diff --git a/autotest/UnicodeInfo/UnicodeInfo.cpp b/autotest/UnicodeInfo/UnicodeInfo.cpp
index 7b924230c..b97ddd9ee 100644
--- a/autotest/UnicodeInfo/UnicodeInfo.cpp
+++ b/autotest/UnicodeInfo/UnicodeInfo.cpp
@@ -9,14 +9,90 @@ dword ScanHex(const char *s)
 
 CONSOLE_APP_MAIN
 {
+//	StdLogSetup(LOG_COUT|LOG_FILE);
+
 	FileIn in(GetDataFile("../../uppbox/Unicode/UnicodeData.txt"));
 
 	while(!in.IsEof()) {
-		DUMP(in.GetLine());
+		Vector<String> h = Split(in.GetLine(), ';', false);
+		if(h.GetCount() > 14) {
+			int code = ScanInt(h[0], NULL, 16);
+			if(code < 200)
+				DUMPC(h);
+			String cat = h[2];
+			int upper = ScanHex(h[12]);
+			int lower = ScanHex(h[13]);
+			DUMPHEX(code);
+			DUMP(cat);
+			DUMPHEX(lower);
+			DUMPHEX(upper);
+			DDUMPHEX(ToUpper(code));
+			DDUMPHEX(ToLower(code));
+			DDUMP(IsUpper(code));
+			DDUMP(IsLower(code));
+			DUMPHEX(ToAscii(code));
+			ASSERT(IsLetter(code) == (*cat == 'L'));
+			if(cat == "Lu")
+				ASSERT(IsUpper(code));
+			if(cat == "Ll")
+				ASSERT(IsLower(code));
+			if(lower)
+				ASSERT(ToLower(code) == lower);
+			if(upper)
+				ASSERT(ToUpper(code) == upper);
+			ASSERT((h[4] == "R") == IsRTL(code));
+
+			Vector<String> decomb;
+			decomb.AppendRange(FilterRange(Split(h[5], ' '), [] (const String& x) { return x[0] != '<'; }));
+			dword first = decomb.GetCount() ? ScanHex(decomb[0]) : 0;
+
+			if(code < 128)
+				ASSERT(ToAscii(code) == code);
+			else {
+				if(first < 128)
+					ASSERT(ToAscii(code) == first);
+				else
+					ASSERT(ToAscii(code) == 0);
+			}
+
+			if(decomb.GetCount() > 1 ||
+			   decomb.GetCount() && code >= 2048 && first != code && first < 128) // for ToAscii...
+			{
+				decomb.SetCount(3, "0");
+				dword comb[3];
+				comb[0] = ScanHex(decomb[0]);
+				comb[1] = ScanHex(decomb[1]);
+				comb[2] = ScanHex(decomb[2]);
+				
+				dword t[MAX_DECOMPOSED];
+				int n = UnicodeDecompose(code, t);
+				ASSERT(comb[0] == t[0]);
+				ASSERT(comb[1] == t[1]);
+				ASSERT(comb[2] == t[2]);
+				
+				DDUMP(n);
+				
+				if(n > 1) {
+					DDUMPHEX(UnicodeCompose(t, n));
+					ASSERT(code == UnicodeCompose(t, n));
+				}
+			}
+
+			DLOG("============");
+		}
 	}
 	
+	for(int i = 0; i < 256; i++) {
+		#define CHT(fn, type) ASSERT(fn(i) == fn((type)i));
+		#define CHK(fn) CHT(fn, dword) CHT(fn, wchar) CHT(fn, char)
+		
+		CHK(IsLetter);
+		CHK(IsUpper);
+		CHK(IsLower);
+		CHK(ToUpper);
+		CHK(ToLower);
+		CHK(ToAscii);
+	}
 	
-	for(int i = 0; i < 0x10ffff; i++)
-		if(IsLetter(i))
-			DDUMPHEX(i);
+	LOG("============= OK");
 }