1 /* Conversion from and to CP1255.
2 Copyright (C) 1998, 2001 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998,
5 and Bruno Haible <haible@clisp.cons.org>, 2001.
7 The GNU C Library is free software; you can redistribute it and/or
8 modify it under the terms of the GNU Lesser General Public
9 License as published by the Free Software Foundation; either
10 version 2.1 of the License, or (at your option) any later version.
12 The GNU C Library is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 Lesser General Public License for more details.
17 You should have received a copy of the GNU Lesser General Public
18 License along with the GNU C Library; if not, write to the Free
19 Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
26 #define NELEMS(arr) (sizeof (arr) / sizeof (arr[0]))
28 /* Definitions used in the body of the `gconv' function. */
29 #define CHARSET_NAME "CP1255//"
30 #define FROM_LOOP from_cp1255
31 #define TO_LOOP to_cp1255
34 #define MIN_NEEDED_FROM 1
35 #define MAX_NEEDED_FROM 1
36 #define MIN_NEEDED_TO 4
37 #define MAX_NEEDED_TO 4
38 #define PREPARE_LOOP \
40 int *statep = &data->__statep->__count;
41 #define EXTRA_LOOP_ARGS , statep
44 /* Since we might have to reset input pointer we must be able to save
45 and restore the state. */
46 #define SAVE_RESET_STATE(Save) \
48 saved_state = *statep; \
53 /* During CP1255 to UCS4 conversion, the COUNT element of the state
54 contains the last UCS4 character, shifted by 3 bits. */
57 /* Since this is a stateful encoding we have to provide code which resets
58 the output state to the initial state. This has to be done during the
60 #define EMIT_SHIFT_TO_INIT \
61 if (data->__statep->__count != 0) \
65 if (__builtin_expect (outbuf + 4 <= outend, 1)) \
67 /* Write out the last character. */ \
68 *((uint32_t *) outbuf)++ = data->__statep->__count >> 3; \
69 data->__statep->__count = 0; \
72 /* We don't have enough room in the output buffer. */ \
73 status = __GCONV_FULL_OUTPUT; \
76 /* We don't use shift states in the TO_DIRECTION. */ \
77 data->__statep->__count = 0; \
81 /* First define the conversion function from CP1255 to UCS4. */
83 static const uint16_t to_ucs4
[128] = {
85 0x20AC, 0, 0x201A, 0x0192, 0x201E, 0x2026, 0x2020, 0x2021,
86 0x02C6, 0x2030, 0, 0x2039, 0, 0, 0, 0,
88 0, 0x2018, 0x2019, 0x201C, 0x201D, 0x2022, 0x2013, 0x2014,
89 0x02DC, 0x2122, 0, 0x203A, 0, 0, 0, 0,
91 0x00A0, 0x00A1, 0x00A2, 0x00A3, 0x20AA, 0x00A5, 0x00A6, 0x00A7,
92 0x00A8, 0x00A9, 0x00D7, 0x00AB, 0x00AC, 0x00AD, 0x00AE, 0x00AF,
94 0x00B0, 0x00B1, 0x00B2, 0x00B3, 0x00B4, 0x00B5, 0x00B6, 0x00B7,
95 0x00B8, 0x00B9, 0x00F7, 0x00BB, 0x00BC, 0x00BD, 0x00BE, 0x00BF,
97 0x05B0, 0x05B1, 0x05B2, 0x05B3, 0x05B4, 0x05B5, 0x05B6, 0x05B7,
98 0x05B8, 0x05B9, 0, 0x05BB, 0x05BC, 0x05BD, 0x05BE, 0x05BF,
100 0x05C0, 0x05C1, 0x05C2, 0x05C3, 0x05F0, 0x05F1, 0x05F2, 0x05F3,
101 0x05F4, 0, 0, 0, 0, 0, 0, 0,
103 0x05D0, 0x05D1, 0x05D2, 0x05D3, 0x05D4, 0x05D5, 0x05D6, 0x05D7,
104 0x05D8, 0x05D9, 0x05DA, 0x05DB, 0x05DC, 0x05DD, 0x05DE, 0x05DF,
106 0x05E0, 0x05E1, 0x05E2, 0x05E3, 0x05E4, 0x05E5, 0x05E6, 0x05E7,
107 0x05E8, 0x05E9, 0x05EA, 0, 0, 0x200E, 0x200F, 0,
110 /* CP1255 contains eight combining characters:
111 0x05b4, 0x05b7, 0x05b8, 0x05b9, 0x05bc, 0x05bf, 0x05c1, 0x05c2. */
113 /* Composition tables for each of the relevant combining characters. */
114 static const struct {
117 } comp_table_data
[] = {
118 #define COMP_TABLE_IDX_05B4 0
119 #define COMP_TABLE_LEN_05B4 1
121 #define COMP_TABLE_IDX_05B7 (COMP_TABLE_IDX_05B4 + COMP_TABLE_LEN_05B4)
122 #define COMP_TABLE_LEN_05B7 2
125 #define COMP_TABLE_IDX_05B8 (COMP_TABLE_IDX_05B7 + COMP_TABLE_LEN_05B7)
126 #define COMP_TABLE_LEN_05B8 1
128 #define COMP_TABLE_IDX_05B9 (COMP_TABLE_IDX_05B8 + COMP_TABLE_LEN_05B8)
129 #define COMP_TABLE_LEN_05B9 1
131 #define COMP_TABLE_IDX_05BC (COMP_TABLE_IDX_05B9 + COMP_TABLE_LEN_05B9)
132 #define COMP_TABLE_LEN_05BC 24
157 #define COMP_TABLE_IDX_05BF (COMP_TABLE_IDX_05BC + COMP_TABLE_LEN_05BC)
158 #define COMP_TABLE_LEN_05BF 3
162 #define COMP_TABLE_IDX_05C1 (COMP_TABLE_IDX_05BF + COMP_TABLE_LEN_05BF)
163 #define COMP_TABLE_LEN_05C1 2
166 #define COMP_TABLE_IDX_05C2 (COMP_TABLE_IDX_05C1 + COMP_TABLE_LEN_05C1)
167 #define COMP_TABLE_LEN_05C2 2
170 #define COMP_TABLE_IDX_END (COMP_TABLE_IDX_05C2 + COMP_TABLE_LEN_05C2)
172 /* Compile-time verification of table size. */
173 typedef int verify1
[(NELEMS (comp_table_data
) == COMP_TABLE_IDX_END
) - 1];
175 static const struct { unsigned int idx
; unsigned int len
; } comp_table
[8] = {
176 { COMP_TABLE_IDX_05B4
, COMP_TABLE_LEN_05B4
},
177 { COMP_TABLE_IDX_05B7
, COMP_TABLE_LEN_05B7
},
178 { COMP_TABLE_IDX_05B8
, COMP_TABLE_LEN_05B8
},
179 { COMP_TABLE_IDX_05B9
, COMP_TABLE_LEN_05B9
},
180 { COMP_TABLE_IDX_05BC
, COMP_TABLE_LEN_05BC
},
181 { COMP_TABLE_IDX_05BF
, COMP_TABLE_LEN_05BF
},
182 { COMP_TABLE_IDX_05C1
, COMP_TABLE_LEN_05C1
},
183 { COMP_TABLE_IDX_05C2
, COMP_TABLE_LEN_05C2
},
186 #define MIN_NEEDED_INPUT MIN_NEEDED_FROM
187 #define MIN_NEEDED_OUTPUT MIN_NEEDED_TO
188 #define MAX_NEEDED_OUTPUT 8
189 #define LOOPFCT FROM_LOOP
192 uint32_t ch = *inptr; \
194 int must_buffer_ch; \
198 ch = to_ucs4[ch - 0x80]; \
199 if (__builtin_expect (ch == L'\0', 0)) \
201 /* This is an illegal character. */ \
202 if (! ignore_errors_p ()) \
204 result = __GCONV_ILLEGAL_INPUT; \
214 /* Determine whether there is a buffered character pending. */ \
215 last_ch = *statep >> 3; \
217 /* We have to buffer ch if it is a possible match in comp_table_data. */ \
218 must_buffer_ch = (ch >= 0x05d0 && ch <= 0x05f2); \
222 if (ch >= 0x05b0 && ch < 0x05c5) \
224 /* See whether last_ch and ch can be combined. */ \
225 unsigned int i, i1, i2; \
254 goto not_combining; \
257 i1 = comp_table[i].idx; \
258 i2 = i1 + comp_table[i].len - 1; \
260 if (last_ch >= comp_table_data[i1].base \
261 && last_ch <= comp_table_data[i2].base) \
265 i = (i1 + i2) >> 1; \
266 if (last_ch == comp_table_data[i].base) \
268 if (last_ch < comp_table_data[i].base) \
271 goto not_combining; \
281 if (last_ch == comp_table_data[i].base) \
283 goto not_combining; \
287 last_ch = comp_table_data[i].composed; \
288 if (last_ch == 0xfb2a || last_ch == 0xfb2b \
289 || last_ch == 0xfb49) \
290 /* Buffer the combined character. */ \
291 *statep = last_ch << 3; \
294 /* Output the combined character. */ \
295 put32 (outptr, last_ch); \
305 /* Output the buffered character. */ \
306 put32 (outptr, last_ch); \
310 /* If we don't have enough room to output ch as well, then deal \
311 with it in another round. */ \
312 if (!must_buffer_ch && __builtin_expect (outptr + 4 > outend, 0)) \
316 if (must_buffer_ch) \
320 put32 (outptr, ch); \
325 #define LOOP_NEED_FLAGS
326 #define EXTRA_LOOP_DECLS , int *statep
327 #include <iconv/loop.c>
330 /* Next, define the conversion function from UCS4 to CP1255. */
332 static const unsigned char from_ucs4
[] = {
333 #define FROM_IDX_00 0
334 0xa0, 0xa1, 0xa2, 0xa3, 0x00, 0xa5, 0xa6, 0xa7, /* 0x00a0-0x00a7 */
335 0xa8, 0xa9, 0x00, 0xab, 0xac, 0xad, 0xae, 0xaf, /* 0x00a8-0x00af */
336 0xb0, 0xb1, 0xb2, 0xb3, 0xb4, 0xb5, 0xb6, 0xb7, /* 0x00b0-0x00b7 */
337 0xb8, 0xb9, 0x00, 0xbb, 0xbc, 0xbd, 0xbe, 0xbf, /* 0x00b8-0x00bf */
338 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00c0-0x00c7 */
339 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00c8-0x00cf */
340 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xaa, /* 0x00d0-0x00d7 */
341 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00d8-0x00df */
342 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00e0-0x00e7 */
343 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00e8-0x00ef */
344 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xba, /* 0x00f0-0x00f7 */
345 #define FROM_IDX_02 (FROM_IDX_00 + 88)
346 0x88, 0x00, /* 0x02c6-0x02c7 */
347 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x02c8-0x02cf */
348 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x02d0-0x02d7 */
349 0x00, 0x00, 0x00, 0x00, 0x98, /* 0x02d8-0x02dc */
350 #define FROM_IDX_05 (FROM_IDX_02 + 23)
351 0xc0, 0xc1, 0xc2, 0xc3, 0xc4, 0xc5, 0xc6, 0xc7, /* 0x05b0-0x05b7 */
352 0xc8, 0xc9, 0x00, 0xcb, 0xcc, 0xcd, 0xce, 0xcf, /* 0x05b8-0x05bf */
353 0xd0, 0xd1, 0xd2, 0xd3, 0x00, 0x00, 0x00, 0x00, /* 0x05c0-0x05c7 */
354 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x05c8-0x05cf */
355 0xe0, 0xe1, 0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, /* 0x05d0-0x05d7 */
356 0xe8, 0xe9, 0xea, 0xeb, 0xec, 0xed, 0xee, 0xef, /* 0x05d8-0x05df */
357 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, /* 0x05e0-0x05e7 */
358 0xf8, 0xf9, 0xfa, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x05e8-0x05ef */
359 0xd4, 0xd5, 0xd6, 0xd7, 0xd8, /* 0x05f0-0x05f4 */
360 #define FROM_IDX_20 (FROM_IDX_05 + 69)
361 0xfd, 0xfe, /* 0x200e-0x200f */
362 0x00, 0x00, 0x00, 0x96, 0x97, 0x00, 0x00, 0x00, /* 0x2010-0x2017 */
363 0x91, 0x92, 0x82, 0x00, 0x93, 0x94, 0x84, 0x00, /* 0x2018-0x201f */
364 0x86, 0x87, 0x95, 0x00, 0x00, 0x00, 0x85, 0x00, /* 0x2020-0x2027 */
365 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2028-0x202f */
366 0x89, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2030-0x2037 */
367 0x00, 0x8b, 0x9b, /* 0x2038-0x203a */
368 #define FROM_IDX_FF (FROM_IDX_20 + 45)
370 /* Compile-time verification of table size. */
371 typedef int verify2
[(NELEMS (from_ucs4
) == FROM_IDX_FF
) - 1];
373 static const unsigned char comb_table
[8] = {
374 0xc4, 0xc7, 0xc8, 0xc9, 0xcc, 0xcf, 0xd1, 0xd2,
377 /* Decomposition table for the relevant Unicode characters. */
378 static const struct {
384 { 0xFB1D, 0x05D9, 0, -1 },
385 { 0xFB1F, 0x05F2, 1, -1 },
386 { 0xFB2A, 0x05E9, 6, -1 },
387 { 0xFB2B, 0x05E9, 7, -1 },
388 { 0xFB2C, 0x05E9, 4, 6 },
389 { 0xFB2D, 0x05E9, 4, 7 },
390 { 0xFB2E, 0x05D0, 1, -1 },
391 { 0xFB2F, 0x05D0, 2, -1 },
392 { 0xFB30, 0x05D0, 4, -1 },
393 { 0xFB31, 0x05D1, 4, -1 },
394 { 0xFB32, 0x05D2, 4, -1 },
395 { 0xFB33, 0x05D3, 4, -1 },
396 { 0xFB34, 0x05D4, 4, -1 },
397 { 0xFB35, 0x05D5, 4, -1 },
398 { 0xFB36, 0x05D6, 4, -1 },
399 { 0xFB38, 0x05D8, 4, -1 },
400 { 0xFB39, 0x05D9, 4, -1 },
401 { 0xFB3A, 0x05DA, 4, -1 },
402 { 0xFB3B, 0x05DB, 4, -1 },
403 { 0xFB3C, 0x05DC, 4, -1 },
404 { 0xFB3E, 0x05DE, 4, -1 },
405 { 0xFB40, 0x05E0, 4, -1 },
406 { 0xFB41, 0x05E1, 4, -1 },
407 { 0xFB43, 0x05E3, 4, -1 },
408 { 0xFB44, 0x05E4, 4, -1 },
409 { 0xFB46, 0x05E6, 4, -1 },
410 { 0xFB47, 0x05E7, 4, -1 },
411 { 0xFB48, 0x05E8, 4, -1 },
412 { 0xFB49, 0x05E9, 4, -1 },
413 { 0xFB4A, 0x05EA, 4, -1 },
414 { 0xFB4B, 0x05D5, 3, -1 },
415 { 0xFB4C, 0x05D1, 5, -1 },
416 { 0xFB4D, 0x05DB, 5, -1 },
417 { 0xFB4E, 0x05E4, 5, -1 },
420 #define MIN_NEEDED_INPUT MIN_NEEDED_TO
421 #define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM
422 #define MAX_NEEDED_OUTPUT 3
423 #define LOOPFCT TO_LOOP
426 uint32_t ch = get32 (inptr); \
437 if (ch >= 0x00a0 && ch < 0x00f8) \
438 res = from_ucs4[ch - 0x00a0 + FROM_IDX_00]; \
439 else if (ch == 0x0192) \
441 else if (ch >= 0x02c6 && ch < 0x02dd) \
442 res = from_ucs4[ch - 0x02c6 + FROM_IDX_02]; \
443 else if (ch >= 0x05b0 && ch < 0x05f5) \
444 res = from_ucs4[ch - 0x05b0 + FROM_IDX_05]; \
445 else if (ch >= 0x200e && ch < 0x203b) \
446 res = from_ucs4[ch - 0x200e + FROM_IDX_20]; \
447 else if (ch == 0x20aa) \
449 else if (ch == 0x20ac) \
451 else if (ch == 0x2122) \
455 UNICODE_TAG_HANDLER (ch, 4); \
459 if (__builtin_expect (res != 0, 1)) \
466 /* Try canonical decomposition. */ \
467 unsigned int i1, i2; \
470 i2 = sizeof (decomp_table) / sizeof (decomp_table[0]) - 1; \
471 if (ch >= decomp_table[i1].composed \
472 && ch <= decomp_table[i2].composed) \
478 i = (i1 + i2) >> 1; \
479 if (ch == decomp_table[i].composed) \
481 if (ch < decomp_table[i].composed) \
494 if (ch == decomp_table[i].composed) \
501 /* Found a canonical decomposition. */ \
502 ch = decomp_table[i].base; \
503 /* ch is one of 0x05d0..0x05d6, 0x05d8..0x05dc, 0x05de, \
504 0x05e0..0x05e1, 0x05e3..0x05e4, 0x05e6..0x05ea, 0x05f2. */ \
505 ch = from_ucs4[ch - 0x05b0 + FROM_IDX_05]; \
508 if (decomp_table[i].comb2 < 0) \
510 /* See whether we have room for two bytes. */ \
511 if (__builtin_expect (outptr + 1 >= outend, 0)) \
513 result = __GCONV_FULL_OUTPUT; \
517 *outptr++ = (unsigned char) ch; \
518 *outptr++ = comb_table[decomp_table[i].comb1]; \
522 /* See whether we have room for three bytes. */ \
523 if (__builtin_expect (outptr + 2 >= outend, 0)) \
525 result = __GCONV_FULL_OUTPUT; \
529 *outptr++ = (unsigned char) ch; \
530 *outptr++ = comb_table[decomp_table[i].comb1]; \
531 *outptr++ = comb_table[decomp_table[i].comb2]; \
539 /* This is an illegal character. */ \
540 STANDARD_ERR_HANDLER (4); \
544 #define LOOP_NEED_FLAGS
545 #define EXTRA_LOOP_DECLS , int *statep
546 #include <iconv/loop.c>
549 /* Now define the toplevel functions. */
550 #include <iconv/skeleton.c>