Exclude long attribute values from tilestats output, as specified

text.cpp

@@ -50,3 +50,75 @@ std::string check_utf8(std::string s) {
 
 	return "";
 }
+
+const char *utf8_next(const char *s, long *c) {
+	if (s == NULL) {
+		*c = -1;
+		return NULL;
+	}
+
+	if (*s == '\0') {
+		*c = -1;
+		return NULL;
+	}
+
+	if ((s[0] & 0x80) == 0x80) {
+		if ((s[0] & 0xE0) == 0xC0) {
+			if ((s[1] & 0xC0) != 0x80) {
+				*c = 0xFFFD;
+				s++;
+			} else {
+				*c = ((long) (s[0] & 0x1F) << 6) | ((long) (s[1] & 0x7F));
+				s += 2;
+			}
+		} else if ((s[0] & 0xF0) == 0xE0) {
+			if ((s[1] & 0xC0) != 0x80 || (s[2] & 0xC0) != 0x80) {
+				*c = 0xFFFD;
+				s++;
+			} else {
+				*c = ((long) (s[0] & 0x0F) << 12) | ((long) (s[1] & 0x7F) << 6) | ((long) (s[2] & 0x7F));
+				s += 3;
+			}
+		} else if ((s[0] & 0xF8) == 0xF0) {
+			if ((s[1] & 0xC0) != 0x80 || (s[2] & 0xC0) != 0x80 || (s[3] & 0xC0) != 0x80) {
+				*c = 0xFFFD;
+				s++;
+			} else {
+				*c = ((long) (s[0] & 0x0F) << 18) | ((long) (s[1] & 0x7F) << 12) | ((long) (s[2] & 0x7F) << 6) | ((long) (s[3] & 0x7F));
+				s += 4;
+			}
+		} else {
+			*c = 0xFFFD;
+			s++;
+		}
+	} else {
+		*c = s[0];
+		s++;
+	}
+
+	return s;
+}
+
+std::string truncate16(std::string const &s, size_t runes) {
+	const char *cp = s.c_str();
+	const char *start = cp;
+	const char *lastgood = cp;
+	size_t len = 0;
+	long c;
+
+	while ((cp = utf8_next(cp, &c)) != NULL) {
+		if (c <= 0xFFFF) {
+			len++;
+		} else {
+			len += 2;
+		}
+
+		if (len <= runes) {
+			lastgood = cp;
+		} else {
+			break;
+		}
+	}
+
+	return std::string(s, 0, lastgood - start);
+}