Mercurial > hg > xemacs-beta
annotate src/unicode.c @ 5891:a0e751d6c3ad
Import the #'clear-string API from GNU, use it in tls.c
src/ChangeLog addition:
2015-04-18 Aidan Kehoe <kehoea@parhasard.net>
* sequence.c (Fclear_string): New, API from GNU. Zero a string's
contents, making sure the text is not kept around even when the
string's data is reallocated because of a changed character
length.
* sequence.c (syms_of_sequence): Make it available to Lisp.
* lisp.h: Make it available to C code.
* tls.c (nss_pk11_password): Use it.
* tls.c (gnutls_pk11_password): Use it.
* tls.c (openssl_password): Use it.
tests/ChangeLog addition:
2015-04-18 Aidan Kehoe <kehoea@parhasard.net>
* automated/lisp-tests.el:
Test #'clear-string, just added. Unfortunately there's no way to
be certain from Lisp that the old password data has been erased
after realloc; it may be worth adding a test to tests.c, but
*we'll be reading memory we shouldn't be*, so that gives me pause.
| author | Aidan Kehoe <kehoea@parhasard.net> |
|---|---|
| date | Sat, 18 Apr 2015 23:00:14 +0100 |
| parents | 65d65b52d608 |
| children | 1b2fdcc3cc5c |
| rev | line source |
|---|---|
| 771 | 1 /* Code to handle Unicode conversion. |
|
4834
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
2 Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005, 2010 Ben Wing. |
| 771 | 3 |
| 4 This file is part of XEmacs. | |
| 5 | |
|
5402
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5157
diff
changeset
|
6 XEmacs is free software: you can redistribute it and/or modify it |
| 771 | 7 under the terms of the GNU General Public License as published by the |
|
5402
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5157
diff
changeset
|
8 Free Software Foundation, either version 3 of the License, or (at your |
|
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5157
diff
changeset
|
9 option) any later version. |
| 771 | 10 |
| 11 XEmacs is distributed in the hope that it will be useful, but WITHOUT | |
| 12 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
| 13 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License | |
| 14 for more details. | |
| 15 | |
| 16 You should have received a copy of the GNU General Public License | |
|
5402
308d34e9f07d
Changed bulk of GPLv2 or later files identified by script
Mats Lidell <matsl@xemacs.org>
parents:
5157
diff
changeset
|
17 along with XEmacs. If not, see <http://www.gnu.org/licenses/>. */ |
| 771 | 18 |
| 19 /* Synched up with: FSF 20.3. Not in FSF. */ | |
| 20 | |
| 21 /* Authorship: | |
| 22 | |
| 23 Current primary author: Ben Wing <ben@xemacs.org> | |
| 24 | |
| 25 Written by Ben Wing <ben@xemacs.org>, June, 2001. | |
| 26 Separated out into this file, August, 2001. | |
| 27 Includes Unicode coding systems, some parts of which have been written | |
| 877 | 28 by someone else. #### Morioka and Hayashi, I think. |
| 771 | 29 |
| 30 As of September 2001, the detection code is here and abstraction of the | |
| 877 | 31 detection system is finished. The unicode detectors have been rewritten |
| 771 | 32 to include multiple levels of likelihood. |
| 33 */ | |
| 34 | |
| 35 #include <config.h> | |
| 36 #include "lisp.h" | |
| 37 | |
| 38 #include "charset.h" | |
| 39 #include "file-coding.h" | |
| 40 #include "opaque.h" | |
| 41 | |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
42 #include "buffer.h" |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
43 #include "rangetab.h" |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
44 #include "extents.h" |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
45 |
| 771 | 46 #include "sysfile.h" |
| 47 | |
| 2367 | 48 /* For more info about how Unicode works under Windows, see intl-win32.c. */ |
| 49 | |
| 50 /* Info about Unicode translation tables [ben]: | |
| 51 | |
| 52 FORMAT: | |
| 53 ------- | |
| 54 | |
| 55 We currently use the following format for tables: | |
| 56 | |
| 57 If dimension == 1, to_unicode_table is a 96-element array of ints | |
| 58 (Unicode code points); else, it's a 96-element array of int * pointers, | |
| 59 each of which points to a 96-element array of ints. If no elements in a | |
| 60 row have been filled in, the pointer will point to a default empty | |
| 61 table; that way, memory usage is more reasonable but lookup still fast. | |
| 62 | |
| 63 -- If from_unicode_levels == 1, from_unicode_table is a 256-element | |
| 64 array of shorts (octet 1 in high byte, octet 2 in low byte; we don't | |
| 65 store Ichars directly to save space). | |
| 66 | |
| 67 -- If from_unicode_levels == 2, from_unicode_table is a 256-element | |
| 68 array of short * pointers, each of which points to a 256-element array | |
| 69 of shorts. | |
| 70 | |
| 71 -- If from_unicode_levels == 3, from_unicode_table is a 256-element | |
| 72 array of short ** pointers, each of which points to a 256-element array | |
| 73 of short * pointers, each of which points to a 256-element array of | |
| 74 shorts. | |
| 75 | |
| 76 -- If from_unicode_levels == 4, same thing but one level deeper. | |
| 77 | |
| 78 Just as for to_unicode_table, we use default tables to fill in all | |
| 79 entries with no values in them. | |
| 80 | |
| 81 #### An obvious space-saving optimization is to use variable-sized | |
| 82 tables, where each table instead of just being a 256-element array, is a | |
| 83 structure with a start value, an end value, and a variable number of | |
| 84 entries (END - START + 1). Only 8 bits are needed for END and START, | |
| 85 and could be stored at the end to avoid alignment problems. However, | |
| 86 before charging off and implementing this, we need to consider whether | |
| 87 it's worth it: | |
| 88 | |
| 89 (1) Most tables will be highly localized in which code points are | |
| 90 defined, heavily reducing the possible memory waste. Before doing any | |
| 91 rewriting, write some code to see how much memory is actually being | |
| 92 wasted (i.e. ratio of empty entries to total # of entries) and only | |
| 93 start rewriting if it's unacceptably high. You have to check over all | |
| 94 charsets. | |
| 95 | |
| 96 (2) Since entries are usually added one at a time, you have to be very | |
| 97 careful when creating the tables to avoid realloc()/free() thrashing in | |
| 98 the common case when you are in an area of high localization and are | |
| 99 going to end up using most entries in the table. You'd certainly want | |
| 100 to allow only certain sizes, not arbitrary ones (probably powers of 2, | |
| 101 where you want the entire block including the START/END values to fit | |
| 102 into a power of 2, minus any malloc overhead if there is any -- there's | |
| 103 none under gmalloc.c, and probably most system malloc() functions are | |
| 104 quite smart nowadays and also have no overhead). You could optimize | |
| 105 somewhat during the in-C initializations, because you can compute the | |
| 106 actual usage of various tables by scanning the entries you're going to | |
| 107 add in a separate pass before adding them. (You could actually do the | |
| 108 same thing when entries are added on the Lisp level by making the | |
| 109 assumption that all the entries will come in one after another before | |
| 110 any use is made of the data. So as they're coming in, you just store | |
| 111 them in a big long list, and the first time you need to retrieve an | |
| 112 entry, you compute the whole table at once.) You'd still have to deal | |
| 113 with the possibility of later entries coming in, though. | |
| 114 | |
| 115 (3) You do lose some speed using START/END values, since you need a | |
| 116 couple of comparisons at each level. This could easily make each single | |
| 117 lookup become 3-4 times slower. The Unicode book considers this a big | |
| 118 issue, and recommends against variable-sized tables for this reason; | |
| 119 however, they almost certainly have in mind applications that primarily | |
| 120 involve conversion of large amounts of data. Most Unicode strings that | |
| 121 are translated in XEmacs are fairly small. The only place where this | |
| 122 might matter is in loading large files -- e.g. a 3-megabyte | |
| 123 Unicode-encoded file. So think about this, and maybe do a trial | |
| 124 implementation where you don't worry too much about the intricacies of | |
| 125 (2) and just implement some basic "multiply by 1.5" trick or something | |
| 126 to do the resizing. There is a very good FAQ on Unicode called | |
| 127 something like the Linux-Unicode How-To (it should be part of the Linux | |
| 128 How-To's, I think), that lists the url of a guy with a whole bunch of | |
| 129 unicode files you can use to stress-test your implementations, and he's | |
| 130 highly likely to have a good multi-megabyte Unicode-encoded file (with | |
| 131 normal text in it -- if you created your own just by creating repeated | |
| 132 strings of letters and numbers, you probably wouldn't get accurate | |
| 133 results). | |
| 134 | |
| 135 INITIALIZATION: | |
| 136 --------------- | |
| 137 | |
| 138 There are advantages and disadvantages to loading the tables at | |
| 139 run-time. | |
| 140 | |
| 141 Advantages: | |
| 142 | |
| 143 They're big, and it's very fast to recreate them (a fraction of a second | |
| 144 on modern processors). | |
| 145 | |
| 146 Disadvantages: | |
| 147 | |
| 148 (1) User-defined charsets: It would be inconvenient to require all | |
| 149 dumped user-defined charsets to be reloaded at init time. | |
| 150 | |
| 151 NB With run-time loading, we load in init-mule-at-startup, in | |
| 152 mule-cmds.el. This is called from startup.el, which is quite late in | |
| 153 the initialization process -- but data-directory isn't set until then. | |
| 154 With dump-time loading, you still can't dump in a Japanese directory | |
| 155 (again, until we move to Unicode internally), but this is not such an | |
| 156 imposition. | |
| 157 | |
| 158 | |
| 159 */ | |
| 160 | |
| 771 | 161 /* #### WARNING! The current sledgehammer routines have a fundamental |
| 162 problem in that they can't handle two characters mapping to a | |
| 163 single Unicode codepoint or vice-versa in a single charset table. | |
| 164 It's not clear there is any way to handle this and still make the | |
| 877 | 165 sledgehammer routines useful. |
| 166 | |
| 167 Inquiring Minds Want To Know Dept: does the above WARNING mean that | |
| 168 _if_ it happens, then it will signal error, or then it will do | |
| 169 something evil and unpredictable? Signaling an error is OK: for | |
| 170 all national standards, the national to Unicode map is an inclusion | |
| 171 (1-to-1). Any character set that does not behave that way is | |
| 1318 | 172 broken according to the Unicode standard. |
| 173 | |
| 2500 | 174 Answer: You will get an ABORT(), since the purpose of the sledgehammer |
| 1318 | 175 routines is self-checking. The above problem with non-1-to-1 mapping |
| 176 occurs in the Big5 tables, as provided by the Unicode Consortium. */ | |
| 877 | 177 |
| 771 | 178 /* #define SLEDGEHAMMER_CHECK_UNICODE */ |
| 179 | |
| 180 /* When MULE is not defined, we may still need some Unicode support -- | |
| 181 in particular, some Windows API's always want Unicode, and the way | |
| 182 we've set up the Unicode encapsulation, we may as well go ahead and | |
| 183 always use the Unicode versions of split API's. (It would be | |
| 184 trickier to not use them, and pointless -- under NT, the ANSI API's | |
| 185 call the Unicode ones anyway, so in the case of structures, we'd be | |
| 186 converting from Unicode to ANSI structures, only to have the OS | |
| 187 convert them back.) */ | |
| 188 | |
| 189 Lisp_Object Qunicode; | |
| 4096 | 190 Lisp_Object Qutf_16, Qutf_8, Qucs_4, Qutf_7, Qutf_32; |
| 771 | 191 Lisp_Object Qneed_bom; |
| 192 | |
| 193 Lisp_Object Qutf_16_little_endian, Qutf_16_bom; | |
| 194 Lisp_Object Qutf_16_little_endian_bom; | |
| 195 | |
| 985 | 196 Lisp_Object Qutf_8_bom; |
| 197 | |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
198 #ifdef MULE |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
199 /* These range tables are not directly accessible from Lisp: */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
200 static Lisp_Object Vunicode_invalid_and_query_skip_chars; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
201 static Lisp_Object Vutf_8_invalid_and_query_skip_chars; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
202 static Lisp_Object Vunicode_query_skip_chars; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
203 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
204 static Lisp_Object Vunicode_query_string, Vunicode_invalid_string, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
205 Vutf_8_invalid_string; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
206 #endif /* MULE */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
207 |
| 3952 | 208 /* See the Unicode FAQ, http://www.unicode.org/faq/utf_bom.html#35 for this |
| 209 algorithm. | |
| 210 | |
| 211 (They also give another, really verbose one, as part of their explanation | |
| 212 of the various planes of the encoding, but we won't use that.) */ | |
| 213 | |
| 214 #define UTF_16_LEAD_OFFSET (0xD800 - (0x10000 >> 10)) | |
| 215 #define UTF_16_SURROGATE_OFFSET (0x10000 - (0xD800 << 10) - 0xDC00) | |
| 216 | |
| 217 #define utf_16_surrogates_to_code(lead, trail) \ | |
| 218 (((lead) << 10) + (trail) + UTF_16_SURROGATE_OFFSET) | |
| 219 | |
| 220 #define CODE_TO_UTF_16_SURROGATES(codepoint, lead, trail) do { \ | |
| 221 int __ctu16s_code = (codepoint); \ | |
| 222 lead = UTF_16_LEAD_OFFSET + (__ctu16s_code >> 10); \ | |
| 223 trail = 0xDC00 + (__ctu16s_code & 0x3FF); \ | |
| 224 } while (0) | |
| 225 | |
| 771 | 226 #ifdef MULE |
| 227 | |
| 3352 | 228 /* Using ints for to_unicode is OK (as long as they are >= 32 bits). |
| 229 In from_unicode, we're converting from Mule characters, which means | |
| 230 that the values being converted to are only 96x96, and we can save | |
| 231 space by using shorts (signedness doesn't matter). */ | |
| 771 | 232 static int *to_unicode_blank_1; |
| 233 static int **to_unicode_blank_2; | |
| 234 | |
| 235 static short *from_unicode_blank_1; | |
| 236 static short **from_unicode_blank_2; | |
| 237 static short ***from_unicode_blank_3; | |
| 238 static short ****from_unicode_blank_4; | |
| 239 | |
| 1204 | 240 static const struct memory_description to_unicode_level_0_desc_1[] = { |
| 771 | 241 { XD_END } |
| 242 }; | |
| 243 | |
| 1204 | 244 static const struct sized_memory_description to_unicode_level_0_desc = { |
| 245 sizeof (int), to_unicode_level_0_desc_1 | |
| 771 | 246 }; |
| 247 | |
| 1204 | 248 static const struct memory_description to_unicode_level_1_desc_1[] = { |
| 2551 | 249 { XD_BLOCK_PTR, 0, 96, { &to_unicode_level_0_desc } }, |
| 771 | 250 { XD_END } |
| 251 }; | |
| 252 | |
| 1204 | 253 static const struct sized_memory_description to_unicode_level_1_desc = { |
| 254 sizeof (void *), to_unicode_level_1_desc_1 | |
| 771 | 255 }; |
| 256 | |
| 1204 | 257 static const struct memory_description to_unicode_description_1[] = { |
| 2551 | 258 { XD_BLOCK_PTR, 1, 96, { &to_unicode_level_0_desc } }, |
| 259 { XD_BLOCK_PTR, 2, 96, { &to_unicode_level_1_desc } }, | |
| 771 | 260 { XD_END } |
| 261 }; | |
| 262 | |
| 263 /* Not static because each charset has a set of to and from tables and | |
| 264 needs to describe them to pdump. */ | |
| 1204 | 265 const struct sized_memory_description to_unicode_description = { |
| 266 sizeof (void *), to_unicode_description_1 | |
| 267 }; | |
| 268 | |
| 2367 | 269 /* Used only for to_unicode_blank_2 */ |
| 270 static const struct memory_description to_unicode_level_2_desc_1[] = { | |
| 2551 | 271 { XD_BLOCK_PTR, 0, 96, { &to_unicode_level_1_desc } }, |
| 2367 | 272 { XD_END } |
| 273 }; | |
| 274 | |
| 1204 | 275 static const struct memory_description from_unicode_level_0_desc_1[] = { |
| 771 | 276 { XD_END } |
| 277 }; | |
| 278 | |
| 1204 | 279 static const struct sized_memory_description from_unicode_level_0_desc = { |
| 280 sizeof (short), from_unicode_level_0_desc_1 | |
| 771 | 281 }; |
| 282 | |
| 1204 | 283 static const struct memory_description from_unicode_level_1_desc_1[] = { |
| 2551 | 284 { XD_BLOCK_PTR, 0, 256, { &from_unicode_level_0_desc } }, |
| 771 | 285 { XD_END } |
| 286 }; | |
| 287 | |
| 1204 | 288 static const struct sized_memory_description from_unicode_level_1_desc = { |
| 289 sizeof (void *), from_unicode_level_1_desc_1 | |
| 771 | 290 }; |
| 291 | |
| 1204 | 292 static const struct memory_description from_unicode_level_2_desc_1[] = { |
| 2551 | 293 { XD_BLOCK_PTR, 0, 256, { &from_unicode_level_1_desc } }, |
| 771 | 294 { XD_END } |
| 295 }; | |
| 296 | |
| 1204 | 297 static const struct sized_memory_description from_unicode_level_2_desc = { |
| 298 sizeof (void *), from_unicode_level_2_desc_1 | |
| 771 | 299 }; |
| 300 | |
| 1204 | 301 static const struct memory_description from_unicode_level_3_desc_1[] = { |
| 2551 | 302 { XD_BLOCK_PTR, 0, 256, { &from_unicode_level_2_desc } }, |
| 771 | 303 { XD_END } |
| 304 }; | |
| 305 | |
| 1204 | 306 static const struct sized_memory_description from_unicode_level_3_desc = { |
| 307 sizeof (void *), from_unicode_level_3_desc_1 | |
| 771 | 308 }; |
| 309 | |
| 1204 | 310 static const struct memory_description from_unicode_description_1[] = { |
| 2551 | 311 { XD_BLOCK_PTR, 1, 256, { &from_unicode_level_0_desc } }, |
| 312 { XD_BLOCK_PTR, 2, 256, { &from_unicode_level_1_desc } }, | |
| 313 { XD_BLOCK_PTR, 3, 256, { &from_unicode_level_2_desc } }, | |
| 314 { XD_BLOCK_PTR, 4, 256, { &from_unicode_level_3_desc } }, | |
| 771 | 315 { XD_END } |
| 316 }; | |
| 317 | |
| 318 /* Not static because each charset has a set of to and from tables and | |
| 319 needs to describe them to pdump. */ | |
| 1204 | 320 const struct sized_memory_description from_unicode_description = { |
| 321 sizeof (void *), from_unicode_description_1 | |
| 771 | 322 }; |
| 323 | |
| 2367 | 324 /* Used only for from_unicode_blank_4 */ |
| 325 static const struct memory_description from_unicode_level_4_desc_1[] = { | |
| 2551 | 326 { XD_BLOCK_PTR, 0, 256, { &from_unicode_level_3_desc } }, |
| 2367 | 327 { XD_END } |
| 328 }; | |
| 329 | |
| 771 | 330 static Lisp_Object_dynarr *unicode_precedence_dynarr; |
| 331 | |
| 1204 | 332 static const struct memory_description lod_description_1[] = { |
| 333 XD_DYNARR_DESC (Lisp_Object_dynarr, &lisp_object_description), | |
| 771 | 334 { XD_END } |
| 335 }; | |
| 336 | |
| 1204 | 337 static const struct sized_memory_description lisp_object_dynarr_description = { |
| 771 | 338 sizeof (Lisp_Object_dynarr), |
| 339 lod_description_1 | |
| 340 }; | |
| 341 | |
| 342 Lisp_Object Vlanguage_unicode_precedence_list; | |
| 343 Lisp_Object Vdefault_unicode_precedence_list; | |
| 344 | |
| 345 Lisp_Object Qignore_first_column; | |
| 346 | |
| 3439 | 347 Lisp_Object Vcurrent_jit_charset; |
| 348 Lisp_Object Qlast_allocated_character; | |
| 349 Lisp_Object Qccl_encode_to_ucs_2; | |
| 350 | |
| 4268 | 351 Lisp_Object Vnumber_of_jit_charsets; |
| 352 Lisp_Object Vlast_jit_charset_final; | |
| 353 Lisp_Object Vcharset_descr; | |
| 354 | |
| 355 | |
| 771 | 356 |
| 357 /************************************************************************/ | |
| 358 /* Unicode implementation */ | |
| 359 /************************************************************************/ | |
| 360 | |
| 361 #define BREAKUP_UNICODE_CODE(val, u1, u2, u3, u4, levels) \ | |
| 362 do { \ | |
| 363 int buc_val = (val); \ | |
| 364 \ | |
| 365 (u1) = buc_val >> 24; \ | |
| 366 (u2) = (buc_val >> 16) & 255; \ | |
| 367 (u3) = (buc_val >> 8) & 255; \ | |
| 368 (u4) = buc_val & 255; \ | |
| 369 (levels) = (buc_val <= 0xFF ? 1 : \ | |
| 370 buc_val <= 0xFFFF ? 2 : \ | |
| 371 buc_val <= 0xFFFFFF ? 3 : \ | |
| 372 4); \ | |
| 373 } while (0) | |
| 374 | |
| 375 static void | |
| 376 init_blank_unicode_tables (void) | |
| 377 { | |
| 378 int i; | |
| 379 | |
| 380 from_unicode_blank_1 = xnew_array (short, 256); | |
| 381 from_unicode_blank_2 = xnew_array (short *, 256); | |
| 382 from_unicode_blank_3 = xnew_array (short **, 256); | |
| 383 from_unicode_blank_4 = xnew_array (short ***, 256); | |
| 384 for (i = 0; i < 256; i++) | |
| 385 { | |
| 877 | 386 /* #### IMWTK: Why does using -1 here work? Simply because there are |
| 1318 | 387 no existing 96x96 charsets? |
| 388 | |
| 389 Answer: I don't understand the concern. -1 indicates there is no | |
| 390 entry for this particular codepoint, which is always the case for | |
| 391 blank tables. */ | |
| 771 | 392 from_unicode_blank_1[i] = (short) -1; |
| 393 from_unicode_blank_2[i] = from_unicode_blank_1; | |
| 394 from_unicode_blank_3[i] = from_unicode_blank_2; | |
| 395 from_unicode_blank_4[i] = from_unicode_blank_3; | |
| 396 } | |
| 397 | |
| 398 to_unicode_blank_1 = xnew_array (int, 96); | |
| 399 to_unicode_blank_2 = xnew_array (int *, 96); | |
| 400 for (i = 0; i < 96; i++) | |
| 401 { | |
| 877 | 402 /* Here -1 is guaranteed OK. */ |
| 771 | 403 to_unicode_blank_1[i] = -1; |
| 404 to_unicode_blank_2[i] = to_unicode_blank_1; | |
| 405 } | |
| 406 } | |
| 407 | |
| 408 static void * | |
| 409 create_new_from_unicode_table (int level) | |
| 410 { | |
| 411 switch (level) | |
| 412 { | |
| 413 /* WARNING: If you are thinking of compressing these, keep in | |
| 414 mind that sizeof (short) does not equal sizeof (short *). */ | |
| 415 case 1: | |
| 416 { | |
| 417 short *newtab = xnew_array (short, 256); | |
| 418 memcpy (newtab, from_unicode_blank_1, 256 * sizeof (short)); | |
| 419 return newtab; | |
| 420 } | |
| 421 case 2: | |
| 422 { | |
| 423 short **newtab = xnew_array (short *, 256); | |
| 424 memcpy (newtab, from_unicode_blank_2, 256 * sizeof (short *)); | |
| 425 return newtab; | |
| 426 } | |
| 427 case 3: | |
| 428 { | |
| 429 short ***newtab = xnew_array (short **, 256); | |
| 430 memcpy (newtab, from_unicode_blank_3, 256 * sizeof (short **)); | |
| 431 return newtab; | |
| 432 } | |
| 433 case 4: | |
| 434 { | |
| 435 short ****newtab = xnew_array (short ***, 256); | |
| 436 memcpy (newtab, from_unicode_blank_4, 256 * sizeof (short ***)); | |
| 437 return newtab; | |
| 438 } | |
| 439 default: | |
| 2500 | 440 ABORT (); |
| 771 | 441 return 0; |
| 442 } | |
| 443 } | |
| 444 | |
| 877 | 445 /* Allocate and blank the tables. |
| 1318 | 446 Loading them up is done by load-unicode-mapping-table. */ |
| 771 | 447 void |
| 448 init_charset_unicode_tables (Lisp_Object charset) | |
| 449 { | |
| 450 if (XCHARSET_DIMENSION (charset) == 1) | |
| 451 { | |
| 452 int *to_table = xnew_array (int, 96); | |
| 453 memcpy (to_table, to_unicode_blank_1, 96 * sizeof (int)); | |
| 454 XCHARSET_TO_UNICODE_TABLE (charset) = to_table; | |
| 455 } | |
| 456 else | |
| 457 { | |
| 458 int **to_table = xnew_array (int *, 96); | |
| 459 memcpy (to_table, to_unicode_blank_2, 96 * sizeof (int *)); | |
| 460 XCHARSET_TO_UNICODE_TABLE (charset) = to_table; | |
| 461 } | |
| 462 | |
| 463 { | |
| 2367 | 464 XCHARSET_FROM_UNICODE_TABLE (charset) = |
| 465 create_new_from_unicode_table (1); | |
| 771 | 466 XCHARSET_FROM_UNICODE_LEVELS (charset) = 1; |
| 467 } | |
| 468 } | |
| 469 | |
| 470 static void | |
| 471 free_from_unicode_table (void *table, int level) | |
| 472 { | |
| 473 int i; | |
| 474 | |
| 475 switch (level) | |
| 476 { | |
| 477 case 2: | |
| 478 { | |
| 479 short **tab = (short **) table; | |
| 480 for (i = 0; i < 256; i++) | |
| 481 { | |
| 482 if (tab[i] != from_unicode_blank_1) | |
| 483 free_from_unicode_table (tab[i], 1); | |
| 484 } | |
| 485 break; | |
| 486 } | |
| 487 case 3: | |
| 488 { | |
| 489 short ***tab = (short ***) table; | |
| 490 for (i = 0; i < 256; i++) | |
| 491 { | |
| 492 if (tab[i] != from_unicode_blank_2) | |
| 493 free_from_unicode_table (tab[i], 2); | |
| 494 } | |
| 495 break; | |
| 496 } | |
| 497 case 4: | |
| 498 { | |
| 499 short ****tab = (short ****) table; | |
| 500 for (i = 0; i < 256; i++) | |
| 501 { | |
| 502 if (tab[i] != from_unicode_blank_3) | |
| 503 free_from_unicode_table (tab[i], 3); | |
| 504 } | |
| 505 break; | |
| 506 } | |
| 507 } | |
| 508 | |
|
4976
16112448d484
Rename xfree(FOO, TYPE) -> xfree(FOO)
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
509 xfree (table); |
| 771 | 510 } |
| 511 | |
| 512 static void | |
| 513 free_to_unicode_table (void *table, int level) | |
| 514 { | |
| 515 if (level == 2) | |
| 516 { | |
| 517 int i; | |
| 518 int **tab = (int **) table; | |
| 519 | |
| 520 for (i = 0; i < 96; i++) | |
| 521 { | |
| 522 if (tab[i] != to_unicode_blank_1) | |
| 523 free_to_unicode_table (tab[i], 1); | |
| 524 } | |
| 525 } | |
| 526 | |
|
4976
16112448d484
Rename xfree(FOO, TYPE) -> xfree(FOO)
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
527 xfree (table); |
| 771 | 528 } |
| 529 | |
| 530 void | |
| 531 free_charset_unicode_tables (Lisp_Object charset) | |
| 532 { | |
| 533 free_to_unicode_table (XCHARSET_TO_UNICODE_TABLE (charset), | |
| 534 XCHARSET_DIMENSION (charset)); | |
| 535 free_from_unicode_table (XCHARSET_FROM_UNICODE_TABLE (charset), | |
| 536 XCHARSET_FROM_UNICODE_LEVELS (charset)); | |
| 537 } | |
| 538 | |
| 539 #ifdef MEMORY_USAGE_STATS | |
| 540 | |
| 541 static Bytecount | |
| 542 compute_from_unicode_table_size_1 (void *table, int level, | |
|
5157
1fae11d56ad2
redo memory-usage mechanism, add way of dynamically initializing Lisp objects
Ben Wing <ben@xemacs.org>
parents:
4976
diff
changeset
|
543 struct usage_stats *stats) |
| 771 | 544 { |
| 545 int i; | |
| 546 Bytecount size = 0; | |
| 547 | |
| 548 switch (level) | |
| 549 { | |
| 550 case 2: | |
| 551 { | |
| 552 short **tab = (short **) table; | |
| 553 for (i = 0; i < 256; i++) | |
| 554 { | |
| 555 if (tab[i] != from_unicode_blank_1) | |
| 556 size += compute_from_unicode_table_size_1 (tab[i], 1, stats); | |
| 557 } | |
| 558 break; | |
| 559 } | |
| 560 case 3: | |
| 561 { | |
| 562 short ***tab = (short ***) table; | |
| 563 for (i = 0; i < 256; i++) | |
| 564 { | |
| 565 if (tab[i] != from_unicode_blank_2) | |
| 566 size += compute_from_unicode_table_size_1 (tab[i], 2, stats); | |
| 567 } | |
| 568 break; | |
| 569 } | |
| 570 case 4: | |
| 571 { | |
| 572 short ****tab = (short ****) table; | |
| 573 for (i = 0; i < 256; i++) | |
| 574 { | |
| 575 if (tab[i] != from_unicode_blank_3) | |
| 576 size += compute_from_unicode_table_size_1 (tab[i], 3, stats); | |
| 577 } | |
| 578 break; | |
| 579 } | |
| 580 } | |
| 581 | |
| 3024 | 582 size += malloced_storage_size (table, |
| 771 | 583 256 * (level == 1 ? sizeof (short) : |
| 584 sizeof (void *)), | |
| 585 stats); | |
| 586 return size; | |
| 587 } | |
| 588 | |
| 589 static Bytecount | |
| 590 compute_to_unicode_table_size_1 (void *table, int level, | |
|
5157
1fae11d56ad2
redo memory-usage mechanism, add way of dynamically initializing Lisp objects
Ben Wing <ben@xemacs.org>
parents:
4976
diff
changeset
|
591 struct usage_stats *stats) |
| 771 | 592 { |
| 593 Bytecount size = 0; | |
| 594 | |
| 595 if (level == 2) | |
| 596 { | |
| 597 int i; | |
| 598 int **tab = (int **) table; | |
| 599 | |
| 600 for (i = 0; i < 96; i++) | |
| 601 { | |
| 602 if (tab[i] != to_unicode_blank_1) | |
| 603 size += compute_to_unicode_table_size_1 (tab[i], 1, stats); | |
| 604 } | |
| 605 } | |
| 606 | |
| 3024 | 607 size += malloced_storage_size (table, |
| 771 | 608 96 * (level == 1 ? sizeof (int) : |
| 609 sizeof (void *)), | |
| 610 stats); | |
| 611 return size; | |
| 612 } | |
| 613 | |
| 614 Bytecount | |
| 615 compute_from_unicode_table_size (Lisp_Object charset, | |
|
5157
1fae11d56ad2
redo memory-usage mechanism, add way of dynamically initializing Lisp objects
Ben Wing <ben@xemacs.org>
parents:
4976
diff
changeset
|
616 struct usage_stats *stats) |
| 771 | 617 { |
| 618 return (compute_from_unicode_table_size_1 | |
| 619 (XCHARSET_FROM_UNICODE_TABLE (charset), | |
| 620 XCHARSET_FROM_UNICODE_LEVELS (charset), | |
| 621 stats)); | |
| 622 } | |
| 623 | |
| 624 Bytecount | |
| 625 compute_to_unicode_table_size (Lisp_Object charset, | |
|
5157
1fae11d56ad2
redo memory-usage mechanism, add way of dynamically initializing Lisp objects
Ben Wing <ben@xemacs.org>
parents:
4976
diff
changeset
|
626 struct usage_stats *stats) |
| 771 | 627 { |
| 628 return (compute_to_unicode_table_size_1 | |
| 629 (XCHARSET_TO_UNICODE_TABLE (charset), | |
| 630 XCHARSET_DIMENSION (charset), | |
| 631 stats)); | |
| 632 } | |
| 633 | |
| 634 #endif | |
| 635 | |
| 636 #ifdef SLEDGEHAMMER_CHECK_UNICODE | |
| 637 | |
| 638 /* "Sledgehammer checks" are checks that verify the self-consistency | |
| 639 of an entire structure every time a change is about to be made or | |
| 640 has been made to the structure. Not fast but a pretty much | |
| 641 sure-fire way of flushing out any incorrectnesses in the algorithms | |
| 642 that create the structure. | |
| 643 | |
| 644 Checking only after a change has been made will speed things up by | |
| 645 a factor of 2, but it doesn't absolutely prove that the code just | |
| 646 checked caused the problem; perhaps it happened elsewhere, either | |
| 647 in some code you forgot to sledgehammer check or as a result of | |
| 648 data corruption. */ | |
| 649 | |
| 650 static void | |
| 651 assert_not_any_blank_table (void *tab) | |
| 652 { | |
| 653 assert (tab != from_unicode_blank_1); | |
| 654 assert (tab != from_unicode_blank_2); | |
| 655 assert (tab != from_unicode_blank_3); | |
| 656 assert (tab != from_unicode_blank_4); | |
| 657 assert (tab != to_unicode_blank_1); | |
| 658 assert (tab != to_unicode_blank_2); | |
| 659 assert (tab); | |
| 660 } | |
| 661 | |
| 662 static void | |
| 663 sledgehammer_check_from_table (Lisp_Object charset, void *table, int level, | |
| 664 int codetop) | |
| 665 { | |
| 666 int i; | |
| 667 | |
| 668 switch (level) | |
| 669 { | |
| 670 case 1: | |
| 671 { | |
| 672 short *tab = (short *) table; | |
| 673 for (i = 0; i < 256; i++) | |
| 674 { | |
| 675 if (tab[i] != -1) | |
| 676 { | |
| 677 Lisp_Object char_charset; | |
| 678 int c1, c2; | |
| 679 | |
| 867 | 680 assert (valid_ichar_p (tab[i])); |
| 681 BREAKUP_ICHAR (tab[i], char_charset, c1, c2); | |
| 771 | 682 assert (EQ (charset, char_charset)); |
| 683 if (XCHARSET_DIMENSION (charset) == 1) | |
| 684 { | |
| 685 int *to_table = | |
| 686 (int *) XCHARSET_TO_UNICODE_TABLE (charset); | |
| 687 assert_not_any_blank_table (to_table); | |
| 688 assert (to_table[c1 - 32] == (codetop << 8) + i); | |
| 689 } | |
| 690 else | |
| 691 { | |
| 692 int **to_table = | |
| 693 (int **) XCHARSET_TO_UNICODE_TABLE (charset); | |
| 694 assert_not_any_blank_table (to_table); | |
| 695 assert_not_any_blank_table (to_table[c1 - 32]); | |
| 696 assert (to_table[c1 - 32][c2 - 32] == (codetop << 8) + i); | |
| 697 } | |
| 698 } | |
| 699 } | |
| 700 break; | |
| 701 } | |
| 702 case 2: | |
| 703 { | |
| 704 short **tab = (short **) table; | |
| 705 for (i = 0; i < 256; i++) | |
| 706 { | |
| 707 if (tab[i] != from_unicode_blank_1) | |
| 708 sledgehammer_check_from_table (charset, tab[i], 1, | |
| 709 (codetop << 8) + i); | |
| 710 } | |
| 711 break; | |
| 712 } | |
| 713 case 3: | |
| 714 { | |
| 715 short ***tab = (short ***) table; | |
| 716 for (i = 0; i < 256; i++) | |
| 717 { | |
| 718 if (tab[i] != from_unicode_blank_2) | |
| 719 sledgehammer_check_from_table (charset, tab[i], 2, | |
| 720 (codetop << 8) + i); | |
| 721 } | |
| 722 break; | |
| 723 } | |
| 724 case 4: | |
| 725 { | |
| 726 short ****tab = (short ****) table; | |
| 727 for (i = 0; i < 256; i++) | |
| 728 { | |
| 729 if (tab[i] != from_unicode_blank_3) | |
| 730 sledgehammer_check_from_table (charset, tab[i], 3, | |
| 731 (codetop << 8) + i); | |
| 732 } | |
| 733 break; | |
| 734 } | |
| 735 default: | |
| 2500 | 736 ABORT (); |
| 771 | 737 } |
| 738 } | |
| 739 | |
| 740 static void | |
| 741 sledgehammer_check_to_table (Lisp_Object charset, void *table, int level, | |
| 742 int codetop) | |
| 743 { | |
| 744 int i; | |
| 745 | |
| 746 switch (level) | |
| 747 { | |
| 748 case 1: | |
| 749 { | |
| 750 int *tab = (int *) table; | |
| 751 | |
| 752 if (XCHARSET_CHARS (charset) == 94) | |
| 753 { | |
| 754 assert (tab[0] == -1); | |
| 755 assert (tab[95] == -1); | |
| 756 } | |
| 757 | |
| 758 for (i = 0; i < 96; i++) | |
| 759 { | |
| 760 if (tab[i] != -1) | |
| 761 { | |
| 762 int u4, u3, u2, u1, levels; | |
| 867 | 763 Ichar ch; |
| 764 Ichar this_ch; | |
| 771 | 765 short val; |
| 766 void *frtab = XCHARSET_FROM_UNICODE_TABLE (charset); | |
| 767 | |
| 768 if (XCHARSET_DIMENSION (charset) == 1) | |
| 867 | 769 this_ch = make_ichar (charset, i + 32, 0); |
| 771 | 770 else |
| 867 | 771 this_ch = make_ichar (charset, codetop + 32, i + 32); |
| 771 | 772 |
| 773 assert (tab[i] >= 0); | |
| 774 BREAKUP_UNICODE_CODE (tab[i], u4, u3, u2, u1, levels); | |
| 775 assert (levels <= XCHARSET_FROM_UNICODE_LEVELS (charset)); | |
| 776 | |
| 777 switch (XCHARSET_FROM_UNICODE_LEVELS (charset)) | |
| 778 { | |
| 779 case 1: val = ((short *) frtab)[u1]; break; | |
| 780 case 2: val = ((short **) frtab)[u2][u1]; break; | |
| 781 case 3: val = ((short ***) frtab)[u3][u2][u1]; break; | |
| 782 case 4: val = ((short ****) frtab)[u4][u3][u2][u1]; break; | |
| 2500 | 783 default: ABORT (); |
| 771 | 784 } |
| 785 | |
| 867 | 786 ch = make_ichar (charset, val >> 8, val & 0xFF); |
| 771 | 787 assert (ch == this_ch); |
| 788 | |
| 789 switch (XCHARSET_FROM_UNICODE_LEVELS (charset)) | |
| 790 { | |
| 791 case 4: | |
| 792 assert_not_any_blank_table (frtab); | |
| 793 frtab = ((short ****) frtab)[u4]; | |
| 794 /* fall through */ | |
| 795 case 3: | |
| 796 assert_not_any_blank_table (frtab); | |
| 797 frtab = ((short ***) frtab)[u3]; | |
| 798 /* fall through */ | |
| 799 case 2: | |
| 800 assert_not_any_blank_table (frtab); | |
| 801 frtab = ((short **) frtab)[u2]; | |
| 802 /* fall through */ | |
| 803 case 1: | |
| 804 assert_not_any_blank_table (frtab); | |
| 805 break; | |
| 2500 | 806 default: ABORT (); |
| 771 | 807 } |
| 808 } | |
| 809 } | |
| 810 break; | |
| 811 } | |
| 812 case 2: | |
| 813 { | |
| 814 int **tab = (int **) table; | |
| 815 | |
| 816 if (XCHARSET_CHARS (charset) == 94) | |
| 817 { | |
| 818 assert (tab[0] == to_unicode_blank_1); | |
| 819 assert (tab[95] == to_unicode_blank_1); | |
| 820 } | |
| 821 | |
| 822 for (i = 0; i < 96; i++) | |
| 823 { | |
| 824 if (tab[i] != to_unicode_blank_1) | |
| 825 sledgehammer_check_to_table (charset, tab[i], 1, i); | |
| 826 } | |
| 827 break; | |
| 828 } | |
| 829 default: | |
| 2500 | 830 ABORT (); |
| 771 | 831 } |
| 832 } | |
| 833 | |
| 834 static void | |
| 835 sledgehammer_check_unicode_tables (Lisp_Object charset) | |
| 836 { | |
| 837 /* verify that the blank tables have not been modified */ | |
| 838 int i; | |
| 839 int from_level = XCHARSET_FROM_UNICODE_LEVELS (charset); | |
| 840 int to_level = XCHARSET_FROM_UNICODE_LEVELS (charset); | |
| 841 | |
| 842 for (i = 0; i < 256; i++) | |
| 843 { | |
| 844 assert (from_unicode_blank_1[i] == (short) -1); | |
| 845 assert (from_unicode_blank_2[i] == from_unicode_blank_1); | |
| 846 assert (from_unicode_blank_3[i] == from_unicode_blank_2); | |
| 847 assert (from_unicode_blank_4[i] == from_unicode_blank_3); | |
| 848 } | |
| 849 | |
| 850 for (i = 0; i < 96; i++) | |
| 851 { | |
| 852 assert (to_unicode_blank_1[i] == -1); | |
| 853 assert (to_unicode_blank_2[i] == to_unicode_blank_1); | |
| 854 } | |
| 855 | |
| 856 assert (from_level >= 1 && from_level <= 4); | |
| 857 | |
| 858 sledgehammer_check_from_table (charset, | |
| 859 XCHARSET_FROM_UNICODE_TABLE (charset), | |
| 860 from_level, 0); | |
| 861 | |
| 862 sledgehammer_check_to_table (charset, | |
| 863 XCHARSET_TO_UNICODE_TABLE (charset), | |
| 864 XCHARSET_DIMENSION (charset), 0); | |
| 865 } | |
| 866 | |
| 867 #endif /* SLEDGEHAMMER_CHECK_UNICODE */ | |
| 868 | |
| 869 static void | |
| 867 | 870 set_unicode_conversion (Ichar chr, int code) |
| 771 | 871 { |
| 872 Lisp_Object charset; | |
| 873 int c1, c2; | |
| 874 | |
| 867 | 875 BREAKUP_ICHAR (chr, charset, c1, c2); |
| 771 | 876 |
| 877 | 877 /* I tried an assert on code > 255 || chr == code, but that fails because |
| 878 Mule gives many Latin characters separate code points for different | |
| 879 ISO 8859 coded character sets. Obvious in hindsight.... */ | |
| 880 assert (!EQ (charset, Vcharset_ascii) || chr == code); | |
| 881 assert (!EQ (charset, Vcharset_latin_iso8859_1) || chr == code); | |
| 882 assert (!EQ (charset, Vcharset_control_1) || chr == code); | |
| 883 | |
| 884 /* This assert is needed because it is simply unimplemented. */ | |
| 771 | 885 assert (!EQ (charset, Vcharset_composite)); |
| 886 | |
| 887 #ifdef SLEDGEHAMMER_CHECK_UNICODE | |
| 888 sledgehammer_check_unicode_tables (charset); | |
| 889 #endif | |
| 890 | |
| 2704 | 891 if (EQ(charset, Vcharset_ascii) || EQ(charset, Vcharset_control_1)) |
| 892 return; | |
| 893 | |
| 771 | 894 /* First, the char -> unicode translation */ |
| 895 | |
| 896 if (XCHARSET_DIMENSION (charset) == 1) | |
| 897 { | |
| 898 int *to_table = (int *) XCHARSET_TO_UNICODE_TABLE (charset); | |
| 899 to_table[c1 - 32] = code; | |
| 900 } | |
| 901 else | |
| 902 { | |
| 903 int **to_table_2 = (int **) XCHARSET_TO_UNICODE_TABLE (charset); | |
| 904 int *to_table_1; | |
| 905 | |
| 906 assert (XCHARSET_DIMENSION (charset) == 2); | |
| 907 to_table_1 = to_table_2[c1 - 32]; | |
| 908 if (to_table_1 == to_unicode_blank_1) | |
| 909 { | |
| 910 to_table_1 = xnew_array (int, 96); | |
| 911 memcpy (to_table_1, to_unicode_blank_1, 96 * sizeof (int)); | |
| 912 to_table_2[c1 - 32] = to_table_1; | |
| 913 } | |
| 914 to_table_1[c2 - 32] = code; | |
| 915 } | |
| 916 | |
| 917 /* Then, unicode -> char: much harder */ | |
| 918 | |
| 919 { | |
| 920 int charset_levels; | |
| 921 int u4, u3, u2, u1; | |
| 922 int code_levels; | |
| 923 BREAKUP_UNICODE_CODE (code, u4, u3, u2, u1, code_levels); | |
| 924 | |
| 925 charset_levels = XCHARSET_FROM_UNICODE_LEVELS (charset); | |
| 926 | |
| 927 /* Make sure the charset's tables have at least as many levels as | |
| 928 the code point has: Note that the charset is guaranteed to have | |
| 929 at least one level, because it was created that way */ | |
| 930 if (charset_levels < code_levels) | |
| 931 { | |
| 932 int i; | |
| 933 | |
| 934 assert (charset_levels > 0); | |
| 935 for (i = 2; i <= code_levels; i++) | |
| 936 { | |
| 937 if (charset_levels < i) | |
| 938 { | |
| 939 void *old_table = XCHARSET_FROM_UNICODE_TABLE (charset); | |
| 940 void *table = create_new_from_unicode_table (i); | |
| 941 XCHARSET_FROM_UNICODE_TABLE (charset) = table; | |
| 942 | |
| 943 switch (i) | |
| 944 { | |
| 945 case 2: | |
| 946 ((short **) table)[0] = (short *) old_table; | |
| 947 break; | |
| 948 case 3: | |
| 949 ((short ***) table)[0] = (short **) old_table; | |
| 950 break; | |
| 951 case 4: | |
| 952 ((short ****) table)[0] = (short ***) old_table; | |
| 953 break; | |
| 2500 | 954 default: ABORT (); |
| 771 | 955 } |
| 956 } | |
| 957 } | |
| 958 | |
| 959 charset_levels = code_levels; | |
| 960 XCHARSET_FROM_UNICODE_LEVELS (charset) = code_levels; | |
| 961 } | |
| 962 | |
| 963 /* Now, make sure there is a non-default table at each level */ | |
| 964 { | |
| 965 int i; | |
| 966 void *table = XCHARSET_FROM_UNICODE_TABLE (charset); | |
| 967 | |
| 968 for (i = charset_levels; i >= 2; i--) | |
| 969 { | |
| 970 switch (i) | |
| 971 { | |
| 972 case 4: | |
| 973 if (((short ****) table)[u4] == from_unicode_blank_3) | |
| 974 ((short ****) table)[u4] = | |
| 975 ((short ***) create_new_from_unicode_table (3)); | |
| 976 table = ((short ****) table)[u4]; | |
| 977 break; | |
| 978 case 3: | |
| 979 if (((short ***) table)[u3] == from_unicode_blank_2) | |
| 980 ((short ***) table)[u3] = | |
| 981 ((short **) create_new_from_unicode_table (2)); | |
| 982 table = ((short ***) table)[u3]; | |
| 983 break; | |
| 984 case 2: | |
| 985 if (((short **) table)[u2] == from_unicode_blank_1) | |
| 986 ((short **) table)[u2] = | |
| 987 ((short *) create_new_from_unicode_table (1)); | |
| 988 table = ((short **) table)[u2]; | |
| 989 break; | |
| 2500 | 990 default: ABORT (); |
| 771 | 991 } |
| 992 } | |
| 993 } | |
| 994 | |
| 995 /* Finally, set the character */ | |
| 996 | |
| 997 { | |
| 998 void *table = XCHARSET_FROM_UNICODE_TABLE (charset); | |
| 999 switch (charset_levels) | |
| 1000 { | |
| 1001 case 1: ((short *) table)[u1] = (c1 << 8) + c2; break; | |
| 1002 case 2: ((short **) table)[u2][u1] = (c1 << 8) + c2; break; | |
| 1003 case 3: ((short ***) table)[u3][u2][u1] = (c1 << 8) + c2; break; | |
| 1004 case 4: ((short ****) table)[u4][u3][u2][u1] = (c1 << 8) + c2; break; | |
| 2500 | 1005 default: ABORT (); |
| 771 | 1006 } |
| 1007 } | |
| 1008 } | |
| 1009 | |
| 1010 #ifdef SLEDGEHAMMER_CHECK_UNICODE | |
| 1011 sledgehammer_check_unicode_tables (charset); | |
| 1012 #endif | |
| 1013 } | |
| 1014 | |
| 788 | 1015 int |
| 867 | 1016 ichar_to_unicode (Ichar chr) |
| 771 | 1017 { |
| 1018 Lisp_Object charset; | |
| 1019 int c1, c2; | |
| 1020 | |
| 867 | 1021 type_checking_assert (valid_ichar_p (chr)); |
| 877 | 1022 /* This shortcut depends on the representation of an Ichar, see text.c. */ |
| 771 | 1023 if (chr < 256) |
| 1024 return (int) chr; | |
| 1025 | |
| 867 | 1026 BREAKUP_ICHAR (chr, charset, c1, c2); |
| 771 | 1027 if (EQ (charset, Vcharset_composite)) |
| 1028 return -1; /* #### don't know how to handle */ | |
| 1029 else if (XCHARSET_DIMENSION (charset) == 1) | |
| 1030 return ((int *) XCHARSET_TO_UNICODE_TABLE (charset))[c1 - 32]; | |
| 1031 else | |
| 1032 return ((int **) XCHARSET_TO_UNICODE_TABLE (charset))[c1 - 32][c2 - 32]; | |
| 1033 } | |
| 1034 | |
| 867 | 1035 static Ichar |
| 3439 | 1036 get_free_codepoint(Lisp_Object charset) |
| 1037 { | |
| 1038 Lisp_Object name = Fcharset_name(charset); | |
| 1039 Lisp_Object zeichen = Fget(name, Qlast_allocated_character, Qnil); | |
| 1040 Ichar res; | |
| 1041 | |
| 1042 /* Only allow this with the 96x96 character sets we are using for | |
| 1043 temporary Unicode support. */ | |
| 1044 assert(2 == XCHARSET_DIMENSION(charset) && 96 == XCHARSET_CHARS(charset)); | |
| 1045 | |
| 1046 if (!NILP(zeichen)) | |
| 1047 { | |
| 1048 int c1, c2; | |
| 1049 | |
| 1050 BREAKUP_ICHAR(XCHAR(zeichen), charset, c1, c2); | |
| 1051 | |
| 1052 if (127 == c1 && 127 == c2) | |
| 1053 { | |
| 1054 /* We've already used the hightest-numbered character in this | |
| 1055 set--tell our caller to create another. */ | |
| 1056 return -1; | |
| 1057 } | |
| 1058 | |
| 1059 if (127 == c2) | |
| 1060 { | |
| 1061 ++c1; | |
| 1062 c2 = 0x20; | |
| 1063 } | |
| 1064 else | |
| 1065 { | |
| 1066 ++c2; | |
| 1067 } | |
| 1068 | |
| 1069 res = make_ichar(charset, c1, c2); | |
| 1070 Fput(name, Qlast_allocated_character, make_char(res)); | |
| 1071 } | |
| 1072 else | |
| 1073 { | |
| 1074 res = make_ichar(charset, 32, 32); | |
| 1075 Fput(name, Qlast_allocated_character, make_char(res)); | |
| 1076 } | |
| 1077 return res; | |
| 1078 } | |
| 1079 | |
| 1080 /* The just-in-time creation of XEmacs characters that correspond to unknown | |
| 1081 Unicode code points happens when: | |
| 1082 | |
| 1083 1. The lookup would otherwise fail. | |
| 1084 | |
| 1085 2. The charsets array is the nil or the default. | |
| 1086 | |
| 1087 If there are no free code points in the just-in-time Unicode character | |
| 1088 set, and the charsets array is the default unicode precedence list, | |
| 1089 create a new just-in-time Unicode character set, add it at the end of the | |
| 1090 unicode precedence list, create the XEmacs character in that character | |
| 1091 set, and return it. */ | |
| 1092 | |
| 1093 static Ichar | |
| 877 | 1094 unicode_to_ichar (int code, Lisp_Object_dynarr *charsets) |
| 771 | 1095 { |
| 1096 int u1, u2, u3, u4; | |
| 1097 int code_levels; | |
| 1098 int i; | |
| 1099 int n = Dynarr_length (charsets); | |
| 1100 | |
| 1101 type_checking_assert (code >= 0); | |
| 877 | 1102 /* This shortcut depends on the representation of an Ichar, see text.c. |
| 1103 Note that it may _not_ be extended to U+00A0 to U+00FF (many ISO 8859 | |
| 893 | 1104 coded character sets have points that map into that region, so this |
| 1105 function is many-valued). */ | |
| 877 | 1106 if (code < 0xA0) |
| 867 | 1107 return (Ichar) code; |
| 771 | 1108 |
| 1109 BREAKUP_UNICODE_CODE (code, u4, u3, u2, u1, code_levels); | |
| 1110 | |
| 1111 for (i = 0; i < n; i++) | |
| 1112 { | |
| 1113 Lisp_Object charset = Dynarr_at (charsets, i); | |
| 1114 int charset_levels = XCHARSET_FROM_UNICODE_LEVELS (charset); | |
| 1115 if (charset_levels >= code_levels) | |
| 1116 { | |
| 1117 void *table = XCHARSET_FROM_UNICODE_TABLE (charset); | |
| 1118 short retval; | |
| 1119 | |
| 1120 switch (charset_levels) | |
| 1121 { | |
| 1122 case 1: retval = ((short *) table)[u1]; break; | |
| 1123 case 2: retval = ((short **) table)[u2][u1]; break; | |
| 1124 case 3: retval = ((short ***) table)[u3][u2][u1]; break; | |
| 1125 case 4: retval = ((short ****) table)[u4][u3][u2][u1]; break; | |
| 2500 | 1126 default: ABORT (); retval = 0; |
| 771 | 1127 } |
| 1128 | |
| 1129 if (retval != -1) | |
| 867 | 1130 return make_ichar (charset, retval >> 8, retval & 0xFF); |
| 771 | 1131 } |
| 1132 } | |
| 3439 | 1133 |
| 1134 /* Only do the magic just-in-time assignment if we're using the default | |
| 1135 list. */ | |
| 1136 if (unicode_precedence_dynarr == charsets) | |
| 1137 { | |
| 1138 if (NILP (Vcurrent_jit_charset) || | |
| 1139 (-1 == (i = get_free_codepoint(Vcurrent_jit_charset)))) | |
| 1140 { | |
| 3452 | 1141 Ibyte setname[32]; |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1142 int number_of_jit_charsets = XFIXNUM (Vnumber_of_jit_charsets); |
| 4268 | 1143 Ascbyte last_jit_charset_final = XCHAR (Vlast_jit_charset_final); |
| 1144 | |
| 1145 /* This final byte shit is, umm, not that cool. */ | |
| 1146 assert (last_jit_charset_final >= 0x30); | |
| 3439 | 1147 |
| 3452 | 1148 /* Assertion added partly because our Win32 layer doesn't |
| 1149 support snprintf; with this, we're sure it won't overflow | |
| 1150 the buffer. */ | |
| 1151 assert(100 > number_of_jit_charsets); | |
| 1152 | |
| 4268 | 1153 qxesprintf(setname, "jit-ucs-charset-%d", number_of_jit_charsets); |
| 1154 | |
| 3439 | 1155 Vcurrent_jit_charset = Fmake_charset |
|
4953
304aebb79cd3
function renamings to track names of char typedefs
Ben Wing <ben@xemacs.org>
parents:
4952
diff
changeset
|
1156 (intern_istring (setname), Vcharset_descr, |
| 3439 | 1157 /* Set encode-as-utf-8 to t, to have this character set written |
| 1158 using UTF-8 escapes in escape-quoted and ctext. This | |
| 1159 sidesteps the fact that our internal character -> Unicode | |
| 1160 mapping is not stable from one invocation to the next. */ | |
| 1161 nconc2 (list2(Qencode_as_utf_8, Qt), | |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1162 nconc2 (list6(Qcolumns, make_fixnum(1), Qchars, make_fixnum(96), |
|
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1163 Qdimension, make_fixnum(2)), |
| 3659 | 1164 list6(Qregistries, Qunicode_registries, |
| 4268 | 1165 Qfinal, make_char(last_jit_charset_final), |
| 3439 | 1166 /* This CCL program is initialised in |
| 1167 unicode.el. */ | |
| 1168 Qccl_program, Qccl_encode_to_ucs_2)))); | |
| 4268 | 1169 |
| 1170 /* Record for the Unicode infrastructure that we've created | |
| 1171 this character set. */ | |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1172 Vnumber_of_jit_charsets = make_fixnum (number_of_jit_charsets + 1); |
| 4268 | 1173 Vlast_jit_charset_final = make_char (last_jit_charset_final + 1); |
| 3439 | 1174 |
| 1175 i = get_free_codepoint(Vcurrent_jit_charset); | |
| 1176 } | |
| 1177 | |
| 1178 if (-1 != i) | |
| 1179 { | |
| 1180 set_unicode_conversion((Ichar)i, code); | |
| 1181 /* No need to add the charset to the end of the list; it's done | |
| 1182 automatically. */ | |
| 1183 } | |
| 1184 } | |
| 1185 return (Ichar) i; | |
| 771 | 1186 } |
| 1187 | |
| 877 | 1188 /* Add charsets to precedence list. |
| 1189 LIST must be a list of charsets. Charsets which are in the list more | |
| 1190 than once are given the precedence implied by their earliest appearance. | |
| 1191 Later appearances are ignored. */ | |
| 771 | 1192 static void |
| 1193 add_charsets_to_precedence_list (Lisp_Object list, int *lbs, | |
| 1194 Lisp_Object_dynarr *dynarr) | |
| 1195 { | |
| 1196 { | |
| 1197 EXTERNAL_LIST_LOOP_2 (elt, list) | |
| 1198 { | |
| 1199 Lisp_Object charset = Fget_charset (elt); | |
| 778 | 1200 int lb = XCHARSET_LEADING_BYTE (charset); |
| 771 | 1201 if (lbs[lb - MIN_LEADING_BYTE] == 0) |
| 1202 { | |
| 877 | 1203 Dynarr_add (dynarr, charset); |
| 771 | 1204 lbs[lb - MIN_LEADING_BYTE] = 1; |
| 1205 } | |
| 1206 } | |
| 1207 } | |
| 1208 } | |
| 1209 | |
| 877 | 1210 /* Rebuild the charset precedence array. |
| 1211 The "charsets preferred for the current language" get highest precedence, | |
| 1212 followed by the "charsets preferred by default", ordered as in | |
| 1213 Vlanguage_unicode_precedence_list and Vdefault_unicode_precedence_list, | |
| 1214 respectively. All remaining charsets follow in an arbitrary order. */ | |
| 771 | 1215 void |
| 1216 recalculate_unicode_precedence (void) | |
| 1217 { | |
| 1218 int lbs[NUM_LEADING_BYTES]; | |
| 1219 int i; | |
| 1220 | |
| 1221 for (i = 0; i < NUM_LEADING_BYTES; i++) | |
| 1222 lbs[i] = 0; | |
| 1223 | |
| 1224 Dynarr_reset (unicode_precedence_dynarr); | |
| 1225 | |
| 1226 add_charsets_to_precedence_list (Vlanguage_unicode_precedence_list, | |
| 1227 lbs, unicode_precedence_dynarr); | |
| 1228 add_charsets_to_precedence_list (Vdefault_unicode_precedence_list, | |
| 1229 lbs, unicode_precedence_dynarr); | |
| 1230 | |
| 1231 for (i = 0; i < NUM_LEADING_BYTES; i++) | |
| 1232 { | |
| 1233 if (lbs[i] == 0) | |
| 1234 { | |
| 826 | 1235 Lisp_Object charset = charset_by_leading_byte (i + MIN_LEADING_BYTE); |
| 771 | 1236 if (!NILP (charset)) |
| 1237 Dynarr_add (unicode_precedence_dynarr, charset); | |
| 1238 } | |
| 1239 } | |
| 1240 } | |
| 1241 | |
| 877 | 1242 DEFUN ("unicode-precedence-list", |
| 1243 Funicode_precedence_list, | |
| 1244 0, 0, 0, /* | |
| 1245 Return the precedence order among charsets used for Unicode decoding. | |
| 1246 | |
| 1247 Value is a list of charsets, which are searched in order for a translation | |
| 1248 matching a given Unicode character. | |
| 1249 | |
| 1250 The highest precedence is given to the language-specific precedence list of | |
| 1251 charsets, defined by `set-language-unicode-precedence-list'. These are | |
| 1252 followed by charsets in the default precedence list, defined by | |
| 1253 `set-default-unicode-precedence-list'. Charsets occurring multiple times are | |
|
5384
3889ef128488
Fix misspelled words, and some grammar, across the entire source tree.
Jerry James <james@xemacs.org>
parents:
5345
diff
changeset
|
1254 given precedence according to their first occurrence in either list. These |
| 877 | 1255 are followed by the remaining charsets, in some arbitrary order. |
| 771 | 1256 |
| 1257 The language-specific precedence list is meant to be set as part of the | |
| 1258 language environment initialization; the default precedence list is meant | |
| 1259 to be set by the user. | |
| 1318 | 1260 |
| 1261 #### NOTE: This interface may be changed. | |
| 771 | 1262 */ |
| 877 | 1263 ()) |
| 1264 { | |
| 1265 int i; | |
| 1266 Lisp_Object list = Qnil; | |
| 1267 | |
| 1268 for (i = Dynarr_length (unicode_precedence_dynarr) - 1; i >= 0; i--) | |
| 1269 list = Fcons (Dynarr_at (unicode_precedence_dynarr, i), list); | |
| 1270 return list; | |
| 1271 } | |
| 1272 | |
| 1273 | |
| 1274 /* #### This interface is wrong. Cyrillic users and Chinese users are going | |
| 1275 to have varying opinions about whether ISO Cyrillic, KOI8-R, or Windows | |
| 1276 1251 should take precedence, and whether Big Five or CNS should take | |
| 1277 precedence, respectively. This means that users are sometimes going to | |
| 1278 want to set Vlanguage_unicode_precedence_list. | |
| 1279 Furthermore, this should be language-local (buffer-local would be a | |
| 1318 | 1280 reasonable approximation). |
| 1281 | |
| 1282 Answer: You are right, this needs rethinking. */ | |
| 877 | 1283 DEFUN ("set-language-unicode-precedence-list", |
| 1284 Fset_language_unicode_precedence_list, | |
| 1285 1, 1, 0, /* | |
| 1286 Set the language-specific precedence of charsets in Unicode decoding. | |
| 1287 LIST is a list of charsets. | |
| 1288 See `unicode-precedence-list' for more information. | |
| 1318 | 1289 |
| 1290 #### NOTE: This interface may be changed. | |
| 877 | 1291 */ |
| 771 | 1292 (list)) |
| 1293 { | |
| 1294 { | |
| 1295 EXTERNAL_LIST_LOOP_2 (elt, list) | |
| 1296 Fget_charset (elt); | |
| 1297 } | |
| 1298 | |
| 1299 Vlanguage_unicode_precedence_list = list; | |
| 1300 recalculate_unicode_precedence (); | |
| 1301 return Qnil; | |
| 1302 } | |
| 1303 | |
| 1304 DEFUN ("language-unicode-precedence-list", | |
| 1305 Flanguage_unicode_precedence_list, | |
| 1306 0, 0, 0, /* | |
| 1307 Return the language-specific precedence list used for Unicode decoding. | |
| 877 | 1308 See `unicode-precedence-list' for more information. |
| 1318 | 1309 |
| 1310 #### NOTE: This interface may be changed. | |
| 771 | 1311 */ |
| 1312 ()) | |
| 1313 { | |
| 1314 return Vlanguage_unicode_precedence_list; | |
| 1315 } | |
| 1316 | |
| 1317 DEFUN ("set-default-unicode-precedence-list", | |
| 1318 Fset_default_unicode_precedence_list, | |
| 1319 1, 1, 0, /* | |
| 1320 Set the default precedence list used for Unicode decoding. | |
| 877 | 1321 This is intended to be set by the user. See |
| 1322 `unicode-precedence-list' for more information. | |
| 1318 | 1323 |
| 1324 #### NOTE: This interface may be changed. | |
| 771 | 1325 */ |
| 1326 (list)) | |
| 1327 { | |
| 1328 { | |
| 1329 EXTERNAL_LIST_LOOP_2 (elt, list) | |
| 1330 Fget_charset (elt); | |
| 1331 } | |
| 1332 | |
| 1333 Vdefault_unicode_precedence_list = list; | |
| 1334 recalculate_unicode_precedence (); | |
| 1335 return Qnil; | |
| 1336 } | |
| 1337 | |
| 1338 DEFUN ("default-unicode-precedence-list", | |
| 1339 Fdefault_unicode_precedence_list, | |
| 1340 0, 0, 0, /* | |
| 1341 Return the default precedence list used for Unicode decoding. | |
| 877 | 1342 See `unicode-precedence-list' for more information. |
| 1318 | 1343 |
| 1344 #### NOTE: This interface may be changed. | |
| 771 | 1345 */ |
| 1346 ()) | |
| 1347 { | |
| 1348 return Vdefault_unicode_precedence_list; | |
| 1349 } | |
| 1350 | |
| 1351 DEFUN ("set-unicode-conversion", Fset_unicode_conversion, | |
| 1352 2, 2, 0, /* | |
| 1353 Add conversion information between Unicode codepoints and characters. | |
| 877 | 1354 Conversions for U+0000 to U+00FF are hardwired to ASCII, Control-1, and |
| 1355 Latin-1. Attempts to set these values will raise an error. | |
| 1356 | |
| 771 | 1357 CHARACTER is one of the following: |
| 1358 | |
| 1359 -- A character (in which case CODE must be a non-negative integer; values | |
| 1360 above 2^20 - 1 are allowed for the purpose of specifying private | |
| 877 | 1361 characters, but are illegal in standard Unicode---they will cause errors |
| 1362 when converted to utf-16) | |
| 771 | 1363 -- A vector of characters (in which case CODE must be a vector of integers |
| 1364 of the same length) | |
| 1365 */ | |
| 1366 (character, code)) | |
| 1367 { | |
| 1368 Lisp_Object charset; | |
| 877 | 1369 int ichar, unicode; |
| 771 | 1370 |
| 1371 CHECK_CHAR (character); | |
|
5307
c096d8051f89
Have NATNUMP give t for positive bignums; check limits appropriately.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5157
diff
changeset
|
1372 |
|
5736
3192994c49ca
Convert C (un)signed long long values to bignums properly.
Jerry James <james@xemacs.org>
parents:
5581
diff
changeset
|
1373 check_integer_range (code, Qzero, make_fixnum (MOST_POSITIVE_FIXNUM)); |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1374 |
|
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1375 unicode = XFIXNUM (code); |
| 877 | 1376 ichar = XCHAR (character); |
| 1377 charset = ichar_charset (ichar); | |
| 1378 | |
| 1379 /* The translations of ASCII, Control-1, and Latin-1 code points are | |
| 1380 hard-coded in ichar_to_unicode and unicode_to_ichar. | |
| 1381 | |
| 1382 Checking unicode < 256 && ichar != unicode is wrong because Mule gives | |
| 1383 many Latin characters code points in a few different character sets. */ | |
| 1384 if ((EQ (charset, Vcharset_ascii) || | |
| 1385 EQ (charset, Vcharset_control_1) || | |
| 1386 EQ (charset, Vcharset_latin_iso8859_1)) | |
| 1387 && unicode != ichar) | |
| 893 | 1388 signal_error (Qinvalid_argument, "Can't change Unicode translation for ASCII, Control-1 or Latin-1 character", |
| 771 | 1389 character); |
| 1390 | |
| 877 | 1391 /* #### Composite characters are not properly implemented yet. */ |
| 1392 if (EQ (charset, Vcharset_composite)) | |
| 1393 signal_error (Qinvalid_argument, "Can't set Unicode translation for Composite char", | |
| 1394 character); | |
| 1395 | |
| 1396 set_unicode_conversion (ichar, unicode); | |
| 771 | 1397 return Qnil; |
| 1398 } | |
| 1399 | |
| 1400 #endif /* MULE */ | |
| 1401 | |
| 800 | 1402 DEFUN ("char-to-unicode", Fchar_to_unicode, 1, 1, 0, /* |
| 771 | 1403 Convert character to Unicode codepoint. |
| 3025 | 1404 When there is no international support (i.e. the `mule' feature is not |
| 877 | 1405 present), this function simply does `char-to-int'. |
| 771 | 1406 */ |
| 1407 (character)) | |
| 1408 { | |
| 1409 CHECK_CHAR (character); | |
| 1410 #ifdef MULE | |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1411 return make_fixnum (ichar_to_unicode (XCHAR (character))); |
| 771 | 1412 #else |
| 1413 return Fchar_to_int (character); | |
| 1414 #endif /* MULE */ | |
| 1415 } | |
| 1416 | |
| 800 | 1417 DEFUN ("unicode-to-char", Funicode_to_char, 1, 2, 0, /* |
| 771 | 1418 Convert Unicode codepoint to character. |
| 1419 CODE should be a non-negative integer. | |
| 1420 If CHARSETS is given, it should be a list of charsets, and only those | |
| 1421 charsets will be consulted, in the given order, for a translation. | |
| 1422 Otherwise, the default ordering of all charsets will be given (see | |
| 1423 `set-unicode-charset-precedence'). | |
| 1424 | |
| 3025 | 1425 When there is no international support (i.e. the `mule' feature is not |
| 877 | 1426 present), this function simply does `int-to-char' and ignores the CHARSETS |
| 1427 argument. | |
| 2622 | 1428 |
| 3439 | 1429 If the CODE would not otherwise be converted to an XEmacs character, and the |
| 1430 list of character sets to be consulted is nil or the default, a new XEmacs | |
| 1431 character will be created for it in one of the `jit-ucs-charset' Mule | |
| 4268 | 1432 character sets, and that character will be returned. |
| 1433 | |
| 1434 This is limited to around 400,000 characters per XEmacs session, though, so | |
| 1435 while normal usage will not be problematic, things like: | |
| 1436 | |
| 1437 \(dotimes (i #x110000) (decode-char 'ucs i)) | |
| 1438 | |
| 1439 will eventually error. The long-term solution to this is Unicode as an | |
| 1440 internal encoding. | |
| 771 | 1441 */ |
| 2333 | 1442 (code, USED_IF_MULE (charsets))) |
| 771 | 1443 { |
| 1444 #ifdef MULE | |
| 1445 Lisp_Object_dynarr *dyn; | |
| 1446 int lbs[NUM_LEADING_BYTES]; | |
| 1447 int c; | |
| 1448 | |
|
5736
3192994c49ca
Convert C (un)signed long long values to bignums properly.
Jerry James <james@xemacs.org>
parents:
5581
diff
changeset
|
1449 check_integer_range (code, Qzero, make_fixnum (MOST_POSITIVE_FIXNUM)); |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1450 c = XFIXNUM (code); |
| 771 | 1451 { |
| 1452 EXTERNAL_LIST_LOOP_2 (elt, charsets) | |
| 1453 Fget_charset (elt); | |
| 1454 } | |
| 1455 | |
| 1456 if (NILP (charsets)) | |
| 1457 { | |
| 877 | 1458 Ichar ret = unicode_to_ichar (c, unicode_precedence_dynarr); |
| 771 | 1459 if (ret == -1) |
| 1460 return Qnil; | |
| 1461 return make_char (ret); | |
| 1462 } | |
| 1463 | |
| 1464 dyn = Dynarr_new (Lisp_Object); | |
| 1465 memset (lbs, 0, NUM_LEADING_BYTES * sizeof (int)); | |
| 1466 add_charsets_to_precedence_list (charsets, lbs, dyn); | |
| 1467 { | |
| 877 | 1468 Ichar ret = unicode_to_ichar (c, dyn); |
| 771 | 1469 Dynarr_free (dyn); |
| 1470 if (ret == -1) | |
| 1471 return Qnil; | |
| 1472 return make_char (ret); | |
| 1473 } | |
| 1474 #else | |
|
5736
3192994c49ca
Convert C (un)signed long long values to bignums properly.
Jerry James <james@xemacs.org>
parents:
5581
diff
changeset
|
1475 check_integer_range (code, Qzero, make_fixnum (MOST_POSITIVE_FIXNUM)); |
| 771 | 1476 return Fint_to_char (code); |
| 1477 #endif /* MULE */ | |
| 1478 } | |
| 1479 | |
| 872 | 1480 #ifdef MULE |
| 1481 | |
| 771 | 1482 static Lisp_Object |
| 1483 cerrar_el_fulano (Lisp_Object fulano) | |
| 1484 { | |
| 1485 FILE *file = (FILE *) get_opaque_ptr (fulano); | |
| 1486 retry_fclose (file); | |
| 1487 return Qnil; | |
| 1488 } | |
| 1489 | |
| 1318 | 1490 DEFUN ("load-unicode-mapping-table", Fload_unicode_mapping_table, |
| 771 | 1491 2, 6, 0, /* |
| 877 | 1492 Load Unicode tables with the Unicode mapping data in FILENAME for CHARSET. |
| 771 | 1493 Data is text, in the form of one translation per line -- charset |
| 1494 codepoint followed by Unicode codepoint. Numbers are decimal or hex | |
| 1495 \(preceded by 0x). Comments are marked with a #. Charset codepoints | |
| 877 | 1496 for two-dimensional charsets have the first octet stored in the |
| 771 | 1497 high 8 bits of the hex number and the second in the low 8 bits. |
| 1498 | |
| 1499 If START and END are given, only charset codepoints within the given | |
| 877 | 1500 range will be processed. (START and END apply to the codepoints in the |
| 1501 file, before OFFSET is applied.) | |
| 771 | 1502 |
| 877 | 1503 If OFFSET is given, that value will be added to all charset codepoints |
| 1504 in the file to obtain the internal charset codepoint. \(We assume | |
| 1505 that octets in the table are in the range 33 to 126 or 32 to 127. If | |
| 1506 you have a table in ku-ten form, with octets in the range 1 to 94, you | |
| 1507 will have to use an offset of 5140, i.e. 0x2020.) | |
| 771 | 1508 |
| 1509 FLAGS, if specified, control further how the tables are interpreted | |
| 877 | 1510 and are used to special-case certain known format deviations in the |
| 1511 Unicode tables or in the charset: | |
| 771 | 1512 |
| 1513 `ignore-first-column' | |
| 877 | 1514 The JIS X 0208 tables have 3 columns of data instead of 2. The first |
| 1515 column contains the Shift-JIS codepoint, which we ignore. | |
| 771 | 1516 `big5' |
| 877 | 1517 The charset codepoints are Big Five codepoints; convert it to the |
| 1518 hacked-up Mule codepoint in `chinese-big5-1' or `chinese-big5-2'. | |
| 771 | 1519 */ |
| 1520 (filename, charset, start, end, offset, flags)) | |
| 1521 { | |
| 1522 int st = 0, en = INT_MAX, of = 0; | |
| 1523 FILE *file; | |
| 1524 struct gcpro gcpro1; | |
| 1525 char line[1025]; | |
| 1526 int fondo = specpdl_depth (); | |
| 1527 int ignore_first_column = 0; | |
| 1528 int big5 = 0; | |
| 1529 | |
| 1530 CHECK_STRING (filename); | |
| 1531 charset = Fget_charset (charset); | |
| 1532 if (!NILP (start)) | |
| 1533 { | |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1534 CHECK_FIXNUM (start); |
|
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1535 st = XFIXNUM (start); |
| 771 | 1536 } |
| 1537 if (!NILP (end)) | |
| 1538 { | |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1539 CHECK_FIXNUM (end); |
|
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1540 en = XFIXNUM (end); |
| 771 | 1541 } |
| 1542 if (!NILP (offset)) | |
| 1543 { | |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1544 CHECK_FIXNUM (offset); |
|
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
1545 of = XFIXNUM (offset); |
| 771 | 1546 } |
| 1547 | |
| 1548 if (!LISTP (flags)) | |
| 1549 flags = list1 (flags); | |
| 1550 | |
| 1551 { | |
| 1552 EXTERNAL_LIST_LOOP_2 (elt, flags) | |
| 1553 { | |
| 1554 if (EQ (elt, Qignore_first_column)) | |
| 1555 ignore_first_column = 1; | |
| 1556 else if (EQ (elt, Qbig5)) | |
| 1557 big5 = 1; | |
| 1558 else | |
| 1559 invalid_constant | |
| 1318 | 1560 ("Unrecognized `load-unicode-mapping-table' flag", elt); |
| 771 | 1561 } |
| 1562 } | |
| 1563 | |
| 1564 GCPRO1 (filename); | |
| 1565 filename = Fexpand_file_name (filename, Qnil); | |
| 1566 file = qxe_fopen (XSTRING_DATA (filename), READ_TEXT); | |
| 1567 if (!file) | |
| 1568 report_file_error ("Cannot open", filename); | |
| 1569 record_unwind_protect (cerrar_el_fulano, make_opaque_ptr (file)); | |
| 1570 while (fgets (line, sizeof (line), file)) | |
| 1571 { | |
| 1572 char *p = line; | |
| 1573 int cp1, cp2, endcount; | |
| 1574 int cp1high, cp1low; | |
| 1575 int dummy; | |
| 1576 | |
| 1577 while (*p) /* erase all comments out of the line */ | |
| 1578 { | |
| 1579 if (*p == '#') | |
| 1580 *p = '\0'; | |
| 1581 else | |
| 1582 p++; | |
| 1583 } | |
| 1584 /* see if line is nothing but whitespace and skip if so */ | |
| 1585 p = line + strspn (line, " \t\n\r\f"); | |
| 1586 if (!*p) | |
| 1587 continue; | |
| 1588 /* NOTE: It appears that MS Windows and Newlib sscanf() have | |
| 1589 different interpretations for whitespace (== "skip all whitespace | |
| 1590 at processing point"): Newlib requires at least one corresponding | |
| 1591 whitespace character in the input, but MS allows none. The | |
| 1592 following would be easier to write if we could count on the MS | |
| 1593 interpretation. | |
| 1594 | |
| 1595 Also, the return value does NOT include %n storage. */ | |
| 1596 if ((!ignore_first_column ? | |
| 1597 sscanf (p, "%i %i%n", &cp1, &cp2, &endcount) < 2 : | |
| 1598 sscanf (p, "%i %i %i%n", &dummy, &cp1, &cp2, &endcount) < 3) | |
| 2367 | 1599 /* #### Temporary code! Cygwin newlib fucked up scanf() handling |
| 1600 of numbers beginning 0x0... starting in 04/2004, in an attempt | |
| 1601 to fix another bug. A partial fix for this was put in in | |
| 1602 06/2004, but as of 10/2004 the value of ENDCOUNT returned in | |
| 1603 such case is still wrong. If this gets fixed soon, remove | |
| 1604 this code. --ben */ | |
| 1605 #ifndef CYGWIN_SCANF_BUG | |
| 1606 || *(p + endcount + strspn (p + endcount, " \t\n\r\f")) | |
| 1607 #endif | |
| 1608 ) | |
| 771 | 1609 { |
| 793 | 1610 warn_when_safe (Qunicode, Qwarning, |
| 771 | 1611 "Unrecognized line in translation file %s:\n%s", |
| 1612 XSTRING_DATA (filename), line); | |
| 1613 continue; | |
| 1614 } | |
| 1615 if (cp1 >= st && cp1 <= en) | |
| 1616 { | |
| 1617 cp1 += of; | |
| 1618 if (cp1 < 0 || cp1 >= 65536) | |
| 1619 { | |
| 1620 out_of_range: | |
| 793 | 1621 warn_when_safe (Qunicode, Qwarning, |
| 1622 "Out of range first codepoint 0x%x in " | |
| 1623 "translation file %s:\n%s", | |
| 771 | 1624 cp1, XSTRING_DATA (filename), line); |
| 1625 continue; | |
| 1626 } | |
| 1627 | |
| 1628 cp1high = cp1 >> 8; | |
| 1629 cp1low = cp1 & 255; | |
| 1630 | |
| 1631 if (big5) | |
| 1632 { | |
| 867 | 1633 Ichar ch = decode_big5_char (cp1high, cp1low); |
| 771 | 1634 if (ch == -1) |
| 793 | 1635 |
| 1636 warn_when_safe (Qunicode, Qwarning, | |
| 1637 "Out of range Big5 codepoint 0x%x in " | |
| 1638 "translation file %s:\n%s", | |
| 771 | 1639 cp1, XSTRING_DATA (filename), line); |
| 1640 else | |
| 1641 set_unicode_conversion (ch, cp2); | |
| 1642 } | |
| 1643 else | |
| 1644 { | |
| 1645 int l1, h1, l2, h2; | |
| 867 | 1646 Ichar emch; |
| 771 | 1647 |
| 1648 switch (XCHARSET_TYPE (charset)) | |
| 1649 { | |
| 1650 case CHARSET_TYPE_94: l1 = 33; h1 = 126; l2 = 0; h2 = 0; break; | |
| 1651 case CHARSET_TYPE_96: l1 = 32; h1 = 127; l2 = 0; h2 = 0; break; | |
| 1652 case CHARSET_TYPE_94X94: l1 = 33; h1 = 126; l2 = 33; h2 = 126; | |
| 1653 break; | |
| 1654 case CHARSET_TYPE_96X96: l1 = 32; h1 = 127; l2 = 32; h2 = 127; | |
| 1655 break; | |
| 2500 | 1656 default: ABORT (); l1 = 0; h1 = 0; l2 = 0; h2 = 0; |
| 771 | 1657 } |
| 1658 | |
| 1659 if (cp1high < l2 || cp1high > h2 || cp1low < l1 || cp1low > h1) | |
| 1660 goto out_of_range; | |
| 1661 | |
| 867 | 1662 emch = (cp1high == 0 ? make_ichar (charset, cp1low, 0) : |
| 1663 make_ichar (charset, cp1high, cp1low)); | |
| 771 | 1664 set_unicode_conversion (emch, cp2); |
| 1665 } | |
| 1666 } | |
| 1667 } | |
| 1668 | |
| 1669 if (ferror (file)) | |
| 1670 report_file_error ("IO error when reading", filename); | |
| 1671 | |
| 1672 unbind_to (fondo); /* close file */ | |
| 1673 UNGCPRO; | |
| 1674 return Qnil; | |
| 1675 } | |
| 1676 | |
| 1677 #endif /* MULE */ | |
| 1678 | |
| 1679 | |
| 1680 /************************************************************************/ | |
| 1681 /* Unicode coding system */ | |
| 1682 /************************************************************************/ | |
| 1683 | |
| 1684 struct unicode_coding_system | |
| 1685 { | |
| 1686 enum unicode_type type; | |
| 1887 | 1687 unsigned int little_endian :1; |
| 1688 unsigned int need_bom :1; | |
| 771 | 1689 }; |
| 1690 | |
| 1691 #define CODING_SYSTEM_UNICODE_TYPE(codesys) \ | |
| 1692 (CODING_SYSTEM_TYPE_DATA (codesys, unicode)->type) | |
| 1693 #define XCODING_SYSTEM_UNICODE_TYPE(codesys) \ | |
| 1694 CODING_SYSTEM_UNICODE_TYPE (XCODING_SYSTEM (codesys)) | |
| 1695 #define CODING_SYSTEM_UNICODE_LITTLE_ENDIAN(codesys) \ | |
| 1696 (CODING_SYSTEM_TYPE_DATA (codesys, unicode)->little_endian) | |
| 1697 #define XCODING_SYSTEM_UNICODE_LITTLE_ENDIAN(codesys) \ | |
| 1698 CODING_SYSTEM_UNICODE_LITTLE_ENDIAN (XCODING_SYSTEM (codesys)) | |
| 1699 #define CODING_SYSTEM_UNICODE_NEED_BOM(codesys) \ | |
| 1700 (CODING_SYSTEM_TYPE_DATA (codesys, unicode)->need_bom) | |
| 1701 #define XCODING_SYSTEM_UNICODE_NEED_BOM(codesys) \ | |
| 1702 CODING_SYSTEM_UNICODE_NEED_BOM (XCODING_SYSTEM (codesys)) | |
| 1703 | |
| 1704 struct unicode_coding_stream | |
| 1705 { | |
| 1706 /* decode */ | |
| 1707 unsigned char counter; | |
| 4096 | 1708 unsigned char indicated_length; |
| 771 | 1709 int seen_char; |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1710 Charcount characters_seen; |
| 771 | 1711 /* encode */ |
| 1712 Lisp_Object current_charset; | |
| 1713 int current_char_boundary; | |
| 1714 int wrote_bom; | |
| 1715 }; | |
| 1716 | |
| 1204 | 1717 static const struct memory_description unicode_coding_system_description[] = { |
| 771 | 1718 { XD_END } |
| 1719 }; | |
| 1720 | |
| 1204 | 1721 DEFINE_CODING_SYSTEM_TYPE_WITH_DATA (unicode); |
| 1722 | |
| 771 | 1723 static void |
| 1724 decode_unicode_char (int ch, unsigned_char_dynarr *dst, | |
| 1887 | 1725 struct unicode_coding_stream *data, |
| 1726 unsigned int ignore_bom) | |
| 771 | 1727 { |
| 1728 if (ch == 0xFEFF && !data->seen_char && ignore_bom) | |
| 1729 ; | |
| 1730 else | |
| 1731 { | |
| 1732 #ifdef MULE | |
| 877 | 1733 Ichar chr = unicode_to_ichar (ch, unicode_precedence_dynarr); |
| 771 | 1734 |
| 1735 if (chr != -1) | |
| 1736 { | |
| 867 | 1737 Ibyte work[MAX_ICHAR_LEN]; |
| 771 | 1738 int len; |
| 1739 | |
| 867 | 1740 len = set_itext_ichar (work, chr); |
| 771 | 1741 Dynarr_add_many (dst, work, len); |
| 1742 } | |
| 1743 else | |
| 1744 { | |
| 1745 Dynarr_add (dst, LEADING_BYTE_JAPANESE_JISX0208); | |
| 1746 Dynarr_add (dst, 34 + 128); | |
| 1747 Dynarr_add (dst, 46 + 128); | |
| 1748 } | |
| 1749 #else | |
| 867 | 1750 Dynarr_add (dst, (Ibyte) ch); |
| 771 | 1751 #endif /* MULE */ |
| 1752 } | |
| 1753 | |
| 1754 data->seen_char = 1; | |
| 1755 } | |
| 1756 | |
| 4096 | 1757 #define DECODE_ERROR_OCTET(octet, dst, data, ignore_bom) \ |
| 1758 decode_unicode_char ((octet) + UNICODE_ERROR_OCTET_RANGE_START, \ | |
| 1759 dst, data, ignore_bom) | |
| 1760 | |
| 1761 static inline void | |
| 1762 indicate_invalid_utf_8 (unsigned char indicated_length, | |
| 1763 unsigned char counter, | |
| 1764 int ch, unsigned_char_dynarr *dst, | |
| 1765 struct unicode_coding_stream *data, | |
| 1766 unsigned int ignore_bom) | |
| 1767 { | |
| 1768 Binbyte stored = indicated_length - counter; | |
| 1769 Binbyte mask = "\x00\x00\xC0\xE0\xF0\xF8\xFC"[indicated_length]; | |
| 1770 | |
| 1771 while (stored > 0) | |
| 1772 { | |
| 1773 DECODE_ERROR_OCTET (((ch >> (6 * (stored - 1))) & 0x3f) | mask, | |
| 1774 dst, data, ignore_bom); | |
| 1775 mask = 0x80, stored--; | |
| 1776 } | |
| 1777 } | |
| 1778 | |
| 771 | 1779 static void |
| 1780 encode_unicode_char_1 (int code, unsigned_char_dynarr *dst, | |
| 4096 | 1781 enum unicode_type type, unsigned int little_endian, |
| 1782 int write_error_characters_as_such) | |
| 771 | 1783 { |
| 1784 switch (type) | |
| 1785 { | |
| 1786 case UNICODE_UTF_16: | |
| 1787 if (little_endian) | |
| 1788 { | |
| 3952 | 1789 if (code < 0x10000) { |
| 1790 Dynarr_add (dst, (unsigned char) (code & 255)); | |
| 1791 Dynarr_add (dst, (unsigned char) ((code >> 8) & 255)); | |
| 4096 | 1792 } else if (write_error_characters_as_such && |
| 1793 code >= UNICODE_ERROR_OCTET_RANGE_START && | |
| 1794 code < (UNICODE_ERROR_OCTET_RANGE_START + 0x100)) | |
| 1795 { | |
| 1796 Dynarr_add (dst, (unsigned char) ((code & 0xFF))); | |
| 1797 } | |
| 1798 else if (code < 0x110000) | |
| 1799 { | |
| 1800 /* Little endian; least significant byte first. */ | |
| 1801 int first, second; | |
| 1802 | |
| 1803 CODE_TO_UTF_16_SURROGATES(code, first, second); | |
| 1804 | |
| 1805 Dynarr_add (dst, (unsigned char) (first & 255)); | |
| 1806 Dynarr_add (dst, (unsigned char) ((first >> 8) & 255)); | |
| 1807 | |
| 1808 Dynarr_add (dst, (unsigned char) (second & 255)); | |
| 1809 Dynarr_add (dst, (unsigned char) ((second >> 8) & 255)); | |
| 1810 } | |
| 1811 else | |
| 1812 { | |
| 1813 /* Not valid Unicode. Pass U+FFFD, least significant byte | |
| 1814 first. */ | |
| 1815 Dynarr_add (dst, (unsigned char) 0xFD); | |
| 1816 Dynarr_add (dst, (unsigned char) 0xFF); | |
| 1817 } | |
| 771 | 1818 } |
| 1819 else | |
| 1820 { | |
| 3952 | 1821 if (code < 0x10000) { |
| 1822 Dynarr_add (dst, (unsigned char) ((code >> 8) & 255)); | |
| 1823 Dynarr_add (dst, (unsigned char) (code & 255)); | |
| 4096 | 1824 } else if (write_error_characters_as_such && |
| 1825 code >= UNICODE_ERROR_OCTET_RANGE_START && | |
| 1826 code < (UNICODE_ERROR_OCTET_RANGE_START + 0x100)) | |
| 1827 { | |
| 1828 Dynarr_add (dst, (unsigned char) ((code & 0xFF))); | |
| 1829 } | |
| 1830 else if (code < 0x110000) | |
| 1831 { | |
| 1832 /* Big endian; most significant byte first. */ | |
| 1833 int first, second; | |
| 1834 | |
| 1835 CODE_TO_UTF_16_SURROGATES(code, first, second); | |
| 1836 | |
| 1837 Dynarr_add (dst, (unsigned char) ((first >> 8) & 255)); | |
| 1838 Dynarr_add (dst, (unsigned char) (first & 255)); | |
| 1839 | |
| 1840 Dynarr_add (dst, (unsigned char) ((second >> 8) & 255)); | |
| 1841 Dynarr_add (dst, (unsigned char) (second & 255)); | |
| 1842 } | |
| 1843 else | |
| 1844 { | |
| 1845 /* Not valid Unicode. Pass U+FFFD, most significant byte | |
| 1846 first. */ | |
| 1847 Dynarr_add (dst, (unsigned char) 0xFF); | |
| 1848 Dynarr_add (dst, (unsigned char) 0xFD); | |
| 1849 } | |
| 771 | 1850 } |
| 1851 break; | |
| 1852 | |
| 1853 case UNICODE_UCS_4: | |
| 4096 | 1854 case UNICODE_UTF_32: |
| 771 | 1855 if (little_endian) |
| 1856 { | |
| 4096 | 1857 if (write_error_characters_as_such && |
| 1858 code >= UNICODE_ERROR_OCTET_RANGE_START && | |
| 1859 code < (UNICODE_ERROR_OCTET_RANGE_START + 0x100)) | |
| 1860 { | |
| 1861 Dynarr_add (dst, (unsigned char) ((code & 0xFF))); | |
| 1862 } | |
| 1863 else | |
| 1864 { | |
| 1865 /* We generate and accept incorrect sequences here, which is | |
| 1866 okay, in the interest of preservation of the user's | |
| 1867 data. */ | |
| 1868 Dynarr_add (dst, (unsigned char) (code & 255)); | |
| 1869 Dynarr_add (dst, (unsigned char) ((code >> 8) & 255)); | |
| 1870 Dynarr_add (dst, (unsigned char) ((code >> 16) & 255)); | |
| 1871 Dynarr_add (dst, (unsigned char) (code >> 24)); | |
| 1872 } | |
| 771 | 1873 } |
| 1874 else | |
| 1875 { | |
| 4096 | 1876 if (write_error_characters_as_such && |
| 1877 code >= UNICODE_ERROR_OCTET_RANGE_START && | |
| 1878 code < (UNICODE_ERROR_OCTET_RANGE_START + 0x100)) | |
| 1879 { | |
| 1880 Dynarr_add (dst, (unsigned char) ((code & 0xFF))); | |
| 1881 } | |
| 1882 else | |
| 1883 { | |
| 1884 /* We generate and accept incorrect sequences here, which is okay, | |
| 1885 in the interest of preservation of the user's data. */ | |
| 1886 Dynarr_add (dst, (unsigned char) (code >> 24)); | |
| 1887 Dynarr_add (dst, (unsigned char) ((code >> 16) & 255)); | |
| 1888 Dynarr_add (dst, (unsigned char) ((code >> 8) & 255)); | |
| 1889 Dynarr_add (dst, (unsigned char) (code & 255)); | |
| 1890 } | |
| 771 | 1891 } |
| 1892 break; | |
| 1893 | |
| 1894 case UNICODE_UTF_8: | |
| 1895 if (code <= 0x7f) | |
| 1896 { | |
| 1897 Dynarr_add (dst, (unsigned char) code); | |
| 1898 } | |
| 1899 else if (code <= 0x7ff) | |
| 1900 { | |
| 1901 Dynarr_add (dst, (unsigned char) ((code >> 6) | 0xc0)); | |
| 1902 Dynarr_add (dst, (unsigned char) ((code & 0x3f) | 0x80)); | |
| 1903 } | |
| 1904 else if (code <= 0xffff) | |
| 1905 { | |
| 1906 Dynarr_add (dst, (unsigned char) ((code >> 12) | 0xe0)); | |
| 1907 Dynarr_add (dst, (unsigned char) (((code >> 6) & 0x3f) | 0x80)); | |
| 1908 Dynarr_add (dst, (unsigned char) ((code & 0x3f) | 0x80)); | |
| 1909 } | |
| 1910 else if (code <= 0x1fffff) | |
| 1911 { | |
| 1912 Dynarr_add (dst, (unsigned char) ((code >> 18) | 0xf0)); | |
| 1913 Dynarr_add (dst, (unsigned char) (((code >> 12) & 0x3f) | 0x80)); | |
| 1914 Dynarr_add (dst, (unsigned char) (((code >> 6) & 0x3f) | 0x80)); | |
| 1915 Dynarr_add (dst, (unsigned char) ((code & 0x3f) | 0x80)); | |
| 1916 } | |
| 1917 else if (code <= 0x3ffffff) | |
| 1918 { | |
| 4096 | 1919 |
| 1920 #if !(UNICODE_ERROR_OCTET_RANGE_START > 0x1fffff \ | |
| 1921 && UNICODE_ERROR_OCTET_RANGE_START < 0x3ffffff) | |
| 1922 #error "This code needs to be rewritten. " | |
| 1923 #endif | |
| 1924 if (write_error_characters_as_such && | |
| 1925 code >= UNICODE_ERROR_OCTET_RANGE_START && | |
| 1926 code < (UNICODE_ERROR_OCTET_RANGE_START + 0x100)) | |
| 1927 { | |
| 1928 Dynarr_add (dst, (unsigned char) ((code & 0xFF))); | |
| 1929 } | |
| 1930 else | |
| 1931 { | |
| 1932 Dynarr_add (dst, (unsigned char) ((code >> 24) | 0xf8)); | |
| 1933 Dynarr_add (dst, (unsigned char) (((code >> 18) & 0x3f) | 0x80)); | |
| 1934 Dynarr_add (dst, (unsigned char) (((code >> 12) & 0x3f) | 0x80)); | |
| 1935 Dynarr_add (dst, (unsigned char) (((code >> 6) & 0x3f) | 0x80)); | |
| 1936 Dynarr_add (dst, (unsigned char) ((code & 0x3f) | 0x80)); | |
| 1937 } | |
| 771 | 1938 } |
| 1939 else | |
| 1940 { | |
| 1941 Dynarr_add (dst, (unsigned char) ((code >> 30) | 0xfc)); | |
| 1942 Dynarr_add (dst, (unsigned char) (((code >> 24) & 0x3f) | 0x80)); | |
| 1943 Dynarr_add (dst, (unsigned char) (((code >> 18) & 0x3f) | 0x80)); | |
| 1944 Dynarr_add (dst, (unsigned char) (((code >> 12) & 0x3f) | 0x80)); | |
| 1945 Dynarr_add (dst, (unsigned char) (((code >> 6) & 0x3f) | 0x80)); | |
| 1946 Dynarr_add (dst, (unsigned char) ((code & 0x3f) | 0x80)); | |
| 1947 } | |
| 1948 break; | |
| 1949 | |
| 2500 | 1950 case UNICODE_UTF_7: ABORT (); |
| 771 | 1951 |
| 2500 | 1952 default: ABORT (); |
| 771 | 1953 } |
| 1954 } | |
| 1955 | |
| 3439 | 1956 /* Also used in mule-coding.c for UTF-8 handling in ISO 2022-oriented |
| 1957 encodings. */ | |
| 1958 void | |
| 2333 | 1959 encode_unicode_char (Lisp_Object USED_IF_MULE (charset), int h, |
| 1960 int USED_IF_MULE (l), unsigned_char_dynarr *dst, | |
| 4096 | 1961 enum unicode_type type, unsigned int little_endian, |
| 1962 int write_error_characters_as_such) | |
| 771 | 1963 { |
| 1964 #ifdef MULE | |
| 867 | 1965 int code = ichar_to_unicode (make_ichar (charset, h & 127, l & 127)); |
| 771 | 1966 |
| 1967 if (code == -1) | |
| 1968 { | |
| 1969 if (type != UNICODE_UTF_16 && | |
| 1970 XCHARSET_DIMENSION (charset) == 2 && | |
| 1971 XCHARSET_CHARS (charset) == 94) | |
| 1972 { | |
| 1973 unsigned char final = XCHARSET_FINAL (charset); | |
| 1974 | |
| 1975 if (('@' <= final) && (final < 0x7f)) | |
| 1976 code = (0xe00000 + (final - '@') * 94 * 94 | |
| 1977 + ((h & 127) - 33) * 94 + (l & 127) - 33); | |
| 1978 else | |
| 1979 code = '?'; | |
| 1980 } | |
| 1981 else | |
| 1982 code = '?'; | |
| 1983 } | |
| 1984 #else | |
| 1985 int code = h; | |
| 1986 #endif /* MULE */ | |
| 1987 | |
| 4096 | 1988 encode_unicode_char_1 (code, dst, type, little_endian, |
| 1989 write_error_characters_as_such); | |
| 771 | 1990 } |
| 1991 | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1992 static Charcount |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1993 unicode_character_tell (struct coding_stream *str) |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1994 { |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1995 if (CODING_STREAM_TYPE_DATA (str, unicode)->counter == 0) |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1996 { |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1997 return CODING_STREAM_TYPE_DATA (str, unicode)->characters_seen; |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1998 } |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
1999 |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2000 return -1; |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2001 } |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2002 |
| 771 | 2003 static Bytecount |
| 2004 unicode_convert (struct coding_stream *str, const UExtbyte *src, | |
| 2005 unsigned_char_dynarr *dst, Bytecount n) | |
| 2006 { | |
| 2007 unsigned int ch = str->ch; | |
| 2008 struct unicode_coding_stream *data = CODING_STREAM_TYPE_DATA (str, unicode); | |
| 2009 enum unicode_type type = | |
| 2010 XCODING_SYSTEM_UNICODE_TYPE (str->codesys); | |
| 1887 | 2011 unsigned int little_endian = |
| 2012 XCODING_SYSTEM_UNICODE_LITTLE_ENDIAN (str->codesys); | |
| 2013 unsigned int ignore_bom = XCODING_SYSTEM_UNICODE_NEED_BOM (str->codesys); | |
| 771 | 2014 Bytecount orign = n; |
| 2015 | |
| 2016 if (str->direction == CODING_DECODE) | |
| 2017 { | |
| 2018 unsigned char counter = data->counter; | |
| 4096 | 2019 unsigned char indicated_length |
| 2020 = data->indicated_length; | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2021 Charcount characters_seen = data->characters_seen; |
| 771 | 2022 |
| 2023 while (n--) | |
| 2024 { | |
| 2025 UExtbyte c = *src++; | |
| 2026 | |
| 2027 switch (type) | |
| 2028 { | |
| 2029 case UNICODE_UTF_8: | |
| 4096 | 2030 if (0 == counter) |
| 2031 { | |
| 2032 if (0 == (c & 0x80)) | |
| 2033 { | |
| 2034 /* ASCII. */ | |
| 2035 decode_unicode_char (c, dst, data, ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2036 characters_seen++; |
| 4096 | 2037 } |
| 2038 else if (0 == (c & 0x40)) | |
| 2039 { | |
| 2040 /* Highest bit set, second highest not--there's | |
| 2041 something wrong. */ | |
| 2042 DECODE_ERROR_OCTET (c, dst, data, ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2043 /* This is a character in the buffer. */ |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2044 characters_seen++; |
| 4096 | 2045 } |
| 2046 else if (0 == (c & 0x20)) | |
| 2047 { | |
| 2048 ch = c & 0x1f; | |
| 2049 counter = 1; | |
| 2050 indicated_length = 2; | |
| 2051 } | |
| 2052 else if (0 == (c & 0x10)) | |
| 2053 { | |
| 2054 ch = c & 0x0f; | |
| 2055 counter = 2; | |
| 2056 indicated_length = 3; | |
| 2057 } | |
| 2058 else if (0 == (c & 0x08)) | |
| 2059 { | |
| 2060 ch = c & 0x0f; | |
| 2061 counter = 3; | |
| 2062 indicated_length = 4; | |
| 2063 } | |
| 2064 else | |
| 2065 { | |
| 2066 /* We don't supports lengths longer than 4 in | |
| 2067 external-format data. */ | |
| 2068 DECODE_ERROR_OCTET (c, dst, data, ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2069 characters_seen++; |
| 4096 | 2070 } |
| 2071 } | |
| 2072 else | |
| 2073 { | |
| 2074 /* counter != 0 */ | |
| 2075 if ((0 == (c & 0x80)) || (0 != (c & 0x40))) | |
| 2076 { | |
| 2077 indicate_invalid_utf_8(indicated_length, | |
| 2078 counter, | |
| 2079 ch, dst, data, ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2080 /* These are characters our receiver will see, not |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2081 actual characters we've seen in the input. */ |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2082 characters_seen += (indicated_length - counter); |
| 4096 | 2083 if (c & 0x80) |
| 2084 { | |
| 2085 DECODE_ERROR_OCTET (c, dst, data, ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2086 characters_seen++; |
| 4096 | 2087 } |
| 2088 else | |
| 2089 { | |
| 2090 /* The character just read is ASCII. Treat it as | |
| 2091 such. */ | |
| 2092 decode_unicode_char (c, dst, data, ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2093 characters_seen++; |
| 4096 | 2094 } |
| 2095 ch = 0; | |
| 2096 counter = 0; | |
| 2097 } | |
| 2098 else | |
| 2099 { | |
| 2100 ch = (ch << 6) | (c & 0x3f); | |
| 2101 counter--; | |
| 2102 /* Just processed the final byte. Emit the character. */ | |
| 2103 if (!counter) | |
| 2104 { | |
| 2105 /* Don't accept over-long sequences, surrogates, | |
| 2106 or codes above #x10FFFF. */ | |
| 2107 if ((ch < 0x80) || | |
| 2108 ((ch < 0x800) && indicated_length > 2) || | |
| 2109 ((ch < 0x10000) && indicated_length > 3) || | |
| 2110 valid_utf_16_surrogate(ch) || (ch > 0x110000)) | |
| 2111 { | |
| 2112 indicate_invalid_utf_8(indicated_length, | |
| 2113 counter, | |
| 2114 ch, dst, data, | |
| 2115 ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2116 characters_seen += (indicated_length - counter); |
| 4096 | 2117 } |
| 2118 else | |
| 2119 { | |
| 2120 decode_unicode_char (ch, dst, data, ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2121 characters_seen++; |
| 4096 | 2122 } |
| 2123 ch = 0; | |
| 2124 } | |
| 2125 } | |
| 771 | 2126 } |
| 2127 break; | |
| 2128 | |
| 2129 case UNICODE_UTF_16: | |
| 3952 | 2130 |
| 771 | 2131 if (little_endian) |
| 2132 ch = (c << counter) | ch; | |
| 2133 else | |
| 2134 ch = (ch << 8) | c; | |
| 4096 | 2135 |
| 771 | 2136 counter += 8; |
| 3952 | 2137 |
| 4096 | 2138 if (16 == counter) |
| 2139 { | |
| 771 | 2140 int tempch = ch; |
| 4096 | 2141 |
| 2142 if (valid_utf_16_first_surrogate(ch)) | |
| 2143 { | |
| 2144 break; | |
| 2145 } | |
| 771 | 2146 ch = 0; |
| 2147 counter = 0; | |
| 2148 decode_unicode_char (tempch, dst, data, ignore_bom); | |
| 2149 } | |
| 4096 | 2150 else if (32 == counter) |
| 3952 | 2151 { |
| 2152 int tempch; | |
| 4096 | 2153 |
|
4583
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2154 if (little_endian) |
| 4096 | 2155 { |
|
4583
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2156 if (!valid_utf_16_last_surrogate(ch >> 16)) |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2157 { |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2158 DECODE_ERROR_OCTET (ch & 0xFF, dst, data, |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2159 ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2160 DECODE_ERROR_OCTET ((ch >> 8) & 0xFF, dst, data, |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2161 ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2162 DECODE_ERROR_OCTET ((ch >> 16) & 0xFF, dst, data, |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2163 ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2164 DECODE_ERROR_OCTET ((ch >> 24) & 0xFF, dst, data, |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2165 ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2166 } |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2167 else |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2168 { |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2169 tempch = utf_16_surrogates_to_code((ch & 0xffff), |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2170 (ch >> 16)); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2171 decode_unicode_char(tempch, dst, data, ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2172 } |
| 4096 | 2173 } |
|
4583
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2174 else |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2175 { |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2176 if (!valid_utf_16_last_surrogate(ch & 0xFFFF)) |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2177 { |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2178 DECODE_ERROR_OCTET ((ch >> 24) & 0xFF, dst, data, |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2179 ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2180 DECODE_ERROR_OCTET ((ch >> 16) & 0xFF, dst, data, |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2181 ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2182 DECODE_ERROR_OCTET ((ch >> 8) & 0xFF, dst, data, |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2183 ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2184 DECODE_ERROR_OCTET (ch & 0xFF, dst, data, |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2185 ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2186 } |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2187 else |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2188 { |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2189 tempch = utf_16_surrogates_to_code((ch >> 16), |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2190 (ch & 0xffff)); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2191 decode_unicode_char(tempch, dst, data, ignore_bom); |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2192 } |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2193 } |
|
2669b1b7e33b
Correct little-endian UTF-16 surrogate handling.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4270
diff
changeset
|
2194 |
| 3952 | 2195 ch = 0; |
| 2196 counter = 0; | |
| 4096 | 2197 } |
| 2198 else | |
| 2199 assert(8 == counter || 24 == counter); | |
| 771 | 2200 break; |
| 2201 | |
| 2202 case UNICODE_UCS_4: | |
| 4096 | 2203 case UNICODE_UTF_32: |
| 771 | 2204 if (little_endian) |
| 2205 ch = (c << counter) | ch; | |
| 2206 else | |
| 2207 ch = (ch << 8) | c; | |
| 2208 counter += 8; | |
| 2209 if (counter == 32) | |
| 2210 { | |
| 4096 | 2211 if (ch > 0x10ffff) |
| 2212 { | |
| 2213 /* ch is not a legal Unicode character. We're fine | |
| 2214 with that in UCS-4, though not in UTF-32. */ | |
| 2215 if (UNICODE_UCS_4 == type && ch < 0x80000000) | |
| 2216 { | |
| 2217 decode_unicode_char (ch, dst, data, ignore_bom); | |
| 2218 } | |
| 2219 else if (little_endian) | |
| 2220 { | |
| 2221 DECODE_ERROR_OCTET (ch & 0xFF, dst, data, | |
| 2222 ignore_bom); | |
| 2223 DECODE_ERROR_OCTET ((ch >> 8) & 0xFF, dst, data, | |
| 2224 ignore_bom); | |
| 2225 DECODE_ERROR_OCTET ((ch >> 16) & 0xFF, dst, data, | |
| 2226 ignore_bom); | |
| 2227 DECODE_ERROR_OCTET ((ch >> 24) & 0xFF, dst, data, | |
| 2228 ignore_bom); | |
| 2229 } | |
| 2230 else | |
| 2231 { | |
| 2232 DECODE_ERROR_OCTET ((ch >> 24) & 0xFF, dst, data, | |
| 2233 ignore_bom); | |
| 2234 DECODE_ERROR_OCTET ((ch >> 16) & 0xFF, dst, data, | |
| 2235 ignore_bom); | |
| 2236 DECODE_ERROR_OCTET ((ch >> 8) & 0xFF, dst, data, | |
| 2237 ignore_bom); | |
| 2238 DECODE_ERROR_OCTET (ch & 0xFF, dst, data, | |
| 2239 ignore_bom); | |
| 2240 } | |
| 2241 } | |
| 2242 else | |
| 2243 { | |
| 2244 decode_unicode_char (ch, dst, data, ignore_bom); | |
| 2245 } | |
| 771 | 2246 ch = 0; |
| 2247 counter = 0; | |
| 2248 } | |
| 2249 break; | |
| 2250 | |
| 2251 case UNICODE_UTF_7: | |
| 2500 | 2252 ABORT (); |
| 771 | 2253 break; |
| 2254 | |
| 2500 | 2255 default: ABORT (); |
| 771 | 2256 } |
| 2257 | |
| 2258 } | |
| 4096 | 2259 |
|
4688
7e54adf407a1
Fix a bug with Unicode error sequences and very short input strings.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4583
diff
changeset
|
2260 if (str->eof && counter) |
| 4096 | 2261 { |
| 2262 switch (type) | |
| 2263 { | |
| 2264 case UNICODE_UTF_8: | |
| 2265 indicate_invalid_utf_8(indicated_length, | |
| 2266 counter, ch, dst, data, | |
| 2267 ignore_bom); | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2268 characters_seen += (indicated_length - counter); |
| 4096 | 2269 break; |
| 2270 | |
| 2271 case UNICODE_UTF_16: | |
| 2272 case UNICODE_UCS_4: | |
| 2273 case UNICODE_UTF_32: | |
| 2274 if (8 == counter) | |
| 2275 { | |
| 2276 DECODE_ERROR_OCTET (ch, dst, data, ignore_bom); | |
| 2277 } | |
| 2278 else if (16 == counter) | |
| 2279 { | |
| 2280 if (little_endian) | |
| 2281 { | |
| 2282 DECODE_ERROR_OCTET (ch & 0xFF, dst, data, ignore_bom); | |
| 2283 DECODE_ERROR_OCTET ((ch >> 8) & 0xFF, dst, data, | |
| 2284 ignore_bom); | |
| 2285 } | |
| 2286 else | |
| 2287 { | |
| 2288 DECODE_ERROR_OCTET ((ch >> 8) & 0xFF, dst, data, | |
| 2289 ignore_bom); | |
| 2290 DECODE_ERROR_OCTET (ch & 0xFF, dst, data, ignore_bom); | |
| 2291 } | |
| 2292 } | |
| 2293 else if (24 == counter) | |
| 2294 { | |
| 2295 if (little_endian) | |
| 2296 { | |
| 2297 DECODE_ERROR_OCTET ((ch >> 16) & 0xFF, dst, data, | |
| 2298 ignore_bom); | |
| 2299 DECODE_ERROR_OCTET (ch & 0xFF, dst, data, ignore_bom); | |
| 2300 DECODE_ERROR_OCTET ((ch >> 8) & 0xFF, dst, data, | |
| 2301 ignore_bom); | |
| 2302 } | |
| 2303 else | |
| 2304 { | |
| 2305 DECODE_ERROR_OCTET ((ch >> 16) & 0xFF, dst, data, | |
| 2306 ignore_bom); | |
| 2307 DECODE_ERROR_OCTET ((ch >> 8) & 0xFF, dst, data, | |
| 2308 ignore_bom); | |
| 2309 DECODE_ERROR_OCTET (ch & 0xFF, dst, data, | |
| 2310 ignore_bom); | |
| 2311 } | |
| 2312 } | |
| 2313 else assert(0); | |
| 2314 break; | |
| 2315 } | |
| 2316 ch = 0; | |
|
4688
7e54adf407a1
Fix a bug with Unicode error sequences and very short input strings.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4583
diff
changeset
|
2317 counter = 0; |
| 4096 | 2318 } |
| 771 | 2319 |
| 2320 data->counter = counter; | |
| 4096 | 2321 data->indicated_length = indicated_length; |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
2322 data->characters_seen = characters_seen; |
| 771 | 2323 } |
| 2324 else | |
| 2325 { | |
| 2326 unsigned char char_boundary = data->current_char_boundary; | |
| 2327 Lisp_Object charset = data->current_charset; | |
| 2328 | |
| 2329 #ifdef ENABLE_COMPOSITE_CHARS | |
| 2330 /* flags for handling composite chars. We do a little switcheroo | |
| 2331 on the source while we're outputting the composite char. */ | |
| 2332 Bytecount saved_n = 0; | |
| 867 | 2333 const Ibyte *saved_src = NULL; |
| 771 | 2334 int in_composite = 0; |
| 2335 | |
| 2336 back_to_square_n: | |
| 2337 #endif /* ENABLE_COMPOSITE_CHARS */ | |
| 2338 | |
| 2339 if (XCODING_SYSTEM_UNICODE_NEED_BOM (str->codesys) && !data->wrote_bom) | |
| 2340 { | |
| 4096 | 2341 encode_unicode_char_1 (0xFEFF, dst, type, little_endian, 1); |
| 771 | 2342 data->wrote_bom = 1; |
| 2343 } | |
| 2344 | |
| 2345 while (n--) | |
| 2346 { | |
| 867 | 2347 Ibyte c = *src++; |
| 771 | 2348 |
| 2349 #ifdef MULE | |
| 826 | 2350 if (byte_ascii_p (c)) |
| 771 | 2351 #endif /* MULE */ |
| 2352 { /* Processing ASCII character */ | |
| 2353 ch = 0; | |
| 2354 encode_unicode_char (Vcharset_ascii, c, 0, dst, type, | |
| 4096 | 2355 little_endian, 1); |
| 771 | 2356 |
| 2357 char_boundary = 1; | |
| 2358 } | |
| 2359 #ifdef MULE | |
| 867 | 2360 else if (ibyte_leading_byte_p (c) || ibyte_leading_byte_p (ch)) |
| 771 | 2361 { /* Processing Leading Byte */ |
| 2362 ch = 0; | |
| 826 | 2363 charset = charset_by_leading_byte (c); |
| 2364 if (leading_byte_prefix_p(c)) | |
| 771 | 2365 ch = c; |
| 2366 char_boundary = 0; | |
| 2367 } | |
| 2368 else | |
| 2369 { /* Processing Non-ASCII character */ | |
| 2370 char_boundary = 1; | |
| 2371 if (EQ (charset, Vcharset_control_1)) | |
| 2704 | 2372 /* See: |
| 2373 | |
| 2374 (Info-goto-node "(internals)Internal String Encoding") | |
| 2375 | |
| 2376 for the rationale behind subtracting #xa0 from the | |
| 2377 character's code. */ | |
| 2378 encode_unicode_char (Vcharset_control_1, c - 0xa0, 0, dst, | |
| 4096 | 2379 type, little_endian, 1); |
| 771 | 2380 else |
| 2381 { | |
| 2382 switch (XCHARSET_REP_BYTES (charset)) | |
| 2383 { | |
| 2384 case 2: | |
| 2385 encode_unicode_char (charset, c, 0, dst, type, | |
| 4096 | 2386 little_endian, 1); |
| 771 | 2387 break; |
| 2388 case 3: | |
| 2389 if (XCHARSET_PRIVATE_P (charset)) | |
| 2390 { | |
| 2391 encode_unicode_char (charset, c, 0, dst, type, | |
| 4096 | 2392 little_endian, 1); |
| 771 | 2393 ch = 0; |
| 2394 } | |
| 2395 else if (ch) | |
| 2396 { | |
| 2397 #ifdef ENABLE_COMPOSITE_CHARS | |
| 2398 if (EQ (charset, Vcharset_composite)) | |
| 2399 { | |
| 2400 if (in_composite) | |
| 2401 { | |
| 2402 /* #### Bother! We don't know how to | |
| 2403 handle this yet. */ | |
| 2404 encode_unicode_char (Vcharset_ascii, '~', 0, | |
| 2405 dst, type, | |
| 4096 | 2406 little_endian, 1); |
| 771 | 2407 } |
| 2408 else | |
| 2409 { | |
| 867 | 2410 Ichar emch = make_ichar (Vcharset_composite, |
| 771 | 2411 ch & 0x7F, |
| 2412 c & 0x7F); | |
| 2413 Lisp_Object lstr = | |
| 2414 composite_char_string (emch); | |
| 2415 saved_n = n; | |
| 2416 saved_src = src; | |
| 2417 in_composite = 1; | |
| 2418 src = XSTRING_DATA (lstr); | |
| 2419 n = XSTRING_LENGTH (lstr); | |
| 2420 } | |
| 2421 } | |
| 2422 else | |
| 2423 #endif /* ENABLE_COMPOSITE_CHARS */ | |
| 2424 encode_unicode_char (charset, ch, c, dst, type, | |
| 4096 | 2425 little_endian, 1); |
| 771 | 2426 ch = 0; |
| 2427 } | |
| 2428 else | |
| 2429 { | |
| 2430 ch = c; | |
| 2431 char_boundary = 0; | |
| 2432 } | |
| 2433 break; | |
| 2434 case 4: | |
| 2435 if (ch) | |
| 2436 { | |
| 2437 encode_unicode_char (charset, ch, c, dst, type, | |
| 4096 | 2438 little_endian, 1); |
| 771 | 2439 ch = 0; |
| 2440 } | |
| 2441 else | |
| 2442 { | |
| 2443 ch = c; | |
| 2444 char_boundary = 0; | |
| 2445 } | |
| 2446 break; | |
| 2447 default: | |
| 2500 | 2448 ABORT (); |
| 771 | 2449 } |
| 2450 } | |
| 2451 } | |
| 2452 #endif /* MULE */ | |
| 2453 } | |
| 2454 | |
| 2455 #ifdef ENABLE_COMPOSITE_CHARS | |
| 2456 if (in_composite) | |
| 2457 { | |
| 2458 n = saved_n; | |
| 2459 src = saved_src; | |
| 2460 in_composite = 0; | |
| 2461 goto back_to_square_n; /* Wheeeeeeeee ..... */ | |
| 2462 } | |
| 2463 #endif /* ENABLE_COMPOSITE_CHARS */ | |
| 2464 | |
| 2465 data->current_char_boundary = char_boundary; | |
| 2466 data->current_charset = charset; | |
| 2467 | |
| 2468 /* La palabra se hizo carne! */ | |
| 2469 /* A palavra fez-se carne! */ | |
| 2470 /* Whatever. */ | |
| 2471 } | |
| 2472 | |
| 2473 str->ch = ch; | |
| 2474 return orign; | |
| 2475 } | |
| 2476 | |
| 2477 /* DEFINE_DETECTOR (utf_7); */ | |
| 2478 DEFINE_DETECTOR (utf_8); | |
| 2479 DEFINE_DETECTOR_CATEGORY (utf_8, utf_8); | |
| 985 | 2480 DEFINE_DETECTOR_CATEGORY (utf_8, utf_8_bom); |
| 771 | 2481 DEFINE_DETECTOR (ucs_4); |
| 2482 DEFINE_DETECTOR_CATEGORY (ucs_4, ucs_4); | |
| 2483 DEFINE_DETECTOR (utf_16); | |
| 2484 DEFINE_DETECTOR_CATEGORY (utf_16, utf_16); | |
| 2485 DEFINE_DETECTOR_CATEGORY (utf_16, utf_16_little_endian); | |
| 2486 DEFINE_DETECTOR_CATEGORY (utf_16, utf_16_bom); | |
| 2487 DEFINE_DETECTOR_CATEGORY (utf_16, utf_16_little_endian_bom); | |
| 2488 | |
| 2489 struct ucs_4_detector | |
| 2490 { | |
| 2491 int in_ucs_4_byte; | |
| 2492 }; | |
| 2493 | |
| 2494 static void | |
| 2495 ucs_4_detect (struct detection_state *st, const UExtbyte *src, | |
| 2496 Bytecount n) | |
| 2497 { | |
| 2498 struct ucs_4_detector *data = DETECTION_STATE_DATA (st, ucs_4); | |
| 2499 | |
| 2500 while (n--) | |
| 2501 { | |
| 2502 UExtbyte c = *src++; | |
| 2503 switch (data->in_ucs_4_byte) | |
| 2504 { | |
| 2505 case 0: | |
| 2506 if (c >= 128) | |
| 2507 { | |
| 2508 DET_RESULT (st, ucs_4) = DET_NEARLY_IMPOSSIBLE; | |
| 2509 return; | |
| 2510 } | |
| 2511 else | |
| 2512 data->in_ucs_4_byte++; | |
| 2513 break; | |
| 2514 case 3: | |
| 2515 data->in_ucs_4_byte = 0; | |
| 2516 break; | |
| 2517 default: | |
| 2518 data->in_ucs_4_byte++; | |
| 2519 } | |
| 2520 } | |
| 2521 | |
| 2522 /* !!#### write this for real */ | |
| 2523 DET_RESULT (st, ucs_4) = DET_AS_LIKELY_AS_UNLIKELY; | |
| 2524 } | |
| 2525 | |
| 2526 struct utf_16_detector | |
| 2527 { | |
| 2528 unsigned int seen_ffff:1; | |
| 2529 unsigned int seen_forward_bom:1; | |
| 2530 unsigned int seen_rev_bom:1; | |
| 2531 int byteno; | |
| 2532 int prev_char; | |
| 2533 int text, rev_text; | |
| 1267 | 2534 int sep, rev_sep; |
| 2535 int num_ascii; | |
| 771 | 2536 }; |
| 2537 | |
| 2538 static void | |
| 2539 utf_16_detect (struct detection_state *st, const UExtbyte *src, | |
| 2540 Bytecount n) | |
| 2541 { | |
| 2542 struct utf_16_detector *data = DETECTION_STATE_DATA (st, utf_16); | |
| 2543 | |
| 2544 while (n--) | |
| 2545 { | |
| 2546 UExtbyte c = *src++; | |
| 2547 int prevc = data->prev_char; | |
| 2548 if (data->byteno == 1 && c == 0xFF && prevc == 0xFE) | |
| 2549 data->seen_forward_bom = 1; | |
| 2550 else if (data->byteno == 1 && c == 0xFE && prevc == 0xFF) | |
| 2551 data->seen_rev_bom = 1; | |
| 2552 | |
| 2553 if (data->byteno & 1) | |
| 2554 { | |
| 2555 if (c == 0xFF && prevc == 0xFF) | |
| 2556 data->seen_ffff = 1; | |
| 2557 if (prevc == 0 | |
| 2558 && (c == '\r' || c == '\n' | |
| 2559 || (c >= 0x20 && c <= 0x7E))) | |
| 2560 data->text++; | |
| 2561 if (c == 0 | |
| 2562 && (prevc == '\r' || prevc == '\n' | |
| 2563 || (prevc >= 0x20 && prevc <= 0x7E))) | |
| 2564 data->rev_text++; | |
| 1267 | 2565 /* #### 0x2028 is LINE SEPARATOR and 0x2029 is PARAGRAPH SEPARATOR. |
| 2566 I used to count these in text and rev_text but that is very bad, | |
| 2567 as 0x2028 is also space + left-paren in ASCII, which is extremely | |
| 2568 common. So, what do we do with these? */ | |
| 771 | 2569 if (prevc == 0x20 && (c == 0x28 || c == 0x29)) |
| 1267 | 2570 data->sep++; |
| 771 | 2571 if (c == 0x20 && (prevc == 0x28 || prevc == 0x29)) |
| 1267 | 2572 data->rev_sep++; |
| 771 | 2573 } |
| 2574 | |
| 1267 | 2575 if ((c >= ' ' && c <= '~') || c == '\n' || c == '\r' || c == '\t' || |
| 2576 c == '\f' || c == '\v') | |
| 2577 data->num_ascii++; | |
| 771 | 2578 data->byteno++; |
| 2579 data->prev_char = c; | |
| 2580 } | |
| 2581 | |
| 2582 { | |
| 2583 int variance_indicates_big_endian = | |
| 2584 (data->text >= 10 | |
| 2585 && (data->rev_text == 0 | |
| 2586 || data->text / data->rev_text >= 10)); | |
| 2587 int variance_indicates_little_endian = | |
| 2588 (data->rev_text >= 10 | |
| 2589 && (data->text == 0 | |
| 2590 || data->rev_text / data->text >= 10)); | |
| 2591 | |
| 2592 if (data->seen_ffff) | |
| 2593 SET_DET_RESULTS (st, utf_16, DET_NEARLY_IMPOSSIBLE); | |
| 2594 else if (data->seen_forward_bom) | |
| 2595 { | |
| 2596 SET_DET_RESULTS (st, utf_16, DET_NEARLY_IMPOSSIBLE); | |
| 2597 if (variance_indicates_big_endian) | |
| 2598 DET_RESULT (st, utf_16_bom) = DET_NEAR_CERTAINTY; | |
| 2599 else if (variance_indicates_little_endian) | |
| 2600 DET_RESULT (st, utf_16_bom) = DET_SOMEWHAT_LIKELY; | |
| 2601 else | |
| 2602 DET_RESULT (st, utf_16_bom) = DET_QUITE_PROBABLE; | |
| 2603 } | |
| 2604 else if (data->seen_forward_bom) | |
| 2605 { | |
| 2606 SET_DET_RESULTS (st, utf_16, DET_NEARLY_IMPOSSIBLE); | |
| 2607 if (variance_indicates_big_endian) | |
| 2608 DET_RESULT (st, utf_16_bom) = DET_NEAR_CERTAINTY; | |
| 2609 else if (variance_indicates_little_endian) | |
| 2610 /* #### may need to rethink */ | |
| 2611 DET_RESULT (st, utf_16_bom) = DET_SOMEWHAT_LIKELY; | |
| 2612 else | |
| 2613 /* #### may need to rethink */ | |
| 2614 DET_RESULT (st, utf_16_bom) = DET_QUITE_PROBABLE; | |
| 2615 } | |
| 2616 else if (data->seen_rev_bom) | |
| 2617 { | |
| 2618 SET_DET_RESULTS (st, utf_16, DET_NEARLY_IMPOSSIBLE); | |
| 2619 if (variance_indicates_little_endian) | |
| 2620 DET_RESULT (st, utf_16_little_endian_bom) = DET_NEAR_CERTAINTY; | |
| 2621 else if (variance_indicates_big_endian) | |
| 2622 /* #### may need to rethink */ | |
| 2623 DET_RESULT (st, utf_16_little_endian_bom) = DET_SOMEWHAT_LIKELY; | |
| 2624 else | |
| 2625 /* #### may need to rethink */ | |
| 2626 DET_RESULT (st, utf_16_little_endian_bom) = DET_QUITE_PROBABLE; | |
| 2627 } | |
| 2628 else if (variance_indicates_big_endian) | |
| 2629 { | |
| 2630 SET_DET_RESULTS (st, utf_16, DET_NEARLY_IMPOSSIBLE); | |
| 2631 DET_RESULT (st, utf_16) = DET_SOMEWHAT_LIKELY; | |
| 2632 DET_RESULT (st, utf_16_little_endian) = DET_SOMEWHAT_UNLIKELY; | |
| 2633 } | |
| 2634 else if (variance_indicates_little_endian) | |
| 2635 { | |
| 2636 SET_DET_RESULTS (st, utf_16, DET_NEARLY_IMPOSSIBLE); | |
| 2637 DET_RESULT (st, utf_16) = DET_SOMEWHAT_UNLIKELY; | |
| 2638 DET_RESULT (st, utf_16_little_endian) = DET_SOMEWHAT_LIKELY; | |
| 2639 } | |
| 2640 else | |
| 1267 | 2641 { |
| 2642 /* #### FUCKME! There should really be an ASCII detector. This | |
| 2643 would rule out the need to have this built-in here as | |
| 2644 well. --ben */ | |
| 1292 | 2645 int pct_ascii = data->byteno ? (100 * data->num_ascii) / data->byteno |
| 2646 : 100; | |
| 1267 | 2647 |
| 2648 if (pct_ascii > 90) | |
| 2649 SET_DET_RESULTS (st, utf_16, DET_QUITE_IMPROBABLE); | |
| 2650 else if (pct_ascii > 75) | |
| 2651 SET_DET_RESULTS (st, utf_16, DET_SOMEWHAT_UNLIKELY); | |
| 2652 else | |
| 2653 SET_DET_RESULTS (st, utf_16, DET_AS_LIKELY_AS_UNLIKELY); | |
| 2654 } | |
| 771 | 2655 } |
| 2656 } | |
| 2657 | |
| 2658 struct utf_8_detector | |
| 2659 { | |
| 985 | 2660 int byteno; |
| 2661 int first_byte; | |
| 2662 int second_byte; | |
| 1267 | 2663 int prev_byte; |
| 771 | 2664 int in_utf_8_byte; |
| 1267 | 2665 int recent_utf_8_sequence; |
| 2666 int seen_bogus_utf8; | |
| 2667 int seen_really_bogus_utf8; | |
| 2668 int seen_2byte_sequence; | |
| 2669 int seen_longer_sequence; | |
| 2670 int seen_iso2022_esc; | |
| 2671 int seen_iso_shift; | |
| 1887 | 2672 unsigned int seen_utf_bom:1; |
| 771 | 2673 }; |
| 2674 | |
| 2675 static void | |
| 2676 utf_8_detect (struct detection_state *st, const UExtbyte *src, | |
| 2677 Bytecount n) | |
| 2678 { | |
| 2679 struct utf_8_detector *data = DETECTION_STATE_DATA (st, utf_8); | |
| 2680 | |
| 2681 while (n--) | |
| 2682 { | |
| 2683 UExtbyte c = *src++; | |
| 985 | 2684 switch (data->byteno) |
| 2685 { | |
| 2686 case 0: | |
| 2687 data->first_byte = c; | |
| 2688 break; | |
| 2689 case 1: | |
| 2690 data->second_byte = c; | |
| 2691 break; | |
| 2692 case 2: | |
| 2693 if (data->first_byte == 0xef && | |
| 2694 data->second_byte == 0xbb && | |
| 2695 c == 0xbf) | |
| 1267 | 2696 data->seen_utf_bom = 1; |
| 985 | 2697 break; |
| 2698 } | |
| 2699 | |
| 771 | 2700 switch (data->in_utf_8_byte) |
| 2701 { | |
| 2702 case 0: | |
| 1267 | 2703 if (data->prev_byte == ISO_CODE_ESC && c >= 0x28 && c <= 0x2F) |
| 2704 data->seen_iso2022_esc++; | |
| 2705 else if (c == ISO_CODE_SI || c == ISO_CODE_SO) | |
| 2706 data->seen_iso_shift++; | |
| 771 | 2707 else if (c >= 0xfc) |
| 2708 data->in_utf_8_byte = 5; | |
| 2709 else if (c >= 0xf8) | |
| 2710 data->in_utf_8_byte = 4; | |
| 2711 else if (c >= 0xf0) | |
| 2712 data->in_utf_8_byte = 3; | |
| 2713 else if (c >= 0xe0) | |
| 2714 data->in_utf_8_byte = 2; | |
| 2715 else if (c >= 0xc0) | |
| 2716 data->in_utf_8_byte = 1; | |
| 2717 else if (c >= 0x80) | |
| 1267 | 2718 data->seen_bogus_utf8++; |
| 2719 if (data->in_utf_8_byte > 0) | |
| 2720 data->recent_utf_8_sequence = data->in_utf_8_byte; | |
| 771 | 2721 break; |
| 2722 default: | |
| 2723 if ((c & 0xc0) != 0x80) | |
| 1267 | 2724 data->seen_really_bogus_utf8++; |
| 2725 else | |
| 771 | 2726 { |
| 1267 | 2727 data->in_utf_8_byte--; |
| 2728 if (data->in_utf_8_byte == 0) | |
| 2729 { | |
| 2730 if (data->recent_utf_8_sequence == 1) | |
| 2731 data->seen_2byte_sequence++; | |
| 2732 else | |
| 2733 { | |
| 2734 assert (data->recent_utf_8_sequence >= 2); | |
| 2735 data->seen_longer_sequence++; | |
| 2736 } | |
| 2737 } | |
| 771 | 2738 } |
| 2739 } | |
| 985 | 2740 |
| 2741 data->byteno++; | |
| 1267 | 2742 data->prev_byte = c; |
| 771 | 2743 } |
| 1267 | 2744 |
| 2745 /* either BOM or no BOM, but not both */ | |
| 2746 SET_DET_RESULTS (st, utf_8, DET_NEARLY_IMPOSSIBLE); | |
| 2747 | |
| 2748 | |
| 2749 if (data->seen_utf_bom) | |
| 2750 DET_RESULT (st, utf_8_bom) = DET_NEAR_CERTAINTY; | |
| 2751 else | |
| 2752 { | |
| 2753 if (data->seen_really_bogus_utf8 || | |
| 2754 data->seen_bogus_utf8 >= 2) | |
| 2755 ; /* bogus */ | |
| 2756 else if (data->seen_bogus_utf8) | |
| 2757 DET_RESULT (st, utf_8) = DET_SOMEWHAT_UNLIKELY; | |
| 2758 else if ((data->seen_longer_sequence >= 5 || | |
| 2759 data->seen_2byte_sequence >= 10) && | |
| 2760 (!(data->seen_iso2022_esc + data->seen_iso_shift) || | |
| 2761 (data->seen_longer_sequence * 2 + data->seen_2byte_sequence) / | |
| 2762 (data->seen_iso2022_esc + data->seen_iso_shift) >= 10)) | |
| 2763 /* heuristics, heuristics, we love heuristics */ | |
| 2764 DET_RESULT (st, utf_8) = DET_QUITE_PROBABLE; | |
| 2765 else if (data->seen_iso2022_esc || | |
| 2766 data->seen_iso_shift >= 3) | |
| 2767 DET_RESULT (st, utf_8) = DET_SOMEWHAT_UNLIKELY; | |
| 2768 else if (data->seen_longer_sequence || | |
| 2769 data->seen_2byte_sequence) | |
| 2770 DET_RESULT (st, utf_8) = DET_SOMEWHAT_LIKELY; | |
| 2771 else if (data->seen_iso_shift) | |
| 2772 DET_RESULT (st, utf_8) = DET_SOMEWHAT_UNLIKELY; | |
| 2773 else | |
| 2774 DET_RESULT (st, utf_8) = DET_AS_LIKELY_AS_UNLIKELY; | |
| 2775 } | |
| 771 | 2776 } |
| 2777 | |
| 2778 static void | |
| 2779 unicode_init_coding_stream (struct coding_stream *str) | |
| 2780 { | |
| 2781 struct unicode_coding_stream *data = | |
| 2782 CODING_STREAM_TYPE_DATA (str, unicode); | |
| 2783 xzero (*data); | |
| 2784 data->current_charset = Qnil; | |
| 2785 } | |
| 2786 | |
| 2787 static void | |
| 2788 unicode_rewind_coding_stream (struct coding_stream *str) | |
| 2789 { | |
| 2790 unicode_init_coding_stream (str); | |
| 2791 } | |
| 2792 | |
| 2793 static int | |
| 2794 unicode_putprop (Lisp_Object codesys, Lisp_Object key, Lisp_Object value) | |
| 2795 { | |
| 3767 | 2796 if (EQ (key, Qunicode_type)) |
| 771 | 2797 { |
| 2798 enum unicode_type type; | |
| 2799 | |
| 2800 if (EQ (value, Qutf_8)) | |
| 2801 type = UNICODE_UTF_8; | |
| 2802 else if (EQ (value, Qutf_16)) | |
| 2803 type = UNICODE_UTF_16; | |
| 2804 else if (EQ (value, Qutf_7)) | |
| 2805 type = UNICODE_UTF_7; | |
| 2806 else if (EQ (value, Qucs_4)) | |
| 2807 type = UNICODE_UCS_4; | |
| 4096 | 2808 else if (EQ (value, Qutf_32)) |
| 2809 type = UNICODE_UTF_32; | |
| 771 | 2810 else |
| 2811 invalid_constant ("Invalid Unicode type", key); | |
| 2812 | |
| 2813 XCODING_SYSTEM_UNICODE_TYPE (codesys) = type; | |
| 2814 } | |
| 2815 else if (EQ (key, Qlittle_endian)) | |
| 2816 XCODING_SYSTEM_UNICODE_LITTLE_ENDIAN (codesys) = !NILP (value); | |
| 2817 else if (EQ (key, Qneed_bom)) | |
| 2818 XCODING_SYSTEM_UNICODE_NEED_BOM (codesys) = !NILP (value); | |
| 2819 else | |
| 2820 return 0; | |
| 2821 return 1; | |
| 2822 } | |
| 2823 | |
| 2824 static Lisp_Object | |
| 2825 unicode_getprop (Lisp_Object coding_system, Lisp_Object prop) | |
| 2826 { | |
| 3767 | 2827 if (EQ (prop, Qunicode_type)) |
| 771 | 2828 { |
| 2829 switch (XCODING_SYSTEM_UNICODE_TYPE (coding_system)) | |
| 2830 { | |
| 2831 case UNICODE_UTF_16: return Qutf_16; | |
| 2832 case UNICODE_UTF_8: return Qutf_8; | |
| 2833 case UNICODE_UTF_7: return Qutf_7; | |
| 2834 case UNICODE_UCS_4: return Qucs_4; | |
| 4096 | 2835 case UNICODE_UTF_32: return Qutf_32; |
| 2500 | 2836 default: ABORT (); |
| 771 | 2837 } |
| 2838 } | |
| 2839 else if (EQ (prop, Qlittle_endian)) | |
| 2840 return XCODING_SYSTEM_UNICODE_LITTLE_ENDIAN (coding_system) ? Qt : Qnil; | |
| 2841 else if (EQ (prop, Qneed_bom)) | |
| 2842 return XCODING_SYSTEM_UNICODE_NEED_BOM (coding_system) ? Qt : Qnil; | |
| 2843 return Qunbound; | |
| 2844 } | |
| 2845 | |
| 2846 static void | |
| 2286 | 2847 unicode_print (Lisp_Object cs, Lisp_Object printcharfun, |
| 2848 int UNUSED (escapeflag)) | |
| 771 | 2849 { |
| 3767 | 2850 write_fmt_string_lisp (printcharfun, "(%s", 1, |
| 2851 unicode_getprop (cs, Qunicode_type)); | |
| 771 | 2852 if (XCODING_SYSTEM_UNICODE_LITTLE_ENDIAN (cs)) |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
2853 write_ascstring (printcharfun, ", little-endian"); |
| 771 | 2854 if (XCODING_SYSTEM_UNICODE_NEED_BOM (cs)) |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
2855 write_ascstring (printcharfun, ", need-bom"); |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
2856 write_ascstring (printcharfun, ")"); |
| 771 | 2857 } |
| 2858 | |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2859 #ifdef MULE |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2860 DEFUN ("set-unicode-query-skip-chars-args", Fset_unicode_query_skip_chars_args, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2861 3, 3, 0, /* |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2862 Specify strings as matching characters known to Unicode coding systems. |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2863 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2864 QUERY-STRING is a string matching characters that can unequivocally be |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2865 encoded by the Unicode coding systems. |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2866 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2867 INVALID-STRING is a string to match XEmacs characters that represent known |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2868 octets on disk, but that are invalid sequences according to Unicode. |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2869 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2870 UTF-8-INVALID-STRING is a more restrictive string to match XEmacs characters |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2871 that are invalid UTF-8 octets. |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2872 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2873 All three strings are in the format accepted by `skip-chars-forward'. |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2874 */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2875 (query_string, invalid_string, utf_8_invalid_string)) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2876 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2877 CHECK_STRING (query_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2878 CHECK_STRING (invalid_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2879 CHECK_STRING (utf_8_invalid_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2880 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2881 Vunicode_query_string = query_string; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2882 Vunicode_invalid_string = invalid_string; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2883 Vutf_8_invalid_string = utf_8_invalid_string; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2884 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2885 return Qnil; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2886 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2887 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2888 static void |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2889 add_lisp_string_to_skip_chars_range (Lisp_Object string, Lisp_Object rtab, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2890 Lisp_Object value) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2891 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2892 Ibyte *p, *pend; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2893 Ichar c; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2894 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2895 p = XSTRING_DATA (string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2896 pend = p + XSTRING_LENGTH (string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2897 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2898 while (p != pend) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2899 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2900 c = itext_ichar (p); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2901 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2902 INC_IBYTEPTR (p); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2903 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2904 if (c == '\\') |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2905 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2906 if (p == pend) break; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2907 c = itext_ichar (p); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2908 INC_IBYTEPTR (p); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2909 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2910 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2911 if (p != pend && *p == '-') |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2912 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2913 Ichar cend; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2914 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2915 /* Skip over the dash. */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2916 p++; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2917 if (p == pend) break; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2918 cend = itext_ichar (p); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2919 |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
2920 Fput_range_table (make_fixnum (c), make_fixnum (cend), value, |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2921 rtab); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2922 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2923 INC_IBYTEPTR (p); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2924 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2925 else |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2926 { |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
2927 Fput_range_table (make_fixnum (c), make_fixnum (c), value, rtab); |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2928 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2929 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2930 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2931 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2932 /* This function wouldn't be necessary if initialised range tables were |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2933 dumped properly; see |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2934 http://mid.gmane.org/18179.49815.622843.336527@parhasard.net . */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2935 static void |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2936 initialize_unicode_query_range_tables_from_strings (void) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2937 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2938 CHECK_STRING (Vunicode_query_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2939 CHECK_STRING (Vunicode_invalid_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2940 CHECK_STRING (Vutf_8_invalid_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2941 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2942 Vunicode_query_skip_chars = Fmake_range_table (Qstart_closed_end_closed); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2943 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2944 add_lisp_string_to_skip_chars_range (Vunicode_query_string, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2945 Vunicode_query_skip_chars, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2946 Qsucceeded); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2947 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2948 Vunicode_invalid_and_query_skip_chars |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2949 = Fcopy_range_table (Vunicode_query_skip_chars); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2950 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2951 add_lisp_string_to_skip_chars_range (Vunicode_invalid_string, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2952 Vunicode_invalid_and_query_skip_chars, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2953 Qinvalid_sequence); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2954 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2955 Vutf_8_invalid_and_query_skip_chars |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2956 = Fcopy_range_table (Vunicode_query_skip_chars); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2957 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2958 add_lisp_string_to_skip_chars_range (Vutf_8_invalid_string, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2959 Vutf_8_invalid_and_query_skip_chars, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2960 Qinvalid_sequence); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2961 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2962 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2963 static Lisp_Object |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2964 unicode_query (Lisp_Object codesys, struct buffer *buf, Charbpos end, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2965 int flags) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2966 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2967 Charbpos pos = BUF_PT (buf), fail_range_start, fail_range_end; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2968 Charbpos pos_byte = BYTE_BUF_PT (buf); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2969 Lisp_Object skip_chars_range_table, result = Qnil; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2970 enum query_coding_failure_reasons failed_reason, |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2971 previous_failed_reason = query_coding_succeeded; |
|
4824
c12b646d84ee
changes to get things to compile under latest cygwin
Ben Wing <ben@xemacs.org>
parents:
4770
diff
changeset
|
2972 int checked_unicode, |
|
c12b646d84ee
changes to get things to compile under latest cygwin
Ben Wing <ben@xemacs.org>
parents:
4770
diff
changeset
|
2973 invalid_lower_limit = UNICODE_ERROR_OCTET_RANGE_START, |
|
c12b646d84ee
changes to get things to compile under latest cygwin
Ben Wing <ben@xemacs.org>
parents:
4770
diff
changeset
|
2974 invalid_upper_limit = -1, |
|
c12b646d84ee
changes to get things to compile under latest cygwin
Ben Wing <ben@xemacs.org>
parents:
4770
diff
changeset
|
2975 unicode_type = XCODING_SYSTEM_UNICODE_TYPE (codesys); |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2976 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2977 if (flags & QUERY_METHOD_HIGHLIGHT && |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2978 /* If we're being called really early, live without highlights getting |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2979 cleared properly: */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2980 !(UNBOUNDP (XSYMBOL (Qquery_coding_clear_highlights)->function))) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2981 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2982 /* It's okay to call Lisp here, the only non-stack object we may have |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2983 allocated up to this point is skip_chars_range_table, and that's |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2984 reachable from its entry in Vfixed_width_query_ranges_cache. */ |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
2985 call3 (Qquery_coding_clear_highlights, make_fixnum (pos), make_fixnum (end), |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2986 wrap_buffer (buf)); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2987 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2988 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2989 if (NILP (Vunicode_query_skip_chars)) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2990 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2991 initialize_unicode_query_range_tables_from_strings(); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2992 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2993 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2994 if (flags & QUERY_METHOD_IGNORE_INVALID_SEQUENCES) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2995 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2996 switch (unicode_type) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2997 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2998 case UNICODE_UTF_8: |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
2999 skip_chars_range_table = Vutf_8_invalid_and_query_skip_chars; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3000 break; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3001 case UNICODE_UTF_7: |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3002 /* #### See above. */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3003 return Qunbound; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3004 break; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3005 default: |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3006 skip_chars_range_table = Vunicode_invalid_and_query_skip_chars; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3007 break; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3008 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3009 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3010 else |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3011 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3012 switch (unicode_type) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3013 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3014 case UNICODE_UTF_8: |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3015 invalid_lower_limit = UNICODE_ERROR_OCTET_RANGE_START + 0x80; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3016 invalid_upper_limit = UNICODE_ERROR_OCTET_RANGE_START + 0xFF; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3017 break; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3018 case UNICODE_UTF_7: |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3019 /* #### Work out what to do here in reality, read the spec and decide |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3020 which octets are invalid. */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3021 return Qunbound; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3022 break; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3023 default: |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3024 invalid_lower_limit = UNICODE_ERROR_OCTET_RANGE_START; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3025 invalid_upper_limit = UNICODE_ERROR_OCTET_RANGE_START + 0xFF; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3026 break; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3027 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3028 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3029 skip_chars_range_table = Vunicode_query_skip_chars; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3030 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3031 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3032 while (pos < end) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3033 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3034 Ichar ch = BYTE_BUF_FETCH_CHAR (buf, pos_byte); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3035 if ((ch < 0x100 ? 1 : |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
3036 (!EQ (Qnil, Fget_range_table (make_fixnum (ch), skip_chars_range_table, |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3037 Qnil))))) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3038 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3039 pos++; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3040 INC_BYTEBPOS (buf, pos_byte); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3041 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3042 else |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3043 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3044 fail_range_start = pos; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3045 while ((pos < end) && |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3046 ((checked_unicode = ichar_to_unicode (ch), |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3047 -1 == checked_unicode |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3048 && (failed_reason = query_coding_unencodable)) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3049 || (!(flags & QUERY_METHOD_IGNORE_INVALID_SEQUENCES) && |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3050 (invalid_lower_limit <= checked_unicode) && |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3051 (checked_unicode <= invalid_upper_limit) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3052 && (failed_reason = query_coding_invalid_sequence))) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3053 && (previous_failed_reason == query_coding_succeeded |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3054 || previous_failed_reason == failed_reason)) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3055 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3056 pos++; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3057 INC_BYTEBPOS (buf, pos_byte); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3058 ch = BYTE_BUF_FETCH_CHAR (buf, pos_byte); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3059 previous_failed_reason = failed_reason; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3060 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3061 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3062 if (fail_range_start == pos) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3063 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3064 /* The character can actually be encoded; move on. */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3065 pos++; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3066 INC_BYTEBPOS (buf, pos_byte); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3067 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3068 else |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3069 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3070 assert (previous_failed_reason == query_coding_invalid_sequence |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3071 || previous_failed_reason == query_coding_unencodable); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3072 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3073 if (flags & QUERY_METHOD_ERRORP) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3074 { |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3075 signal_error_2 |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3076 (Qtext_conversion_error, |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3077 "Cannot encode using coding system", |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3078 make_string_from_buffer (buf, fail_range_start, |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3079 pos - fail_range_start), |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3080 XCODING_SYSTEM_NAME (codesys)); |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3081 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3082 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3083 if (NILP (result)) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3084 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3085 result = Fmake_range_table (Qstart_closed_end_open); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3086 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3087 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3088 fail_range_end = pos; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3089 |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
3090 Fput_range_table (make_fixnum (fail_range_start), |
|
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
3091 make_fixnum (fail_range_end), |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3092 (previous_failed_reason |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3093 == query_coding_unencodable ? |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3094 Qunencodable : Qinvalid_sequence), |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3095 result); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3096 previous_failed_reason = query_coding_succeeded; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3097 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3098 if (flags & QUERY_METHOD_HIGHLIGHT) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3099 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3100 Lisp_Object extent |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
3101 = Fmake_extent (make_fixnum (fail_range_start), |
|
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
3102 make_fixnum (fail_range_end), |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3103 wrap_buffer (buf)); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3104 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3105 Fset_extent_priority |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
3106 (extent, make_fixnum (2 + mouse_highlight_priority)); |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3107 Fset_extent_face (extent, Qquery_coding_warning_face); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3108 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3109 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3110 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3111 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3112 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3113 return result; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3114 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3115 #else /* !MULE */ |
|
4770
b9aaf2a18957
Add missing return value type to unicode_query.
Stephen J. Turnbull <stephen@xemacs.org>
parents:
4690
diff
changeset
|
3116 static Lisp_Object |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3117 unicode_query (Lisp_Object UNUSED (codesys), |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3118 struct buffer * UNUSED (buf), |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3119 Charbpos UNUSED (end), int UNUSED (flags)) |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3120 { |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3121 return Qnil; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3122 } |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3123 #endif |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3124 |
| 771 | 3125 int |
| 2286 | 3126 dfc_coding_system_is_unicode ( |
| 3127 #ifdef WIN32_ANY | |
| 3128 Lisp_Object codesys | |
| 3129 #else | |
| 3130 Lisp_Object UNUSED (codesys) | |
| 3131 #endif | |
| 3132 ) | |
| 771 | 3133 { |
| 1315 | 3134 #ifdef WIN32_ANY |
| 771 | 3135 codesys = Fget_coding_system (codesys); |
| 3136 return (EQ (XCODING_SYSTEM_TYPE (codesys), Qunicode) && | |
| 3137 XCODING_SYSTEM_UNICODE_TYPE (codesys) == UNICODE_UTF_16 && | |
| 3138 XCODING_SYSTEM_UNICODE_LITTLE_ENDIAN (codesys)); | |
| 3139 | |
| 3140 #else | |
| 3141 return 0; | |
| 3142 #endif | |
| 3143 } | |
| 3144 | |
| 3145 | |
| 3146 /************************************************************************/ | |
| 3147 /* Initialization */ | |
| 3148 /************************************************************************/ | |
| 3149 | |
| 3150 void | |
| 3151 syms_of_unicode (void) | |
| 3152 { | |
| 3153 #ifdef MULE | |
| 877 | 3154 DEFSUBR (Funicode_precedence_list); |
| 771 | 3155 DEFSUBR (Fset_language_unicode_precedence_list); |
| 3156 DEFSUBR (Flanguage_unicode_precedence_list); | |
| 3157 DEFSUBR (Fset_default_unicode_precedence_list); | |
| 3158 DEFSUBR (Fdefault_unicode_precedence_list); | |
| 3159 DEFSUBR (Fset_unicode_conversion); | |
| 3160 | |
| 1318 | 3161 DEFSUBR (Fload_unicode_mapping_table); |
| 771 | 3162 |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3163 DEFSUBR (Fset_unicode_query_skip_chars_args); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3164 |
| 3439 | 3165 DEFSYMBOL (Qccl_encode_to_ucs_2); |
| 3166 DEFSYMBOL (Qlast_allocated_character); | |
| 771 | 3167 DEFSYMBOL (Qignore_first_column); |
| 3659 | 3168 |
| 3169 DEFSYMBOL (Qunicode_registries); | |
| 771 | 3170 #endif /* MULE */ |
| 3171 | |
| 800 | 3172 DEFSUBR (Fchar_to_unicode); |
| 3173 DEFSUBR (Funicode_to_char); | |
| 771 | 3174 |
| 3175 DEFSYMBOL (Qunicode); | |
| 3176 DEFSYMBOL (Qucs_4); | |
| 3177 DEFSYMBOL (Qutf_16); | |
| 4096 | 3178 DEFSYMBOL (Qutf_32); |
| 771 | 3179 DEFSYMBOL (Qutf_8); |
| 3180 DEFSYMBOL (Qutf_7); | |
| 3181 | |
| 3182 DEFSYMBOL (Qneed_bom); | |
| 3183 | |
| 3184 DEFSYMBOL (Qutf_16); | |
| 3185 DEFSYMBOL (Qutf_16_little_endian); | |
| 3186 DEFSYMBOL (Qutf_16_bom); | |
| 3187 DEFSYMBOL (Qutf_16_little_endian_bom); | |
| 985 | 3188 |
| 3189 DEFSYMBOL (Qutf_8); | |
| 3190 DEFSYMBOL (Qutf_8_bom); | |
| 771 | 3191 } |
| 3192 | |
| 3193 void | |
| 3194 coding_system_type_create_unicode (void) | |
| 3195 { | |
| 3196 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA (unicode, "unicode-coding-system-p"); | |
| 3197 CODING_SYSTEM_HAS_METHOD (unicode, print); | |
| 3198 CODING_SYSTEM_HAS_METHOD (unicode, convert); | |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3199 CODING_SYSTEM_HAS_METHOD (unicode, query); |
| 771 | 3200 CODING_SYSTEM_HAS_METHOD (unicode, init_coding_stream); |
| 3201 CODING_SYSTEM_HAS_METHOD (unicode, rewind_coding_stream); | |
| 3202 CODING_SYSTEM_HAS_METHOD (unicode, putprop); | |
| 3203 CODING_SYSTEM_HAS_METHOD (unicode, getprop); | |
| 3204 | |
|
5776
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
3205 CODING_SYSTEM_HAS_METHOD (unicode, character_tell); |
|
65d65b52d608
Pass character count from coding systems to buffer insertion code.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5736
diff
changeset
|
3206 |
| 771 | 3207 INITIALIZE_DETECTOR (utf_8); |
| 3208 DETECTOR_HAS_METHOD (utf_8, detect); | |
| 3209 INITIALIZE_DETECTOR_CATEGORY (utf_8, utf_8); | |
| 985 | 3210 INITIALIZE_DETECTOR_CATEGORY (utf_8, utf_8_bom); |
| 771 | 3211 |
| 3212 INITIALIZE_DETECTOR (ucs_4); | |
| 3213 DETECTOR_HAS_METHOD (ucs_4, detect); | |
| 3214 INITIALIZE_DETECTOR_CATEGORY (ucs_4, ucs_4); | |
| 3215 | |
| 3216 INITIALIZE_DETECTOR (utf_16); | |
| 3217 DETECTOR_HAS_METHOD (utf_16, detect); | |
| 3218 INITIALIZE_DETECTOR_CATEGORY (utf_16, utf_16); | |
| 3219 INITIALIZE_DETECTOR_CATEGORY (utf_16, utf_16_little_endian); | |
| 3220 INITIALIZE_DETECTOR_CATEGORY (utf_16, utf_16_bom); | |
| 3221 INITIALIZE_DETECTOR_CATEGORY (utf_16, utf_16_little_endian_bom); | |
| 3222 } | |
| 3223 | |
| 3224 void | |
| 3225 reinit_coding_system_type_create_unicode (void) | |
| 3226 { | |
| 3227 REINITIALIZE_CODING_SYSTEM_TYPE (unicode); | |
| 3228 } | |
| 3229 | |
| 3230 void | |
| 3231 vars_of_unicode (void) | |
| 3232 { | |
| 3233 Fprovide (intern ("unicode")); | |
| 3234 | |
| 3235 #ifdef MULE | |
| 4270 | 3236 staticpro (&Vnumber_of_jit_charsets); |
|
5581
56144c8593a8
Mechanically change INT to FIXNUM in our sources.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5474
diff
changeset
|
3237 Vnumber_of_jit_charsets = make_fixnum (0); |
| 4270 | 3238 staticpro (&Vlast_jit_charset_final); |
| 3239 Vlast_jit_charset_final = make_char (0x30); | |
| 3240 staticpro (&Vcharset_descr); | |
| 3241 Vcharset_descr | |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3242 = build_defer_string ("Mule charset for otherwise unknown Unicode code points."); |
| 4270 | 3243 |
| 771 | 3244 staticpro (&Vlanguage_unicode_precedence_list); |
| 3245 Vlanguage_unicode_precedence_list = Qnil; | |
| 3246 | |
| 3247 staticpro (&Vdefault_unicode_precedence_list); | |
| 3248 Vdefault_unicode_precedence_list = Qnil; | |
| 3249 | |
| 3250 unicode_precedence_dynarr = Dynarr_new (Lisp_Object); | |
| 2367 | 3251 dump_add_root_block_ptr (&unicode_precedence_dynarr, |
| 771 | 3252 &lisp_object_dynarr_description); |
| 2367 | 3253 |
| 3659 | 3254 |
| 3255 | |
| 2367 | 3256 init_blank_unicode_tables (); |
| 3257 | |
| 3439 | 3258 staticpro (&Vcurrent_jit_charset); |
| 3259 Vcurrent_jit_charset = Qnil; | |
| 3260 | |
| 2367 | 3261 /* Note that the "block" we are describing is a single pointer, and hence |
| 3262 we could potentially use dump_add_root_block_ptr(). However, given | |
| 3263 the way the descriptions are written, we couldn't use them, and would | |
| 3264 have to write new descriptions for each of the pointers below, since | |
| 3265 we would have to make use of a description with an XD_BLOCK_ARRAY | |
| 3266 in it. */ | |
| 3267 | |
| 3268 dump_add_root_block (&to_unicode_blank_1, sizeof (void *), | |
| 3269 to_unicode_level_1_desc_1); | |
| 3270 dump_add_root_block (&to_unicode_blank_2, sizeof (void *), | |
| 3271 to_unicode_level_2_desc_1); | |
| 3272 | |
| 3273 dump_add_root_block (&from_unicode_blank_1, sizeof (void *), | |
| 3274 from_unicode_level_1_desc_1); | |
| 3275 dump_add_root_block (&from_unicode_blank_2, sizeof (void *), | |
| 3276 from_unicode_level_2_desc_1); | |
| 3277 dump_add_root_block (&from_unicode_blank_3, sizeof (void *), | |
| 3278 from_unicode_level_3_desc_1); | |
| 3279 dump_add_root_block (&from_unicode_blank_4, sizeof (void *), | |
| 3280 from_unicode_level_4_desc_1); | |
| 3659 | 3281 |
| 3282 DEFVAR_LISP ("unicode-registries", &Qunicode_registries /* | |
| 3283 Vector describing the X11 registries searched when using fallback fonts. | |
| 3284 | |
| 3285 "Fallback fonts" here includes by default those fonts used by redisplay when | |
| 3286 displaying charsets for which the `encode-as-utf-8' property is true, and | |
| 3287 those used when no font matching the charset's registries property has been | |
| 3288 found (that is, they're probably Mule-specific charsets like Ethiopic or | |
| 3289 IPA.) | |
| 3290 */ ); | |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3291 Qunicode_registries = vector1(build_ascstring("iso10646-1")); |
|
4690
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3292 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3293 /* Initialised in lisp/mule/general-late.el, by a call to |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3294 #'set-unicode-query-skip-chars-args. Or at least they would be, but we |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3295 can't do this at dump time right now, initialised range tables aren't |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3296 dumped properly. */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3297 staticpro (&Vunicode_invalid_and_query_skip_chars); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3298 Vunicode_invalid_and_query_skip_chars = Qnil; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3299 staticpro (&Vutf_8_invalid_and_query_skip_chars); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3300 Vutf_8_invalid_and_query_skip_chars = Qnil; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3301 staticpro (&Vunicode_query_skip_chars); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3302 Vunicode_query_skip_chars = Qnil; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3303 |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3304 /* If we could dump the range table above these wouldn't be necessary: */ |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3305 staticpro (&Vunicode_query_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3306 Vunicode_query_string = Qnil; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3307 staticpro (&Vunicode_invalid_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3308 Vunicode_invalid_string = Qnil; |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3309 staticpro (&Vutf_8_invalid_string); |
|
257b468bf2ca
Move the #'query-coding-region implementation to C.
Aidan Kehoe <kehoea@parhasard.net>
parents:
4688
diff
changeset
|
3310 Vutf_8_invalid_string = Qnil; |
| 771 | 3311 #endif /* MULE */ |
| 3312 } | |
|
4834
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3313 |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3314 void |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3315 complex_vars_of_unicode (void) |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3316 { |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3317 /* We used to define this in unicode.el. But we need it early for |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3318 Cygwin 1.7 -- used in LOCAL_FILE_FORMAT_TO_TSTR() et al. */ |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3319 Fmake_coding_system_internal |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3320 (Qutf_8, Qunicode, |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4834
diff
changeset
|
3321 build_defer_string ("UTF-8"), |
|
5345
db326b8fe982
Use Ben's recently-introduced listu (), where appropriate.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5307
diff
changeset
|
3322 listu (Qdocumentation, |
|
db326b8fe982
Use Ben's recently-introduced listu (), where appropriate.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5307
diff
changeset
|
3323 build_defer_string ( |
|
4834
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3324 "UTF-8 Unicode encoding -- ASCII-compatible 8-bit variable-width encoding\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3325 "sharing the following principles with the Mule-internal encoding:\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3326 "\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3327 " -- All ASCII characters (codepoints 0 through 127) are represented\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3328 " by themselves (i.e. using one byte, with the same value as the\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3329 " ASCII codepoint), and these bytes are disjoint from bytes\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3330 " representing non-ASCII characters.\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3331 "\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3332 " This means that any 8-bit clean application can safely process\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3333 " UTF-8-encoded text as it were ASCII, with no corruption (e.g. a\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3334 " '/' byte is always a slash character, never the second byte of\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3335 " some other character, as with Big5, so a pathname encoded in\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3336 " UTF-8 can safely be split up into components and reassembled\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3337 " again using standard ASCII processes).\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3338 "\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3339 " -- Leading bytes and non-leading bytes in the encoding of a\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3340 " character are disjoint, so moving backwards is easy.\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3341 "\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3342 " -- Given only the leading byte, you know how many following bytes\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3343 " are present.\n" |
|
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3344 ), |
|
5345
db326b8fe982
Use Ben's recently-introduced listu (), where appropriate.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5307
diff
changeset
|
3345 Qmnemonic, build_ascstring ("UTF8"), |
|
db326b8fe982
Use Ben's recently-introduced listu (), where appropriate.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5307
diff
changeset
|
3346 Qunicode_type, Qutf_8, |
|
db326b8fe982
Use Ben's recently-introduced listu (), where appropriate.
Aidan Kehoe <kehoea@parhasard.net>
parents:
5307
diff
changeset
|
3347 Qunbound)); |
|
4834
b3ea9c582280
Use new cygwin_conv_path API with Cygwin 1.7 for converting names between Win32 and POSIX, UTF-8-aware, with attendant changes elsewhere
Ben Wing <ben@xemacs.org>
parents:
4824
diff
changeset
|
3348 } |