Add custom mapping files used for generating some of the charmap

codecs.

3 files changed, 873 insertions, 0 deletions

diff --git a/Tools/unicode/python-mappings/CP1140.TXT b/Tools/unicode/python-mappings/CP1140.TXT
new file mode 100644
index 0000000..ede5cbf
--- /dev/null
+++ b/Tools/unicode/python-mappings/CP1140.TXT
@@ -0,0 +1,291 @@
+#
+#       Name:             CP1140
+#	Unicode version:  3.2
+#	Table version:    1.0
+#	Table format:     Format A
+#	Date:             2005-10-25
+#	Authors:          Marc-Andre Lemburg <mal@egenix.com>
+#
+#       This encoding is a modified CP037 encoding (with added Euro
+#       currency sign).
+#
+#       (c) Copyright Marc-Andre Lemburg, 2005.
+#           Licensed to PSF under a Contributor Agreement.
+#
+#       Based on the file 
+#       ftp://ftp.unicode.org/Public/MAPPINGS/VENDORS/MICSFT/EBCDIC/CP037.TXT
+#       which is:
+#
+#	Copyright (c) 2002 Unicode, Inc.  All Rights reserved.
+#
+#	This file is provided as-is by Unicode, Inc. (The Unicode Consortium).
+#	No claims are made as to fitness for any particular purpose.  No
+#	warranties of any kind are expressed or implied.  The recipient
+#	agrees to determine applicability of information provided.  If this
+#	file has been provided on optical media by Unicode, Inc., the sole
+#	remedy for any claim will be exchange of defective media within 90
+#	days of receipt.
+#
+#	Unicode, Inc. hereby grants the right to freely use the information
+#	supplied in this file in the creation of products supporting the
+#	Unicode Standard, and to make copies of this file in any form for
+#	internal or external distribution as long as this notice remains
+#	attached.
+#
+0x00	0x0000	#NULL
+0x01	0x0001	#START OF HEADING
+0x02	0x0002	#START OF TEXT
+0x03	0x0003	#END OF TEXT
+0x04	0x009C	#CONTROL
+0x05	0x0009	#HORIZONTAL TABULATION
+0x06	0x0086	#CONTROL
+0x07	0x007F	#DELETE
+0x08	0x0097	#CONTROL
+0x09	0x008D	#CONTROL
+0x0A	0x008E	#CONTROL
+0x0B	0x000B	#VERTICAL TABULATION
+0x0C	0x000C	#FORM FEED
+0x0D	0x000D	#CARRIAGE RETURN
+0x0E	0x000E	#SHIFT OUT
+0x0F	0x000F	#SHIFT IN
+0x10	0x0010	#DATA LINK ESCAPE
+0x11	0x0011	#DEVICE CONTROL ONE
+0x12	0x0012	#DEVICE CONTROL TWO
+0x13	0x0013	#DEVICE CONTROL THREE
+0x14	0x009D	#CONTROL
+0x15	0x0085	#CONTROL
+0x16	0x0008	#BACKSPACE
+0x17	0x0087	#CONTROL
+0x18	0x0018	#CANCEL
+0x19	0x0019	#END OF MEDIUM
+0x1A	0x0092	#CONTROL
+0x1B	0x008F	#CONTROL
+0x1C	0x001C	#FILE SEPARATOR
+0x1D	0x001D	#GROUP SEPARATOR
+0x1E	0x001E	#RECORD SEPARATOR
+0x1F	0x001F	#UNIT SEPARATOR
+0x20	0x0080	#CONTROL
+0x21	0x0081	#CONTROL
+0x22	0x0082	#CONTROL
+0x23	0x0083	#CONTROL
+0x24	0x0084	#CONTROL
+0x25	0x000A	#LINE FEED
+0x26	0x0017	#END OF TRANSMISSION BLOCK
+0x27	0x001B	#ESCAPE
+0x28	0x0088	#CONTROL
+0x29	0x0089	#CONTROL
+0x2A	0x008A	#CONTROL
+0x2B	0x008B	#CONTROL
+0x2C	0x008C	#CONTROL
+0x2D	0x0005	#ENQUIRY
+0x2E	0x0006	#ACKNOWLEDGE
+0x2F	0x0007	#BELL
+0x30	0x0090	#CONTROL
+0x31	0x0091	#CONTROL
+0x32	0x0016	#SYNCHRONOUS IDLE
+0x33	0x0093	#CONTROL
+0x34	0x0094	#CONTROL
+0x35	0x0095	#CONTROL
+0x36	0x0096	#CONTROL
+0x37	0x0004	#END OF TRANSMISSION
+0x38	0x0098	#CONTROL
+0x39	0x0099	#CONTROL
+0x3A	0x009A	#CONTROL
+0x3B	0x009B	#CONTROL
+0x3C	0x0014	#DEVICE CONTROL FOUR
+0x3D	0x0015	#NEGATIVE ACKNOWLEDGE
+0x3E	0x009E	#CONTROL
+0x3F	0x001A	#SUBSTITUTE
+0x40	0x0020	#SPACE
+0x41	0x00A0	#NO-BREAK SPACE
+0x42	0x00E2	#LATIN SMALL LETTER A WITH CIRCUMFLEX
+0x43	0x00E4	#LATIN SMALL LETTER A WITH DIAERESIS
+0x44	0x00E0	#LATIN SMALL LETTER A WITH GRAVE
+0x45	0x00E1	#LATIN SMALL LETTER A WITH ACUTE
+0x46	0x00E3	#LATIN SMALL LETTER A WITH TILDE
+0x47	0x00E5	#LATIN SMALL LETTER A WITH RING ABOVE
+0x48	0x00E7	#LATIN SMALL LETTER C WITH CEDILLA
+0x49	0x00F1	#LATIN SMALL LETTER N WITH TILDE
+0x4A	0x00A2	#CENT SIGN
+0x4B	0x002E	#FULL STOP
+0x4C	0x003C	#LESS-THAN SIGN
+0x4D	0x0028	#LEFT PARENTHESIS
+0x4E	0x002B	#PLUS SIGN
+0x4F	0x007C	#VERTICAL LINE
+0x50	0x0026	#AMPERSAND
+0x51	0x00E9	#LATIN SMALL LETTER E WITH ACUTE
+0x52	0x00EA	#LATIN SMALL LETTER E WITH CIRCUMFLEX
+0x53	0x00EB	#LATIN SMALL LETTER E WITH DIAERESIS
+0x54	0x00E8	#LATIN SMALL LETTER E WITH GRAVE
+0x55	0x00ED	#LATIN SMALL LETTER I WITH ACUTE
+0x56	0x00EE	#LATIN SMALL LETTER I WITH CIRCUMFLEX
+0x57	0x00EF	#LATIN SMALL LETTER I WITH DIAERESIS
+0x58	0x00EC	#LATIN SMALL LETTER I WITH GRAVE
+0x59	0x00DF	#LATIN SMALL LETTER SHARP S (GERMAN)
+0x5A	0x0021	#EXCLAMATION MARK
+0x5B	0x0024	#DOLLAR SIGN
+0x5C	0x002A	#ASTERISK
+0x5D	0x0029	#RIGHT PARENTHESIS
+0x5E	0x003B	#SEMICOLON
+0x5F	0x00AC	#NOT SIGN
+0x60	0x002D	#HYPHEN-MINUS
+0x61	0x002F	#SOLIDUS
+0x62	0x00C2	#LATIN CAPITAL LETTER A WITH CIRCUMFLEX
+0x63	0x00C4	#LATIN CAPITAL LETTER A WITH DIAERESIS
+0x64	0x00C0	#LATIN CAPITAL LETTER A WITH GRAVE
+0x65	0x00C1	#LATIN CAPITAL LETTER A WITH ACUTE
+0x66	0x00C3	#LATIN CAPITAL LETTER A WITH TILDE
+0x67	0x00C5	#LATIN CAPITAL LETTER A WITH RING ABOVE
+0x68	0x00C7	#LATIN CAPITAL LETTER C WITH CEDILLA
+0x69	0x00D1	#LATIN CAPITAL LETTER N WITH TILDE
+0x6A	0x00A6	#BROKEN BAR
+0x6B	0x002C	#COMMA
+0x6C	0x0025	#PERCENT SIGN
+0x6D	0x005F	#LOW LINE
+0x6E	0x003E	#GREATER-THAN SIGN
+0x6F	0x003F	#QUESTION MARK
+0x70	0x00F8	#LATIN SMALL LETTER O WITH STROKE
+0x71	0x00C9	#LATIN CAPITAL LETTER E WITH ACUTE
+0x72	0x00CA	#LATIN CAPITAL LETTER E WITH CIRCUMFLEX
+0x73	0x00CB	#LATIN CAPITAL LETTER E WITH DIAERESIS
+0x74	0x00C8	#LATIN CAPITAL LETTER E WITH GRAVE
+0x75	0x00CD	#LATIN CAPITAL LETTER I WITH ACUTE
+0x76	0x00CE	#LATIN CAPITAL LETTER I WITH CIRCUMFLEX
+0x77	0x00CF	#LATIN CAPITAL LETTER I WITH DIAERESIS
+0x78	0x00CC	#LATIN CAPITAL LETTER I WITH GRAVE
+0x79	0x0060	#GRAVE ACCENT
+0x7A	0x003A	#COLON
+0x7B	0x0023	#NUMBER SIGN
+0x7C	0x0040	#COMMERCIAL AT
+0x7D	0x0027	#APOSTROPHE
+0x7E	0x003D	#EQUALS SIGN
+0x7F	0x0022	#QUOTATION MARK
+0x80	0x00D8	#LATIN CAPITAL LETTER O WITH STROKE
+0x81	0x0061	#LATIN SMALL LETTER A
+0x82	0x0062	#LATIN SMALL LETTER B
+0x83	0x0063	#LATIN SMALL LETTER C
+0x84	0x0064	#LATIN SMALL LETTER D
+0x85	0x0065	#LATIN SMALL LETTER E
+0x86	0x0066	#LATIN SMALL LETTER F
+0x87	0x0067	#LATIN SMALL LETTER G
+0x88	0x0068	#LATIN SMALL LETTER H
+0x89	0x0069	#LATIN SMALL LETTER I
+0x8A	0x00AB	#LEFT-POINTING DOUBLE ANGLE QUOTATION MARK
+0x8B	0x00BB	#RIGHT-POINTING DOUBLE ANGLE QUOTATION MARK
+0x8C	0x00F0	#LATIN SMALL LETTER ETH (ICELANDIC)
+0x8D	0x00FD	#LATIN SMALL LETTER Y WITH ACUTE
+0x8E	0x00FE	#LATIN SMALL LETTER THORN (ICELANDIC)
+0x8F	0x00B1	#PLUS-MINUS SIGN
+0x90	0x00B0	#DEGREE SIGN
+0x91	0x006A	#LATIN SMALL LETTER J
+0x92	0x006B	#LATIN SMALL LETTER K
+0x93	0x006C	#LATIN SMALL LETTER L
+0x94	0x006D	#LATIN SMALL LETTER M
+0x95	0x006E	#LATIN SMALL LETTER N
+0x96	0x006F	#LATIN SMALL LETTER O
+0x97	0x0070	#LATIN SMALL LETTER P
+0x98	0x0071	#LATIN SMALL LETTER Q
+0x99	0x0072	#LATIN SMALL LETTER R
+0x9A	0x00AA	#FEMININE ORDINAL INDICATOR
+0x9B	0x00BA	#MASCULINE ORDINAL INDICATOR
+0x9C	0x00E6	#LATIN SMALL LIGATURE AE
+0x9D	0x00B8	#CEDILLA
+0x9E	0x00C6	#LATIN CAPITAL LIGATURE AE
+#0x9F	0x00A4	#CURRENCY SIGN
+0x9F    0x20AC   # EURO SIGN
+0xA0	0x00B5	#MICRO SIGN
+0xA1	0x007E	#TILDE
+0xA2	0x0073	#LATIN SMALL LETTER S
+0xA3	0x0074	#LATIN SMALL LETTER T
+0xA4	0x0075	#LATIN SMALL LETTER U
+0xA5	0x0076	#LATIN SMALL LETTER V
+0xA6	0x0077	#LATIN SMALL LETTER W
+0xA7	0x0078	#LATIN SMALL LETTER X
+0xA8	0x0079	#LATIN SMALL LETTER Y
+0xA9	0x007A	#LATIN SMALL LETTER Z
+0xAA	0x00A1	#INVERTED EXCLAMATION MARK
+0xAB	0x00BF	#INVERTED QUESTION MARK
+0xAC	0x00D0	#LATIN CAPITAL LETTER ETH (ICELANDIC)
+0xAD	0x00DD	#LATIN CAPITAL LETTER Y WITH ACUTE
+0xAE	0x00DE	#LATIN CAPITAL LETTER THORN (ICELANDIC)
+0xAF	0x00AE	#REGISTERED SIGN
+0xB0	0x005E	#CIRCUMFLEX ACCENT
+0xB1	0x00A3	#POUND SIGN
+0xB2	0x00A5	#YEN SIGN
+0xB3	0x00B7	#MIDDLE DOT
+0xB4	0x00A9	#COPYRIGHT SIGN
+0xB5	0x00A7	#SECTION SIGN
+0xB6	0x00B6	#PILCROW SIGN
+0xB7	0x00BC	#VULGAR FRACTION ONE QUARTER
+0xB8	0x00BD	#VULGAR FRACTION ONE HALF
+0xB9	0x00BE	#VULGAR FRACTION THREE QUARTERS
+0xBA	0x005B	#LEFT SQUARE BRACKET
+0xBB	0x005D	#RIGHT SQUARE BRACKET
+0xBC	0x00AF	#MACRON
+0xBD	0x00A8	#DIAERESIS
+0xBE	0x00B4	#ACUTE ACCENT
+0xBF	0x00D7	#MULTIPLICATION SIGN
+0xC0	0x007B	#LEFT CURLY BRACKET
+0xC1	0x0041	#LATIN CAPITAL LETTER A
+0xC2	0x0042	#LATIN CAPITAL LETTER B
+0xC3	0x0043	#LATIN CAPITAL LETTER C
+0xC4	0x0044	#LATIN CAPITAL LETTER D
+0xC5	0x0045	#LATIN CAPITAL LETTER E
+0xC6	0x0046	#LATIN CAPITAL LETTER F
+0xC7	0x0047	#LATIN CAPITAL LETTER G
+0xC8	0x0048	#LATIN CAPITAL LETTER H
+0xC9	0x0049	#LATIN CAPITAL LETTER I
+0xCA	0x00AD	#SOFT HYPHEN
+0xCB	0x00F4	#LATIN SMALL LETTER O WITH CIRCUMFLEX
+0xCC	0x00F6	#LATIN SMALL LETTER O WITH DIAERESIS
+0xCD	0x00F2	#LATIN SMALL LETTER O WITH GRAVE
+0xCE	0x00F3	#LATIN SMALL LETTER O WITH ACUTE
+0xCF	0x00F5	#LATIN SMALL LETTER O WITH TILDE
+0xD0	0x007D	#RIGHT CURLY BRACKET
+0xD1	0x004A	#LATIN CAPITAL LETTER J
+0xD2	0x004B	#LATIN CAPITAL LETTER K
+0xD3	0x004C	#LATIN CAPITAL LETTER L
+0xD4	0x004D	#LATIN CAPITAL LETTER M
+0xD5	0x004E	#LATIN CAPITAL LETTER N
+0xD6	0x004F	#LATIN CAPITAL LETTER O
+0xD7	0x0050	#LATIN CAPITAL LETTER P
+0xD8	0x0051	#LATIN CAPITAL LETTER Q
+0xD9	0x0052	#LATIN CAPITAL LETTER R
+0xDA	0x00B9	#SUPERSCRIPT ONE
+0xDB	0x00FB	#LATIN SMALL LETTER U WITH CIRCUMFLEX
+0xDC	0x00FC	#LATIN SMALL LETTER U WITH DIAERESIS
+0xDD	0x00F9	#LATIN SMALL LETTER U WITH GRAVE
+0xDE	0x00FA	#LATIN SMALL LETTER U WITH ACUTE
+0xDF	0x00FF	#LATIN SMALL LETTER Y WITH DIAERESIS
+0xE0	0x005C	#REVERSE SOLIDUS
+0xE1	0x00F7	#DIVISION SIGN
+0xE2	0x0053	#LATIN CAPITAL LETTER S
+0xE3	0x0054	#LATIN CAPITAL LETTER T
+0xE4	0x0055	#LATIN CAPITAL LETTER U
+0xE5	0x0056	#LATIN CAPITAL LETTER V
+0xE6	0x0057	#LATIN CAPITAL LETTER W
+0xE7	0x0058	#LATIN CAPITAL LETTER X
+0xE8	0x0059	#LATIN CAPITAL LETTER Y
+0xE9	0x005A	#LATIN CAPITAL LETTER Z
+0xEA	0x00B2	#SUPERSCRIPT TWO
+0xEB	0x00D4	#LATIN CAPITAL LETTER O WITH CIRCUMFLEX
+0xEC	0x00D6	#LATIN CAPITAL LETTER O WITH DIAERESIS
+0xED	0x00D2	#LATIN CAPITAL LETTER O WITH GRAVE
+0xEE	0x00D3	#LATIN CAPITAL LETTER O WITH ACUTE
+0xEF	0x00D5	#LATIN CAPITAL LETTER O WITH TILDE
+0xF0	0x0030	#DIGIT ZERO
+0xF1	0x0031	#DIGIT ONE
+0xF2	0x0032	#DIGIT TWO
+0xF3	0x0033	#DIGIT THREE
+0xF4	0x0034	#DIGIT FOUR
+0xF5	0x0035	#DIGIT FIVE
+0xF6	0x0036	#DIGIT SIX
+0xF7	0x0037	#DIGIT SEVEN
+0xF8	0x0038	#DIGIT EIGHT
+0xF9	0x0039	#DIGIT NINE
+0xFA	0x00B3	#SUPERSCRIPT THREE
+0xFB	0x00DB	#LATIN CAPITAL LETTER U WITH CIRCUMFLEX
+0xFC	0x00DC	#LATIN CAPITAL LETTER U WITH DIAERESIS
+0xFD	0x00D9	#LATIN CAPITAL LETTER U WITH GRAVE
+0xFE	0x00DA	#LATIN CAPITAL LETTER U WITH ACUTE
+0xFF	0x009F	#CONTROL
diff --git a/Tools/unicode/python-mappings/KOI8-U.TXT b/Tools/unicode/python-mappings/KOI8-U.TXT
new file mode 100644
index 0000000..fe9bf6a
--- /dev/null
+++ b/Tools/unicode/python-mappings/KOI8-U.TXT
@@ -0,0 +1,298 @@
+#
+#	Name:             KOI8-U (RFC2319) to Unicode
+#	Unicode version:  3.2
+#	Table version:    1.0
+#	Table format:     Format A
+#	Date:             2005-10-25
+#	Authors:          Marc-Andre Lemburg <mal@egenix.com>
+#
+#       See RFC2319 for details. This encoding is a modified KOI8-R
+#       encoding.
+#
+#       (c) Copyright Marc-Andre Lemburg, 2005.
+#           Licensed to PSF under a Contributor Agreement.
+#
+#       Based on the file 
+#       ftp://ftp.unicode.org/Public/MAPPINGS/VENDORS/MISC/KOI8-R.TXT
+#       which is:
+#
+#	Copyright (c) 1991-1999 Unicode, Inc.  All Rights reserved.
+#
+#	This file is provided as-is by Unicode, Inc. (The Unicode Consortium).
+#	No claims are made as to fitness for any particular purpose.  No
+#	warranties of any kind are expressed or implied.  The recipient
+#	agrees to determine applicability of information provided.  If this
+#	file has been provided on optical media by Unicode, Inc., the sole
+#	remedy for any claim will be exchange of defective media within 90
+#	days of receipt.
+#
+#	Unicode, Inc. hereby grants the right to freely use the information
+#	supplied in this file in the creation of products supporting the
+#	Unicode Standard, and to make copies of this file in any form for
+#	internal or external distribution as long as this notice remains
+#	attached.
+#
+0x00	0x0000	#	NULL
+0x01	0x0001	#	START OF HEADING
+0x02	0x0002	#	START OF TEXT
+0x03	0x0003	#	END OF TEXT
+0x04	0x0004	#	END OF TRANSMISSION
+0x05	0x0005	#	ENQUIRY
+0x06	0x0006	#	ACKNOWLEDGE
+0x07	0x0007	#	BELL
+0x08	0x0008	#	BACKSPACE
+0x09	0x0009	#	HORIZONTAL TABULATION
+0x0A	0x000A	#	LINE FEED
+0x0B	0x000B	#	VERTICAL TABULATION
+0x0C	0x000C	#	FORM FEED
+0x0D	0x000D	#	CARRIAGE RETURN
+0x0E	0x000E	#	SHIFT OUT
+0x0F	0x000F	#	SHIFT IN
+0x10	0x0010	#	DATA LINK ESCAPE
+0x11	0x0011	#	DEVICE CONTROL ONE
+0x12	0x0012	#	DEVICE CONTROL TWO
+0x13	0x0013	#	DEVICE CONTROL THREE
+0x14	0x0014	#	DEVICE CONTROL FOUR
+0x15	0x0015	#	NEGATIVE ACKNOWLEDGE
+0x16	0x0016	#	SYNCHRONOUS IDLE
+0x17	0x0017	#	END OF TRANSMISSION BLOCK
+0x18	0x0018	#	CANCEL
+0x19	0x0019	#	END OF MEDIUM
+0x1A	0x001A	#	SUBSTITUTE
+0x1B	0x001B	#	ESCAPE
+0x1C	0x001C	#	FILE SEPARATOR
+0x1D	0x001D	#	GROUP SEPARATOR
+0x1E	0x001E	#	RECORD SEPARATOR
+0x1F	0x001F	#	UNIT SEPARATOR
+0x20	0x0020	#	SPACE
+0x21	0x0021	#	EXCLAMATION MARK
+0x22	0x0022	#	QUOTATION MARK
+0x23	0x0023	#	NUMBER SIGN
+0x24	0x0024	#	DOLLAR SIGN
+0x25	0x0025	#	PERCENT SIGN
+0x26	0x0026	#	AMPERSAND
+0x27	0x0027	#	APOSTROPHE
+0x28	0x0028	#	LEFT PARENTHESIS
+0x29	0x0029	#	RIGHT PARENTHESIS
+0x2A	0x002A	#	ASTERISK
+0x2B	0x002B	#	PLUS SIGN
+0x2C	0x002C	#	COMMA
+0x2D	0x002D	#	HYPHEN-MINUS
+0x2E	0x002E	#	FULL STOP
+0x2F	0x002F	#	SOLIDUS
+0x30	0x0030	#	DIGIT ZERO
+0x31	0x0031	#	DIGIT ONE
+0x32	0x0032	#	DIGIT TWO
+0x33	0x0033	#	DIGIT THREE
+0x34	0x0034	#	DIGIT FOUR
+0x35	0x0035	#	DIGIT FIVE
+0x36	0x0036	#	DIGIT SIX
+0x37	0x0037	#	DIGIT SEVEN
+0x38	0x0038	#	DIGIT EIGHT
+0x39	0x0039	#	DIGIT NINE
+0x3A	0x003A	#	COLON
+0x3B	0x003B	#	SEMICOLON
+0x3C	0x003C	#	LESS-THAN SIGN
+0x3D	0x003D	#	EQUALS SIGN
+0x3E	0x003E	#	GREATER-THAN SIGN
+0x3F	0x003F	#	QUESTION MARK
+0x40	0x0040	#	COMMERCIAL AT
+0x41	0x0041	#	LATIN CAPITAL LETTER A
+0x42	0x0042	#	LATIN CAPITAL LETTER B
+0x43	0x0043	#	LATIN CAPITAL LETTER C
+0x44	0x0044	#	LATIN CAPITAL LETTER D
+0x45	0x0045	#	LATIN CAPITAL LETTER E
+0x46	0x0046	#	LATIN CAPITAL LETTER F
+0x47	0x0047	#	LATIN CAPITAL LETTER G
+0x48	0x0048	#	LATIN CAPITAL LETTER H
+0x49	0x0049	#	LATIN CAPITAL LETTER I
+0x4A	0x004A	#	LATIN CAPITAL LETTER J
+0x4B	0x004B	#	LATIN CAPITAL LETTER K
+0x4C	0x004C	#	LATIN CAPITAL LETTER L
+0x4D	0x004D	#	LATIN CAPITAL LETTER M
+0x4E	0x004E	#	LATIN CAPITAL LETTER N
+0x4F	0x004F	#	LATIN CAPITAL LETTER O
+0x50	0x0050	#	LATIN CAPITAL LETTER P
+0x51	0x0051	#	LATIN CAPITAL LETTER Q
+0x52	0x0052	#	LATIN CAPITAL LETTER R
+0x53	0x0053	#	LATIN CAPITAL LETTER S
+0x54	0x0054	#	LATIN CAPITAL LETTER T
+0x55	0x0055	#	LATIN CAPITAL LETTER U
+0x56	0x0056	#	LATIN CAPITAL LETTER V
+0x57	0x0057	#	LATIN CAPITAL LETTER W
+0x58	0x0058	#	LATIN CAPITAL LETTER X
+0x59	0x0059	#	LATIN CAPITAL LETTER Y
+0x5A	0x005A	#	LATIN CAPITAL LETTER Z
+0x5B	0x005B	#	LEFT SQUARE BRACKET
+0x5C	0x005C	#	REVERSE SOLIDUS
+0x5D	0x005D	#	RIGHT SQUARE BRACKET
+0x5E	0x005E	#	CIRCUMFLEX ACCENT
+0x5F	0x005F	#	LOW LINE
+0x60	0x0060	#	GRAVE ACCENT
+0x61	0x0061	#	LATIN SMALL LETTER A
+0x62	0x0062	#	LATIN SMALL LETTER B
+0x63	0x0063	#	LATIN SMALL LETTER C
+0x64	0x0064	#	LATIN SMALL LETTER D
+0x65	0x0065	#	LATIN SMALL LETTER E
+0x66	0x0066	#	LATIN SMALL LETTER F
+0x67	0x0067	#	LATIN SMALL LETTER G
+0x68	0x0068	#	LATIN SMALL LETTER H
+0x69	0x0069	#	LATIN SMALL LETTER I
+0x6A	0x006A	#	LATIN SMALL LETTER J
+0x6B	0x006B	#	LATIN SMALL LETTER K
+0x6C	0x006C	#	LATIN SMALL LETTER L
+0x6D	0x006D	#	LATIN SMALL LETTER M
+0x6E	0x006E	#	LATIN SMALL LETTER N
+0x6F	0x006F	#	LATIN SMALL LETTER O
+0x70	0x0070	#	LATIN SMALL LETTER P
+0x71	0x0071	#	LATIN SMALL LETTER Q
+0x72	0x0072	#	LATIN SMALL LETTER R
+0x73	0x0073	#	LATIN SMALL LETTER S
+0x74	0x0074	#	LATIN SMALL LETTER T
+0x75	0x0075	#	LATIN SMALL LETTER U
+0x76	0x0076	#	LATIN SMALL LETTER V
+0x77	0x0077	#	LATIN SMALL LETTER W
+0x78	0x0078	#	LATIN SMALL LETTER X
+0x79	0x0079	#	LATIN SMALL LETTER Y
+0x7A	0x007A	#	LATIN SMALL LETTER Z
+0x7B	0x007B	#	LEFT CURLY BRACKET
+0x7C	0x007C	#	VERTICAL LINE
+0x7D	0x007D	#	RIGHT CURLY BRACKET
+0x7E	0x007E	#	TILDE
+0x7F	0x007F	#	DELETE
+0x80	0x2500	#	BOX DRAWINGS LIGHT HORIZONTAL
+0x81	0x2502	#	BOX DRAWINGS LIGHT VERTICAL
+0x82	0x250C	#	BOX DRAWINGS LIGHT DOWN AND RIGHT
+0x83	0x2510	#	BOX DRAWINGS LIGHT DOWN AND LEFT
+0x84	0x2514	#	BOX DRAWINGS LIGHT UP AND RIGHT
+0x85	0x2518	#	BOX DRAWINGS LIGHT UP AND LEFT
+0x86	0x251C	#	BOX DRAWINGS LIGHT VERTICAL AND RIGHT
+0x87	0x2524	#	BOX DRAWINGS LIGHT VERTICAL AND LEFT
+0x88	0x252C	#	BOX DRAWINGS LIGHT DOWN AND HORIZONTAL
+0x89	0x2534	#	BOX DRAWINGS LIGHT UP AND HORIZONTAL
+0x8A	0x253C	#	BOX DRAWINGS LIGHT VERTICAL AND HORIZONTAL
+0x8B	0x2580	#	UPPER HALF BLOCK
+0x8C	0x2584	#	LOWER HALF BLOCK
+0x8D	0x2588	#	FULL BLOCK
+0x8E	0x258C	#	LEFT HALF BLOCK
+0x8F	0x2590	#	RIGHT HALF BLOCK
+0x90	0x2591	#	LIGHT SHADE
+0x91	0x2592	#	MEDIUM SHADE
+0x92	0x2593	#	DARK SHADE
+0x93	0x2320	#	TOP HALF INTEGRAL
+0x94	0x25A0	#	BLACK SQUARE
+0x95	0x2219	#	BULLET OPERATOR
+0x96	0x221A	#	SQUARE ROOT
+0x97	0x2248	#	ALMOST EQUAL TO
+0x98	0x2264	#	LESS-THAN OR EQUAL TO
+0x99	0x2265	#	GREATER-THAN OR EQUAL TO
+0x9A	0x00A0	#	NO-BREAK SPACE
+0x9B	0x2321	#	BOTTOM HALF INTEGRAL
+0x9C	0x00B0	#	DEGREE SIGN
+0x9D	0x00B2	#	SUPERSCRIPT TWO
+0x9E	0x00B7	#	MIDDLE DOT
+0x9F	0x00F7	#	DIVISION SIGN
+0xA0	0x2550	#	BOX DRAWINGS DOUBLE HORIZONTAL
+0xA1	0x2551	#	BOX DRAWINGS DOUBLE VERTICAL
+0xA2	0x2552	#	BOX DRAWINGS DOWN SINGLE AND RIGHT DOUBLE
+0xA3	0x0451	#	CYRILLIC SMALL LETTER IO
+#0xA4	0x2553	#	BOX DRAWINGS DOWN DOUBLE AND RIGHT SINGLE
+0xA4    0x0454  #       CYRILLIC SMALL LETTER UKRAINIAN IE
+0xA5	0x2554	#	BOX DRAWINGS DOUBLE DOWN AND RIGHT
+#0xA6	0x2555	#	BOX DRAWINGS DOWN SINGLE AND LEFT DOUBLE
+0xA6    0x0456  #       CYRILLIC SMALL LETTER BYELORUSSIAN-UKRAINIAN I
+#0xA7	0x2556	#	BOX DRAWINGS DOWN DOUBLE AND LEFT SINGLE
+0xA7    0x0457  #       CYRILLIC SMALL LETTER YI (UKRAINIAN)
+0xA8	0x2557	#	BOX DRAWINGS DOUBLE DOWN AND LEFT
+0xA9	0x2558	#	BOX DRAWINGS UP SINGLE AND RIGHT DOUBLE
+0xAA	0x2559	#	BOX DRAWINGS UP DOUBLE AND RIGHT SINGLE
+0xAB	0x255A	#	BOX DRAWINGS DOUBLE UP AND RIGHT
+0xAC	0x255B	#	BOX DRAWINGS UP SINGLE AND LEFT DOUBLE
+#0xAD	0x255C	#	BOX DRAWINGS UP DOUBLE AND LEFT SINGLE
+0xAD    0x0491  #       CYRILLIC SMALL LETTER UKRAINIAN GHE WITH UPTURN
+0xAE	0x255D	#	BOX DRAWINGS DOUBLE UP AND LEFT
+0xAF	0x255E	#	BOX DRAWINGS VERTICAL SINGLE AND RIGHT DOUBLE
+0xB0	0x255F	#	BOX DRAWINGS VERTICAL DOUBLE AND RIGHT SINGLE
+0xB1	0x2560	#	BOX DRAWINGS DOUBLE VERTICAL AND RIGHT
+0xB2	0x2561	#	BOX DRAWINGS VERTICAL SINGLE AND LEFT DOUBLE
+0xB3	0x0401	#	CYRILLIC CAPITAL LETTER IO
+#0xB4	0x2562	#	BOX DRAWINGS VERTICAL DOUBLE AND LEFT SINGLE
+0xB4    0x0404  #       CYRILLIC CAPITAL LETTER UKRAINIAN IE
+0xB5	0x2563	#	BOX DRAWINGS DOUBLE VERTICAL AND LEFT
+#0xB6	0x2564	#	BOX DRAWINGS DOWN SINGLE AND HORIZONTAL DOUBLE
+0xB6    0x0406  #       CYRILLIC CAPITAL LETTER BYELORUSSIAN-UKRAINIAN I
+#0xB7	0x2565	#	BOX DRAWINGS DOWN DOUBLE AND HORIZONTAL SINGLE
+0xB7    0x0407  #       CYRILLIC CAPITAL LETTER YI (UKRAINIAN)
+0xB8	0x2566	#	BOX DRAWINGS DOUBLE DOWN AND HORIZONTAL
+0xB9	0x2567	#	BOX DRAWINGS UP SINGLE AND HORIZONTAL DOUBLE
+0xBA	0x2568	#	BOX DRAWINGS UP DOUBLE AND HORIZONTAL SINGLE
+0xBB	0x2569	#	BOX DRAWINGS DOUBLE UP AND HORIZONTAL
+0xBC	0x256A	#	BOX DRAWINGS VERTICAL SINGLE AND HORIZONTAL DOUBLE
+#0xBD	0x256B	#	BOX DRAWINGS VERTICAL DOUBLE AND HORIZONTAL SINGLE
+0xBD    0x0490  #       CYRILLIC CAPITAL LETTER UKRAINIAN GHE WITH UPTURN
+0xBE	0x256C	#	BOX DRAWINGS DOUBLE VERTICAL AND HORIZONTAL
+0xBF	0x00A9	#	COPYRIGHT SIGN
+0xC0	0x044E	#	CYRILLIC SMALL LETTER YU
+0xC1	0x0430	#	CYRILLIC SMALL LETTER A
+0xC2	0x0431	#	CYRILLIC SMALL LETTER BE
+0xC3	0x0446	#	CYRILLIC SMALL LETTER TSE
+0xC4	0x0434	#	CYRILLIC SMALL LETTER DE
+0xC5	0x0435	#	CYRILLIC SMALL LETTER IE
+0xC6	0x0444	#	CYRILLIC SMALL LETTER EF
+0xC7	0x0433	#	CYRILLIC SMALL LETTER GHE
+0xC8	0x0445	#	CYRILLIC SMALL LETTER HA
+0xC9	0x0438	#	CYRILLIC SMALL LETTER I
+0xCA	0x0439	#	CYRILLIC SMALL LETTER SHORT I
+0xCB	0x043A	#	CYRILLIC SMALL LETTER KA
+0xCC	0x043B	#	CYRILLIC SMALL LETTER EL
+0xCD	0x043C	#	CYRILLIC SMALL LETTER EM
+0xCE	0x043D	#	CYRILLIC SMALL LETTER EN
+0xCF	0x043E	#	CYRILLIC SMALL LETTER O
+0xD0	0x043F	#	CYRILLIC SMALL LETTER PE
+0xD1	0x044F	#	CYRILLIC SMALL LETTER YA
+0xD2	0x0440	#	CYRILLIC SMALL LETTER ER
+0xD3	0x0441	#	CYRILLIC SMALL LETTER ES
+0xD4	0x0442	#	CYRILLIC SMALL LETTER TE
+0xD5	0x0443	#	CYRILLIC SMALL LETTER U
+0xD6	0x0436	#	CYRILLIC SMALL LETTER ZHE
+0xD7	0x0432	#	CYRILLIC SMALL LETTER VE
+0xD8	0x044C	#	CYRILLIC SMALL LETTER SOFT SIGN
+0xD9	0x044B	#	CYRILLIC SMALL LETTER YERU
+0xDA	0x0437	#	CYRILLIC SMALL LETTER ZE
+0xDB	0x0448	#	CYRILLIC SMALL LETTER SHA
+0xDC	0x044D	#	CYRILLIC SMALL LETTER E
+0xDD	0x0449	#	CYRILLIC SMALL LETTER SHCHA
+0xDE	0x0447	#	CYRILLIC SMALL LETTER CHE
+0xDF	0x044A	#	CYRILLIC SMALL LETTER HARD SIGN
+0xE0	0x042E	#	CYRILLIC CAPITAL LETTER YU
+0xE1	0x0410	#	CYRILLIC CAPITAL LETTER A
+0xE2	0x0411	#	CYRILLIC CAPITAL LETTER BE
+0xE3	0x0426	#	CYRILLIC CAPITAL LETTER TSE
+0xE4	0x0414	#	CYRILLIC CAPITAL LETTER DE
+0xE5	0x0415	#	CYRILLIC CAPITAL LETTER IE
+0xE6	0x0424	#	CYRILLIC CAPITAL LETTER EF
+0xE7	0x0413	#	CYRILLIC CAPITAL LETTER GHE
+0xE8	0x0425	#	CYRILLIC CAPITAL LETTER HA
+0xE9	0x0418	#	CYRILLIC CAPITAL LETTER I
+0xEA	0x0419	#	CYRILLIC CAPITAL LETTER SHORT I
+0xEB	0x041A	#	CYRILLIC CAPITAL LETTER KA
+0xEC	0x041B	#	CYRILLIC CAPITAL LETTER EL
+0xED	0x041C	#	CYRILLIC CAPITAL LETTER EM
+0xEE	0x041D	#	CYRILLIC CAPITAL LETTER EN
+0xEF	0x041E	#	CYRILLIC CAPITAL LETTER O
+0xF0	0x041F	#	CYRILLIC CAPITAL LETTER PE
+0xF1	0x042F	#	CYRILLIC CAPITAL LETTER YA
+0xF2	0x0420	#	CYRILLIC CAPITAL LETTER ER
+0xF3	0x0421	#	CYRILLIC CAPITAL LETTER ES
+0xF4	0x0422	#	CYRILLIC CAPITAL LETTER TE
+0xF5	0x0423	#	CYRILLIC CAPITAL LETTER U
+0xF6	0x0416	#	CYRILLIC CAPITAL LETTER ZHE
+0xF7	0x0412	#	CYRILLIC CAPITAL LETTER VE
+0xF8	0x042C	#	CYRILLIC CAPITAL LETTER SOFT SIGN
+0xF9	0x042B	#	CYRILLIC CAPITAL LETTER YERU
+0xFA	0x0417	#	CYRILLIC CAPITAL LETTER ZE
+0xFB	0x0428	#	CYRILLIC CAPITAL LETTER SHA
+0xFC	0x042D	#	CYRILLIC CAPITAL LETTER E
+0xFD	0x0429	#	CYRILLIC CAPITAL LETTER SHCHA
+0xFE	0x0427	#	CYRILLIC CAPITAL LETTER CHE
+0xFF	0x042A	#	CYRILLIC CAPITAL LETTER HARD SIGN
diff --git a/Tools/unicode/python-mappings/TIS-620.TXT b/Tools/unicode/python-mappings/TIS-620.TXT
new file mode 100644
index 0000000..c2e90c2
--- /dev/null
+++ b/Tools/unicode/python-mappings/TIS-620.TXT
@@ -0,0 +1,284 @@
+#
+#	Name:             TIS-620
+#	Unicode version:  3.2
+#	Table version:    1.0
+#	Table format:     Format A
+#	Date:             2005-10-25
+#	Authors:          Marc-Andre Lemburg <mal@egenix.com>
+#
+#       According to
+#       ftp://ftp.unicode.org/Public/MAPPINGS/ISO8859/8859-11.TXT the
+#       TIS-620 is the identical to ISO_8859-11 with the 0xA0
+#       (no-break space) mapping removed.
+#
+#       (c) Copyright Marc-Andre Lemburg, 2005.
+#           Licensed to PSF under a Contributor Agreement.
+#
+#       Based on the file 
+#       ftp://ftp.unicode.org/Public/MAPPINGS/ISO8859/8859-11.TXT
+#       which is:
+#
+#	Copyright (c) 2002 Unicode, Inc.  All Rights reserved.
+#
+#	This file is provided as-is by Unicode, Inc. (The Unicode Consortium).
+#	No claims are made as to fitness for any particular purpose.  No
+#	warranties of any kind are expressed or implied.  The recipient
+#	agrees to determine applicability of information provided.  If this
+#	file has been provided on optical media by Unicode, Inc., the sole
+#	remedy for any claim will be exchange of defective media within 90
+#	days of receipt.
+#
+#	Unicode, Inc. hereby grants the right to freely use the information
+#	supplied in this file in the creation of products supporting the
+#	Unicode Standard, and to make copies of this file in any form for
+#	internal or external distribution as long as this notice remains
+#	attached.
+#
+0x00	0x0000	#	NULL
+0x01	0x0001	#	START OF HEADING
+0x02	0x0002	#	START OF TEXT
+0x03	0x0003	#	END OF TEXT
+0x04	0x0004	#	END OF TRANSMISSION
+0x05	0x0005	#	ENQUIRY
+0x06	0x0006	#	ACKNOWLEDGE
+0x07	0x0007	#	BELL
+0x08	0x0008	#	BACKSPACE
+0x09	0x0009	#	HORIZONTAL TABULATION
+0x0A	0x000A	#	LINE FEED
+0x0B	0x000B	#	VERTICAL TABULATION
+0x0C	0x000C	#	FORM FEED
+0x0D	0x000D	#	CARRIAGE RETURN
+0x0E	0x000E	#	SHIFT OUT
+0x0F	0x000F	#	SHIFT IN
+0x10	0x0010	#	DATA LINK ESCAPE
+0x11	0x0011	#	DEVICE CONTROL ONE
+0x12	0x0012	#	DEVICE CONTROL TWO
+0x13	0x0013	#	DEVICE CONTROL THREE
+0x14	0x0014	#	DEVICE CONTROL FOUR
+0x15	0x0015	#	NEGATIVE ACKNOWLEDGE
+0x16	0x0016	#	SYNCHRONOUS IDLE
+0x17	0x0017	#	END OF TRANSMISSION BLOCK
+0x18	0x0018	#	CANCEL
+0x19	0x0019	#	END OF MEDIUM
+0x1A	0x001A	#	SUBSTITUTE
+0x1B	0x001B	#	ESCAPE
+0x1C	0x001C	#	FILE SEPARATOR
+0x1D	0x001D	#	GROUP SEPARATOR
+0x1E	0x001E	#	RECORD SEPARATOR
+0x1F	0x001F	#	UNIT SEPARATOR
+0x20	0x0020	#	SPACE
+0x21	0x0021	#	EXCLAMATION MARK
+0x22	0x0022	#	QUOTATION MARK
+0x23	0x0023	#	NUMBER SIGN
+0x24	0x0024	#	DOLLAR SIGN
+0x25	0x0025	#	PERCENT SIGN
+0x26	0x0026	#	AMPERSAND
+0x27	0x0027	#	APOSTROPHE
+0x28	0x0028	#	LEFT PARENTHESIS
+0x29	0x0029	#	RIGHT PARENTHESIS
+0x2A	0x002A	#	ASTERISK
+0x2B	0x002B	#	PLUS SIGN
+0x2C	0x002C	#	COMMA
+0x2D	0x002D	#	HYPHEN-MINUS
+0x2E	0x002E	#	FULL STOP
+0x2F	0x002F	#	SOLIDUS
+0x30	0x0030	#	DIGIT ZERO
+0x31	0x0031	#	DIGIT ONE
+0x32	0x0032	#	DIGIT TWO
+0x33	0x0033	#	DIGIT THREE
+0x34	0x0034	#	DIGIT FOUR
+0x35	0x0035	#	DIGIT FIVE
+0x36	0x0036	#	DIGIT SIX
+0x37	0x0037	#	DIGIT SEVEN
+0x38	0x0038	#	DIGIT EIGHT
+0x39	0x0039	#	DIGIT NINE
+0x3A	0x003A	#	COLON
+0x3B	0x003B	#	SEMICOLON
+0x3C	0x003C	#	LESS-THAN SIGN
+0x3D	0x003D	#	EQUALS SIGN
+0x3E	0x003E	#	GREATER-THAN SIGN
+0x3F	0x003F	#	QUESTION MARK
+0x40	0x0040	#	COMMERCIAL AT
+0x41	0x0041	#	LATIN CAPITAL LETTER A
+0x42	0x0042	#	LATIN CAPITAL LETTER B
+0x43	0x0043	#	LATIN CAPITAL LETTER C
+0x44	0x0044	#	LATIN CAPITAL LETTER D
+0x45	0x0045	#	LATIN CAPITAL LETTER E
+0x46	0x0046	#	LATIN CAPITAL LETTER F
+0x47	0x0047	#	LATIN CAPITAL LETTER G
+0x48	0x0048	#	LATIN CAPITAL LETTER H
+0x49	0x0049	#	LATIN CAPITAL LETTER I
+0x4A	0x004A	#	LATIN CAPITAL LETTER J
+0x4B	0x004B	#	LATIN CAPITAL LETTER K
+0x4C	0x004C	#	LATIN CAPITAL LETTER L
+0x4D	0x004D	#	LATIN CAPITAL LETTER M
+0x4E	0x004E	#	LATIN CAPITAL LETTER N
+0x4F	0x004F	#	LATIN CAPITAL LETTER O
+0x50	0x0050	#	LATIN CAPITAL LETTER P
+0x51	0x0051	#	LATIN CAPITAL LETTER Q
+0x52	0x0052	#	LATIN CAPITAL LETTER R
+0x53	0x0053	#	LATIN CAPITAL LETTER S
+0x54	0x0054	#	LATIN CAPITAL LETTER T
+0x55	0x0055	#	LATIN CAPITAL LETTER U
+0x56	0x0056	#	LATIN CAPITAL LETTER V
+0x57	0x0057	#	LATIN CAPITAL LETTER W
+0x58	0x0058	#	LATIN CAPITAL LETTER X
+0x59	0x0059	#	LATIN CAPITAL LETTER Y
+0x5A	0x005A	#	LATIN CAPITAL LETTER Z
+0x5B	0x005B	#	LEFT SQUARE BRACKET
+0x5C	0x005C	#	REVERSE SOLIDUS
+0x5D	0x005D	#	RIGHT SQUARE BRACKET
+0x5E	0x005E	#	CIRCUMFLEX ACCENT
+0x5F	0x005F	#	LOW LINE
+0x60	0x0060	#	GRAVE ACCENT
+0x61	0x0061	#	LATIN SMALL LETTER A
+0x62	0x0062	#	LATIN SMALL LETTER B
+0x63	0x0063	#	LATIN SMALL LETTER C
+0x64	0x0064	#	LATIN SMALL LETTER D
+0x65	0x0065	#	LATIN SMALL LETTER E
+0x66	0x0066	#	LATIN SMALL LETTER F
+0x67	0x0067	#	LATIN SMALL LETTER G
+0x68	0x0068	#	LATIN SMALL LETTER H
+0x69	0x0069	#	LATIN SMALL LETTER I
+0x6A	0x006A	#	LATIN SMALL LETTER J
+0x6B	0x006B	#	LATIN SMALL LETTER K
+0x6C	0x006C	#	LATIN SMALL LETTER L
+0x6D	0x006D	#	LATIN SMALL LETTER M
+0x6E	0x006E	#	LATIN SMALL LETTER N
+0x6F	0x006F	#	LATIN SMALL LETTER O
+0x70	0x0070	#	LATIN SMALL LETTER P
+0x71	0x0071	#	LATIN SMALL LETTER Q
+0x72	0x0072	#	LATIN SMALL LETTER R
+0x73	0x0073	#	LATIN SMALL LETTER S
+0x74	0x0074	#	LATIN SMALL LETTER T
+0x75	0x0075	#	LATIN SMALL LETTER U
+0x76	0x0076	#	LATIN SMALL LETTER V
+0x77	0x0077	#	LATIN SMALL LETTER W
+0x78	0x0078	#	LATIN SMALL LETTER X
+0x79	0x0079	#	LATIN SMALL LETTER Y
+0x7A	0x007A	#	LATIN SMALL LETTER Z
+0x7B	0x007B	#	LEFT CURLY BRACKET
+0x7C	0x007C	#	VERTICAL LINE
+0x7D	0x007D	#	RIGHT CURLY BRACKET
+0x7E	0x007E	#	TILDE
+0x7F	0x007F	#	DELETE
+0x80	0x0080	#	<control>
+0x81	0x0081	#	<control>
+0x82	0x0082	#	<control>
+0x83	0x0083	#	<control>
+0x84	0x0084	#	<control>
+0x85	0x0085	#	<control>
+0x86	0x0086	#	<control>
+0x87	0x0087	#	<control>
+0x88	0x0088	#	<control>
+0x89	0x0089	#	<control>
+0x8A	0x008A	#	<control>
+0x8B	0x008B	#	<control>
+0x8C	0x008C	#	<control>
+0x8D	0x008D	#	<control>
+0x8E	0x008E	#	<control>
+0x8F	0x008F	#	<control>
+0x90	0x0090	#	<control>
+0x91	0x0091	#	<control>
+0x92	0x0092	#	<control>
+0x93	0x0093	#	<control>
+0x94	0x0094	#	<control>
+0x95	0x0095	#	<control>
+0x96	0x0096	#	<control>
+0x97	0x0097	#	<control>
+0x98	0x0098	#	<control>
+0x99	0x0099	#	<control>
+0x9A	0x009A	#	<control>
+0x9B	0x009B	#	<control>
+0x9C	0x009C	#	<control>
+0x9D	0x009D	#	<control>
+0x9E	0x009E	#	<control>
+0x9F	0x009F	#	<control>
+#0xA0	0x00A0	#	NO-BREAK SPACE
+0xA1	0x0E01	#	THAI CHARACTER KO KAI
+0xA2	0x0E02	#	THAI CHARACTER KHO KHAI
+0xA3	0x0E03	#	THAI CHARACTER KHO KHUAT
+0xA4	0x0E04	#	THAI CHARACTER KHO KHWAI
+0xA5	0x0E05	#	THAI CHARACTER KHO KHON
+0xA6	0x0E06	#	THAI CHARACTER KHO RAKHANG
+0xA7	0x0E07	#	THAI CHARACTER NGO NGU
+0xA8	0x0E08	#	THAI CHARACTER CHO CHAN
+0xA9	0x0E09	#	THAI CHARACTER CHO CHING
+0xAA	0x0E0A	#	THAI CHARACTER CHO CHANG
+0xAB	0x0E0B	#	THAI CHARACTER SO SO
+0xAC	0x0E0C	#	THAI CHARACTER CHO CHOE
+0xAD	0x0E0D	#	THAI CHARACTER YO YING
+0xAE	0x0E0E	#	THAI CHARACTER DO CHADA
+0xAF	0x0E0F	#	THAI CHARACTER TO PATAK
+0xB0	0x0E10	#	THAI CHARACTER THO THAN
+0xB1	0x0E11	#	THAI CHARACTER THO NANGMONTHO
+0xB2	0x0E12	#	THAI CHARACTER THO PHUTHAO
+0xB3	0x0E13	#	THAI CHARACTER NO NEN
+0xB4	0x0E14	#	THAI CHARACTER DO DEK
+0xB5	0x0E15	#	THAI CHARACTER TO TAO
+0xB6	0x0E16	#	THAI CHARACTER THO THUNG
+0xB7	0x0E17	#	THAI CHARACTER THO THAHAN
+0xB8	0x0E18	#	THAI CHARACTER THO THONG
+0xB9	0x0E19	#	THAI CHARACTER NO NU
+0xBA	0x0E1A	#	THAI CHARACTER BO BAIMAI
+0xBB	0x0E1B	#	THAI CHARACTER PO PLA
+0xBC	0x0E1C	#	THAI CHARACTER PHO PHUNG
+0xBD	0x0E1D	#	THAI CHARACTER FO FA
+0xBE	0x0E1E	#	THAI CHARACTER PHO PHAN
+0xBF	0x0E1F	#	THAI CHARACTER FO FAN
+0xC0	0x0E20	#	THAI CHARACTER PHO SAMPHAO
+0xC1	0x0E21	#	THAI CHARACTER MO MA
+0xC2	0x0E22	#	THAI CHARACTER YO YAK
+0xC3	0x0E23	#	THAI CHARACTER RO RUA
+0xC4	0x0E24	#	THAI CHARACTER RU
+0xC5	0x0E25	#	THAI CHARACTER LO LING
+0xC6	0x0E26	#	THAI CHARACTER LU
+0xC7	0x0E27	#	THAI CHARACTER WO WAEN
+0xC8	0x0E28	#	THAI CHARACTER SO SALA
+0xC9	0x0E29	#	THAI CHARACTER SO RUSI
+0xCA	0x0E2A	#	THAI CHARACTER SO SUA
+0xCB	0x0E2B	#	THAI CHARACTER HO HIP
+0xCC	0x0E2C	#	THAI CHARACTER LO CHULA
+0xCD	0x0E2D	#	THAI CHARACTER O ANG
+0xCE	0x0E2E	#	THAI CHARACTER HO NOKHUK
+0xCF	0x0E2F	#	THAI CHARACTER PAIYANNOI
+0xD0	0x0E30	#	THAI CHARACTER SARA A
+0xD1	0x0E31	#	THAI CHARACTER MAI HAN-AKAT
+0xD2	0x0E32	#	THAI CHARACTER SARA AA
+0xD3	0x0E33	#	THAI CHARACTER SARA AM
+0xD4	0x0E34	#	THAI CHARACTER SARA I
+0xD5	0x0E35	#	THAI CHARACTER SARA II
+0xD6	0x0E36	#	THAI CHARACTER SARA UE
+0xD7	0x0E37	#	THAI CHARACTER SARA UEE
+0xD8	0x0E38	#	THAI CHARACTER SARA U
+0xD9	0x0E39	#	THAI CHARACTER SARA UU
+0xDA	0x0E3A	#	THAI CHARACTER PHINTHU
+0xDF	0x0E3F	#	THAI CURRENCY SYMBOL BAHT
+0xE0	0x0E40	#	THAI CHARACTER SARA E
+0xE1	0x0E41	#	THAI CHARACTER SARA AE
+0xE2	0x0E42	#	THAI CHARACTER SARA O
+0xE3	0x0E43	#	THAI CHARACTER SARA AI MAIMUAN
+0xE4	0x0E44	#	THAI CHARACTER SARA AI MAIMALAI
+0xE5	0x0E45	#	THAI CHARACTER LAKKHANGYAO
+0xE6	0x0E46	#	THAI CHARACTER MAIYAMOK
+0xE7	0x0E47	#	THAI CHARACTER MAITAIKHU
+0xE8	0x0E48	#	THAI CHARACTER MAI EK
+0xE9	0x0E49	#	THAI CHARACTER MAI THO
+0xEA	0x0E4A	#	THAI CHARACTER MAI TRI
+0xEB	0x0E4B	#	THAI CHARACTER MAI CHATTAWA
+0xEC	0x0E4C	#	THAI CHARACTER THANTHAKHAT
+0xED	0x0E4D	#	THAI CHARACTER NIKHAHIT
+0xEE	0x0E4E	#	THAI CHARACTER YAMAKKAN
+0xEF	0x0E4F	#	THAI CHARACTER FONGMAN
+0xF0	0x0E50	#	THAI DIGIT ZERO
+0xF1	0x0E51	#	THAI DIGIT ONE
+0xF2	0x0E52	#	THAI DIGIT TWO
+0xF3	0x0E53	#	THAI DIGIT THREE
+0xF4	0x0E54	#	THAI DIGIT FOUR
+0xF5	0x0E55	#	THAI DIGIT FIVE
+0xF6	0x0E56	#	THAI DIGIT SIX
+0xF7	0x0E57	#	THAI DIGIT SEVEN
+0xF8	0x0E58	#	THAI DIGIT EIGHT
+0xF9	0x0E59	#	THAI DIGIT NINE
+0xFA	0x0E5A	#	THAI CHARACTER ANGKHANKHU
+0xFB	0x0E5B	#	THAI CHARACTER KHOMUT