libiconv: port over simplified iconv from musl

27 files changed, 811 insertions, 835 deletions

diff --git a/libiconv/iconv.c b/libiconv/iconv.c
index c028f0a51..ec01f381d 100644
--- a/libiconv/iconv.c
+++ b/libiconv/iconv.c
@@ -1,3 +1,32 @@
+/* 
+ * Copyright © 2018 Waldemar Brodkorb <wbx@uclibc-ng.org>
+ * Simplified port of iconv.c from musl C library including 
+ * parts of libiconv-tiny.
+ */
+
+/* Copyright © 2005-2018 Rich Felker, et al.
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+
+*/
+
 #include <iconv.h>
 #include <errno.h>
 #include <wchar.h>
@@ -5,7 +34,6 @@
 #include <strings.h>
 #include <stdlib.h>
 #include <limits.h>
-
 #include <dirent.h>
 #include <fcntl.h>
 #include <sys/mman.h>
@@ -13,124 +41,177 @@
 #include <unistd.h>
 #include <stdint.h>
 
-/* builtin charmaps */
-#include "charmaps.h"
-
-/* only 0-7 are valid as dest charset */
-#define UTF_16BE    000
-#define UTF_16LE    001
-#define UTF_32BE    002
-#define UTF_32LE    003
-#define WCHAR_T     004
-#define UTF_8       005
-#define US_ASCII    006
-#define LATIN_1     007
-
-/* additional charsets with algorithmic conversion */
-#define LATIN_9     010
-#define TIS_620     011
-#define JIS_0201    012
-#define EUC_JP      013
-
-/* some programs like php need this */
-int _libiconv_version = _LIBICONV_VERSION;
-
-/* these must match the constants above */
-static const unsigned char charsets[] =
-	"\005" "UTF-8"      "\0"
-	"\004" "WCHAR_T"    "\0"
-	"\000" "UTF-16BE"   "\0"
-	"\001" "UTF-16LE"   "\0"
-	"\002" "UTF-32BE"   "\0"
-	"\003" "UTF-32LE"   "\0"
-	"\006" "ASCII"      "\0"
-	"\006" "US-ASCII"   "\0"
-	"\007" "ISO-8859-1" "\0"
-	"\007" "LATIN1"     "\0"
-	"\010" "ISO-8859-15""\0"
-	"\010" "LATIN9"     "\0"
-	"\011" "ISO-8859-11""\0"
-	"\011" "TIS-620"    "\0"
-	"\012" "JIS-0201"   "\0"
-	"\013" "EUC-JP"     "\0"
-	"\377";
-
-/* separate identifiers for sbcs/dbcs/etc map type */
-#define UCS2_8BIT   000
-#define UCS3_8BIT   001
-#define EUC_JP      002
-#define EUC_TW      003
-#define SHIFT_JIS   004
-#define BIG5        005
-#define GBK         006
-
-/* FIXME: these are not implemented yet
-// EUC:   A1-FE A1-FE
-// GBK:   81-FE 40-7E,80-FE
-// Big5:  A1-FE 40-7E,A1-FE
-*/
+#define UTF_32BE    0300
+#define UTF_16LE    0301
+#define UTF_16BE    0302
+#define UTF_32LE    0303
+#define UCS2BE      0304
+#define UCS2LE      0305
+#define WCHAR_T     0306
+#define US_ASCII    0307
+#define UTF_8       0310
+#define UTF_16      0312
+#define UTF_32      0313
+#define UCS2        0314
+#define EUC_JP      0320
+#define SHIFT_JIS   0321
+#define ISO2022_JP  0322
+#define GB18030     0330
+#define GBK         0331
+#define GB2312      0332
+#define BIG5        0340
+#define EUC_KR      0350
+
+/* Definitions of charmaps. Each charmap consists of:
+ * 1. Empty-string-terminated list of null-terminated aliases.
+ * 2. Special type code or number of elided quads of entries.
+ * 3. Character table (size determined by field 2), consisting
+ *    of 5 bytes for every 4 characters, interpreted as 10-bit
+ *    indices into the legacy_chars table. */
+
+static const unsigned char charmaps[] =
+"utf8\0char\0\0\310"
+"wchart\0\0\306"
+"ucs2be\0\0\304"
+"ucs2le\0\0\305"
+"utf16be\0\0\302"
+"utf16le\0\0\301"
+"ucs4be\0utf32be\0\0\300"
+"ucs4le\0utf32le\0\0\303"
+"ascii\0usascii\0iso646\0iso646us\0\0\307"
+"utf16\0\0\312"
+"ucs4\0utf32\0\0\313"
+"ucs2\0\0\314"
+"eucjp\0\0\320"
+"shiftjis\0sjis\0\0\321"
+"iso2022jp\0\0\322"
+"gb18030\0\0\330"
+"gbk\0\0\331"
+"gb2312\0\0\332"
+"big5\0bigfive\0cp950\0big5hkscs\0\0\340"
+"euckr\0ksc5601\0ksx1001\0cp949\0\0\350"
+#include "codepages.h"
+;
+
+/* Table of characters that appear in legacy 8-bit codepages,
+ * limited to 1024 slots (10 bit indices). The first 256 entries
+ * are elided since those characters are obviously all included. */
+static const unsigned short legacy_chars[] = {
+#include "legacychars.h"
+};
+
+static const unsigned short jis0208[84][94] = {
+#include "jis0208.h"
+};
 
-static const unsigned short maplen[] = {
-	[UCS2_8BIT] = 4+ 2* 128,
-	[UCS3_8BIT] = 4+ 3* 128,
-	[EUC_JP]    = 4+ 2* 94*94,
-	[SHIFT_JIS] = 4+ 2* 94*94,
-	[BIG5]      = 4+ 2* 94*157,
-	[GBK]       = 4+ 2* 126*190,
-	[EUC_TW]    = 4+ 2* 2*94*94,
+static const unsigned short rev_jis[] = {
+#include "revjis.h"
 };
 
-static int find_charmap(const char *name)
+static int fuzzycmp(const unsigned char *a, const unsigned char *b)
 {
-	int i;
-	for (i = 0; i < (sizeof(charmaps) / sizeof(charmaps[0])); i++)
-		if (!strcasecmp(charmaps[i].name, name))
-			return i;
-	return -1;
+	for (; *a && *b; a++, b++) {
+		while (*a && (*a|32U)-'a'>26 && *a-'0'>10U) a++;
+		if ((*a|32U) != *b) return 1;
+	}
+	return *a != *b;
 }
 
-static int find_charset(const char *name)
+static size_t find_charmap(const void *name)
 {
 	const unsigned char *s;
-	for (s=charsets; *s<0xff && strcasecmp(s+1, name); s+=strlen(s)+1);
-	return *s;
+	if (!*(char *)name) name=charmaps; /* "utf8" */
+	for (s=charmaps; *s; ) {
+		if (!fuzzycmp(name, s)) {
+			for (; *s; s+=strlen((void *)s)+1);
+			return s+1-charmaps;
+		}
+		s += strlen((void *)s)+1;
+		if (!*s) {
+			if (s[1] > 0200) s+=2;
+			else s+=2+(64U-s[1])*5;
+		}
+	}
+	return -1;
 }
 
-iconv_t iconv_open(const char *to, const char *from)
+struct stateful_cd {
+	iconv_t base_cd;
+	unsigned state;
+};
+
+static iconv_t combine_to_from(size_t t, size_t f)
 {
-	unsigned f, t;
-	int m;
+	return (void *)(f<<16 | t<<1 | 1);
+}
 
-	if ((t = find_charset(to)) > 8)
-		return -1;
+static size_t extract_from(iconv_t cd)
+{
+	return (size_t)cd >> 16;
+}
 
-	if ((f = find_charset(from)) < 255)
-		return 0 | (t<<1) | (f<<8);
+static size_t extract_to(iconv_t cd)
+{
+	return (size_t)cd >> 1 & 0x7fff;
+}
 
-	if ((m = find_charmap(from)) > -1)
-		return 1 | (t<<1) | (m<<8);
+iconv_t iconv_open(const char *to, const char *from)
+{
+	size_t f, t;
+	struct stateful_cd *scd;
+
+	if ((t = find_charmap(to))==-1
+	 || (f = find_charmap(from))==-1
+	 || (charmaps[t] >= 0330)) {
+		errno = EINVAL;
+		return (iconv_t)-1;
+	}
+	iconv_t cd = combine_to_from(t, f);
+
+	switch (charmaps[f]) {
+	case UTF_16:
+	case UTF_32:
+	case UCS2:
+	case ISO2022_JP:
+		scd = malloc(sizeof *scd);
+		if (!scd) return (iconv_t)-1;
+		scd->base_cd = cd;
+		scd->state = 0;
+		cd = (iconv_t)scd;
+	}
 
-	return -1;
+	return cd;
 }
 
-int iconv_close(iconv_t cd)
+static unsigned get_16(const unsigned char *s, int e)
 {
-	return 0;
+	e &= 1;
+	return s[e]<<8 | s[1-e];
 }
 
-static inline wchar_t get_16(const unsigned char *s, int endian)
+static void put_16(unsigned char *s, unsigned c, int e)
 {
-	endian &= 1;
-	return s[endian]<<8 | s[endian^1];
+	e &= 1;
+	s[e] = c>>8;
+	s[1-e] = c;
 }
 
-static inline void put_16(unsigned char *s, wchar_t c, int endian)
+static unsigned get_32(const unsigned char *s, int e)
 {
-	endian &= 1;
-	s[endian] = c>>8;
-	s[endian^1] = c;
+	e &= 3;
+	return s[e]+0U<<24 | s[e^1]<<16 | s[e^2]<<8 | s[e^3];
 }
 
+static void put_32(unsigned char *s, unsigned c, int e)
+{
+	e &= 3;
+	s[e^0] = c>>24;
+	s[e^1] = c>>16;
+	s[e^2] = c>>8;
+	s[e^3] = c;
+}
+
+
 static inline int utf8enc_wchar(char *outb, wchar_t c)
 {
 	if (c <= 0x7F) {
@@ -246,133 +327,211 @@ static inline int utf8dec_wchar(wchar_t *c, unsigned char *in, size_t inb)
 	return -1;
 }
 
-static inline char latin9_translit(wchar_t c)
+static unsigned legacy_map(const unsigned char *map, unsigned c)
+{
+	if (c < 4*map[-1]) return c;
+	unsigned x = c - 4*map[-1];
+	x = map[x*5/4]>>2*x%8 | map[x*5/4+1]<<8-2*x%8 & 1023;
+	return x < 256 ? x : legacy_chars[x-256];
+}
+
+static unsigned uni_to_jis(unsigned c)
 {
-	/* a number of trivial iso-8859-15 <> utf-8 transliterations */
-	switch (c) {
-	case 0x20AC: return 0xA4; /* Euro */
-	case 0x0160: return 0xA6; /* S caron */
-	case 0x0161: return 0xA8; /* s caron */
-	case 0x017D: return 0xB4; /* Z caron */
-	case 0x017E: return 0xB8; /* z caron */
-	case 0x0152: return 0xBC; /* OE */
-	case 0x0153: return 0xBD; /* oe */
-	case 0x0178: return 0xBE; /* Y diaeresis */
-	default:     return '?';
+	unsigned nel = sizeof rev_jis / sizeof *rev_jis;
+	unsigned d, j, i, b = 0;
+	for (;;) {
+		i = nel/2;
+		j = rev_jis[b+i];
+		d = jis0208[j/256][j%256];
+		if (d==c) return j + 0x2121;
+		else if (nel == 1) return 0;
+		else if (c < d)
+			nel /= 2;
+		else {
+			b += i;
+			nel -= nel/2;
+		}
 	}
 }
 
-size_t iconv(iconv_t cd, char **in, size_t *inb, char **out, size_t *outb)
+size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restrict out, size_t *restrict outb)
 {
 	size_t x=0;
-	unsigned char to = (cd>>1)&127;
-	unsigned char from = 255;
-	const unsigned char *map = 0;
+	struct stateful_cd *scd=0;
+	if (!((size_t)cd & 1)) {
+		scd = (void *)cd;
+		cd = scd->base_cd;
+	}
+	unsigned to = extract_to(cd);
+	unsigned from = extract_from(cd);
+	const unsigned char *map = charmaps+from+1;
+	const unsigned char *tomap = charmaps+to+1;
 	char tmp[MB_LEN_MAX];
-	wchar_t c, d;
+	unsigned c, d;
 	size_t k, l;
 	int err;
+	unsigned char type = map[-1];
+	unsigned char totype = tomap[-1];
 
 	if (!in || !*in || !*inb) return 0;
 
-	if (cd & 1)
-		map = charmaps[cd>>8].map;
-	else
-		from = cd>>8;
-
 	for (; *inb; *in+=l, *inb-=l) {
 		c = *(unsigned char *)*in;
 		l = 1;
-		if (from >= UTF_8 && c < 0x80) goto charok;
-		switch (from) {
-		case WCHAR_T:
-			l = sizeof(wchar_t);
-			if (*inb < l) goto starved;
-			c = *(wchar_t *)*in;
-			break;
+
+		switch (type) {
 		case UTF_8:
+			if (c < 128) break;
 			l = utf8dec_wchar(&c, *in, *inb);
 			if (!l) l++;
 			else if (l == (size_t)-1) goto ilseq;
 			else if (l == (size_t)-2) goto starved;
 			break;
 		case US_ASCII:
-			goto ilseq;
-		case LATIN_9:
-			if ((unsigned)c - 0xa4 <= 0xbe - 0xa4) {
-				static const unsigned char map[] = {
-					0, 0x60, 0, 0x61, 0, 0, 0, 0, 0, 0, 0,
-					0, 0, 0, 0, 0x7d, 0, 0, 0, 0x7e, 0, 0, 0,
-					0x52, 0x53, 0x78
-				};
-				if (c == 0xa4) c = 0x20ac;
-				else if (map[c-0xa5]) c = 0x100 | map[c-0xa5];
-			}
-		case LATIN_1:
-			goto charok;
-		case TIS_620:
-			if (c >= 0xa1) c += 0x0e01-0xa1;
-			goto charok;
-		case JIS_0201:
-			if (c >= 0xa1) {
-				if (c <= 0xdf) c += 0xff61-0xa1;
-				else goto ilseq;
+			if (c >= 128) goto ilseq;
+			break;
+		case WCHAR_T:
+			l = sizeof(wchar_t);
+			if (*inb < l) goto starved;
+			c = *(wchar_t *)*in;
+			if (0) {
+		case UTF_32BE:
+		case UTF_32LE:
+			l = 4;
+			if (*inb < 4) goto starved;
+			c = get_32((void *)*in, type);
 			}
-			goto charok;
+			if (c-0xd800u < 0x800u || c >= 0x110000u) goto ilseq;
+			break;
+		case UCS2BE:
+		case UCS2LE:
 		case UTF_16BE:
 		case UTF_16LE:
 			l = 2;
 			if (*inb < 2) goto starved;
-			c = get_16(*in, from);
+			c = get_16((void *)*in, type);
 			if ((unsigned)(c-0xdc00) < 0x400) goto ilseq;
 			if ((unsigned)(c-0xd800) < 0x400) {
+				if (type-UCS2BE < 2U) goto ilseq;
 				l = 4;
 				if (*inb < 4) goto starved;
-				d = get_16(*in + 2, from);
-				if ((unsigned)(c-0xdc00) >= 0x400) goto ilseq;
-				c = ((c-0xd800)<<10) | (d-0xdc00);
+				d = get_16((void *)(*in + 2), type);
+				if ((unsigned)(d-0xdc00) >= 0x400) goto ilseq;
+				c = ((c-0xd7c0)<<10) + (d-0xdc00);
 			}
 			break;
-		case UTF_32BE:
-		case UTF_32LE:
-			l = 4;
-			if (*inb < 4) goto starved;
-			// FIXME
-			// c = get_32(*in, from);
-			break;
-		default:
-			/* only support ascii supersets */
-			if (c < 0x80) break;
-			switch (map[0]) {
-			case UCS2_8BIT:
-				c -= 0x80;
+		case UCS2:
+		case UTF_16:
+			l = 0;
+			if (!scd->state) {
+				if (*inb < 2) goto starved;
+				c = get_16((void *)*in, 0);
+				scd->state = type==UCS2
+					? c==0xfffe ? UCS2LE : UCS2BE
+					: c==0xfffe ? UTF_16LE : UTF_16BE;
+				if (c == 0xfffe || c == 0xfeff)
+					l = 2;
+			}
+			type = scd->state;
+			continue;
+		case UTF_32:
+			l = 0;
+			if (!scd->state) {
+				if (*inb < 4) goto starved;
+				c = get_32((void *)*in, 0);
+				scd->state = c==0xfffe0000 ? UTF_32LE : UTF_32BE;
+				if (c == 0xfffe0000 || c == 0xfeff)
+					l = 4;
+			}
+			type = scd->state;
+			continue;
+		case SHIFT_JIS:
+			if (c < 128) break;
+			if (c-0xa1 <= 0xdf-0xa1) {
+				c += 0xff61-0xa1;
 				break;
-			case EUC_JP:
-				if ((unsigned)c - 0xa1 >= 94) goto ilseq;
-				if ((unsigned)in[0][1] - 0xa1 >= 94) goto ilseq;
-				c = (c-0xa1)*94 + (in[0][1]-0xa1);
-				l = 2;
+			}
+			l = 2;
+			if (*inb < 2) goto starved;
+			d = *((unsigned char *)*in + 1);
+			if (c-129 <= 159-129) c -= 129;
+			else if (c-224 <= 239-224) c -= 193;
+			else goto ilseq;
+			c *= 2;
+			if (d-64 <= 158-64) {
+				if (d==127) goto ilseq;
+				if (d>127) d--;
+				d -= 64;
+			} else if (d-159 <= 252-159) {
+				c++;
+				d -= 159;
+			}
+			c = jis0208[c][d];
+			if (!c) goto ilseq;
+			break;
+		case EUC_JP:
+			if (c < 128) break;
+			l = 2;
+			if (*inb < 2) goto starved;
+			d = *((unsigned char *)*in + 1);
+			if (c==0x8e) {
+				c = d;
+				if (c-0xa1 > 0xdf-0xa1) goto ilseq;
+				c += 0xff61 - 0xa1;
 				break;
-			case SHIFT_JIS:
-				if ((unsigned)c - 0xa1 <= 0xdf-0xa1) {
-					c += 0xff61-0xa1;
-					goto charok;
+			}
+			c -= 0xa1;
+			d -= 0xa1;
+			if (c >= 84 || d >= 94) goto ilseq;
+			c = jis0208[c][d];
+			if (!c) goto ilseq;
+			break;
+		case ISO2022_JP:
+			if (c >= 128) goto ilseq;
+			if (c == '\033') {
+				l = 3;
+				if (*inb < 3) goto starved;
+				c = *((unsigned char *)*in + 1);
+				d = *((unsigned char *)*in + 2);
+				if (c != '(' && c != '$') goto ilseq;
+				switch (128*(c=='$') + d) {
+				case 'B': scd->state=0; continue;
+				case 'J': scd->state=1; continue;
+				case 'I': scd->state=4; continue;
+				case 128+'@': scd->state=2; continue;
+				case 128+'B': scd->state=3; continue;
 				}
-				// FIXME...
+				goto ilseq;
+			}
+			switch (scd->state) {
+			case 1:
+				if (c=='\\') c = 0xa5;
+				if (c=='~') c = 0x203e;
+				break;
+			case 2:
+			case 3:
 				l = 2;
+				if (*inb < 2) goto starved;
+				d = *((unsigned char *)*in + 1);
+				c -= 0x21;
+				d -= 0x21;
+				if (c >= 84 || d >= 94) goto ilseq;
+				c = jis0208[c][d];
+				if (!c) goto ilseq;
+				break;
+			case 4:
+				if (c-0x60 < 0x1f) goto ilseq;
+				if (c-0x21 < 0x5e) c += 0xff61-0x21;
 				break;
-			default:
-				goto badf;
 			}
-			c = get_16(map + 4 + 2*c, 0);
-			if (c == 0xffff) goto ilseq;
-			goto charok;
+			break;
+		default:
+			if (!c) break;
+			c = legacy_map(map, c);
+			if (!c) goto ilseq;
 		}
 
-		if ((unsigned)c - 0xd800 < 0x800 || (unsigned)c >= 0x110000)
-			goto ilseq;
-charok:
-		switch (to) {
+		switch (totype) {
 		case WCHAR_T:
 			if (*outb < sizeof(wchar_t)) goto toobig;
 			*(wchar_t *)*out = c;
@@ -389,36 +548,124 @@ charok:
 			*outb -= k;
 			break;
 		case US_ASCII:
-			if (c > 0x7f) c = 0xfffd;
-			/* fall thru and count replacement in latin1 case */
-		case LATIN_9:
-			if (c >= 0x100 && c != 0xfffd)
-				c = latin9_translit(c);
-			/* fall through */
-		case LATIN_1:
-			if (!*outb) goto toobig;
-			if (c < 0x100) **out = c;
-			else x++, **out = '*'; //FIXME: translit?
-			++*out;
-			--*outb;
+			if (c > 0x7f) subst: x++, c='*';
+		default:
+			if (*outb < 1) goto toobig;
+			if (c<256 && c==legacy_map(tomap, c)) {
+			revout:
+				*(*out)++ = c;
+				*outb -= 1;
+				break;
+			}
+			d = c;
+			for (c=4*totype; c<256; c++) {
+				if (d == legacy_map(tomap, c)) {
+					goto revout;
+				}
+			}
+			goto subst;
+		case SHIFT_JIS:
+			if (c < 128) goto revout;
+			if (c == 0xa5) {
+				x++;
+				c = '\\';
+				goto revout;
+			}
+			if (c == 0x203e) {
+				x++;
+				c = '~';
+				goto revout;
+			}
+			if (c-0xff61 <= 0xdf-0xa1) {
+				c += 0xa1 - 0xff61;
+				goto revout;
+			}
+			c = uni_to_jis(c);
+			if (!c) goto subst;
+			if (*outb < 2) goto toobig;
+			d = c%256;
+			c = c/256;
+			*(*out)++ = (c+1)/2 + (c<95 ? 112 : 176);
+			*(*out)++ = c%2 ? d + 31 + d/96 : d + 126;
+			*outb -= 2;
+			break;
+		case EUC_JP:
+			if (c < 128) goto revout;
+			if (c-0xff61 <= 0xdf-0xa1) {
+				c += 0x0e00 + 0x21 - 0xff61;
+			} else {
+				c = uni_to_jis(c);
+			}
+			if (!c) goto subst;
+			if (*outb < 2) goto toobig;
+			*(*out)++ = c/256 + 0x80;
+			*(*out)++ = c%256 + 0x80;
+			*outb -= 2;
 			break;
+		case ISO2022_JP:
+			if (c < 128) goto revout;
+			if (c-0xff61 <= 0xdf-0xa1 || c==0xa5 || c==0x203e) {
+				if (*outb < 7) goto toobig;
+				*(*out)++ = '\033';
+				*(*out)++ = '(';
+				if (c==0xa5) {
+					*(*out)++ = 'J';
+					*(*out)++ = '\\';
+				} else if (c==0x203e) {
+					*(*out)++ = 'J';
+					*(*out)++ = '~';
+				} else {
+					*(*out)++ = 'I';
+					*(*out)++ = c-0xff61+0x21;
+				}
+				*(*out)++ = '\033';
+				*(*out)++ = '(';
+				*(*out)++ = 'B';
+				*outb -= 7;
+				break;
+			}
+			c = uni_to_jis(c);
+			if (!c) goto subst;
+			if (*outb < 8) goto toobig;
+			*(*out)++ = '\033';
+			*(*out)++ = '$';
+			*(*out)++ = 'B';
+			*(*out)++ = c/256;
+			*(*out)++ = c%256;
+			*(*out)++ = '\033';
+			*(*out)++ = '(';
+			*(*out)++ = 'B';
+			*outb -= 8;
+			break;
+		case UCS2:
+			totype = UCS2BE;
+		case UCS2BE:
+		case UCS2LE:
+		case UTF_16:
 		case UTF_16BE:
 		case UTF_16LE:
-			if (c < 0x10000) {
+			if (c < 0x10000 || totype-UCS2BE < 2U) {
+				if (c >= 0x10000) c = 0xFFFD;
 				if (*outb < 2) goto toobig;
-				put_16(*out, c, to);
+				put_16((void *)*out, c, totype);
 				*out += 2;
 				*outb -= 2;
 				break;
 			}
 			if (*outb < 4) goto toobig;
-			put_16(*out, (c>>10)|0xd800, to);
-			put_16(*out + 2, (c&0x3ff)|0xdc00, to);
+			c -= 0x10000;
+			put_16((void *)*out, (c>>10)|0xd800, totype);
+			put_16((void *)(*out + 2), (c&0x3ff)|0xdc00, totype);
+			*out += 4;
+			*outb -= 4;
+			break;
+		case UTF_32BE:
+		case UTF_32LE:
+			if (*outb < 4) goto toobig;
+			put_32((void *)*out, c, totype);
 			*out += 4;
 			*outb -= 4;
 			break;
-		default:
-			goto badf;
 		}
 	}
 	return x;
@@ -426,17 +673,20 @@ ilseq:
 	err = EILSEQ;
 	x = -1;
 	goto end;
-badf:
-	err = EBADF;
-	x = -1;
-	goto end;
 toobig:
 	err = E2BIG;
 	x = -1;
 	goto end;
 starved:
 	err = EINVAL;
+	x = -1;
 end:
 	errno = err;
 	return x;
 }
+
+int iconv_close(iconv_t cd)
+{
+	return 0;
+}
+
diff --git a/libiconv/include/charmaps.h b/libiconv/include/charmaps.h
deleted file mode 100644
index 083cd699f..000000000
--- a/libiconv/include/charmaps.h
+++ /dev/null
@@ -1,80 +0,0 @@
-#include "charmaps/iso-8859-2.h"
-#include "charmaps/iso-8859-10.h"
-#include "charmaps/windows-874.h"
-#include "charmaps/windows-1250.h"
-#include "charmaps/koi8-r.h"
-
-#ifdef ALL_CHARSETS
-#include "charmaps/iso-8859-3.h"
-#include "charmaps/iso-8859-4.h"
-#include "charmaps/iso-8859-5.h"
-#include "charmaps/iso-8859-6.h"
-#include "charmaps/iso-8859-7.h"
-#include "charmaps/iso-8859-8.h"
-#include "charmaps/iso-8859-9.h"
-#include "charmaps/iso-8859-13.h"
-#include "charmaps/iso-8859-14.h"
-#include "charmaps/iso-8859-16.h"
-#include "charmaps/windows-1251.h"
-#include "charmaps/windows-1252.h"
-#include "charmaps/windows-1253.h"
-#include "charmaps/windows-1254.h"
-#include "charmaps/windows-1255.h"
-#include "charmaps/windows-1256.h"
-#include "charmaps/windows-1257.h"
-#include "charmaps/windows-1258.h"
-#endif
-
-
-struct charmap {
-	const char name[13];
-	const unsigned char *map;
-};
-
-static struct charmap charmaps[] = {
-	{ "ISO-8859-2",   map_iso_8859_2   },
-	{ "ISO-8859-10",  map_iso_8859_10  },
-
-#ifdef ALL_CHARSETS
-	{ "ISO-8859-3",   map_iso_8859_3   },
-	{ "ISO-8859-4",   map_iso_8859_4   },
-	{ "ISO-8859-5",   map_iso_8859_5   },
-	{ "ISO-8859-6",   map_iso_8859_6   },
-	{ "ISO-8859-7",   map_iso_8859_7   },
-	{ "ISO-8859-8",   map_iso_8859_8   },
-	{ "ISO-8859-9",   map_iso_8859_9   },
-	{ "ISO-8859-13",  map_iso_8859_13  },
-	{ "ISO-8859-14",  map_iso_8859_14  },
-	{ "ISO-8859-16",  map_iso_8859_16  },
-#endif
-
-	{ "WINDOWS-874",  map_windows_874  },
-	{ "WINDOWS-1250", map_windows_1250 },
-
-#ifdef ALL_CHARSETS
-	{ "WINDOWS-1251", map_windows_1251 },
-	{ "WINDOWS-1252", map_windows_1252 },
-	{ "WINDOWS-1253", map_windows_1253 },
-	{ "WINDOWS-1254", map_windows_1254 },
-	{ "WINDOWS-1255", map_windows_1255 },
-	{ "WINDOWS-1256", map_windows_1256 },
-	{ "WINDOWS-1257", map_windows_1257 },
-	{ "WINDOWS-1258", map_windows_1258 },
-#endif
-
-	{ "KOI8-R",       map_koi8_r       },
-
-	/* Aliases */
-	{ "LATIN2",       map_iso_8859_2   },
-	{ "LATIN6",       map_iso_8859_10  },
-
-#ifdef ALL_CHARSETS
-	{ "ARABIC",       map_iso_8859_6   },
-	{ "CYRILLIC",     map_iso_8859_5   },
-	{ "GREEK",        map_iso_8859_7   },
-	{ "HEBREW",       map_iso_8859_8   },
-	{ "LATIN3",       map_iso_8859_3   },
-	{ "LATIN4",       map_iso_8859_4   },
-	{ "LATIN5",       map_iso_8859_9   },
-#endif
-};
diff --git a/libiconv/include/charmaps/iso-8859-10.h b/libiconv/include/charmaps/iso-8859-10.h
deleted file mode 100644
index cb93d027d..000000000
--- a/libiconv/include/charmaps/iso-8859-10.h
+++ /dev/null
@@ -1,24 +0,0 @@
-static const unsigned char map_iso_8859_10[] = {
-	0x00, 0x00, 0x00, 0x00, 0x00, 0x80, 0x00, 0x81, 0x00, 0x82, 0x00, 0x83,
-	0x00, 0x84, 0x00, 0x85, 0x00, 0x86, 0x00, 0x87, 0x00, 0x88, 0x00, 0x89,
-	0x00, 0x8a, 0x00, 0x8b, 0x00, 0x8c, 0x00, 0x8d, 0x00, 0x8e, 0x00, 0x8f,
-	0x00, 0x90, 0x00, 0x91, 0x00, 0x92, 0x00, 0x93, 0x00, 0x94, 0x00, 0x95,
-	0x00, 0x96, 0x00, 0x97, 0x00, 0x98, 0x00, 0x99, 0x00, 0x9a, 0x00, 0x9b,
-	0x00, 0x9c, 0x00, 0x9d, 0x00, 0x9e, 0x00, 0x9f, 0x00, 0xa0, 0x01, 0x04,
-	0x01, 0x12, 0x01, 0x22, 0x01, 0x2a, 0x01, 0x28, 0x01, 0x36, 0x00, 0xa7,
-	0x01, 0x3b, 0x01, 0x10, 0x01, 0x60, 0x01, 0x66, 0x01, 0x7d, 0x00, 0xad,
-	0x01, 0x6a, 0x01, 0x4a, 0x00, 0xb0, 0x01, 0x05, 0x01, 0x13, 0x01, 0x23,
-	0x01, 0x2b, 0x01, 0x29, 0x01, 0x37, 0x00, 0xb7, 0x01, 0x3c, 0x01, 0x11,
-	0x01, 0x61, 0x01, 0x67, 0x01, 0x7e, 0x20, 0x15, 0x01, 0x6b, 0x01, 0x4b,
-	0x01, 0x00, 0x00, 0xc1, 0x00, 0xc2, 0x00, 0xc3, 0x00, 0xc4, 0x00, 0xc5,
-	0x00, 0xc6, 0x01, 0x2e, 0x01, 0x0c, 0x00, 0xc9, 0x01, 0x18, 0x00, 0xcb,
-	0x01, 0x16, 0x00, 0xcd, 0x00, 0xce, 0x00, 0xcf, 0x00, 0xd0, 0x01, 0x45,
-	0x01, 0x4c, 0x00, 0xd3, 0x00, 0xd4, 0x00, 0xd5, 0x00, 0xd6, 0x01, 0x68,
-	0x00, 0xd8, 0x01, 0x72, 0x00, 0xda, 0x00, 0xdb, 0x00, 0xdc, 0x00, 0xdd,
-	0x00, 0xde, 0x00, 0xdf, 0x01, 0x01, 0x00, 0xe1, 0x00, 0xe2, 0x00, 0xe3,
-	0x00, 0xe4, 0x00, 0xe5, 0x00, 0xe6, 0x01, 0x2f, 0x01, 0x0d, 0x00, 0xe9,
-	0x01, 0x19, 0x00, 0xeb, 0x01, 0x17, 0x00, 0xed, 0x00, 0xee, 0x00, 0xef,
-	0x00, 0xf0, 0x01, 0x46, 0x01, 0x4d, 0x00, 0xf3, 0x00, 0xf4, 0x00, 0xf5,
-	0x00, 0xf6, 0x01, 0x69, 0x00, 0xf8, 0x01, 0x73, 0x00, 0xfa, 0x00, 0xfb,
-	0x00, 0xfc, 0x00, 0xfd, 0x00, 0xfe, 0x01, 0x38
-};
diff --git a/libiconv/include/charmaps/iso-8859-13.h b/libiconv/include/charmaps/iso-8859-13.h
deleted file mode 100644
index b44d1b434..000000000
--- a/libiconv/include/charmaps/iso-8859-13.h
+++ /dev/null
@@ -1,24 +0,0 @@
-static const unsigned char map_iso_8859_13[] = {
-	0x00, 0x00, 0x00, 0x00, 0x00, 0x80, 0x00, 0x81, 0x00, 0x82, 0x00, 0x83,
-	0x00, 0x84, 0x00, 0x85, 0x00, 0x86, 0x00, 0x87, 0x00, 0x88, 0x00, 0x89,
-	0x00, 0x8a, 0x00, 0x8b, 0x00, 0x8c, 0x00, 0x8d, 0x00, 0x8e, 0x00, 0x8f,
-	0x00, 0x90, 0x00, 0x91, 0x00, 0x92, 0x00, 0x93, 0x00, 0x94, 0x00, 0x95,
-	0x00, 0x96, 0x00, 0x97, 0x00, 0x98, 0x00, 0x99, 0x00, 0x9a, 0x00, 0x9b,
-	0x00, 0x9c, 0x00, 0x9d, 0x00, 0x9e, 0x00, 0x9f, 0x00, 0xa0, 0x20, 0x1d,
-	0x00, 0xa2, 0x00, 0xa3, 0x00, 0xa4, 0x20, 0x1e, 0x00, 0xa6, 0x00, 0xa7,
-	0x00, 0xd8, 0x00, 0xa9, 0x01, 0x56, 0x00, 0xab, 0x00, 0xac, 0x00, 0xad,
-	0x00, 0xae, 0x00, 0xc6, 0x00, 0xb0, 0x00, 0xb1, 0x00, 0xb2, 0x00, 0xb3,
-	0x20, 0x1c, 0x00, 0xb5, 0x00, 0xb6, 0x00, 0xb7, 0x00, 0xf8, 0x00, 0xb9,
-	0x01, 0x57, 0x00, 0xbb, 0x00, 0xbc, 0x00, 0xbd, 0x00, 0xbe, 0x00, 0xe6,
-	0x01, 0x04, 0x01, 0x2e, 0x01, 0x00, 0x01, 0x06, 0x00, 0xc4, 0x00, 0xc5,
-	0x01, 0x18, 0x01, 0x12, 0x01, 0x0c, 0x00, 0xc9, 0x01, 0x79, 0x01, 0x16,
-	0x01, 0x22, 0x01, 0x36, 0x01, 0x2a, 0x01, 0x3b, 0x01, 0x60, 0x01, 0x43,
-	0x01, 0x45, 0x00, 0xd3, 0x01, 0x4c, 0x00, 0xd5, 0x00, 0xd6, 0x00, 0xd7,
-	0x01, 0x72, 0x01, 0x41, 0x01, 0x5a, 0x01, 0x6a, 0x00, 0xdc, 0x01, 0x7b,
-	0x01, 0x7d, 0x00, 0xdf, 0x01, 0x05, 0x01, 0x2f, 0x01, 0x01, 0x01, 0x07,
-	0x00, 0xe4, 0x00, 0xe5, 0x01, 0x19, 0x01, 0x13, 0x01, 0x0d, 0x00, 0xe9,
-	0x01, 0x7a, 0x01, 0x17, 0x01, 0x23, 0x01, 0x37, 0x01, 0x2b, 0x01, 0x3c,
-	0x01, 0x61, 0x01, 0x44, 0x01, 0x46, 0x00, 0xf3, 0x01, 0x4d, 0x00, 0xf5,
-	0x00, 0xf6, 0x00, 0xf7, 0x01, 0x73, 0x01, 0x42, 0x01, 0x5b, 0x01, 0x6b,
-	0x00, 0xfc, 0x01, 0x7c, 0x01, 0x7e, 0x20, 0x19
-};
diff --git a/libiconv/include/charmaps/iso-8859-14.h b/libiconv/include/charmaps/iso-8859-14.h
deleted file mode 100644
index 76cead98c..000000000
--- a/