1 /* Conversion from and to CP1255.
2 Copyright (C) 1998, 2001, 2002, 2004 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, see
19 <http://www.gnu.org/licenses/>. */
25 #define NELEMS(arr) (sizeof (arr) / sizeof (arr[0]))
27 /* Definitions used in the body of the `gconv' function. */
28 #define CHARSET_NAME "CP1255//"
29 #define FROM_LOOP from_cp1255
30 #define TO_LOOP to_cp1255
33 #define FROM_LOOP_MIN_NEEDED_FROM 1
34 #define FROM_LOOP_MAX_NEEDED_FROM 1
35 #define FROM_LOOP_MIN_NEEDED_TO 4
36 #define FROM_LOOP_MAX_NEEDED_TO 4
37 #define TO_LOOP_MIN_NEEDED_FROM 4
38 #define TO_LOOP_MAX_NEEDED_FROM 4
39 #define TO_LOOP_MIN_NEEDED_TO 1
40 #define TO_LOOP_MAX_NEEDED_TO 3
41 #define PREPARE_LOOP \
43 int *statep = &data->__statep->__count;
44 #define EXTRA_LOOP_ARGS , statep
47 /* Since we might have to reset input pointer we must be able to save
48 and restore the state. */
49 #define SAVE_RESET_STATE(Save) \
51 saved_state = *statep; \
56 /* During CP1255 to UCS4 conversion, the COUNT element of the state
57 contains the last UCS4 character, shifted by 3 bits. */
60 /* Since this is a stateful encoding we have to provide code which resets
61 the output state to the initial state. This has to be done during the
63 #define EMIT_SHIFT_TO_INIT \
64 if (data->__statep->__count != 0) \
68 if (__builtin_expect (outbuf + 4 <= outend, 1)) \
70 /* Write out the last character. */ \
71 *((uint32_t *) outbuf) = data->__statep->__count >> 3; \
72 outbuf += sizeof (uint32_t); \
73 data->__statep->__count = 0; \
76 /* We don't have enough room in the output buffer. */ \
77 status = __GCONV_FULL_OUTPUT; \
80 /* We don't use shift states in the TO_DIRECTION. */ \
81 data->__statep->__count = 0; \
85 /* First define the conversion function from CP1255 to UCS4. */
87 static const uint16_t to_ucs4
[128] = {
89 0x20AC, 0, 0x201A, 0x0192, 0x201E, 0x2026, 0x2020, 0x2021,
90 0x02C6, 0x2030, 0, 0x2039, 0, 0, 0, 0,
92 0, 0x2018, 0x2019, 0x201C, 0x201D, 0x2022, 0x2013, 0x2014,
93 0x02DC, 0x2122, 0, 0x203A, 0, 0, 0, 0,
95 0x00A0, 0x00A1, 0x00A2, 0x00A3, 0x20AA, 0x00A5, 0x00A6, 0x00A7,
96 0x00A8, 0x00A9, 0x00D7, 0x00AB, 0x00AC, 0x00AD, 0x00AE, 0x00AF,
98 0x00B0, 0x00B1, 0x00B2, 0x00B3, 0x00B4, 0x00B5, 0x00B6, 0x00B7,
99 0x00B8, 0x00B9, 0x00F7, 0x00BB, 0x00BC, 0x00BD, 0x00BE, 0x00BF,
101 0x05B0, 0x05B1, 0x05B2, 0x05B3, 0x05B4, 0x05B5, 0x05B6, 0x05B7,
102 0x05B8, 0x05B9, 0, 0x05BB, 0x05BC, 0x05BD, 0x05BE, 0x05BF,
104 0x05C0, 0x05C1, 0x05C2, 0x05C3, 0x05F0, 0x05F1, 0x05F2, 0x05F3,
105 0x05F4, 0, 0, 0, 0, 0, 0, 0,
107 0x05D0, 0x05D1, 0x05D2, 0x05D3, 0x05D4, 0x05D5, 0x05D6, 0x05D7,
108 0x05D8, 0x05D9, 0x05DA, 0x05DB, 0x05DC, 0x05DD, 0x05DE, 0x05DF,
110 0x05E0, 0x05E1, 0x05E2, 0x05E3, 0x05E4, 0x05E5, 0x05E6, 0x05E7,
111 0x05E8, 0x05E9, 0x05EA, 0, 0, 0x200E, 0x200F, 0,
114 /* CP1255 contains eight combining characters:
115 0x05b4, 0x05b7, 0x05b8, 0x05b9, 0x05bc, 0x05bf, 0x05c1, 0x05c2. */
117 /* Composition tables for each of the relevant combining characters. */
118 static const struct {
121 } comp_table_data
[] = {
122 #define COMP_TABLE_IDX_05B4 0
123 #define COMP_TABLE_LEN_05B4 1
125 #define COMP_TABLE_IDX_05B7 (COMP_TABLE_IDX_05B4 + COMP_TABLE_LEN_05B4)
126 #define COMP_TABLE_LEN_05B7 2
129 #define COMP_TABLE_IDX_05B8 (COMP_TABLE_IDX_05B7 + COMP_TABLE_LEN_05B7)
130 #define COMP_TABLE_LEN_05B8 1
132 #define COMP_TABLE_IDX_05B9 (COMP_TABLE_IDX_05B8 + COMP_TABLE_LEN_05B8)
133 #define COMP_TABLE_LEN_05B9 1
135 #define COMP_TABLE_IDX_05BC (COMP_TABLE_IDX_05B9 + COMP_TABLE_LEN_05B9)
136 #define COMP_TABLE_LEN_05BC 24
161 #define COMP_TABLE_IDX_05BF (COMP_TABLE_IDX_05BC + COMP_TABLE_LEN_05BC)
162 #define COMP_TABLE_LEN_05BF 3
166 #define COMP_TABLE_IDX_05C1 (COMP_TABLE_IDX_05BF + COMP_TABLE_LEN_05BF)
167 #define COMP_TABLE_LEN_05C1 2
170 #define COMP_TABLE_IDX_05C2 (COMP_TABLE_IDX_05C1 + COMP_TABLE_LEN_05C1)
171 #define COMP_TABLE_LEN_05C2 2
174 #define COMP_TABLE_IDX_END (COMP_TABLE_IDX_05C2 + COMP_TABLE_LEN_05C2)
176 /* Compile-time verification of table size. */
177 typedef int verify1
[(NELEMS (comp_table_data
) == COMP_TABLE_IDX_END
) - 1];
179 static const struct { unsigned int idx
; unsigned int len
; } comp_table
[8] = {
180 { COMP_TABLE_IDX_05B4
, COMP_TABLE_LEN_05B4
},
181 { COMP_TABLE_IDX_05B7
, COMP_TABLE_LEN_05B7
},
182 { COMP_TABLE_IDX_05B8
, COMP_TABLE_LEN_05B8
},
183 { COMP_TABLE_IDX_05B9
, COMP_TABLE_LEN_05B9
},
184 { COMP_TABLE_IDX_05BC
, COMP_TABLE_LEN_05BC
},
185 { COMP_TABLE_IDX_05BF
, COMP_TABLE_LEN_05BF
},
186 { COMP_TABLE_IDX_05C1
, COMP_TABLE_LEN_05C1
},
187 { COMP_TABLE_IDX_05C2
, COMP_TABLE_LEN_05C2
},
190 #define MIN_NEEDED_INPUT FROM_LOOP_MIN_NEEDED_FROM
191 #define MAX_NEEDED_INPUT FROM_LOOP_MAX_NEEDED_FROM
192 #define MIN_NEEDED_OUTPUT FROM_LOOP_MIN_NEEDED_TO
193 #define MAX_NEEDED_OUTPUT FROM_LOOP_MAX_NEEDED_TO
194 #define LOOPFCT FROM_LOOP
197 uint32_t ch = *inptr; \
199 int must_buffer_ch; \
203 ch = to_ucs4[ch - 0x80]; \
204 if (__builtin_expect (ch == L'\0', 0)) \
206 /* This is an illegal character. */ \
207 STANDARD_FROM_LOOP_ERR_HANDLER (1); \
211 /* Determine whether there is a buffered character pending. */ \
212 last_ch = *statep >> 3; \
214 /* We have to buffer ch if it is a possible match in comp_table_data. */ \
215 must_buffer_ch = (ch >= 0x05d0 && ch <= 0x05f2); \
219 if (ch >= 0x05b0 && ch < 0x05c5) \
221 /* See whether last_ch and ch can be combined. */ \
222 unsigned int i, i1, i2; \
251 goto not_combining; \
254 i1 = comp_table[i].idx; \
255 i2 = i1 + comp_table[i].len - 1; \
257 if (last_ch >= comp_table_data[i1].base \
258 && last_ch <= comp_table_data[i2].base) \
262 i = (i1 + i2) >> 1; \
263 if (last_ch == comp_table_data[i].base) \
265 if (last_ch < comp_table_data[i].base) \
268 goto not_combining; \
278 if (last_ch == comp_table_data[i].base) \
280 goto not_combining; \
284 last_ch = comp_table_data[i].composed; \
285 if (last_ch == 0xfb2a || last_ch == 0xfb2b \
286 || last_ch == 0xfb49) \
287 /* Buffer the combined character. */ \
288 *statep = last_ch << 3; \
291 /* Output the combined character. */ \
292 put32 (outptr, last_ch); \
302 /* Output the buffered character. */ \
303 put32 (outptr, last_ch); \
307 /* If we don't have enough room to output ch as well, then deal \
308 with it in another round. */ \
309 if (!must_buffer_ch && __builtin_expect (outptr + 4 > outend, 0)) \
313 if (must_buffer_ch) \
317 put32 (outptr, ch); \
322 #define LOOP_NEED_FLAGS
323 #define EXTRA_LOOP_DECLS , int *statep
324 #define ONEBYTE_BODY \
328 uint32_t ch = to_ucs4[c - 0x80]; \
329 if (ch == L'\0' || (ch >= 0x05d0 && ch <= 0x05f2)) \
333 #include <iconv/loop.c>
336 /* Next, define the conversion function from UCS4 to CP1255. */
338 static const unsigned char from_ucs4
[] = {
339 #define FROM_IDX_00 0
340 0xa0, 0xa1, 0xa2, 0xa3, 0x00, 0xa5, 0xa6, 0xa7, /* 0x00a0-0x00a7 */
341 0xa8, 0xa9, 0x00, 0xab, 0xac, 0xad, 0xae, 0xaf, /* 0x00a8-0x00af */
342 0xb0, 0xb1, 0xb2, 0xb3, 0xb4, 0xb5, 0xb6, 0xb7, /* 0x00b0-0x00b7 */
343 0xb8, 0xb9, 0x00, 0xbb, 0xbc, 0xbd, 0xbe, 0xbf, /* 0x00b8-0x00bf */
344 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00c0-0x00c7 */
345 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00c8-0x00cf */
346 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xaa, /* 0x00d0-0x00d7 */
347 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00d8-0x00df */
348 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00e0-0x00e7 */
349 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00e8-0x00ef */
350 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xba, /* 0x00f0-0x00f7 */
351 #define FROM_IDX_02 (FROM_IDX_00 + 88)
352 0x88, 0x00, /* 0x02c6-0x02c7 */
353 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x02c8-0x02cf */
354 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x02d0-0x02d7 */
355 0x00, 0x00, 0x00, 0x00, 0x98, /* 0x02d8-0x02dc */
356 #define FROM_IDX_05 (FROM_IDX_02 + 23)
357 0xc0, 0xc1, 0xc2, 0xc3, 0xc4, 0xc5, 0xc6, 0xc7, /* 0x05b0-0x05b7 */
358 0xc8, 0xc9, 0x00, 0xcb, 0xcc, 0xcd, 0xce, 0xcf, /* 0x05b8-0x05bf */
359 0xd0, 0xd1, 0xd2, 0xd3, 0x00, 0x00, 0x00, 0x00, /* 0x05c0-0x05c7 */
360 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x05c8-0x05cf */
361 0xe0, 0xe1, 0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, /* 0x05d0-0x05d7 */
362 0xe8, 0xe9, 0xea, 0xeb, 0xec, 0xed, 0xee, 0xef, /* 0x05d8-0x05df */
363 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, /* 0x05e0-0x05e7 */
364 0xf8, 0xf9, 0xfa, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x05e8-0x05ef */
365 0xd4, 0xd5, 0xd6, 0xd7, 0xd8, /* 0x05f0-0x05f4 */
366 #define FROM_IDX_20 (FROM_IDX_05 + 69)
367 0xfd, 0xfe, /* 0x200e-0x200f */
368 0x00, 0x00, 0x00, 0x96, 0x97, 0x00, 0x00, 0x00, /* 0x2010-0x2017 */
369 0x91, 0x92, 0x82, 0x00, 0x93, 0x94, 0x84, 0x00, /* 0x2018-0x201f */
370 0x86, 0x87, 0x95, 0x00, 0x00, 0x00, 0x85, 0x00, /* 0x2020-0x2027 */
371 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2028-0x202f */
372 0x89, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2030-0x2037 */
373 0x00, 0x8b, 0x9b, /* 0x2038-0x203a */
374 #define FROM_IDX_FF (FROM_IDX_20 + 45)
376 /* Compile-time verification of table size. */
377 typedef int verify2
[(NELEMS (from_ucs4
) == FROM_IDX_FF
) - 1];
379 static const unsigned char comb_table
[8] = {
380 0xc4, 0xc7, 0xc8, 0xc9, 0xcc, 0xcf, 0xd1, 0xd2,
383 /* Decomposition table for the relevant Unicode characters. */
384 static const struct {
390 { 0xFB1D, 0x05D9, 0, -1 },
391 { 0xFB1F, 0x05F2, 1, -1 },
392 { 0xFB2A, 0x05E9, 6, -1 },
393 { 0xFB2B, 0x05E9, 7, -1 },
394 { 0xFB2C, 0x05E9, 4, 6 },
395 { 0xFB2D, 0x05E9, 4, 7 },
396 { 0xFB2E, 0x05D0, 1, -1 },
397 { 0xFB2F, 0x05D0, 2, -1 },
398 { 0xFB30, 0x05D0, 4, -1 },
399 { 0xFB31, 0x05D1, 4, -1 },
400 { 0xFB32, 0x05D2, 4, -1 },
401 { 0xFB33, 0x05D3, 4, -1 },
402 { 0xFB34, 0x05D4, 4, -1 },
403 { 0xFB35, 0x05D5, 4, -1 },
404 { 0xFB36, 0x05D6, 4, -1 },
405 { 0xFB38, 0x05D8, 4, -1 },
406 { 0xFB39, 0x05D9, 4, -1 },
407 { 0xFB3A, 0x05DA, 4, -1 },
408 { 0xFB3B, 0x05DB, 4, -1 },
409 { 0xFB3C, 0x05DC, 4, -1 },
410 { 0xFB3E, 0x05DE, 4, -1 },
411 { 0xFB40, 0x05E0, 4, -1 },
412 { 0xFB41, 0x05E1, 4, -1 },
413 { 0xFB43, 0x05E3, 4, -1 },
414 { 0xFB44, 0x05E4, 4, -1 },
415 { 0xFB46, 0x05E6, 4, -1 },
416 { 0xFB47, 0x05E7, 4, -1 },
417 { 0xFB48, 0x05E8, 4, -1 },
418 { 0xFB49, 0x05E9, 4, -1 },
419 { 0xFB4A, 0x05EA, 4, -1 },
420 { 0xFB4B, 0x05D5, 3, -1 },
421 { 0xFB4C, 0x05D1, 5, -1 },
422 { 0xFB4D, 0x05DB, 5, -1 },
423 { 0xFB4E, 0x05E4, 5, -1 },
426 #define MIN_NEEDED_INPUT TO_LOOP_MIN_NEEDED_FROM
427 #define MAX_NEEDED_INPUT TO_LOOP_MAX_NEEDED_FROM
428 #define MIN_NEEDED_OUTPUT TO_LOOP_MIN_NEEDED_TO
429 #define MAX_NEEDED_OUTPUT TO_LOOP_MAX_NEEDED_TO
430 #define LOOPFCT TO_LOOP
433 uint32_t ch = get32 (inptr); \
444 if (ch >= 0x00a0 && ch < 0x00f8) \
445 res = from_ucs4[ch - 0x00a0 + FROM_IDX_00]; \
446 else if (ch == 0x0192) \
448 else if (ch >= 0x02c6 && ch < 0x02dd) \
449 res = from_ucs4[ch - 0x02c6 + FROM_IDX_02]; \
450 else if (ch >= 0x05b0 && ch < 0x05f5) \
451 res = from_ucs4[ch - 0x05b0 + FROM_IDX_05]; \
452 else if (ch >= 0x200e && ch < 0x203b) \
453 res = from_ucs4[ch - 0x200e + FROM_IDX_20]; \
454 else if (ch == 0x20aa) \
456 else if (ch == 0x20ac) \
458 else if (ch == 0x2122) \
462 UNICODE_TAG_HANDLER (ch, 4); \
466 if (__builtin_expect (res != 0, 1)) \
473 /* Try canonical decomposition. */ \
474 unsigned int i1, i2; \
477 i2 = sizeof (decomp_table) / sizeof (decomp_table[0]) - 1; \
478 if (ch >= decomp_table[i1].composed \
479 && ch <= decomp_table[i2].composed) \
485 i = (i1 + i2) >> 1; \
486 if (ch == decomp_table[i].composed) \
488 if (ch < decomp_table[i].composed) \
501 if (ch == decomp_table[i].composed) \
508 /* Found a canonical decomposition. */ \
509 ch = decomp_table[i].base; \
510 /* ch is one of 0x05d0..0x05d6, 0x05d8..0x05dc, 0x05de, \
511 0x05e0..0x05e1, 0x05e3..0x05e4, 0x05e6..0x05ea, 0x05f2. */ \
512 ch = from_ucs4[ch - 0x05b0 + FROM_IDX_05]; \
515 if (decomp_table[i].comb2 < 0) \
517 /* See whether we have room for two bytes. */ \
518 if (__builtin_expect (outptr + 1 >= outend, 0)) \
520 result = __GCONV_FULL_OUTPUT; \
524 *outptr++ = (unsigned char) ch; \
525 *outptr++ = comb_table[decomp_table[i].comb1]; \
529 /* See whether we have room for three bytes. */ \
530 if (__builtin_expect (outptr + 2 >= outend, 0)) \
532 result = __GCONV_FULL_OUTPUT; \
536 *outptr++ = (unsigned char) ch; \
537 *outptr++ = comb_table[decomp_table[i].comb1]; \
538 *outptr++ = comb_table[decomp_table[i].comb2]; \
546 /* This is an illegal character. */ \
547 STANDARD_TO_LOOP_ERR_HANDLER (4); \
551 #define LOOP_NEED_FLAGS
552 #define EXTRA_LOOP_DECLS , int *statep
553 #include <iconv/loop.c>
556 /* Now define the toplevel functions. */
557 #include <iconv/skeleton.c>