (*p != *q)
++			    break;
++			l++, p++, q++;
++		    }
++		}
++	} else
++#endif
+ 	for (l = 0, p = str, q = md->str;
+ 	     l < len && l < md->len && p[ind] == q[ind];
+ 	     l++, p += add, q += add);
diff --git a/shells/zsh+euc_hack/files/patch-Src::Zle::zle.c b/shells/zsh+euc_hack/files/patch-Src::Zle::zle.c
new file mode 100644
index 000000000000..f0e1e59e93ec
--- /dev/null
+++ b/shells/zsh+euc_hack/files/patch-Src::Zle::zle.c
@@ -0,0 +1,11 @@
+--- Src/Zle/zle.h.orig	Tue May 23 17:20:57 2000
++++ Src/Zle/zle.h	Thu May  9 06:49:30 2002
+@@ -194,3 +194,8 @@
+ /* Invalidate the completion list. */
+ 
+ #define invalidatelist() runhookdef(INVALIDATELISTHOOK, NULL)
++
++#ifdef ZSH_EUC
++#define _MB1 0x01
++#define _MB2 0x02
++#endif
diff --git a/shells/zsh+euc_hack/files/patch-Src::Zle::zle_main.c b/shells/zsh+euc_hack/files/patch-Src::Zle::zle_main.c
new file mode 100644
index 000000000000..2f4ca58d6109
--- /dev/null
+++ b/shells/zsh+euc_hack/files/patch-Src::Zle::zle_main.c
@@ -0,0 +1,199 @@
+--- Src/Zle/zle_main.c.orig	Fri May 18 00:56:13 2001
++++ Src/Zle/zle_main.c	Thu May  9 07:53:36 2002
+@@ -141,6 +141,163 @@
+ static int delayzsetterm;
+ #endif
+ 
++#ifdef ZSH_EUC
++/**/
++mod_export int locale_is_euc;
++
++#  if defined(__uxps__) || defined(sgi)  || defined(aix) || defined(__CYGWIN__) || defined(linux)
++char STRLANGEUCJP[]	= { 'j', 'a', '_', 'J', 'P', '.', 'E', 'U', 'C', '\0' };
++char STRLANGEUCKR[]	= { 'k', 'o', '_', 'K', 'R', '.', 'E', 'U', 'C', '\0' };
++#   if defined(__uxps__)
++char STRLANGEUCJPB[]	= { 'j', 'a', 'p', 'a', 'n', '\0' };
++char STRLANGEUCKRB[]	= { 'k', 'o', 'r', 'e', 'a', '\0' };
++#   elif defined(linux)
++char STRLANGEUCJPB[]	= { 'j', 'a', '_', 'J', 'P', '.', 'u', 'j', 'i', 's',
++			    '\0' };
++char STRLANGEUCKRB[]	= { 'k', 'o', '_', 'K', 'R', '.', 'e', 'u', 'c', '\0' };
++#   elif defined(aix)
++char STRLANGEUCJPB[]	= { 'j', 'a', '_', 'J', 'P', '\0' };
++char STRLANGEUCKRB[]	= { 'k', 'o', '_', 'K', 'R', '\0' };
++#   else
++char STRLANGEUCJPB[]	= { '\0' };
++char STRLANGEUCKRB[]	= { '\0' };
++#   endif
++char STRLANGSJIS[]	= { 'j', 'a', '_', 'J', 'P', '.', 'S', 'J', 'I', 'S',
++			    '\0' };
++char STRLANGSJISB[]	= { '\0' };
++char STRLANGBIG5[]	= { 'z', 'h', '_', 'T', 'W', '.', 'B', 'i', 'g', '5',
++			    '\0' };
++char STRLANGEUCZH[]	= { '\0' };
++char STRLANGEUCZHB[]	= { '\0' };
++#  elif defined(__FreeBSD__)
++char STRLANGEUCJP[]	= { 'j', 'a', '_', 'J', 'P', '.', 'e', 'u', 'c', 'J',
++			    'P', '\0' };
++char STRLANGEUCJPB[]	= { 'j', 'a', '_', 'J', 'P', '.', 'E', 'U', 'C', '\0' };
++char STRLANGEUCKR[]	= { 'k', 'o', '_', 'K', 'R', '.', 'e', 'u', 'c', 'K',
++			    'R', '\0' };
++char STRLANGEUCKRB[]	= { 'k', 'o', '_', 'K', 'R', '.', 'E', 'U', 'C', '\0' };
++char STRLANGEUCZH[]	= { 'z', 'h', '_', 'C', 'N', '.', 'e', 'u', 'c', 'C',
++			    'N', '\0' };
++char STRLANGEUCZHB[]	= { 'z', 'h', '_', 'C', 'N', '.', 'E', 'U', 'C', '\0' };
++char STRLANGSJIS[]	= { 'j', 'a', '_', 'J', 'P', '.', 'S', 'J', 'I', 'S',
++			    '\0' };
++char STRLANGSJISB[]	= { 'j', 'a', '_', 'J', 'P', '.', 'S', 'h', 'i', 'f',
++			    't', '_', 'J', 'I', 'S', '\0' };
++char STRLANGBIG5[]	= { 'z', 'h', '_', 'T', 'W', '.', 'B', 'i', 'g', '5',
++			    '\0' };
++#  elif defined(__uxpm__)
++char STRLANGEUCJP[]	= { 'j', 'a', 'p', 'a', 'n', '\0' };
++char STRLANGEUCKR[]	= { 'k', 'o', 'r', 'e', 'a', '\0' };
++char STRLANGEUCZH[]	= { '\0' };
++char STRLANGEUCJPB[]	= { '\0' };
++char STRLANGEUCKRB[]	= { '\0' };
++char STRLANGEUCZHB[]	= { '\0' };
++char STRLANGSJIS[]	= { '\0' };
++char STRLANGSJISB[]	= { '\0' };
++char STRLANGBIG5[]	= { '\0' };
++#  elif defined(SOLARIS2)
++char STRLANGEUCJP[]	= { 'j', 'a', '\0' };
++char STRLANGEUCKR[]	= { 'k', 'o', '\0' };
++char STRLANGEUCZH[]	= { '\0' };
++char STRLANGEUCJPB[]	= { 'j', 'a', 'p', 'a', 'n', 'e', 's', 'e', '\0' };
++char STRLANGEUCKRB[]	= { 'k', 'o', 'r', 'e', 'a', 'n', '\0' };
++char STRLANGEUCZHB[]	= { '\0' };
++char STRLANGSJIS[]	= { '\0' };
++char STRLANGSJISB[]	= { '\0' };
++char STRLANGBIG5[]	= { '\0' };
++#  elif defined(hpux)
++char STRLANGEUCJP[]	= { 'j', 'a', '_', 'J', 'P', '.', 'e', 'u', 'c', 'J', 'P' };
++char STRLANGEUCKR[]	= { 'k', 'o', '_', 'K', 'R', '.', 'e', 'u', 'c', 'K', 'R' };
++char STRLANGEUCZH[]	= { '\0' };
++char STRLANGEUCJPB[]	= { '\0' };
++char STRLANGEUCKRB[]	= { '\0' };
++char STRLANGEUCZHB[]	= { '\0' };
++char STRLANGSJIS[]	= { '\0' };
++char STRLANGSJISB[]	= { '\0' };
++char STRLANGBIG5[]	= { '\0' };
++#  else
++char STRLANGEUCJP[]	= { '\0' };
++char STRLANGEUCKR[]	= { '\0' };
++char STRLANGEUCZH[]	= { '\0' };
++char STRLANGEUCJPB[]	= { '\0' };
++char STRLANGEUCKRB[]	= { '\0' };
++char STRLANGEUCZHB[]	= { '\0' };
++char STRLANGSJIS[]	= { '\0' };
++char STRLANGSJISB[]	= { '\0' };
++char STRLANGBIG5[]	= { '\0' };
++#  endif
++
++/**/
++mod_export unsigned char _mbmap_euc[256] = {
++/* first byte	0x8e,0xa0 - 0xf4 */
++/* second byte 	0xa0 - 0xfe */
++/* 0 - 7f all 0 */
++    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
++    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
++    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
++    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
++/*  80		81		82		83 */
++    0,		0,		0,		0,
++/*  84		85		86		87 */
++    0,		0,		0,		0,
++/*  88		89		8a		8b */
++    0,		0,		0,		0,
++/*  8c		8d		8e		8f */
++    0,		0,		_MB1,		0,
++/* 90 - 9f all 0 */
++    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
++/*  a0		a1		a2		a3 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  a4		a5		a6		a7 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  a8		a9		aa		ab */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  ac		ad		ae		af */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  b0		b1		b2		b3 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  b4		b5		b6		b7 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  b8		b9		ba		bb */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  bc		bd		be		bf */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  c0		c1		c2		c3 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  c4		c5		c6		c7 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  c8		c9		ca		cb */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  cc		cd		ce		cf */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  d0		d1		d2		d3 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  d4		d5		d6		d7 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  d8		d9		da		db */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  dc		dd		de		df */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  e0		e1		e2		e3 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  e4		e5		e6		e7 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  e8		e9		ea		eb */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  ec		ed		ee		ef */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  f0		f1		f2		f3 */
++    _MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,	_MB1|_MB2,
++/*  f4		f5		f6		f7 */
++    _MB1|_MB2,	_MB2,		_MB2,		_MB2,
++/*  f8		f9		fa		fb */
++    _MB2,	_MB2,		_MB2,		_MB2,
++/*  fc		fd		fe		ff */
++    _MB2,	_MB2,		_MB2,		0,
++};
++
++#endif
++
+ /* set up terminal */
+ 
+ /**/
+@@ -1117,6 +1274,32 @@
+ int
+ setup_(Module m)
+ {
++#ifdef ZSH_EUC
++    char *euc_locales[] = {
++	STRLANGEUCJP,
++	STRLANGEUCKR,
++	STRLANGEUCZH,
++	STRLANGEUCJPB,
++	STRLANGEUCKRB,
++	STRLANGEUCZHB,
++	NULL
++    };
++    char **p;
++    char *tcp;
++
++    locale_is_euc = 0;
++
++    if ((tcp = setlocale("LC_CTYPE", NULL)) != NULL ||
++      (tcp = getenv("LANG")) != NULL) {
++	for (p = euc_locales; *p != NULL; p++) {
++	    if (strcmp(tcp, *p) == 0) {
++		locale_is_euc = 1;
++		break;
++	    }
++	}
++    }
++#endif
++
+     /* Set up editor entry points */
+     trashzleptr = trashzle;
+     refreshptr = zrefresh;
diff --git a/shells/zsh+euc_hack/files/patch-Src::Zle::zle_move.c b/shells/zsh+euc_hack/files/patch-Src::Zle::zle_move.c
new file mode 100644
index 000000000000..b4215fbe64cc
--- /dev/null
+++ b/shells/zsh+euc_hack/files/patch-Src::Zle::zle_move.c
@@ -0,0 +1,38 @@
+--- Src/Zle/zle_move.c.orig	Sat Jul  3 22:18:00 1999
++++ Src/Zle/zle_move.c	Thu May  9 17:55:46 2002
+@@ -159,6 +159,17 @@
+ int
+ forwardchar(char **args)
+ {
++#ifdef ZSH_EUC
++    if (locale_is_euc) {
++	if (zmult == 1) {
++	    if (_mbmap_euc[line[cs]] & _MB1 &&
++	      cs+1 <= ll &&
++	      _mbmap_euc[line[cs+1]] & _MB2) {
++		cs++;
++	    }
++	}
++    }
++#endif
+     cs += zmult;
+     if (cs > ll)
+ 	cs = ll;
+@@ -171,6 +182,17 @@
+ int
+ backwardchar(char **args)
+ {
++#ifdef ZSH_EUC
++    if (locale_is_euc) {
++	if (zmult == 1) {
++	    if (_mbmap_euc[line[cs-1]] & _MB2 &&
++	      cs-2 >=0 &&
++	      _mbmap_euc[line[cs-2]] & _MB1) {
++		cs--;
++	    }
++	}
++    }
++#endif
+     cs -= zmult;
+     if (cs > ll)
+ 	cs = ll;
diff --git a/shells/zsh+euc_hack/files/patch-Src::Zle::zle_refresh.c b/shells/zsh+euc_hack/files/patch-Src::Zle::zle_refresh.c
new file mode 100644
index 000000000000..47d7aa6fa867
--- /dev/null
+++ b/shells/zsh+euc_hack/files/patch-Src::Zle::zle_refresh.c
@@ -0,0 +1,47 @@
+--- Src/Zle/zle_refresh.c.orig	Mon Sep 10 19:48:51 2001
++++ Src/Zle/zle_refresh.c	Thu May  9 08:36:47 2002
+@@ -647,7 +647,11 @@
+ static void
+ refreshline(int ln)
+ {
++#ifdef ZSH_EUC
++    unsigned char *nl, *ol, *p1;	/* line buffer pointers		 */
++#else
+     char *nl, *ol, *p1;		/* line buffer pointers			 */
++#endif
+     int ccs = 0,		/* temporary count for cursor position	 */
+ 	char_ins = 0,		/* number of characters inserted/deleted */
+ 	col_cleareol,		/* clear to end-of-line from this column */
+@@ -753,7 +757,32 @@
+     for (;;) {
+ 	if (*nl && *ol && nl[1] == ol[1]) /* skip only if second chars match */
+ 	/* skip past all matching characters */
++#ifdef ZSH_EUC
++	{
++	    if (locale_is_euc) {
++		for (; *nl && (*nl == *ol); nl++, ol++, ccs++) {
++		    if (_mbmap_euc[*nl] & _MB1) {
++			if (*(ol+1) == '\0')
++			    continue;
++			if (_mbmap_euc[*(nl+1)] & _MB2) {
++			    if (*(nl+1) != *(ol+1))
++				break;
++			    else {
++				nl++;
++				ol++;
++				ccs++;
++			    }
++			} else if (*(nl+1) == '\0') {
++			    continue;
++			}
++		    }
++		}
++	    } else
++#endif
+ 	    for (; *nl && (*nl == *ol); nl++, ol++, ccs++) ;
++#ifdef ZSH_EUC
++	}
++#endif
+ 
+ 	if (!*nl) {
+ 	    if (ccs == winw && hasam && char_ins > 0 && ins_last
diff --git a/shells/zsh+euc_hack/files/patch-Src::Zle::zle_utils.c b/shells/zsh+euc_hack/files/patch-Src::Zle::zle_utils.c
new file mode 100644
index 000000000000..6c61530ec524
--- /dev/null
+++ b/shells/zsh+euc_hack/files/patch-Src::Zle::zle_utils.c
@@ -0,0 +1,20 @@
+--- Src/Zle/zle_utils.c.orig	Mon Mar 26 17:58:34 2001
++++ Src/Zle/zle_utils.c	Thu May  9 17:57:20 2002
+@@ -197,6 +197,17 @@
+ mod_export void
+ backdel(int ct)
+ {
++#ifdef ZSH_EUC
++    if (locale_is_euc) {
++	if (ct == 1) {
++	    if (_mbmap_euc[line[cs-1]] & _MB2 &&
++	      cs-2 >=0 &&
++	      _mbmap_euc[line[cs-2]] & _MB1) {
++		ct = 2;
++	    }
++	}
++    }
++#endif
+     shiftchars(cs -= ct, ct);
+ }
+ 
diff --git a/shells/zsh+euc_hack/files/patch-config.h.in b/shells/zsh+euc_hack/files/patch-config.h.in
new file mode 100644
index 000000000000..50ebd1b2fe8a
--- /dev/null
+++ b/shells/zsh+euc_hack/files/patch-config.h.in
@@ -0,0 +1,9 @@
+--- config.h.in.orig	Tue Oct 16 21:46:09 2001
++++ config.h.in	Thu May  9 06:49:30 2002
+@@ -756,3 +756,6 @@
+ 
+ /* Define if you have the socket library (-lsocket).  */
+ #undef HAVE_LIBSOCKET
++
++/* zsh euc completion */
++#define ZSH_EUC 1
diff --git a/shells/zsh+euc_hack/pkg-comment b/shells/zsh+euc_hack/pkg-comment
new file mode 100644
index 000000000000..b2897090b46b
--- /dev/null
+++ b/shells/zsh+euc_hack/pkg-comment
@@ -0,0 +1 @@
+The Z shell with EUC encoding support