diff --git a/java/src/com/android/inputmethod/latin/ExpandableDictionary.java b/java/src/com/android/inputmethod/latin/ExpandableDictionary.java
index a67c919..2666573 100644
--- a/java/src/com/android/inputmethod/latin/ExpandableDictionary.java
+++ b/java/src/com/android/inputmethod/latin/ExpandableDictionary.java
@@ -728,172 +728,206 @@
      * to their base characters.  If c is in range, BASE_CHARS[c] == c
      * if c is not a combined character, or the base character if it
      * is combined.
+     *
+     * cf. native/jni/src/utils/char_utils.cpp
      */
     private static final char BASE_CHARS[] = {
-        0x0000, 0x0001, 0x0002, 0x0003, 0x0004, 0x0005, 0x0006, 0x0007,
-        0x0008, 0x0009, 0x000a, 0x000b, 0x000c, 0x000d, 0x000e, 0x000f,
-        0x0010, 0x0011, 0x0012, 0x0013, 0x0014, 0x0015, 0x0016, 0x0017,
-        0x0018, 0x0019, 0x001a, 0x001b, 0x001c, 0x001d, 0x001e, 0x001f,
-        0x0020, 0x0021, 0x0022, 0x0023, 0x0024, 0x0025, 0x0026, 0x0027,
-        0x0028, 0x0029, 0x002a, 0x002b, 0x002c, 0x002d, 0x002e, 0x002f,
-        0x0030, 0x0031, 0x0032, 0x0033, 0x0034, 0x0035, 0x0036, 0x0037,
-        0x0038, 0x0039, 0x003a, 0x003b, 0x003c, 0x003d, 0x003e, 0x003f,
-        0x0040, 0x0041, 0x0042, 0x0043, 0x0044, 0x0045, 0x0046, 0x0047,
-        0x0048, 0x0049, 0x004a, 0x004b, 0x004c, 0x004d, 0x004e, 0x004f,
-        0x0050, 0x0051, 0x0052, 0x0053, 0x0054, 0x0055, 0x0056, 0x0057,
-        0x0058, 0x0059, 0x005a, 0x005b, 0x005c, 0x005d, 0x005e, 0x005f,
-        0x0060, 0x0061, 0x0062, 0x0063, 0x0064, 0x0065, 0x0066, 0x0067,
-        0x0068, 0x0069, 0x006a, 0x006b, 0x006c, 0x006d, 0x006e, 0x006f,
-        0x0070, 0x0071, 0x0072, 0x0073, 0x0074, 0x0075, 0x0076, 0x0077,
-        0x0078, 0x0079, 0x007a, 0x007b, 0x007c, 0x007d, 0x007e, 0x007f,
-        0x0080, 0x0081, 0x0082, 0x0083, 0x0084, 0x0085, 0x0086, 0x0087,
-        0x0088, 0x0089, 0x008a, 0x008b, 0x008c, 0x008d, 0x008e, 0x008f,
-        0x0090, 0x0091, 0x0092, 0x0093, 0x0094, 0x0095, 0x0096, 0x0097,
-        0x0098, 0x0099, 0x009a, 0x009b, 0x009c, 0x009d, 0x009e, 0x009f,
-        0x0020, 0x00a1, 0x00a2, 0x00a3, 0x00a4, 0x00a5, 0x00a6, 0x00a7,
-        0x0020, 0x00a9, 0x0061, 0x00ab, 0x00ac, 0x00ad, 0x00ae, 0x0020,
-        0x00b0, 0x00b1, 0x0032, 0x0033, 0x0020, 0x03bc, 0x00b6, 0x00b7,
-        0x0020, 0x0031, 0x006f, 0x00bb, 0x0031, 0x0031, 0x0033, 0x00bf,
-        0x0041, 0x0041, 0x0041, 0x0041, 0x0041, 0x0041, 0x00c6, 0x0043,
-        0x0045, 0x0045, 0x0045, 0x0045, 0x0049, 0x0049, 0x0049, 0x0049,
-        0x00d0, 0x004e, 0x004f, 0x004f, 0x004f, 0x004f, 0x004f, 0x00d7,
-        0x004f, 0x0055, 0x0055, 0x0055, 0x0055, 0x0059, 0x00de, 0x0073, // Manually changed d8 to 4f
-                                                                        // Manually changed df to 73
-        0x0061, 0x0061, 0x0061, 0x0061, 0x0061, 0x0061, 0x00e6, 0x0063,
-        0x0065, 0x0065, 0x0065, 0x0065, 0x0069, 0x0069, 0x0069, 0x0069,
-        0x00f0, 0x006e, 0x006f, 0x006f, 0x006f, 0x006f, 0x006f, 0x00f7,
-        0x006f, 0x0075, 0x0075, 0x0075, 0x0075, 0x0079, 0x00fe, 0x0079, // Manually changed f8 to 6f
-        0x0041, 0x0061, 0x0041, 0x0061, 0x0041, 0x0061, 0x0043, 0x0063,
-        0x0043, 0x0063, 0x0043, 0x0063, 0x0043, 0x0063, 0x0044, 0x0064,
-        0x0110, 0x0111, 0x0045, 0x0065, 0x0045, 0x0065, 0x0045, 0x0065,
-        0x0045, 0x0065, 0x0045, 0x0065, 0x0047, 0x0067, 0x0047, 0x0067,
-        0x0047, 0x0067, 0x0047, 0x0067, 0x0048, 0x0068, 0x0126, 0x0127,
-        0x0049, 0x0069, 0x0049, 0x0069, 0x0049, 0x0069, 0x0049, 0x0069,
-        0x0049, 0x0131, 0x0049, 0x0069, 0x004a, 0x006a, 0x004b, 0x006b,
-        0x0138, 0x004c, 0x006c, 0x004c, 0x006c, 0x004c, 0x006c, 0x004c,
-        0x006c, 0x0141, 0x0142, 0x004e, 0x006e, 0x004e, 0x006e, 0x004e,
-        0x006e, 0x02bc, 0x014a, 0x014b, 0x004f, 0x006f, 0x004f, 0x006f,
-        0x004f, 0x006f, 0x0152, 0x0153, 0x0052, 0x0072, 0x0052, 0x0072,
-        0x0052, 0x0072, 0x0053, 0x0073, 0x0053, 0x0073, 0x0053, 0x0073,
-        0x0053, 0x0073, 0x0054, 0x0074, 0x0054, 0x0074, 0x0166, 0x0167,
-        0x0055, 0x0075, 0x0055, 0x0075, 0x0055, 0x0075, 0x0055, 0x0075,
-        0x0055, 0x0075, 0x0055, 0x0075, 0x0057, 0x0077, 0x0059, 0x0079,
-        0x0059, 0x005a, 0x007a, 0x005a, 0x007a, 0x005a, 0x007a, 0x0073,
-        0x0180, 0x0181, 0x0182, 0x0183, 0x0184, 0x0185, 0x0186, 0x0187,
-        0x0188, 0x0189, 0x018a, 0x018b, 0x018c, 0x018d, 0x018e, 0x018f,
-        0x0190, 0x0191, 0x0192, 0x0193, 0x0194, 0x0195, 0x0196, 0x0197,
-        0x0198, 0x0199, 0x019a, 0x019b, 0x019c, 0x019d, 0x019e, 0x019f,
-        0x004f, 0x006f, 0x01a2, 0x01a3, 0x01a4, 0x01a5, 0x01a6, 0x01a7,
-        0x01a8, 0x01a9, 0x01aa, 0x01ab, 0x01ac, 0x01ad, 0x01ae, 0x0055,
-        0x0075, 0x01b1, 0x01b2, 0x01b3, 0x01b4, 0x01b5, 0x01b6, 0x01b7,
-        0x01b8, 0x01b9, 0x01ba, 0x01bb, 0x01bc, 0x01bd, 0x01be, 0x01bf,
-        0x01c0, 0x01c1, 0x01c2, 0x01c3, 0x0044, 0x0044, 0x0064, 0x004c,
-        0x004c, 0x006c, 0x004e, 0x004e, 0x006e, 0x0041, 0x0061, 0x0049,
-        0x0069, 0x004f, 0x006f, 0x0055, 0x0075, 0x00dc, 0x00fc, 0x00dc,
-        0x00fc, 0x00dc, 0x00fc, 0x00dc, 0x00fc, 0x01dd, 0x00c4, 0x00e4,
-        0x0226, 0x0227, 0x00c6, 0x00e6, 0x01e4, 0x01e5, 0x0047, 0x0067,
-        0x004b, 0x006b, 0x004f, 0x006f, 0x01ea, 0x01eb, 0x01b7, 0x0292,
-        0x006a, 0x0044, 0x0044, 0x0064, 0x0047, 0x0067, 0x01f6, 0x01f7,
-        0x004e, 0x006e, 0x00c5, 0x00e5, 0x00c6, 0x00e6, 0x00d8, 0x00f8,
-        0x0041, 0x0061, 0x0041, 0x0061, 0x0045, 0x0065, 0x0045, 0x0065,
-        0x0049, 0x0069, 0x0049, 0x0069, 0x004f, 0x006f, 0x004f, 0x006f,
-        0x0052, 0x0072, 0x0052, 0x0072, 0x0055, 0x0075, 0x0055, 0x0075,
-        0x0053, 0x0073, 0x0054, 0x0074, 0x021c, 0x021d, 0x0048, 0x0068,
-        0x0220, 0x0221, 0x0222, 0x0223, 0x0224, 0x0225, 0x0041, 0x0061,
-        0x0045, 0x0065, 0x00d6, 0x00f6, 0x00d5, 0x00f5, 0x004f, 0x006f,
-        0x022e, 0x022f, 0x0059, 0x0079, 0x0234, 0x0235, 0x0236, 0x0237,
-        0x0238, 0x0239, 0x023a, 0x023b, 0x023c, 0x023d, 0x023e, 0x023f,
-        0x0240, 0x0241, 0x0242, 0x0243, 0x0244, 0x0245, 0x0246, 0x0247,
-        0x0248, 0x0249, 0x024a, 0x024b, 0x024c, 0x024d, 0x024e, 0x024f,
-        0x0250, 0x0251, 0x0252, 0x0253, 0x0254, 0x0255, 0x0256, 0x0257,
-        0x0258, 0x0259, 0x025a, 0x025b, 0x025c, 0x025d, 0x025e, 0x025f,
-        0x0260, 0x0261, 0x0262, 0x0263, 0x0264, 0x0265, 0x0266, 0x0267,
-        0x0268, 0x0269, 0x026a, 0x026b, 0x026c, 0x026d, 0x026e, 0x026f,
-        0x0270, 0x0271, 0x0272, 0x0273, 0x0274, 0x0275, 0x0276, 0x0277,
-        0x0278, 0x0279, 0x027a, 0x027b, 0x027c, 0x027d, 0x027e, 0x027f,
-        0x0280, 0x0281, 0x0282, 0x0283, 0x0284, 0x0285, 0x0286, 0x0287,
-        0x0288, 0x0289, 0x028a, 0x028b, 0x028c, 0x028d, 0x028e, 0x028f,
-        0x0290, 0x0291, 0x0292, 0x0293, 0x0294, 0x0295, 0x0296, 0x0297,
-        0x0298, 0x0299, 0x029a, 0x029b, 0x029c, 0x029d, 0x029e, 0x029f,
-        0x02a0, 0x02a1, 0x02a2, 0x02a3, 0x02a4, 0x02a5, 0x02a6, 0x02a7,
-        0x02a8, 0x02a9, 0x02aa, 0x02ab, 0x02ac, 0x02ad, 0x02ae, 0x02af,
-        0x0068, 0x0266, 0x006a, 0x0072, 0x0279, 0x027b, 0x0281, 0x0077,
-        0x0079, 0x02b9, 0x02ba, 0x02bb, 0x02bc, 0x02bd, 0x02be, 0x02bf,
-        0x02c0, 0x02c1, 0x02c2, 0x02c3, 0x02c4, 0x02c5, 0x02c6, 0x02c7,
-        0x02c8, 0x02c9, 0x02ca, 0x02cb, 0x02cc, 0x02cd, 0x02ce, 0x02cf,
-        0x02d0, 0x02d1, 0x02d2, 0x02d3, 0x02d4, 0x02d5, 0x02d6, 0x02d7,
-        0x0020, 0x0020, 0x0020, 0x0020, 0x0020, 0x0020, 0x02de, 0x02df,
-        0x0263, 0x006c, 0x0073, 0x0078, 0x0295, 0x02e5, 0x02e6, 0x02e7,
-        0x02e8, 0x02e9, 0x02ea, 0x02eb, 0x02ec, 0x02ed, 0x02ee, 0x02ef,
-        0x02f0, 0x02f1, 0x02f2, 0x02f3, 0x02f4, 0x02f5, 0x02f6, 0x02f7,
-        0x02f8, 0x02f9, 0x02fa, 0x02fb, 0x02fc, 0x02fd, 0x02fe, 0x02ff,
-        0x0300, 0x0301, 0x0302, 0x0303, 0x0304, 0x0305, 0x0306, 0x0307,
-        0x0308, 0x0309, 0x030a, 0x030b, 0x030c, 0x030d, 0x030e, 0x030f,
-        0x0310, 0x0311, 0x0312, 0x0313, 0x0314, 0x0315, 0x0316, 0x0317,
-        0x0318, 0x0319, 0x031a, 0x031b, 0x031c, 0x031d, 0x031e, 0x031f,
-        0x0320, 0x0321, 0x0322, 0x0323, 0x0324, 0x0325, 0x0326, 0x0327,
-        0x0328, 0x0329, 0x032a, 0x032b, 0x032c, 0x032d, 0x032e, 0x032f,
-        0x0330, 0x0331, 0x0332, 0x0333, 0x0334, 0x0335, 0x0336, 0x0337,
-        0x0338, 0x0339, 0x033a, 0x033b, 0x033c, 0x033d, 0x033e, 0x033f,
-        0x0300, 0x0301, 0x0342, 0x0313, 0x0308, 0x0345, 0x0346, 0x0347,
-        0x0348, 0x0349, 0x034a, 0x034b, 0x034c, 0x034d, 0x034e, 0x034f,
-        0x0350, 0x0351, 0x0352, 0x0353, 0x0354, 0x0355, 0x0356, 0x0357,
-        0x0358, 0x0359, 0x035a, 0x035b, 0x035c, 0x035d, 0x035e, 0x035f,
-        0x0360, 0x0361, 0x0362, 0x0363, 0x0364, 0x0365, 0x0366, 0x0367,
-        0x0368, 0x0369, 0x036a, 0x036b, 0x036c, 0x036d, 0x036e, 0x036f,
-        0x0370, 0x0371, 0x0372, 0x0373, 0x02b9, 0x0375, 0x0376, 0x0377,
-        0x0378, 0x0379, 0x0020, 0x037b, 0x037c, 0x037d, 0x003b, 0x037f,
-        0x0380, 0x0381, 0x0382, 0x0383, 0x0020, 0x00a8, 0x0391, 0x00b7,
-        0x0395, 0x0397, 0x0399, 0x038b, 0x039f, 0x038d, 0x03a5, 0x03a9,
-        0x03ca, 0x0391, 0x0392, 0x0393, 0x0394, 0x0395, 0x0396, 0x0397,
-        0x0398, 0x0399, 0x039a, 0x039b, 0x039c, 0x039d, 0x039e, 0x039f,
-        0x03a0, 0x03a1, 0x03a2, 0x03a3, 0x03a4, 0x03a5, 0x03a6, 0x03a7,
-        0x03a8, 0x03a9, 0x0399, 0x03a5, 0x03b1, 0x03b5, 0x03b7, 0x03b9,
-        0x03cb, 0x03b1, 0x03b2, 0x03b3, 0x03b4, 0x03b5, 0x03b6, 0x03b7,
-        0x03b8, 0x03b9, 0x03ba, 0x03bb, 0x03bc, 0x03bd, 0x03be, 0x03bf,
-        0x03c0, 0x03c1, 0x03c2, 0x03c3, 0x03c4, 0x03c5, 0x03c6, 0x03c7,
-        0x03c8, 0x03c9, 0x03b9, 0x03c5, 0x03bf, 0x03c5, 0x03c9, 0x03cf,
-        0x03b2, 0x03b8, 0x03a5, 0x03d2, 0x03d2, 0x03c6, 0x03c0, 0x03d7,
-        0x03d8, 0x03d9, 0x03da, 0x03db, 0x03dc, 0x03dd, 0x03de, 0x03df,
-        0x03e0, 0x03e1, 0x03e2, 0x03e3, 0x03e4, 0x03e5, 0x03e6, 0x03e7,
-        0x03e8, 0x03e9, 0x03ea, 0x03eb, 0x03ec, 0x03ed, 0x03ee, 0x03ef,
-        0x03ba, 0x03c1, 0x03c2, 0x03f3, 0x0398, 0x03b5, 0x03f6, 0x03f7,
-        0x03f8, 0x03a3, 0x03fa, 0x03fb, 0x03fc, 0x03fd, 0x03fe, 0x03ff,
-        0x0415, 0x0415, 0x0402, 0x0413, 0x0404, 0x0405, 0x0406, 0x0406,
-        0x0408, 0x0409, 0x040a, 0x040b, 0x041a, 0x0418, 0x0423, 0x040f,
-        0x0410, 0x0411, 0x0412, 0x0413, 0x0414, 0x0415, 0x0416, 0x0417,
-        0x0418, 0x0418, 0x041a, 0x041b, 0x041c, 0x041d, 0x041e, 0x041f,
-        0x0420, 0x0421, 0x0422, 0x0423, 0x0424, 0x0425, 0x0426, 0x0427,
-        0x0428, 0x0429, 0x042a, 0x042b, 0x042c, 0x042d, 0x042e, 0x042f,
-        0x0430, 0x0431, 0x0432, 0x0433, 0x0434, 0x0435, 0x0436, 0x0437,
-        0x0438, 0x0438, 0x043a, 0x043b, 0x043c, 0x043d, 0x043e, 0x043f,
-        0x0440, 0x0441, 0x0442, 0x0443, 0x0444, 0x0445, 0x0446, 0x0447,
-        0x0448, 0x0449, 0x044a, 0x044b, 0x044c, 0x044d, 0x044e, 0x044f,
-        0x0435, 0x0435, 0x0452, 0x0433, 0x0454, 0x0455, 0x0456, 0x0456,
-        0x0458, 0x0459, 0x045a, 0x045b, 0x043a, 0x0438, 0x0443, 0x045f,
-        0x0460, 0x0461, 0x0462, 0x0463, 0x0464, 0x0465, 0x0466, 0x0467,
-        0x0468, 0x0469, 0x046a, 0x046b, 0x046c, 0x046d, 0x046e, 0x046f,
-        0x0470, 0x0471, 0x0472, 0x0473, 0x0474, 0x0475, 0x0474, 0x0475,
-        0x0478, 0x0479, 0x047a, 0x047b, 0x047c, 0x047d, 0x047e, 0x047f,
-        0x0480, 0x0481, 0x0482, 0x0483, 0x0484, 0x0485, 0x0486, 0x0487,
-        0x0488, 0x0489, 0x048a, 0x048b, 0x048c, 0x048d, 0x048e, 0x048f,
-        0x0490, 0x0491, 0x0492, 0x0493, 0x0494, 0x0495, 0x0496, 0x0497,
-        0x0498, 0x0499, 0x049a, 0x049b, 0x049c, 0x049d, 0x049e, 0x049f,
-        0x04a0, 0x04a1, 0x04a2, 0x04a3, 0x04a4, 0x04a5, 0x04a6, 0x04a7,
-        0x04a8, 0x04a9, 0x04aa, 0x04ab, 0x04ac, 0x04ad, 0x04ae, 0x04af,
-        0x04b0, 0x04b1, 0x04b2, 0x04b3, 0x04b4, 0x04b5, 0x04b6, 0x04b7,
-        0x04b8, 0x04b9, 0x04ba, 0x04bb, 0x04bc, 0x04bd, 0x04be, 0x04bf,
-        0x04c0, 0x0416, 0x0436, 0x04c3, 0x04c4, 0x04c5, 0x04c6, 0x04c7,
-        0x04c8, 0x04c9, 0x04ca, 0x04cb, 0x04cc, 0x04cd, 0x04ce, 0x04cf,
-        0x0410, 0x0430, 0x0410, 0x0430, 0x04d4, 0x04d5, 0x0415, 0x0435,
-        0x04d8, 0x04d9, 0x04d8, 0x04d9, 0x0416, 0x0436, 0x0417, 0x0437,
-        0x04e0, 0x04e1, 0x0418, 0x0438, 0x0418, 0x0438, 0x041e, 0x043e,
-        0x04e8, 0x04e9, 0x04e8, 0x04e9, 0x042d, 0x044d, 0x0423, 0x0443,
-        0x0423, 0x0443, 0x0423, 0x0443, 0x0427, 0x0447, 0x04f6, 0x04f7,
-        0x042b, 0x044b, 0x04fa, 0x04fb, 0x04fc, 0x04fd, 0x04fe, 0x04ff,
+        /* U+0000 */ 0x0000, 0x0001, 0x0002, 0x0003, 0x0004, 0x0005, 0x0006, 0x0007,
+        /* U+0008 */ 0x0008, 0x0009, 0x000A, 0x000B, 0x000C, 0x000D, 0x000E, 0x000F,
+        /* U+0010 */ 0x0010, 0x0011, 0x0012, 0x0013, 0x0014, 0x0015, 0x0016, 0x0017,
+        /* U+0018 */ 0x0018, 0x0019, 0x001A, 0x001B, 0x001C, 0x001D, 0x001E, 0x001F,
+        /* U+0020 */ 0x0020, 0x0021, 0x0022, 0x0023, 0x0024, 0x0025, 0x0026, 0x0027,
+        /* U+0028 */ 0x0028, 0x0029, 0x002A, 0x002B, 0x002C, 0x002D, 0x002E, 0x002F,
+        /* U+0030 */ 0x0030, 0x0031, 0x0032, 0x0033, 0x0034, 0x0035, 0x0036, 0x0037,
+        /* U+0038 */ 0x0038, 0x0039, 0x003A, 0x003B, 0x003C, 0x003D, 0x003E, 0x003F,
+        /* U+0040 */ 0x0040, 0x0041, 0x0042, 0x0043, 0x0044, 0x0045, 0x0046, 0x0047,
+        /* U+0048 */ 0x0048, 0x0049, 0x004A, 0x004B, 0x004C, 0x004D, 0x004E, 0x004F,
+        /* U+0050 */ 0x0050, 0x0051, 0x0052, 0x0053, 0x0054, 0x0055, 0x0056, 0x0057,
+        /* U+0058 */ 0x0058, 0x0059, 0x005A, 0x005B, 0x005C, 0x005D, 0x005E, 0x005F,
+        /* U+0060 */ 0x0060, 0x0061, 0x0062, 0x0063, 0x0064, 0x0065, 0x0066, 0x0067,
+        /* U+0068 */ 0x0068, 0x0069, 0x006A, 0x006B, 0x006C, 0x006D, 0x006E, 0x006F,
+        /* U+0070 */ 0x0070, 0x0071, 0x0072, 0x0073, 0x0074, 0x0075, 0x0076, 0x0077,
+        /* U+0078 */ 0x0078, 0x0079, 0x007A, 0x007B, 0x007C, 0x007D, 0x007E, 0x007F,
+        /* U+0080 */ 0x0080, 0x0081, 0x0082, 0x0083, 0x0084, 0x0085, 0x0086, 0x0087,
+        /* U+0088 */ 0x0088, 0x0089, 0x008A, 0x008B, 0x008C, 0x008D, 0x008E, 0x008F,
+        /* U+0090 */ 0x0090, 0x0091, 0x0092, 0x0093, 0x0094, 0x0095, 0x0096, 0x0097,
+        /* U+0098 */ 0x0098, 0x0099, 0x009A, 0x009B, 0x009C, 0x009D, 0x009E, 0x009F,
+        /* U+00A0 */ 0x0020, 0x00A1, 0x00A2, 0x00A3, 0x00A4, 0x00A5, 0x00A6, 0x00A7,
+        /* U+00A8 */ 0x0020, 0x00A9, 0x0061, 0x00AB, 0x00AC, 0x00AD, 0x00AE, 0x0020,
+        /* U+00B0 */ 0x00B0, 0x00B1, 0x0032, 0x0033, 0x0020, 0x03BC, 0x00B6, 0x00B7,
+        /* U+00B8 */ 0x0020, 0x0031, 0x006F, 0x00BB, 0x0031, 0x0031, 0x0033, 0x00BF,
+        /* U+00C0 */ 0x0041, 0x0041, 0x0041, 0x0041, 0x0041, 0x0041, 0x00C6, 0x0043,
+        /* U+00C8 */ 0x0045, 0x0045, 0x0045, 0x0045, 0x0049, 0x0049, 0x0049, 0x0049,
+        /* U+00D0 */ 0x00D0, 0x004E, 0x004F, 0x004F, 0x004F, 0x004F, 0x004F, 0x00D7,
+        /* U+00D8 */ 0x004F, 0x0055, 0x0055, 0x0055, 0x0055, 0x0059, 0x00DE, 0x0073,
+            // U+00D8: Manually changed from 00D8 to 004F
+              // TODO: Check if it's really acceptable to consider Ø a diacritical variant of O
+            // U+00DF: Manually changed from 00DF to 0073
+        /* U+00E0 */ 0x0061, 0x0061, 0x0061, 0x0061, 0x0061, 0x0061, 0x00E6, 0x0063,
+        /* U+00E8 */ 0x0065, 0x0065, 0x0065, 0x0065, 0x0069, 0x0069, 0x0069, 0x0069,
+        /* U+00F0 */ 0x00F0, 0x006E, 0x006F, 0x006F, 0x006F, 0x006F, 0x006F, 0x00F7,
+        /* U+00F8 */ 0x006F, 0x0075, 0x0075, 0x0075, 0x0075, 0x0079, 0x00FE, 0x0079,
+            // U+00F8: Manually changed from 00F8 to 006F
+              // TODO: Check if it's really acceptable to consider ø a diacritical variant of o
+        /* U+0100 */ 0x0041, 0x0061, 0x0041, 0x0061, 0x0041, 0x0061, 0x0043, 0x0063,
+        /* U+0108 */ 0x0043, 0x0063, 0x0043, 0x0063, 0x0043, 0x0063, 0x0044, 0x0064,
+        /* U+0110 */ 0x0110, 0x0111, 0x0045, 0x0065, 0x0045, 0x0065, 0x0045, 0x0065,
+        /* U+0118 */ 0x0045, 0x0065, 0x0045, 0x0065, 0x0047, 0x0067, 0x0047, 0x0067,
+        /* U+0120 */ 0x0047, 0x0067, 0x0047, 0x0067, 0x0048, 0x0068, 0x0126, 0x0127,
+        /* U+0128 */ 0x0049, 0x0069, 0x0049, 0x0069, 0x0049, 0x0069, 0x0049, 0x0069,
+        /* U+0130 */ 0x0049, 0x0131, 0x0049, 0x0069, 0x004A, 0x006A, 0x004B, 0x006B,
+        /* U+0138 */ 0x0138, 0x004C, 0x006C, 0x004C, 0x006C, 0x004C, 0x006C, 0x004C,
+        /* U+0140 */ 0x006C, 0x004C, 0x006C, 0x004E, 0x006E, 0x004E, 0x006E, 0x004E,
+            // U+0141: Manually changed from 0141 to 004C
+            // U+0142: Manually changed from 0142 to 006C
+        /* U+0148 */ 0x006E, 0x02BC, 0x014A, 0x014B, 0x004F, 0x006F, 0x004F, 0x006F,
+        /* U+0150 */ 0x004F, 0x006F, 0x0152, 0x0153, 0x0052, 0x0072, 0x0052, 0x0072,
+        /* U+0158 */ 0x0052, 0x0072, 0x0053, 0x0073, 0x0053, 0x0073, 0x0053, 0x0073,
+        /* U+0160 */ 0x0053, 0x0073, 0x0054, 0x0074, 0x0054, 0x0074, 0x0166, 0x0167,
+        /* U+0168 */ 0x0055, 0x0075, 0x0055, 0x0075, 0x0055, 0x0075, 0x0055, 0x0075,
+        /* U+0170 */ 0x0055, 0x0075, 0x0055, 0x0075, 0x0057, 0x0077, 0x0059, 0x0079,
+        /* U+0178 */ 0x0059, 0x005A, 0x007A, 0x005A, 0x007A, 0x005A, 0x007A, 0x0073,
+        /* U+0180 */ 0x0180, 0x0181, 0x0182, 0x0183, 0x0184, 0x0185, 0x0186, 0x0187,
+        /* U+0188 */ 0x0188, 0x0189, 0x018A, 0x018B, 0x018C, 0x018D, 0x018E, 0x018F,
+        /* U+0190 */ 0x0190, 0x0191, 0x0192, 0x0193, 0x0194, 0x0195, 0x0196, 0x0197,
+        /* U+0198 */ 0x0198, 0x0199, 0x019A, 0x019B, 0x019C, 0x019D, 0x019E, 0x019F,
+        /* U+01A0 */ 0x004F, 0x006F, 0x01A2, 0x01A3, 0x01A4, 0x01A5, 0x01A6, 0x01A7,
+        /* U+01A8 */ 0x01A8, 0x01A9, 0x01AA, 0x01AB, 0x01AC, 0x01AD, 0x01AE, 0x0055,
+        /* U+01B0 */ 0x0075, 0x01B1, 0x01B2, 0x01B3, 0x01B4, 0x01B5, 0x01B6, 0x01B7,
+        /* U+01B8 */ 0x01B8, 0x01B9, 0x01BA, 0x01BB, 0x01BC, 0x01BD, 0x01BE, 0x01BF,
+        /* U+01C0 */ 0x01C0, 0x01C1, 0x01C2, 0x01C3, 0x0044, 0x0044, 0x0064, 0x004C,
+        /* U+01C8 */ 0x004C, 0x006C, 0x004E, 0x004E, 0x006E, 0x0041, 0x0061, 0x0049,
+        /* U+01D0 */ 0x0069, 0x004F, 0x006F, 0x0055, 0x0075, 0x0055, 0x0075, 0x0055,
+            // U+01D5: Manually changed from 00DC to 0055
+            // U+01D6: Manually changed from 00FC to 0075
+            // U+01D7: Manually changed from 00DC to 0055
+        /* U+01D8 */ 0x0075, 0x0055, 0x0075, 0x0055, 0x0075, 0x01DD, 0x0041, 0x0061,
+            // U+01D8: Manually changed from 00FC to 0075
+            // U+01D9: Manually changed from 00DC to 0055
+            // U+01DA: Manually changed from 00FC to 0075
+            // U+01DB: Manually changed from 00DC to 0055
+            // U+01DC: Manually changed from 00FC to 0075
+            // U+01DE: Manually changed from 00C4 to 0041
+            // U+01DF: Manually changed from 00E4 to 0061
+        /* U+01E0 */ 0x0041, 0x0061, 0x00C6, 0x00E6, 0x01E4, 0x01E5, 0x0047, 0x0067,
+            // U+01E0: Manually changed from 0226 to 0041
+            // U+01E1: Manually changed from 0227 to 0061
+        /* U+01E8 */ 0x004B, 0x006B, 0x004F, 0x006F, 0x004F, 0x006F, 0x01B7, 0x0292,
+            // U+01EC: Manually changed from 01EA to 004F
+            // U+01ED: Manually changed from 01EB to 006F
+        /* U+01F0 */ 0x006A, 0x0044, 0x0044, 0x0064, 0x0047, 0x0067, 0x01F6, 0x01F7,
+        /* U+01F8 */ 0x004E, 0x006E, 0x0041, 0x0061, 0x00C6, 0x00E6, 0x004F, 0x006F,
+            // U+01FA: Manually changed from 00C5 to 0041
+            // U+01FB: Manually changed from 00E5 to 0061
+            // U+01FE: Manually changed from 00D8 to 004F
+              // TODO: Check if it's really acceptable to consider Ø a diacritical variant of O
+            // U+01FF: Manually changed from 00F8 to 006F
+              // TODO: Check if it's really acceptable to consider ø a diacritical variant of o
+        /* U+0200 */ 0x0041, 0x0061, 0x0041, 0x0061, 0x0045, 0x0065, 0x0045, 0x0065,
+        /* U+0208 */ 0x0049, 0x0069, 0x0049, 0x0069, 0x004F, 0x006F, 0x004F, 0x006F,
+        /* U+0210 */ 0x0052, 0x0072, 0x0052, 0x0072, 0x0055, 0x0075, 0x0055, 0x0075,
+        /* U+0218 */ 0x0053, 0x0073, 0x0054, 0x0074, 0x021C, 0x021D, 0x0048, 0x0068,
+        /* U+0220 */ 0x0220, 0x0221, 0x0222, 0x0223, 0x0224, 0x0225, 0x0041, 0x0061,
+        /* U+0228 */ 0x0045, 0x0065, 0x004F, 0x006F, 0x004F, 0x006F, 0x004F, 0x006F,
+            // U+022A: Manually changed from 00D6 to 004F
+            // U+022B: Manually changed from 00F6 to 006F
+            // U+022C: Manually changed from 00D5 to 004F
+            // U+022D: Manually changed from 00F5 to 006F
+        /* U+0230 */ 0x004F, 0x006F, 0x0059, 0x0079, 0x0234, 0x0235, 0x0236, 0x0237,
+            // U+0230: Manually changed from 022E to 004F
+            // U+0231: Manually changed from 022F to 006F
+        /* U+0238 */ 0x0238, 0x0239, 0x023A, 0x023B, 0x023C, 0x023D, 0x023E, 0x023F,
+        /* U+0240 */ 0x0240, 0x0241, 0x0242, 0x0243, 0x0244, 0x0245, 0x0246, 0x0247,
+        /* U+0248 */ 0x0248, 0x0249, 0x024A, 0x024B, 0x024C, 0x024D, 0x024E, 0x024F,
+        /* U+0250 */ 0x0250, 0x0251, 0x0252, 0x0253, 0x0254, 0x0255, 0x0256, 0x0257,
+        /* U+0258 */ 0x0258, 0x0259, 0x025A, 0x025B, 0x025C, 0x025D, 0x025E, 0x025F,
+        /* U+0260 */ 0x0260, 0x0261, 0x0262, 0x0263, 0x0264, 0x0265, 0x0266, 0x0267,
+        /* U+0268 */ 0x0268, 0x0269, 0x026A, 0x026B, 0x026C, 0x026D, 0x026E, 0x026F,
+        /* U+0270 */ 0x0270, 0x0271, 0x0272, 0x0273, 0x0274, 0x0275, 0x0276, 0x0277,
+        /* U+0278 */ 0x0278, 0x0279, 0x027A, 0x027B, 0x027C, 0x027D, 0x027E, 0x027F,
+        /* U+0280 */ 0x0280, 0x0281, 0x0282, 0x0283, 0x0284, 0x0285, 0x0286, 0x0287,
+        /* U+0288 */ 0x0288, 0x0289, 0x028A, 0x028B, 0x028C, 0x028D, 0x028E, 0x028F,
+        /* U+0290 */ 0x0290, 0x0291, 0x0292, 0x0293, 0x0294, 0x0295, 0x0296, 0x0297,
+        /* U+0298 */ 0x0298, 0x0299, 0x029A, 0x029B, 0x029C, 0x029D, 0x029E, 0x029F,
+        /* U+02A0 */ 0x02A0, 0x02A1, 0x02A2, 0x02A3, 0x02A4, 0x02A5, 0x02A6, 0x02A7,
+        /* U+02A8 */ 0x02A8, 0x02A9, 0x02AA, 0x02AB, 0x02AC, 0x02AD, 0x02AE, 0x02AF,
+        /* U+02B0 */ 0x0068, 0x0266, 0x006A, 0x0072, 0x0279, 0x027B, 0x0281, 0x0077,
+        /* U+02B8 */ 0x0079, 0x02B9, 0x02BA, 0x02BB, 0x02BC, 0x02BD, 0x02BE, 0x02BF,
+        /* U+02C0 */ 0x02C0, 0x02C1, 0x02C2, 0x02C3, 0x02C4, 0x02C5, 0x02C6, 0x02C7,
+        /* U+02C8 */ 0x02C8, 0x02C9, 0x02CA, 0x02CB, 0x02CC, 0x02CD, 0x02CE, 0x02CF,
+        /* U+02D0 */ 0x02D0, 0x02D1, 0x02D2, 0x02D3, 0x02D4, 0x02D5, 0x02D6, 0x02D7,
+        /* U+02D8 */ 0x0020, 0x0020, 0x0020, 0x0020, 0x0020, 0x0020, 0x02DE, 0x02DF,
+        /* U+02E0 */ 0x0263, 0x006C, 0x0073, 0x0078, 0x0295, 0x02E5, 0x02E6, 0x02E7,
+        /* U+02E8 */ 0x02E8, 0x02E9, 0x02EA, 0x02EB, 0x02EC, 0x02ED, 0x02EE, 0x02EF,
+        /* U+02F0 */ 0x02F0, 0x02F1, 0x02F2, 0x02F3, 0x02F4, 0x02F5, 0x02F6, 0x02F7,
+        /* U+02F8 */ 0x02F8, 0x02F9, 0x02FA, 0x02FB, 0x02FC, 0x02FD, 0x02FE, 0x02FF,
+        /* U+0300 */ 0x0300, 0x0301, 0x0302, 0x0303, 0x0304, 0x0305, 0x0306, 0x0307,
+        /* U+0308 */ 0x0308, 0x0309, 0x030A, 0x030B, 0x030C, 0x030D, 0x030E, 0x030F,
+        /* U+0310 */ 0x0310, 0x0311, 0x0312, 0x0313, 0x0314, 0x0315, 0x0316, 0x0317,
+        /* U+0318 */ 0x0318, 0x0319, 0x031A, 0x031B, 0x031C, 0x031D, 0x031E, 0x031F,
+        /* U+0320 */ 0x0320, 0x0321, 0x0322, 0x0323, 0x0324, 0x0325, 0x0326, 0x0327,
+        /* U+0328 */ 0x0328, 0x0329, 0x032A, 0x032B, 0x032C, 0x032D, 0x032E, 0x032F,
+        /* U+0330 */ 0x0330, 0x0331, 0x0332, 0x0333, 0x0334, 0x0335, 0x0336, 0x0337,
+        /* U+0338 */ 0x0338, 0x0339, 0x033A, 0x033B, 0x033C, 0x033D, 0x033E, 0x033F,
+        /* U+0340 */ 0x0300, 0x0301, 0x0342, 0x0313, 0x0308, 0x0345, 0x0346, 0x0347,
+        /* U+0348 */ 0x0348, 0x0349, 0x034A, 0x034B, 0x034C, 0x034D, 0x034E, 0x034F,
+        /* U+0350 */ 0x0350, 0x0351, 0x0352, 0x0353, 0x0354, 0x0355, 0x0356, 0x0357,
+        /* U+0358 */ 0x0358, 0x0359, 0x035A, 0x035B, 0x035C, 0x035D, 0x035E, 0x035F,
+        /* U+0360 */ 0x0360, 0x0361, 0x0362, 0x0363, 0x0364, 0x0365, 0x0366, 0x0367,
+        /* U+0368 */ 0x0368, 0x0369, 0x036A, 0x036B, 0x036C, 0x036D, 0x036E, 0x036F,
+        /* U+0370 */ 0x0370, 0x0371, 0x0372, 0x0373, 0x02B9, 0x0375, 0x0376, 0x0377,
+        /* U+0378 */ 0x0378, 0x0379, 0x0020, 0x037B, 0x037C, 0x037D, 0x003B, 0x037F,
+        /* U+0380 */ 0x0380, 0x0381, 0x0382, 0x0383, 0x0020, 0x00A8, 0x0391, 0x00B7,
+        /* U+0388 */ 0x0395, 0x0397, 0x0399, 0x038B, 0x039F, 0x038D, 0x03A5, 0x03A9,
+        /* U+0390 */ 0x03CA, 0x0391, 0x0392, 0x0393, 0x0394, 0x0395, 0x0396, 0x0397,
+        /* U+0398 */ 0x0398, 0x0399, 0x039A, 0x039B, 0x039C, 0x039D, 0x039E, 0x039F,
+        /* U+03A0 */ 0x03A0, 0x03A1, 0x03A2, 0x03A3, 0x03A4, 0x03A5, 0x03A6, 0x03A7,
+        /* U+03A8 */ 0x03A8, 0x03A9, 0x0399, 0x03A5, 0x03B1, 0x03B5, 0x03B7, 0x03B9,
+        /* U+03B0 */ 0x03CB, 0x03B1, 0x03B2, 0x03B3, 0x03B4, 0x03B5, 0x03B6, 0x03B7,
+        /* U+03B8 */ 0x03B8, 0x03B9, 0x03BA, 0x03BB, 0x03BC, 0x03BD, 0x03BE, 0x03BF,
+        /* U+03C0 */ 0x03C0, 0x03C1, 0x03C2, 0x03C3, 0x03C4, 0x03C5, 0x03C6, 0x03C7,
+        /* U+03C8 */ 0x03C8, 0x03C9, 0x03B9, 0x03C5, 0x03BF, 0x03C5, 0x03C9, 0x03CF,
+        /* U+03D0 */ 0x03B2, 0x03B8, 0x03A5, 0x03D2, 0x03D2, 0x03C6, 0x03C0, 0x03D7,
+        /* U+03D8 */ 0x03D8, 0x03D9, 0x03DA, 0x03DB, 0x03DC, 0x03DD, 0x03DE, 0x03DF,
+        /* U+03E0 */ 0x03E0, 0x03E1, 0x03E2, 0x03E3, 0x03E4, 0x03E5, 0x03E6, 0x03E7,
+        /* U+03E8 */ 0x03E8, 0x03E9, 0x03EA, 0x03EB, 0x03EC, 0x03ED, 0x03EE, 0x03EF,
+        /* U+03F0 */ 0x03BA, 0x03C1, 0x03C2, 0x03F3, 0x0398, 0x03B5, 0x03F6, 0x03F7,
+        /* U+03F8 */ 0x03F8, 0x03A3, 0x03FA, 0x03FB, 0x03FC, 0x03FD, 0x03FE, 0x03FF,
+        /* U+0400 */ 0x0415, 0x0415, 0x0402, 0x0413, 0x0404, 0x0405, 0x0406, 0x0406,
+        /* U+0408 */ 0x0408, 0x0409, 0x040A, 0x040B, 0x041A, 0x0418, 0x0423, 0x040F,
+        /* U+0410 */ 0x0410, 0x0411, 0x0412, 0x0413, 0x0414, 0x0415, 0x0416, 0x0417,
+        /* U+0418 */ 0x0418, 0x0419, 0x041A, 0x041B, 0x041C, 0x041D, 0x041E, 0x041F,
+            // U+0419: Manually changed from 0418 to 0419
+        /* U+0420 */ 0x0420, 0x0421, 0x0422, 0x0423, 0x0424, 0x0425, 0x0426, 0x0427,
+        /* U+0428 */ 0x0428, 0x0429, 0x042C, 0x042B, 0x042C, 0x042D, 0x042E, 0x042F,
+            // U+042A: Manually changed from 042A to 042C
+        /* U+0430 */ 0x0430, 0x0431, 0x0432, 0x0433, 0x0434, 0x0435, 0x0436, 0x0437,
+        /* U+0438 */ 0x0438, 0x0439, 0x043A, 0x043B, 0x043C, 0x043D, 0x043E, 0x043F,
+            // U+0439: Manually changed from 0438 to 0439
+        /* U+0440 */ 0x0440, 0x0441, 0x0442, 0x0443, 0x0444, 0x0445, 0x0446, 0x0447,
+        /* U+0448 */ 0x0448, 0x0449, 0x044C, 0x044B, 0x044C, 0x044D, 0x044E, 0x044F,
+            // U+044A: Manually changed from 044A to 044C
+        /* U+0450 */ 0x0435, 0x0435, 0x0452, 0x0433, 0x0454, 0x0455, 0x0456, 0x0456,
+        /* U+0458 */ 0x0458, 0x0459, 0x045A, 0x045B, 0x043A, 0x0438, 0x0443, 0x045F,
+        /* U+0460 */ 0x0460, 0x0461, 0x0462, 0x0463, 0x0464, 0x0465, 0x0466, 0x0467,
+        /* U+0468 */ 0x0468, 0x0469, 0x046A, 0x046B, 0x046C, 0x046D, 0x046E, 0x046F,
+        /* U+0470 */ 0x0470, 0x0471, 0x0472, 0x0473, 0x0474, 0x0475, 0x0474, 0x0475,
+        /* U+0478 */ 0x0478, 0x0479, 0x047A, 0x047B, 0x047C, 0x047D, 0x047E, 0x047F,
+        /* U+0480 */ 0x0480, 0x0481, 0x0482, 0x0483, 0x0484, 0x0485, 0x0486, 0x0487,
+        /* U+0488 */ 0x0488, 0x0489, 0x048A, 0x048B, 0x048C, 0x048D, 0x048E, 0x048F,
+        /* U+0490 */ 0x0490, 0x0491, 0x0492, 0x0493, 0x0494, 0x0495, 0x0496, 0x0497,
+        /* U+0498 */ 0x0498, 0x0499, 0x049A, 0x049B, 0x049C, 0x049D, 0x049E, 0x049F,
+        /* U+04A0 */ 0x04A0, 0x04A1, 0x04A2, 0x04A3, 0x04A4, 0x04A5, 0x04A6, 0x04A7,
+        /* U+04A8 */ 0x04A8, 0x04A9, 0x04AA, 0x04AB, 0x04AC, 0x04AD, 0x04AE, 0x04AF,
+        /* U+04B0 */ 0x04B0, 0x04B1, 0x04B2, 0x04B3, 0x04B4, 0x04B5, 0x04B6, 0x04B7,
+        /* U+04B8 */ 0x04B8, 0x04B9, 0x04BA, 0x04BB, 0x04BC, 0x04BD, 0x04BE, 0x04BF,
+        /* U+04C0 */ 0x04C0, 0x0416, 0x0436, 0x04C3, 0x04C4, 0x04C5, 0x04C6, 0x04C7,
+        /* U+04C8 */ 0x04C8, 0x04C9, 0x04CA, 0x04CB, 0x04CC, 0x04CD, 0x04CE, 0x04CF,
+        /* U+04D0 */ 0x0410, 0x0430, 0x0410, 0x0430, 0x04D4, 0x04D5, 0x0415, 0x0435,
+        /* U+04D8 */ 0x04D8, 0x04D9, 0x04D8, 0x04D9, 0x0416, 0x0436, 0x0417, 0x0437,
+        /* U+04E0 */ 0x04E0, 0x04E1, 0x0418, 0x0438, 0x0418, 0x0438, 0x041E, 0x043E,
+        /* U+04E8 */ 0x04E8, 0x04E9, 0x04E8, 0x04E9, 0x042D, 0x044D, 0x0423, 0x0443,
+        /* U+04F0 */ 0x0423, 0x0443, 0x0423, 0x0443, 0x0427, 0x0447, 0x04F6, 0x04F7,
+        /* U+04F8 */ 0x042B, 0x044B, 0x04FA, 0x04FB, 0x04FC, 0x04FD, 0x04FE, 0x04FF,
     };
-
-    // generated with:
-    // cat UnicodeData.txt | perl -e 'while (<>) { @foo = split(/;/); $foo[5] =~ s/<.*> //; $base[hex($foo[0])] = hex($foo[5]);} for ($i = 0; $i < 0x500; $i += 8) { for ($j = $i; $j < $i + 8; $j++) { printf("0x%04x, ", $base[$j] ? $base[$j] : $j)}; print "\n"; }'
-
 }