X-Git-Url: http://nsz.repo.hu/git/?a=blobdiff_plain;f=src%2Flocale%2Ficonv.c;h=1eeea94e0a38753691269b978f105b5e65ee33b8;hb=4cd8b4725907651f329e2f96d428c4e3521643f8;hp=508d322e5e38927da68e31e0ab137f2299415086;hpb=26710be7148d51a954c4fe4acedab90d161e609a;p=musl

diff --git a/src/locale/iconv.c b/src/locale/iconv.c
index 508d322e..1eeea94e 100644
--- a/src/locale/iconv.c
+++ b/src/locale/iconv.c
@@ -5,6 +5,7 @@
 #include <stdlib.h>
 #include <limits.h>
 #include <stdint.h>
+#include "locale_impl.h"
 
 #define UTF_32BE    0300
 #define UTF_16LE    0301
@@ -20,12 +21,8 @@
 #define GB18030     0330
 #define GBK         0331
 #define GB2312      0332
-
-/* 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 BIG5        0340
+#define EUC_KR      0350
 
 /* Definitions of charmaps. Each charmap consists of:
  * 1. Empty-string-terminated list of null-terminated aliases.
@@ -33,7 +30,7 @@
  * 3. Character table (size determined by field 2). */
 
 static const unsigned char charmaps[] =
-"utf8\0\0\310"
+"utf8\0char\0\0\310"
 "wchart\0\0\306"
 "ucs2\0ucs2be\0\0\304"
 "ucs2le\0\0\305"
@@ -47,6 +44,8 @@ static const unsigned char charmaps[] =
 "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"
 ;
 
@@ -62,6 +61,18 @@ static const unsigned short gb18030[126][190] = {
 #include "gb18030.h"
 };
 
+static const unsigned short big5[89][157] = {
+#include "big5.h"
+};
+
+static const unsigned short hkscs[] = {
+#include "hkscs.h"
+};
+
+static const unsigned short ksc[93][94] = {
+#include "ksc.h"
+};
+
 static int fuzzycmp(const unsigned char *a, const unsigned char *b)
 {
 	for (; *a && *b; a++, b++) {
@@ -74,6 +85,7 @@ static int fuzzycmp(const unsigned char *a, const unsigned char *b)
 static size_t find_charmap(const void *name)
 {
 	const unsigned char *s;
+	if (!*(char *)name) name=charmaps; /* "utf8" */
 	for (s=charmaps; *s; ) {
 		if (!fuzzycmp(name, s)) {
 			for (; *s; s+=strlen((void *)s)+1);
@@ -139,7 +151,7 @@ static void put_32(unsigned char *s, unsigned c, int e)
 #define mbrtowc_utf8 mbrtowc
 #define wctomb_utf8 wctomb
 
-size_t iconv(iconv_t cd0, char **in, size_t *inb, char **out, size_t *outb)
+size_t iconv(iconv_t cd0, char **restrict in, size_t *restrict inb, char **restrict out, size_t *restrict outb)
 {
 	size_t x=0;
 	unsigned long cd = (unsigned long)cd0;
@@ -154,9 +166,12 @@ size_t iconv(iconv_t cd0, char **in, size_t *inb, char **out, size_t *outb)
 	int err;
 	unsigned char type = map[-1];
 	unsigned char totype = tomap[-1];
+	locale_t *ploc = &CURRENT_LOCALE, loc = *ploc;
 
 	if (!in || !*in || !*inb) return 0;
 
+	*ploc = UTF8_LOCALE;
+
 	for (; *inb; *in+=l, *inb-=l) {
 		c = *(unsigned char *)*in;
 		l = 1;
@@ -278,6 +293,75 @@ size_t iconv(iconv_t cd0, char **in, size_t *inb, char **out, size_t *outb)
 			if (d>63) d--;
 			c = gb18030[c][d];
 			break;
+		case BIG5:
+			l = 2;
+			if (*inb < 2) goto starved;
+			d = *((unsigned char *)*in + 1);
+			if (d-0x40>=0xff-0x40 || d-0x7f<0xa1-0x7f) goto ilseq;
+			d -= 0x40;
+			if (d > 0x3e) d -= 0x22;
+			if (c-0xa1>=0xfa-0xa1) {
+				if (c-0x87>=0xff-0x87) goto ilseq;
+				if (c < 0xa1) c -= 0x87;
+				else c -= 0x87 + (0xfa-0xa1);
+				c = (hkscs[4867+(c*157+d)/16]>>(c*157+d)%16)%2<<17
+					| hkscs[c*157+d];
+				/* A few HKSCS characters map to pairs of UCS
+				 * characters. These are mapped to surrogate
+				 * range in the hkscs table then hard-coded
+				 * here. Ugly, yes. */
+				if (c/256 == 0xdc) {
+					if (totype-0300U > 8) k = 2;
+					else k = "\10\4\4\10\4\4\10\2\4"[totype-0300];
+					if (k > *outb) goto toobig;
+					x += iconv((iconv_t)(uintptr_t)to,
+						&(char *){"\303\212\314\204"
+						"\303\212\314\214"
+						"\303\252\314\204"
+						"\303\252\314\214"
+						+c%256}, &(size_t){4},
+						out, outb);
+					continue;
+				}
+				if (!c) goto ilseq;
+				break;
+			}
+			c -= 0xa1;
+			c = big5[c][d]|(c==0x27&&(d==0x3a||d==0x3c||d==0x42))<<17;
+			if (!c) goto ilseq;
+			break;
+		case EUC_KR:
+			l = 2;
+			if (*inb < 2) goto starved;
+			d = *((unsigned char *)*in + 1);
+			c -= 0xa1;
+			d -= 0xa1;
+			if (c >= 93 || d >= 94) {
+				c += (0xa1-0x81);
+				d += 0xa1;
+				if (c >= 93 || c>=0xc6-0x81 && d>0x52)
+					goto ilseq;
+				if (d-'A'<26) d = d-'A';
+				else if (d-'a'<26) d = d-'a'+26;
+				else if (d-0x81<0xff-0x81) d = d-0x81+52;
+				else goto ilseq;
+				if (c < 0x20) c = 178*c + d;
+				else c = 178*0x20 + 84*(c-0x20) + d;
+				c += 0xac00;
+				for (d=0xac00; d<=c; ) {
+					k = 0;
+					for (int i=0; i<93; i++)
+						for (int j=0; j<94; j++)
+							if (ksc[i][j]-d <= c-d)
+								k++;
+					d = c+1;
+					c += k;
+				}
+				break;
+			}
+			c = ksc[c][d];
+			if (!c) goto ilseq;
+			break;
 		default:
 			if (c < 128+type) break;
 			c -= 128+type;
@@ -316,8 +400,8 @@ size_t iconv(iconv_t cd0, char **in, size_t *inb, char **out, size_t *outb)
 			}
 			d = c;
 			for (c=0; c<128-totype; c++) {
-				if (d == legacy_chars[ map[c*5/4]>>2*c%8 |
-					map[c*5/4+1]<<8-2*c%8 & 1023 ]) {
+				if (d == legacy_chars[ tomap[c*5/4]>>2*c%8 |
+					tomap[c*5/4+1]<<8-2*c%8 & 1023 ]) {
 					c += 128;
 					goto revout;
 				}
@@ -351,6 +435,7 @@ size_t iconv(iconv_t cd0, char **in, size_t *inb, char **out, size_t *outb)
 			break;
 		}
 	}
+	*ploc = loc;
 	return x;
 ilseq:
 	err = EILSEQ;
@@ -365,5 +450,6 @@ starved:
 	x = -1;
 end:
 	errno = err;
+	*ploc = loc;
 	return x;
 }