first commit

charset.c

@@ -0,0 +1,139 @@
+#include "charset.h"
+#include <string.h>
+
+static int is_utf16_le(const unsigned char *in, size_t len)
+{
+	return len >= 2 && in[0] == 0xFF && in[1] == 0xFE;
+}
+
+static int is_utf16_be(const unsigned char *in, size_t len)
+{
+	return len >= 2 && in[0] == 0xFE && in[1] == 0xFF;
+}
+
+static int is_utf8(const unsigned char *in, size_t len)
+{
+	size_t i;
+	int has_multibyte = 0;
+
+	for (i = 0; i < len; i++) {
+		if (in[i] < 0x80)
+			continue;
+		has_multibyte = 1;
+		if ((in[i] & 0xE0) == 0xC0) {
+			if (i + 1 >= len || (in[i+1] & 0xC0) != 0x80)
+				return 0;
+			i += 1;
+		} else if ((in[i] & 0xF0) == 0xE0) {
+			if (i + 2 >= len || (in[i+1] & 0xC0) != 0x80 ||
+			    (in[i+2] & 0xC0) != 0x80)
+				return 0;
+			i += 2;
+		} else if ((in[i] & 0xF8) == 0xF0) {
+			if (i + 3 >= len || (in[i+1] & 0xC0) != 0x80 ||
+			    (in[i+2] & 0xC0) != 0x80 || (in[i+3] & 0xC0) != 0x80)
+				return 0;
+			i += 3;
+		} else {
+			return 0;
+		}
+	}
+	return has_multibyte;
+}
+
+static int utf8_to_codepoint(const unsigned char *in, size_t remaining,
+                             unsigned int *cp)
+{
+	if (in[0] < 0x80) {
+		*cp = in[0];
+		return 1;
+	} else if ((in[0] & 0xE0) == 0xC0) {
+		if (remaining < 2) return 0;
+		*cp = ((in[0] & 0x1F) << 6) | (in[1] & 0x3F);
+		return 2;
+	} else if ((in[0] & 0xF0) == 0xE0) {
+		if (remaining < 3) return 0;
+		*cp = ((in[0] & 0x0F) << 12) | ((in[1] & 0x3F) << 6) |
+		      (in[2] & 0x3F);
+		return 3;
+	} else if ((in[0] & 0xF8) == 0xF0) {
+		if (remaining < 4) return 0;
+		*cp = ((in[0] & 0x07) << 18) | ((in[1] & 0x3F) << 12) |
+		      ((in[2] & 0x3F) << 6) | (in[3] & 0x3F);
+		return 4;
+	}
+	return 0;
+}
+
+static int utf16le_to_iso8859_1(const unsigned char *in, size_t in_len,
+                                char *out, size_t out_size)
+{
+	size_t i, o = 0;
+
+	for (i = 2; i + 1 < in_len && o + 1 < out_size; i += 2) {
+		unsigned int cp = in[i] | (in[i+1] << 8);
+		out[o++] = cp <= 0xFF ? (char)cp : '?';
+	}
+	out[o] = '\0';
+	return (int)o;
+}
+
+static int utf16be_to_iso8859_1(const unsigned char *in, size_t in_len,
+                                char *out, size_t out_size)
+{
+	size_t i, o = 0;
+
+	for (i = 2; i + 1 < in_len && o + 1 < out_size; i += 2) {
+		unsigned int cp = (in[i] << 8) | in[i+1];
+		out[o++] = cp <= 0xFF ? (char)cp : '?';
+	}
+	out[o] = '\0';
+	return (int)o;
+}
+
+static int utf8_to_iso8859_1(const unsigned char *in, size_t in_len,
+                             char *out, size_t out_size)
+{
+	size_t i = 0, o = 0;
+
+	while (i < in_len && o + 1 < out_size) {
+		unsigned int cp;
+		int consumed = utf8_to_codepoint(in + i, in_len - i, &cp);
+		if (consumed == 0) {
+			out[o++] = '?';
+			i++;
+			continue;
+		}
+		out[o++] = cp <= 0xFF ? (char)cp : '?';
+		i += consumed;
+	}
+	out[o] = '\0';
+	return (int)o;
+}
+
+int to_iso8859_1(const unsigned char *in, size_t in_len,
+                 char *out, size_t out_size)
+{
+	if (!in || !out || out_size == 0)
+		return 0;
+
+	if (in_len == 0) {
+		out[0] = '\0';
+		return 0;
+	}
+
+	if (is_utf16_le(in, in_len))
+		return utf16le_to_iso8859_1(in, in_len, out, out_size);
+
+	if (is_utf16_be(in, in_len))
+		return utf16be_to_iso8859_1(in, in_len, out, out_size);
+
+	if (is_utf8(in, in_len))
+		return utf8_to_iso8859_1(in, in_len, out, out_size);
+
+	/* Already ISO-8859-1 / ASCII — passthrough */
+	size_t copy = in_len < out_size - 1 ? in_len : out_size - 1;
+	memcpy(out, in, copy);
+	out[copy] = '\0';
+	return (int)copy;
+}