xref: /trunk/main/sal/textenc/tcvteas1.tab (revision 11594ab6) 
1/**************************************************************
2 *
3 * Licensed to the Apache Software Foundation (ASF) under one
4 * or more contributor license agreements.  See the NOTICE file
5 * distributed with this work for additional information
6 * regarding copyright ownership.  The ASF licenses this file
7 * to you under the Apache License, Version 2.0 (the
8 * "License"); you may not use this file except in compliance
9 * with the License.  You may obtain a copy of the License at
10 *
11 *   http://www.apache.org/licenses/LICENSE-2.0
12 *
13 * Unless required by applicable law or agreed to in writing,
14 * software distributed under the License is distributed on an
15 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
16 * KIND, either express or implied.  See the License for the
17 * specific language governing permissions and limitations
18 * under the License.
19 *
20 *************************************************************/
21
22
23
24/* ======================================================================= */
25
26/* This file contain the tables for all 1 byte charsets, for the */
27/* following scripts: asia (Thai, Vietnamese) and other exotic scripts */
28
29/* ======================================================================= */
30
31/* MS-874 */
32/* Windows/Dos Standard CharSet for Thai */
33/* 1-Byte, 0x00-0x7F ASCII ohne Ausnahme */
34/* Convert-Tables: mappings/vendors/micsft/pc/cp874.txt from 04/15/98 Version 2.00 */
35/* Last-Changes from us: */
36
37/* ----------------------------------------------------------------------- */
38
39#define MS874UNI_START                  0x80
40#define MS874UNI_END                    0xFB
41static sal_uInt16 const aImplMS874ToUniTab[MS874UNI_END - MS874UNI_START + 1] =
42{
43/*       0       1       2       3       4       5       6       7 */
44/*       8       9       A       B       C       D       E       F */
45    0x20AC,      0,      0,      0,      0, 0x2026,      0,      0, /* 0x80 */
46         0,      0,      0,      0,      0,      0,      0,      0, /* 0x80 */
47         0, 0x2018, 0x2019, 0x201C, 0x201D, 0x2022, 0x2013, 0x2014, /* 0x90 */
48         0,      0,      0,      0,      0,      0,      0,      0, /* 0x90 */
49    0x00A0, 0x0E01, 0x0E02, 0x0E03, 0x0E04, 0x0E05, 0x0E06, 0x0E07, /* 0xA0 */
50    0x0E08, 0x0E09, 0x0E0A, 0x0E0B, 0x0E0C, 0x0E0D, 0x0E0E, 0x0E0F, /* 0xA0 */
51    0x0E10, 0x0E11, 0x0E12, 0x0E13, 0x0E14, 0x0E15, 0x0E16, 0x0E17, /* 0xB0 */
52    0x0E18, 0x0E19, 0x0E1A, 0x0E1B, 0x0E1C, 0x0E1D, 0x0E1E, 0x0E1F, /* 0xB0 */
53    0x0E20, 0x0E21, 0x0E22, 0x0E23, 0x0E24, 0x0E25, 0x0E26, 0x0E27, /* 0xC0 */
54    0x0E28, 0x0E29, 0x0E2A, 0x0E2B, 0x0E2C, 0x0E2D, 0x0E2E, 0x0E2F, /* 0xC0 */
55    0x0E30, 0x0E31, 0x0E32, 0x0E33, 0x0E34, 0x0E35, 0x0E36, 0x0E37, /* 0xD0 */
56    0x0E38, 0x0E39, 0x0E3A,      0,      0,      0,      0, 0x0E3F, /* 0xD0 */
57    0x0E40, 0x0E41, 0x0E42, 0x0E43, 0x0E44, 0x0E45, 0x0E46, 0x0E47, /* 0xE0 */
58    0x0E48, 0x0E49, 0x0E4A, 0x0E4B, 0x0E4C, 0x0E4D, 0x0E4E, 0x0E4F, /* 0xE0 */
59    0x0E50, 0x0E51, 0x0E52, 0x0E53, 0x0E54, 0x0E55, 0x0E56, 0x0E57, /* 0xF0 */
60    0x0E58, 0x0E59, 0x0E5A, 0x0E5B                                  /* 0xF0 */
61};
62
63/* ----------------------------------------------------------------------- */
64
65#define MS874CHAR_START                 0x0E01
66#define MS874CHAR_END                   0x0E5B
67static sal_uChar const aImplMS874ToCharTab[MS874CHAR_END - MS874CHAR_START + 1] =
68{
69/*     0     1     2     3     4     5     6     7 */
70/*     8     9     A     B     C     D     E     F */
71          0xA1, 0xA2, 0xA3, 0xA4, 0xA5, 0xA6, 0xA7, /* 0x0E00 */
72    0xA8, 0xA9, 0xAA, 0xAB, 0xAC, 0xAD, 0xAE, 0xAF, /* 0x0E00 */
73    0xB0, 0xB1, 0xB2, 0xB3, 0xB4, 0xB5, 0xB6, 0xB7, /* 0x0E10 */
74    0xB8, 0xB9, 0xBA, 0xBB, 0xBC, 0xBD, 0xBE, 0xBF, /* 0x0E10 */
75    0xC0, 0xC1, 0xC2, 0xC3, 0xC4, 0xC5, 0xC6, 0xC7, /* 0x0E20 */
76    0xC8, 0xC9, 0xCA, 0xCB, 0xCC, 0xCD, 0xCE, 0xCF, /* 0x0E20 */
77    0xD0, 0xD1, 0xD2, 0xD3, 0xD4, 0xD5, 0xD6, 0xD7, /* 0x0E30 */
78    0xD8, 0xD9, 0xDA,    0,    0,    0,    0, 0xDF, /* 0x0E30 */
79    0xE0, 0xE1, 0xE2, 0xE3, 0xE4, 0xE5, 0xE6, 0xE7, /* 0x0E40 */
80    0xE8, 0xE9, 0xEA, 0xEB, 0xEC, 0xED, 0xEE, 0xEF, /* 0x0E40 */
81    0xF0, 0xF1, 0xF2, 0xF3, 0xF4, 0xF5, 0xF6, 0xF7, /* 0x0E50 */
82    0xF8, 0xF9, 0xFA, 0xFB                          /* 0x0E50 */
83};
84
85/* ----------------------------------------------------------------------- */
86
87#define MS874TOCHARTABEX_COUNT          10
88static ImplUniCharTabData const aImplMS874ToCharTabEx[MS874TOCHARTABEX_COUNT] =
89{
90  { 0x00A0, 0xA0, 0 },
91  { 0x2013, 0x96, 0 },
92  { 0x2014, 0x97, 0 },
93  { 0x2018, 0x91, 0 },
94  { 0x2019, 0x92, 0 },
95  { 0x201C, 0x93, 0 },
96  { 0x201D, 0x94, 0 },
97  { 0x2022, 0x95, 0 },
98  { 0x2026, 0x85, 0 },
99  { 0x20AC, 0x80, 0 }
100};
101
102/* ----------------------------------------------------------------------- */
103
104static ImplByteConvertData const aImplMS874ByteCvtData =
105{
106    aImplMS874ToUniTab,
107    NULL,
108    MS874UNI_START, MS874UNI_END,
109    NOTABUNI_START, NOTABUNI_END,
110    aImplMS874ToCharTab,
111    NULL,
112    aImplMS874ToCharTabEx,
113    MS874CHAR_START, MS874CHAR_END,
114    NOTABCHAR_START, NOTABCHAR_END,
115    MS874TOCHARTABEX_COUNT
116};
117
118/* ----------------------------------------------------------------------- */
119
120static ImplTextEncodingData const aImplMS874TextEncodingData
121    = { { &aImplMS874ByteCvtData,
122          ImplCharToUnicode,
123          ImplUnicodeToChar,
124          NULL,
125          NULL,
126          NULL,
127          NULL,
128          NULL,
129          NULL },
130        1,
131        1,
132        1,
133        222,
134        "iso8859-1", /* TODO! correct? */
135        "windows-874",
136        RTL_TEXTENCODING_INFO_ASCII | RTL_TEXTENCODING_INFO_MIME };
137    /* WIN/DOS/OS2, SCRIPT_THAI, pc code page 874, mac encoding 21 */
138
139/* ======================================================================= */
140
141/* TIS 620-2533
142 *
143 * A good source of information is <http://www.inet.co.th/cyberclub/trin/
144 * thairef/index.html> as of 18 Mar 2002.
145 *
146 * Single byte encoding, from which MS874 is derived (although it is the other
147 * way around in this implementation):
148 *
149 * 0x00--9F map to U+0000--009F
150 * 0xA0 is questionable (unassigned or U+00A0 NO BREAK SPACE), to ease
151 *      implementation, it maps to U+00A0
152 * 0xA1--DA map to U+0E01--0E3A (TIS 620, same for MS874)
153 * 0xDB--DE are unassigned (TIS 620, same for MS874)
154 * 0xDF--FB map to U+0E3F--0E5B (TIS 620, same for MS874)
155 * 0xFC--FF are unassigned (TIS 620, same for MS874)
156 */
157
158static ImplByteConvertData const aImplTis620ByteCvtData =
159{
160    aImplMS874ToUniTab + (0xA0 - MS874UNI_START),
161    aImpl8090SameToUniTab,
162    0xA0, MS874UNI_END,
163    SAME8090UNI_START, SAME8090UNI_END,
164    aImplMS874ToCharTab,
165    aImpl8090SameToCharTab,
166    aImplMS874ToCharTabEx,
167    MS874CHAR_START, MS874CHAR_END,
168    SAME8090CHAR_START, SAME8090CHAR_END,
169    1
170};
171
172static ImplTextEncodingData const aImplTis620TextEncodingData
173    = { { &aImplTis620ByteCvtData,
174          ImplCharToUnicode,
175          ImplUnicodeToChar,
176          NULL,
177          NULL,
178          NULL,
179          NULL,
180          NULL,
181          NULL },
182        1,
183        1,
184        1,
185        222,
186        "iso8859-1", /* TODO! correct? */
187        "TIS-620",
188        RTL_TEXTENCODING_INFO_ASCII | RTL_TEXTENCODING_INFO_MIME };
189
190/* ======================================================================= */
191
192/* MS-1258 */
193/* Windows Standard CharSet for Vietnamease */
194/* 1-Byte, 0x00-0x7F ASCII ohne Ausnahme */
195/* Convert-Tables: mappings/vendors/micsft/windows/cp1258.txt from 04/15/98 Version 2.01 */
196/* Last-Changes from us: */
197
198/* ----------------------------------------------------------------------- */
199
200#define MS1258UNI_START                 0x80
201#define MS1258UNI_END                   0xFF
202static sal_uInt16 const aImplMS1258ToUniTab[MS1258UNI_END - MS1258UNI_START + 1] =
203{
204/*       0       1       2       3       4       5       6       7 */
205/*       8       9       A       B       C       D       E       F */
206    0x20AC,      0, 0x201A, 0x0192, 0x201E, 0x2026, 0x2020, 0x2021, /* 0x80 */
207    0x02C6, 0x2030,      0, 0x2039, 0x0152,      0,      0,      0, /* 0x80 */
208         0, 0x2018, 0x2019, 0x201C, 0x201D, 0x2022, 0x2013, 0x2014, /* 0x90 */
209    0x02DC, 0x2122,      0, 0x203A, 0x0153,      0,      0, 0x0178, /* 0x90 */
210    0x00A0, 0x00A1, 0x00A2, 0x00A3, 0x00A4, 0x00A5, 0x00A6, 0x00A7, /* 0xA0 */
211    0x00A8, 0x00A9, 0x00AA, 0x00AB, 0x00AC, 0x00AD, 0x00AE, 0x00AF, /* 0xA0 */
212    0x00B0, 0x00B1, 0x00B2, 0x00B3, 0x00B4, 0x00B5, 0x00B6, 0x00B7, /* 0xB0 */
213    0x00B8, 0x00B9, 0x00BA, 0x00BB, 0x00BC, 0x00BD, 0x00BE, 0x00BF, /* 0xB0 */
214    0x00C0, 0x00C1, 0x00C2, 0x0102, 0x00C4, 0x00C5, 0x00C6, 0x00C7, /* 0xC0 */
215    0x00C8, 0x00C9, 0x00CA, 0x00CB, 0x0300, 0x00CD, 0x00CE, 0x00CF, /* 0xC0 */
216    0x0110, 0x00D1, 0x0309, 0x00D3, 0x00D4, 0x01A0, 0x00D6, 0x00D7, /* 0xD0 */
217    0x00D8, 0x00D9, 0x00DA, 0x00DB, 0x00DC, 0x01AF, 0x0303, 0x00DF, /* 0xD0 */
218    0x00E0, 0x00E1, 0x00E2, 0x0103, 0x00E4, 0x00E5, 0x00E6, 0x00E7, /* 0xE0 */
219    0x00E8, 0x00E9, 0x00EA, 0x00EB, 0x0301, 0x00ED, 0x00EE, 0x00EF, /* 0xE0 */
220    0x0111, 0x00F1, 0x0323, 0x00F3, 0x00F4, 0x01A1, 0x00F6, 0x00F7, /* 0xF0 */
221    0x00F8, 0x00F9, 0x00FA, 0x00FB, 0x00FC, 0x01B0, 0x20AB, 0x00FF  /* 0xF0 */
222};
223
224/* ----------------------------------------------------------------------- */
225
226#define MS1258CHAR_START                0x00A0
227#define MS1258CHAR_END                  0x00FF
228static sal_uChar const aImplMS1258ToCharTab[MS1258CHAR_END - MS1258CHAR_START + 1] =
229{
230/*     0     1     2     3     4     5     6     7 */
231/*     8     9     A     B     C     D     E     F */
232    0xA0, 0xA1, 0xA2, 0xA3, 0xA4, 0xA5, 0xA6, 0xA7, /* 0x00A0 */
233    0xA8, 0xA9, 0xAA, 0xAB, 0xAC, 0xAD, 0xAE, 0xAF, /* 0x00A0 */
234    0xB0, 0xB1, 0xB2, 0xB3, 0xB4, 0xB5, 0xB6, 0xB7, /* 0x00B0 */
235    0xB8, 0xB9, 0xBA, 0xBB, 0xBC, 0xBD, 0xBE, 0xBF, /* 0x00B0 */
236    0xC0, 0xC1, 0xC2,    0, 0xC4, 0xC5, 0xC6, 0xC7, /* 0x00C0 */
237    0xC8, 0xC9, 0xCA, 0xCB,    0, 0xCD, 0xCE, 0xCF, /* 0x00C0 */
238       0, 0xD1,    0, 0xD3, 0xD4,    0, 0xD6, 0xD7, /* 0x00D0 */
239    0xD8, 0xD9, 0xDA, 0xDB, 0xDC,    0,    0, 0xDF, /* 0x00D0 */
240    0xE0, 0xE1, 0xE2,    0, 0xE4, 0xE5, 0xE6, 0xE7, /* 0x00E0 */
241    0xE8, 0xE9, 0xEA, 0xEB,    0, 0xED, 0xEE, 0xEF, /* 0x00E0 */
242       0, 0xF1,    0, 0xF3, 0xF4,    0, 0xF6, 0xF7, /* 0x00F0 */
243    0xF8, 0xF9, 0xFA, 0xFB, 0xFC,    0,    0, 0xFF  /* 0x00F0 */
244};
245
246/* ----------------------------------------------------------------------- */
247
248/* The list of Vietnamese-relevant precomposed Unicode characters that map to
249   MS1258 base + combining modifier are derived from the information present at
250   <http://vietunicode.sourceforge.net/charset/>: */
251#define MS1258TOCHARTABEX_COUNT         141
252static ImplUniCharTabData const aImplMS1258ToCharTabEx[MS1258TOCHARTABEX_COUNT] =
253{
254  { 0x00C3, 0x41, 0xDE },
255  { 0x00CC, 0x49, 0xCC },
256  { 0x00D2, 0x4F, 0xCC },
257  { 0x00D5, 0x4F, 0xDE },
258  { 0x00DD, 0x59, 0xEC },
259  { 0x00E3, 0x61, 0xDE },
260  { 0x00EC, 0x69, 0xCC },
261  { 0x00F2, 0x6F, 0xCC },
262  { 0x00F5, 0x6F, 0xDE },
263  { 0x00FD, 0x79, 0xCC },
264  { 0x0102, 0xC3, 0 },
265  { 0x0103, 0xE3, 0 },
266  { 0x0110, 0xD0, 0 },
267  { 0x0111, 0xF0, 0 },
268  { 0x0128, 0x49, 0xDE },
269  { 0x0129, 0x69, 0xDE },
270  { 0x0152, 0x8C, 0 },
271  { 0x0153, 0x9C, 0 },
272  { 0x0168, 0x55, 0xDE },
273  { 0x0169, 0x75, 0xDE },
274  { 0x0178, 0x9F, 0 },
275  { 0x0192, 0x83, 0 },
276  { 0x01A0, 0xD5, 0 },
277  { 0x01A1, 0xF5, 0 },
278  { 0x01AF, 0xDD, 0 },
279  { 0x01B0, 0xFD, 0 },
280  { 0x02C6, 0x88, 0 },
281  { 0x02DC, 0x98, 0 },
282  { 0x0300, 0xCC, 0 },
283  { 0x0301, 0xEC, 0 },
284  { 0x0303, 0xDE, 0 },
285  { 0x0309, 0xD2, 0 },
286  { 0x0323, 0xF2, 0 },
287  { 0x1EA0, 0x41, 0xF2 },
288  { 0x1EA1, 0x61, 0xF2 },
289  { 0x1EA2, 0x41, 0xD2 },
290  { 0x1EA3, 0x61, 0xD2 },
291  { 0x1EA4, 0xC2, 0xEC },
292  { 0x1EA5, 0xE2, 0xEC },
293  { 0x1EA6, 0xC2, 0xCC },
294  { 0x1EA7, 0xE2, 0xCC },
295  { 0x1EA8, 0xC2, 0xD2 },
296  { 0x1EA9, 0xE2, 0xD2 },
297  { 0x1EAA, 0xC2, 0xDE },
298  { 0x1EAB, 0xE2, 0xDE },
299  { 0x1EAC, 0xC2, 0xF2 },
300  { 0x1EAD, 0xE2, 0xF2 },
301  { 0x1EAE, 0xC3, 0xEC },
302  { 0x1EAF, 0xE3, 0xEC },
303  { 0x1EB0, 0xC3, 0xCC },
304  { 0x1EB1, 0xE3, 0xCC },
305  { 0x1EB2, 0xC3, 0xD2 },
306  { 0x1EB3, 0xE3, 0xD2 },
307  { 0x1EB4, 0xC3, 0xDE },
308  { 0x1EB5, 0xE3, 0xDE },
309  { 0x1EB6, 0xC3, 0xF2 },
310  { 0x1EB7, 0xE3, 0xF2 },
311  { 0x1EB8, 0x45, 0xF2 },
312  { 0x1EB9, 0x65, 0xF2 },
313  { 0x1EBA, 0x45, 0xD2 },
314  { 0x1EBB, 0x65, 0xD2 },
315  { 0x1EBC, 0x45, 0xDE },
316  { 0x1EBD, 0x65, 0xDE },
317  { 0x1EBE, 0xCA, 0xEC },
318  { 0x1EBF, 0xEA, 0xEC },
319  { 0x1EC0, 0xCA, 0xCC },
320  { 0x1EC1, 0xEA, 0xCC },
321  { 0x1EC2, 0xCA, 0xD2 },
322  { 0x1EC3, 0xEA, 0xD2 },
323  { 0x1EC4, 0xCA, 0xDE },
324  { 0x1EC5, 0xEA, 0xDE },
325  { 0x1EC6, 0xCA, 0xF2 },
326  { 0x1EC7, 0xEA, 0xF2 },
327  { 0x1EC8, 0x49, 0xD2 },
328  { 0x1EC9, 0x69, 0xD2 },
329  { 0x1ECA, 0x49, 0xF2 },
330  { 0x1ECB, 0x69, 0xF2 },
331  { 0x1ECC, 0x4F, 0xF2 },
332  { 0x1ECD, 0x6F, 0xF2 },
333  { 0x1ECE, 0x4F, 0xD2 },
334  { 0x1ECF, 0x6F, 0xD2 },
335  { 0x1ED0, 0xD4, 0xEC },
336  { 0x1ED1, 0xF4, 0xEC },
337  { 0x1ED2, 0xD4, 0xCC },
338  { 0x1ED3, 0xF4, 0xCC },
339  { 0x1ED4, 0xD4, 0xD2 },
340  { 0x1ED5, 0xF4, 0xD2 },
341  { 0x1ED6, 0xD4, 0xDE },
342  { 0x1ED7, 0xF4, 0xDE },
343  { 0x1ED8, 0xD4, 0xF2 },
344  { 0x1ED9, 0xF4, 0xF2 },
345  { 0x1EDA, 0xD5, 0xEC },
346  { 0x1EDB, 0xF5, 0xEC },
347  { 0x1EDC, 0xD5, 0xCC },
348  { 0x1EDD, 0xF5, 0xCC },
349  { 0x1EDE, 0xD5, 0xD2 },
350  { 0x1EDF, 0xF5, 0xD2 },
351  { 0x1EE0, 0xD5, 0xDE },
352  { 0x1EE1, 0xF5, 0xDE },
353  { 0x1EE2, 0xD5, 0xF2 },
354  { 0x1EE3, 0xF5, 0xF2 },
355  { 0x1EE4, 0x55, 0xF2 },
356  { 0x1EE5, 0x75, 0xF2 },
357  { 0x1EE6, 0x55, 0xD2 },
358  { 0x1EE7, 0x75, 0xD2 },
359  { 0x1EE8, 0xDD, 0xEC },
360  { 0x1EE9, 0xFD, 0xEC },
361  { 0x1EEA, 0xDD, 0xCC },
362  { 0x1EEB, 0xFD, 0xCC },
363  { 0x1EEC, 0xDD, 0xD2 },
364  { 0x1EED, 0xFD, 0xD2 },
365  { 0x1EEE, 0xDD, 0xDE },
366  { 0x1EEF, 0xFD, 0xDE },
367  { 0x1EF0, 0xDD, 0xF2 },
368  { 0x1EF1, 0xFD, 0xF2 },
369  { 0x1EF2, 0x59, 0xCC },
370  { 0x1EF3, 0x79, 0xCC },
371  { 0x1EF4, 0x59, 0xF2 },
372  { 0x1EF5, 0x79, 0xF2 },
373  { 0x1EF6, 0x59, 0xD2 },
374  { 0x1EF7, 0x79, 0xD2 },
375  { 0x1EF8, 0x59, 0xDE },
376  { 0x1EF9, 0x79, 0xDE },
377  { 0x2013, 0x96, 0 },
378  { 0x2014, 0x97, 0 },
379  { 0x2018, 0x91, 0 },
380  { 0x2019, 0x92, 0 },
381  { 0x201A, 0x82, 0 },
382  { 0x201C, 0x93, 0 },
383  { 0x201D, 0x94, 0 },
384  { 0x201E, 0x84, 0 },
385  { 0x2020, 0x86, 0 },
386  { 0x2021, 0x87, 0 },
387  { 0x2022, 0x95, 0 },
388  { 0x2026, 0x85, 0 },
389  { 0x2030, 0x89, 0 },
390  { 0x2039, 0x8B, 0 },
391  { 0x203A, 0x9B, 0 },
392  { 0x20AB, 0xFE, 0 },
393  { 0x20AC, 0x80, 0 },
394  { 0x2122, 0x99, 0 }
395};
396
397/* ----------------------------------------------------------------------- */
398
399static ImplByteConvertData const aImplMS1258ByteCvtData =
400{
401    aImplMS1258ToUniTab,
402    NULL,
403    MS1258UNI_START, MS1258UNI_END,
404    NOTABUNI_START, NOTABUNI_END,
405    aImplMS1258ToCharTab,
406    NULL,
407    aImplMS1258ToCharTabEx,
408    MS1258CHAR_START, MS1258CHAR_END,
409    NOTABCHAR_START, NOTABCHAR_END,
410    MS1258TOCHARTABEX_COUNT
411};
412
413/* ----------------------------------------------------------------------- */
414
415static ImplTextEncodingData const aImplMS1258TextEncodingData
416    = { { &aImplMS1258ByteCvtData,
417          ImplCharToUnicode,
418          ImplUnicodeToChar,
419          NULL,
420          NULL,
421          NULL,
422          NULL,
423          NULL,
424          NULL },
425        1,
426        2,
427        1,
428        163,
429        "iso8859-1", /* TODO! correct? */
430        "windows-1258",
431        RTL_TEXTENCODING_INFO_ASCII | RTL_TEXTENCODING_INFO_MIME };
432    /* WIN, SCRIPT_VIETNAMESE, mac encoding 30 */
433