X-Git-Url: http://nsz.repo.hu/git/?a=blobdiff_plain;f=string_rep.h;h=13a167462aec90712a94e5b832de380b917f1cc8;hb=ebb77232f9baed801f3a073741ff7a023acea485;hp=e24b40c496527adeb71f0e9ddf2328f5d89ec79e;hpb=6d7d61de445629e1590d3a2ba2990ef7a651368a;p=cparser

diff --git a/string_rep.h b/string_rep.h
index e24b40c..13a1674 100644
--- a/string_rep.h
+++ b/string_rep.h
@@ -20,18 +20,66 @@
 #ifndef STRING_REP_H
 #define STRING_REP_H
 
-#include <wchar.h>
-
-typedef wchar_t wchar_rep_t;
+#include <assert.h>
+#include <stdlib.h>
 
 typedef struct string_t {
-	const char *begin;
-	size_t      size;
+	const char *begin; /**< UTF-8 encoded string, the last character is
+						* guaranteed to be 0 */
+	size_t      size;  /**< size of string in bytes (not characters) */
 } string_t;
 
-typedef struct wide_string_t {
-	const wchar_rep_t *begin;
-	size_t             size;
-} wide_string_t;
+typedef unsigned int utf32;
+#define UTF32_PRINTF_FORMAT "%u"
+
+/**
+ * "parse" an utf8 character from a string.
+ * Warning: This function only works for valid utf-8 inputs. The behaviour
+ * is undefined for invalid utf-8 input.
+ *
+ * @param p    A pointer to a pointer into the string. The pointer
+ *             is incremented for each consumed char
+ */
+static inline utf32 read_utf8_char(const char **p)
+{
+	const unsigned char *c      = (const unsigned char *) *p;
+	utf32                result;
+
+	if ((*c & 0x80) == 0) {
+		/* 1 character encoding: 0b0??????? */
+		result = *c++;
+	} else if ((*c & 0xE0) == 0xC0) {
+		/* 2 character encoding: 0b110?????, 0b10?????? */
+		result = *c++ & 0x1F;
+		result = (result << 6) | (*c++ & 0x3F);
+	} else if ((*c & 0xF0) == 0xE0) {
+		/* 3 character encoding: 0b1110????, 0b10??????, 0b10?????? */
+		result = *c++ & 0x0F;
+		result = (result << 6) | (*c++ & 0x3F);
+		result = (result << 6) | (*c++ & 0x3F);
+	} else {
+		/* 4 character enc.: 0b11110???, 0b10??????, 0b10??????, 0b10?????? */
+		assert((*c & 0xF8) == 0xF0);
+		result = *c++ & 0x07;
+		result = (result << 6) | (*c++ & 0x3F);
+		result = (result << 6) | (*c++ & 0x3F);
+		result = (result << 6) | (*c++ & 0x3F);
+	}
+
+	*p = (const char*) c;
+	return result;
+}
+
+static inline size_t wstrlen(const string_t *string)
+{
+	size_t      result = 0;
+	const char *p      = string->begin;
+	const char *end    = p + string->size;
+	while (p < end) {
+		read_utf8_char(&p);
+		++result;
+	}
+	return result;
+}
 
 #endif