initial commit

1 files changed, 230 insertions, 0 deletions

diff --git a/iconv.c b/iconv.c
new file mode 100644
index 0000000..533da0a
--- /dev/null
+++ b/iconv.c
@@ -0,0 +1,230 @@
+
+#include <errno.h>
+#include <ctype.h>
+
+#define __need_size_t
+#include <stddef.h>
+#include <string.h>
+
+#include "iconv.h"
+
+#ifndef UNUSED
+#define UNUSED __attribute__ ((__unused__))
+#endif
+
+
+static int decode_ascii(int *scalar, const unsigned char *inbuf, int count);
+static int decode_iso_8859_1(int *scalar, const unsigned char *inbuf, 
+	int count);
+static int decode_utf_8(int *scalar, const unsigned char *inbuf, int count);
+
+static int encode_ascii(int scalar, unsigned char *outbuf);
+static int encode_iso_8859_1(int scalar, unsigned char *outbuf);
+static int encode_utf_8(int scalar, unsigned char *outbuf);
+
+enum {
+	ASCII = 0,
+	UTF_8,
+	ISO_8859_1
+};
+
+struct converter {
+	int code;
+	const char *name;
+	int (*decode)(int *scalar, const unsigned char *inbuf, int count);
+	int (*encode)(int scalar, unsigned char *outbuf);
+};
+
+struct converter supported_codesets[] = {
+	{ 
+		.code = ASCII,
+		.name = "ASCII",
+		.decode = &decode_ascii,
+		.encode = &encode_ascii,
+	}, { 
+		.code = UTF_8,
+		.name = "UTF-8",
+		.decode = &decode_utf_8,
+		.encode = &encode_utf_8,
+	}, {
+		.code = ISO_8859_1,
+		.name = "ISO-8859-1",
+		.decode = &decode_iso_8859_1,
+		.encode = &encode_iso_8859_1
+	}, {
+		.code = -1,
+	}
+};
+
+/* ASCII */
+static int decode_ascii(int *scalar, const unsigned char *inbuf, 
+	int UNUSED count)
+{
+	if (*inbuf >= 0x80) {
+		errno = EILSEQ;
+		return -1;
+	}
+	*scalar = *inbuf;
+	return 1;
+}
+
+static int encode_ascii(int scalar, unsigned char *outbuf)
+{
+	if (scalar >= 0x80)
+		return 0;
+	*outbuf = scalar;
+	return 1;
+}
+
+/* ISO-8859-1 */
+static int decode_iso_8859_1(int *scalar, const unsigned char *inbuf,
+	int UNUSED count)
+{
+	*scalar = *inbuf;
+	return 1;
+}
+
+static int encode_iso_8859_1(int scalar, unsigned char *outbuf)
+{
+	if (scalar > 0xff)
+		return 0;
+	*outbuf = scalar;
+	return 1;
+}
+
+
+/* UTF-8 */
+/* scalar value			utf-8
+ * 00000000 00000000 0zzzzzzz	0zzzzzzz
+ * 00000000 00000yyy yyzzzzzz 	110yyyyy 10zzzzzz
+ * 00000000 xxxxyyyy yyzzzzzz 	1110xxxx 10yyyyyy 10zzzzzz
+ * 000wwwxx xxxxyyyy yyzzzzzz 	11110www 10xxxxxx 10yyyyyy 10zzzzzz
+ *
+ * This implementation only supports the 2 first variants
+ */
+static int decode_utf_8(int *scalar, const unsigned char *inbuf, int count)
+{
+	int i = inbuf[0];
+
+	if (i < 0x80) {
+		*scalar = i;
+		return 1;
+	}
+
+	if (count < 2) {
+		errno = EINVAL;
+		return -1;
+	}
+
+	i = (i << 8) | inbuf[1];
+	if ((i & 0xe0c0) != 0xc080) {
+		errno = EILSEQ;
+		return -1;
+	}
+	
+	*scalar = ((i & 0x1700) >> 2) | (i & 0x3f);
+	return 2;
+}
+
+static int encode_utf_8(int scalar, unsigned char *outbuf)
+{
+	if (scalar < 0x80) {
+		*outbuf = scalar;
+		return 1;
+	}
+
+	if (scalar > 0x7ff)
+		return 0;
+
+	outbuf[0] = (scalar >> 6) | 0xc0;
+	outbuf[1] = (scalar & 0x3f) | 0x80;
+	return 2;
+}
+
+static void toupperstr(char *p)
+{
+	while (p && *p) {
+		if (*p == '_')
+		       *p = '-';
+		*p = toupper(*p);
+		p++;
+	}
+}
+
+static int find_converter(const char *str)
+{
+	int i;
+	char buf[16];
+	strncpy(buf, str, sizeof(buf));
+	buf[15] = '\0';
+	toupperstr(buf);
+	for (i = 0; supported_codesets[i].code != -1; i++) {
+		if (strcmp(buf, supported_codesets[i].name) == 0)
+			return i;
+	}
+	return -1;
+}
+
+iconv_t iconv_open(const char *tocode, const char *fromcode)
+{
+	int src, dest;
+	iconv_t cd = (iconv_t) -1;
+
+	if (tocode == NULL || fromcode == NULL) {
+		errno = EINVAL;
+		goto return_error;
+	}
+	
+	src = find_converter(fromcode);
+	dest = find_converter(tocode);
+	if (src == -1 || dest == -1) {
+		errno = EINVAL;
+		goto return_error;
+	}
+
+	cd = (iconv_t) (dest << 8) | src;
+
+return_error:
+	return cd;
+}	
+	
+int iconv_close(iconv_t UNUSED cd)
+{
+	return 0;
+}
+
+size_t iconv(iconv_t cd, const char **inbuf, size_t *insize,
+		char **outbuf, size_t *outsize)
+{
+	struct converter *in = &supported_codesets[cd & 0xff];
+	struct converter *out = &supported_codesets[cd >> 8];
+	size_t nonidentical = 0;
+	if (!inbuf || !*inbuf || !outbuf || !*outbuf)
+		return 0;
+	while (*insize) {
+		int scalar, infwd, outfwd;
+		infwd = in->decode(&scalar, (unsigned char *)*inbuf, *insize);
+		if (infwd < 0)
+			goto ret_error;
+		
+		outfwd = out->encode(scalar, (unsigned char *)*outbuf);
+		if (outfwd > *outsize) {
+			errno = E2BIG;
+			goto ret_error;
+		}
+
+		if (**inbuf != scalar || outfwd == 0)
+			nonidentical++;
+
+		*inbuf += infwd;		
+		*insize -= infwd;
+		*outbuf += outfwd;
+		*outsize -= outfwd;
+	}
+	if (*outsize)
+		*outbuf = '\0';
+	return nonidentical;
+
+ret_error:
+	return -1;
+}