Mercurial > hg > xemacs-beta
annotate src/text.h @ 5008:cad59a0a3b19
Add license information from Marcus Thiessel.
See xemacs-beta message <20100208091453.25900@gmx.net>.
| author | Jerry James <james@xemacs.org> |
|---|---|
| date | Tue, 09 Feb 2010 09:50:49 -0700 |
| parents | 3c3c1d139863 |
| children | ae48681c47fa 46cf825f6158 |
| rev | line source |
|---|---|
| 771 | 1 /* Header file for text manipulation primitives and macros. |
| 2 Copyright (C) 1985-1995 Free Software Foundation, Inc. | |
| 3 Copyright (C) 1995 Sun Microsystems, Inc. | |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
4 Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005, 2010 Ben Wing. |
| 771 | 5 |
| 6 This file is part of XEmacs. | |
| 7 | |
| 8 XEmacs is free software; you can redistribute it and/or modify it | |
| 9 under the terms of the GNU General Public License as published by the | |
| 10 Free Software Foundation; either version 2, or (at your option) any | |
| 11 later version. | |
| 12 | |
| 13 XEmacs is distributed in the hope that it will be useful, but WITHOUT | |
| 14 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
| 15 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License | |
| 16 for more details. | |
| 17 | |
| 18 You should have received a copy of the GNU General Public License | |
| 19 along with XEmacs; see the file COPYING. If not, write to | |
| 20 the Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
| 21 Boston, MA 02111-1307, USA. */ | |
| 22 | |
| 23 /* Synched up with: FSF 19.30. */ | |
| 24 | |
| 25 /* Authorship: | |
| 26 | |
| 27 Mostly written by Ben Wing, starting around 1995. | |
| 28 Current TO_IN/EXTERNAL_FORMAT macros written by Martin Buchholz, | |
| 29 designed by Ben Wing based on earlier macros by Ben Wing. | |
| 30 Separated out June 18, 2000 from buffer.h into text.h. | |
| 31 */ | |
| 32 | |
| 33 #ifndef INCLUDED_text_h_ | |
| 34 #define INCLUDED_text_h_ | |
| 35 | |
| 912 | 36 #ifdef HAVE_WCHAR_H |
| 771 | 37 #include <wchar.h> |
| 912 | 38 #else |
| 1257 | 39 size_t wcslen (const wchar_t *); |
| 912 | 40 #endif |
| 1204 | 41 #ifndef HAVE_STRLWR |
| 1257 | 42 char *strlwr (char *); |
| 1204 | 43 #endif |
| 44 #ifndef HAVE_STRUPR | |
| 1257 | 45 char *strupr (char *); |
| 1204 | 46 #endif |
| 771 | 47 |
| 1743 | 48 BEGIN_C_DECLS |
| 1650 | 49 |
| 771 | 50 /* ---------------------------------------------------------------------- */ |
| 51 /* Super-basic character properties */ | |
| 52 /* ---------------------------------------------------------------------- */ | |
| 53 | |
| 54 /* These properties define the specifics of how our current encoding fits | |
| 55 in the basic model used for the encoding. Because this model is the same | |
| 56 as is used for UTF-8, all these properties could be defined for it, too. | |
| 57 This would instantly make the rest of this file work with UTF-8 (with | |
| 58 the exception of a few called functions that would need to be redefined). | |
| 59 | |
| 60 (UTF-2000 implementers, take note!) | |
| 61 */ | |
| 62 | |
| 63 /* If you want more than this, you need to include charset.h */ | |
| 64 | |
| 65 #ifndef MULE | |
| 66 | |
| 826 | 67 #define rep_bytes_by_first_byte(fb) 1 |
| 68 #define byte_ascii_p(byte) 1 | |
| 867 | 69 #define MAX_ICHAR_LEN 1 |
| 771 | 70 |
| 71 #else /* MULE */ | |
| 72 | |
| 73 /* These are carefully designed to work if BYTE is signed or unsigned. */ | |
| 74 /* Note that SPC and DEL are considered ASCII, not control. */ | |
| 75 | |
| 826 | 76 #define byte_ascii_p(byte) (((byte) & ~0x7f) == 0) |
| 77 #define byte_c0_p(byte) (((byte) & ~0x1f) == 0) | |
| 78 #define byte_c1_p(byte) (((byte) & ~0x1f) == 0x80) | |
| 771 | 79 |
| 80 /* Does BYTE represent the first byte of a character? */ | |
| 81 | |
| 826 | 82 #ifdef ERROR_CHECK_TEXT |
| 83 | |
| 84 DECLARE_INLINE_HEADER ( | |
| 85 int | |
| 867 | 86 ibyte_first_byte_p_1 (int byte, const char *file, int line) |
| 826 | 87 ) |
| 88 { | |
| 89 assert_at_line (byte >= 0 && byte < 256, file, line); | |
| 90 return byte < 0xA0; | |
| 91 } | |
| 92 | |
| 867 | 93 #define ibyte_first_byte_p(byte) \ |
| 94 ibyte_first_byte_p_1 (byte, __FILE__, __LINE__) | |
| 826 | 95 |
| 96 #else | |
| 97 | |
| 867 | 98 #define ibyte_first_byte_p(byte) ((byte) < 0xA0) |
| 826 | 99 |
| 100 #endif | |
| 101 | |
| 102 #ifdef ERROR_CHECK_TEXT | |
| 771 | 103 |
| 104 /* Does BYTE represent the first byte of a multi-byte character? */ | |
| 105 | |
| 826 | 106 DECLARE_INLINE_HEADER ( |
| 107 int | |
| 867 | 108 ibyte_leading_byte_p_1 (int byte, const char *file, int line) |
| 826 | 109 ) |
| 110 { | |
| 111 assert_at_line (byte >= 0 && byte < 256, file, line); | |
| 112 return byte_c1_p (byte); | |
| 113 } | |
| 114 | |
| 867 | 115 #define ibyte_leading_byte_p(byte) \ |
| 116 ibyte_leading_byte_p_1 (byte, __FILE__, __LINE__) | |
| 826 | 117 |
| 118 #else | |
| 119 | |
| 867 | 120 #define ibyte_leading_byte_p(byte) byte_c1_p (byte) |
| 826 | 121 |
| 122 #endif | |
| 771 | 123 |
| 124 /* Table of number of bytes in the string representation of a character | |
| 125 indexed by the first byte of that representation. | |
| 126 | |
| 127 This value can be derived in other ways -- e.g. something like | |
| 826 | 128 XCHARSET_REP_BYTES (charset_by_leading_byte (first_byte)) |
| 771 | 129 but it's faster this way. */ |
| 1632 | 130 extern MODULE_API const Bytecount rep_bytes_by_first_byte[0xA0]; |
| 771 | 131 |
| 132 /* Number of bytes in the string representation of a character. */ | |
| 788 | 133 |
| 800 | 134 #ifdef ERROR_CHECK_TEXT |
| 788 | 135 |
| 826 | 136 DECLARE_INLINE_HEADER ( |
| 137 Bytecount | |
| 138 rep_bytes_by_first_byte_1 (int fb, const char *file, int line) | |
| 139 ) | |
| 771 | 140 { |
| 826 | 141 assert_at_line (fb >= 0 && fb < 0xA0, file, line); |
| 771 | 142 return rep_bytes_by_first_byte[fb]; |
| 143 } | |
| 144 | |
| 826 | 145 #define rep_bytes_by_first_byte(fb) \ |
| 146 rep_bytes_by_first_byte_1 (fb, __FILE__, __LINE__) | |
| 788 | 147 |
| 800 | 148 #else /* ERROR_CHECK_TEXT */ |
| 788 | 149 |
| 826 | 150 #define rep_bytes_by_first_byte(fb) (rep_bytes_by_first_byte[fb]) |
| 788 | 151 |
| 800 | 152 #endif /* ERROR_CHECK_TEXT */ |
| 788 | 153 |
| 826 | 154 /* Is this character represented by more than one byte in a string in the |
| 155 default format? */ | |
| 156 | |
| 867 | 157 #define ichar_multibyte_p(c) ((c) >= 0x80) |
| 158 | |
| 159 #define ichar_ascii_p(c) (!ichar_multibyte_p (c)) | |
| 826 | 160 |
| 161 /* Maximum number of bytes per Emacs character when represented as text, in | |
| 162 any format. | |
| 163 */ | |
| 771 | 164 |
| 867 | 165 #define MAX_ICHAR_LEN 4 |
| 771 | 166 |
| 826 | 167 #endif /* not MULE */ |
| 168 | |
| 2367 | 169 /* For more discussion, see text.c, "handling non-default formats" */ |
| 170 | |
| 826 | 171 typedef enum internal_format |
| 172 { | |
| 173 FORMAT_DEFAULT, | |
| 174 FORMAT_8_BIT_FIXED, | |
| 175 FORMAT_16_BIT_FIXED, /* not implemented */ | |
| 176 FORMAT_32_BIT_FIXED /* not implemented */ | |
| 177 } Internal_Format; | |
| 178 | |
| 179 #ifdef MULE | |
| 180 /* "OBJECT" below will usually be a buffer, string, or nil. This needs to | |
| 181 be passed in because the interpretation of 8-bit-fixed and 16-bit-fixed | |
| 182 values may depend on the buffer, e.g. depending on what language the | |
| 183 text in the buffer is in. */ | |
| 184 | |
| 867 | 185 /* True if Ichar CH can be represented in 8-bit-fixed format. */ |
| 186 #define ichar_8_bit_fixed_p(ch, object) (((ch) & ~0xff) == 0) | |
| 187 /* Convert Ichar CH to an 8-bit int, as will be stored in the buffer. */ | |
| 188 #define ichar_to_raw_8_bit_fixed(ch, object) ((Ibyte) (ch)) | |
| 826 | 189 /* Convert the other way. */ |
| 867 | 190 #define raw_8_bit_fixed_to_ichar(ch, object) ((Ichar) (ch)) |
| 191 | |
| 192 #define ichar_16_bit_fixed_p(ch, object) (((ch) & ~0xffff) == 0) | |
| 193 /* Convert Ichar CH to a 16-bit int, as will be stored in the buffer. */ | |
| 194 #define ichar_to_raw_16_bit_fixed(ch, object) ((UINT_16_BIT) (ch)) | |
| 826 | 195 /* Convert the other way. */ |
| 867 | 196 #define raw_16_bit_fixed_to_ichar(ch, object) ((Ichar) (ch)) |
| 197 | |
| 198 /* Convert Ichar CH to a 32-bit int, as will be stored in the buffer. */ | |
| 199 #define ichar_to_raw_32_bit_fixed(ch, object) ((UINT_32_BIT) (ch)) | |
| 826 | 200 /* Convert the other way. */ |
| 867 | 201 #define raw_32_bit_fixed_to_ichar(ch, object) ((Ichar) (ch)) |
| 826 | 202 |
| 203 /* Return the "raw value" of a character as stored in the buffer. In the | |
| 204 default format, this is just the same as the character. In fixed-width | |
| 205 formats, this is the actual value in the buffer, which will be limited | |
| 206 to the range as established by the format. This is used when searching | |
| 207 for a character in a buffer -- it's faster to convert the character to | |
| 208 the raw value and look for that, than repeatedly convert each raw value | |
| 209 in the buffer into a character. */ | |
| 210 | |
| 211 DECLARE_INLINE_HEADER ( | |
| 867 | 212 Raw_Ichar |
| 2286 | 213 ichar_to_raw (Ichar ch, Internal_Format fmt, |
| 214 Lisp_Object UNUSED (object)) | |
| 826 | 215 ) |
| 216 { | |
| 217 switch (fmt) | |
| 218 { | |
| 219 case FORMAT_DEFAULT: | |
| 867 | 220 return (Raw_Ichar) ch; |
| 826 | 221 case FORMAT_16_BIT_FIXED: |
| 867 | 222 text_checking_assert (ichar_16_bit_fixed_p (ch, object)); |
| 223 return (Raw_Ichar) ichar_to_raw_16_bit_fixed (ch, object); | |
| 826 | 224 case FORMAT_32_BIT_FIXED: |
| 867 | 225 return (Raw_Ichar) ichar_to_raw_32_bit_fixed (ch, object); |
| 826 | 226 default: |
| 227 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 867 | 228 text_checking_assert (ichar_8_bit_fixed_p (ch, object)); |
| 229 return (Raw_Ichar) ichar_to_raw_8_bit_fixed (ch, object); | |
| 826 | 230 } |
| 231 } | |
| 232 | |
| 233 /* Return whether CH is representable in the given format in the given | |
| 234 object. */ | |
| 235 | |
| 236 DECLARE_INLINE_HEADER ( | |
| 237 int | |
| 2286 | 238 ichar_fits_in_format (Ichar ch, Internal_Format fmt, |
| 239 Lisp_Object UNUSED (object)) | |
| 826 | 240 ) |
| 241 { | |
| 242 switch (fmt) | |
| 243 { | |
| 244 case FORMAT_DEFAULT: | |
| 245 return 1; | |
| 246 case FORMAT_16_BIT_FIXED: | |
| 867 | 247 return ichar_16_bit_fixed_p (ch, object); |
| 826 | 248 case FORMAT_32_BIT_FIXED: |
| 249 return 1; | |
| 250 default: | |
| 251 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 867 | 252 return ichar_8_bit_fixed_p (ch, object); |
| 826 | 253 } |
| 254 } | |
| 255 | |
| 256 /* Assuming the formats are the same, return whether the two objects | |
| 257 represent text in exactly the same way. */ | |
| 258 | |
| 259 DECLARE_INLINE_HEADER ( | |
| 260 int | |
| 2286 | 261 objects_have_same_internal_representation (Lisp_Object UNUSED (srcobj), |
| 262 Lisp_Object UNUSED (dstobj)) | |
| 826 | 263 ) |
| 264 { | |
| 265 /* &&#### implement this properly when we allow per-object format | |
| 266 differences */ | |
| 267 return 1; | |
| 268 } | |
| 269 | |
| 270 #else | |
| 271 | |
| 867 | 272 #define ichar_to_raw(ch, fmt, object) ((Raw_Ichar) (ch)) |
| 273 #define ichar_fits_in_format(ch, fmt, object) 1 | |
| 826 | 274 #define objects_have_same_internal_representation(srcobj, dstobj) 1 |
| 275 | |
| 771 | 276 #endif /* MULE */ |
| 277 | |
| 1632 | 278 MODULE_API int dfc_coding_system_is_unicode (Lisp_Object codesys); |
| 771 | 279 |
| 280 DECLARE_INLINE_HEADER ( | |
| 281 Bytecount dfc_external_data_len (const void *ptr, Lisp_Object codesys) | |
| 282 ) | |
| 283 { | |
| 284 if (dfc_coding_system_is_unicode (codesys)) | |
| 285 return sizeof (wchar_t) * wcslen ((wchar_t *) ptr); | |
| 286 else | |
| 287 return strlen ((char *) ptr); | |
| 288 } | |
| 289 | |
| 290 | |
| 291 /************************************************************************/ | |
| 292 /* */ | |
| 293 /* working with raw internal-format data */ | |
| 294 /* */ | |
| 295 /************************************************************************/ | |
| 296 | |
| 826 | 297 /* |
| 298 Use the following functions/macros on contiguous text in any of the | |
| 299 internal formats. Those that take a format arg work on all internal | |
| 300 formats; the others work only on the default (variable-width under Mule) | |
| 301 format. If the text you're operating on is known to come from a buffer, | |
| 302 use the buffer-level functions in buffer.h, which automatically know the | |
| 303 correct format and handle the gap. | |
| 304 | |
| 305 Some terminology: | |
| 306 | |
| 867 | 307 "itext" appearing in the macros means "internal-format text" -- type |
| 308 `Ibyte *'. Operations on such pointers themselves, rather than on the | |
| 309 text being pointed to, have "itext" instead of "itext" in the macro | |
| 310 name. "ichar" in the macro names means an Ichar -- the representation | |
| 826 | 311 of a character as a single integer rather than a series of bytes, as part |
| 867 | 312 of "itext". Many of the macros below are for converting between the |
| 826 | 313 two representations of characters. |
| 314 | |
| 867 | 315 Note also that we try to consistently distinguish between an "Ichar" and |
| 826 | 316 a Lisp character. Stuff working with Lisp characters often just says |
| 867 | 317 "char", so we consistently use "Ichar" when that's what we're working |
| 826 | 318 with. */ |
| 319 | |
| 320 /* The three golden rules of macros: | |
| 771 | 321 |
| 322 1) Anything that's an lvalue can be evaluated more than once. | |
| 826 | 323 |
| 324 2) Macros where anything else can be evaluated more than once should | |
| 325 have the word "unsafe" in their name (exceptions may be made for | |
| 326 large sets of macros that evaluate arguments of certain types more | |
| 327 than once, e.g. struct buffer * arguments, when clearly indicated in | |
| 328 the macro documentation). These macros are generally meant to be | |
| 329 called only by other macros that have already stored the calling | |
| 330 values in temporary variables. | |
| 331 | |
| 332 3) Nothing else can be evaluated more than once. Use inline | |
| 771 | 333 functions, if necessary, to prevent multiple evaluation. |
| 826 | 334 |
| 335 NOTE: The functions and macros below are given full prototypes in their | |
| 336 docs, even when the implementation is a macro. In such cases, passing | |
| 337 an argument of a type other than expected will produce undefined | |
| 338 results. Also, given that macros can do things functions can't (in | |
| 339 particular, directly modify arguments as if they were passed by | |
| 340 reference), the declaration syntax has been extended to include the | |
| 341 call-by-reference syntax from C++, where an & after a type indicates | |
| 342 that the argument is an lvalue and is passed by reference, i.e. the | |
| 343 function can modify its value. (This is equivalent in C to passing a | |
| 344 pointer to the argument, but without the need to explicitly worry about | |
| 345 pointers.) | |
| 346 | |
| 347 When to capitalize macros: | |
| 348 | |
| 349 -- Capitalize macros doing stuff obviously impossible with (C) | |
| 350 functions, e.g. directly modifying arguments as if they were passed by | |
| 351 reference. | |
| 352 | |
| 353 -- Capitalize macros that evaluate *any* argument more than once regardless | |
| 354 of whether that's "allowed" (e.g. buffer arguments). | |
| 355 | |
| 356 -- Capitalize macros that directly access a field in a Lisp_Object or | |
| 357 its equivalent underlying structure. In such cases, access through the | |
| 358 Lisp_Object precedes the macro with an X, and access through the underlying | |
| 359 structure doesn't. | |
| 360 | |
| 361 -- Capitalize certain other basic macros relating to Lisp_Objects; e.g. | |
| 362 FRAMEP, CHECK_FRAME, etc. | |
| 363 | |
| 364 -- Try to avoid capitalizing any other macros. | |
| 771 | 365 */ |
| 366 | |
| 367 /* ---------------------------------------------------------------------- */ | |
| 867 | 368 /* Working with itext's (pointers to internally-formatted text) */ |
| 771 | 369 /* ---------------------------------------------------------------------- */ |
| 370 | |
| 867 | 371 /* Given an itext, does it point to the beginning of a character? |
| 826 | 372 */ |
| 373 | |
| 771 | 374 #ifdef MULE |
| 867 | 375 # define valid_ibyteptr_p(ptr) ibyte_first_byte_p (* (ptr)) |
| 771 | 376 #else |
| 867 | 377 # define valid_ibyteptr_p(ptr) 1 |
| 771 | 378 #endif |
| 379 | |
| 867 | 380 /* If error-checking is enabled, assert that the given itext points to |
| 826 | 381 the beginning of a character. Otherwise, do nothing. |
| 382 */ | |
| 383 | |
| 867 | 384 #define assert_valid_ibyteptr(ptr) text_checking_assert (valid_ibyteptr_p (ptr)) |
| 385 | |
| 386 /* Given a itext (assumed to point at the beginning of a character), | |
| 826 | 387 modify that pointer so it points to the beginning of the next character. |
| 388 | |
| 867 | 389 Note that INC_IBYTEPTR() and DEC_IBYTEPTR() have to be written in |
| 390 completely separate ways. INC_IBYTEPTR() cannot use the DEC_IBYTEPTR() | |
| 771 | 391 trick of looking for a valid first byte because it might run off |
| 867 | 392 the end of the string. DEC_IBYTEPTR() can't use the INC_IBYTEPTR() |
| 771 | 393 method because it doesn't have easy access to the first byte of |
| 394 the character it's moving over. */ | |
| 395 | |
| 867 | 396 #define INC_IBYTEPTR(ptr) do { \ |
| 397 assert_valid_ibyteptr (ptr); \ | |
| 826 | 398 (ptr) += rep_bytes_by_first_byte (* (ptr)); \ |
| 399 } while (0) | |
| 400 | |
| 1204 | 401 #define INC_IBYTEPTR_FMT(ptr, fmt) \ |
| 402 do { \ | |
| 403 Internal_Format __icf_fmt = (fmt); \ | |
| 404 switch (__icf_fmt) \ | |
| 405 { \ | |
| 406 case FORMAT_DEFAULT: \ | |
| 407 INC_IBYTEPTR (ptr); \ | |
| 408 break; \ | |
| 409 case FORMAT_16_BIT_FIXED: \ | |
| 410 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); \ | |
| 411 (ptr) += 2; \ | |
| 412 break; \ | |
| 413 case FORMAT_32_BIT_FIXED: \ | |
| 414 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); \ | |
| 415 (ptr) += 4; \ | |
| 416 break; \ | |
| 417 default: \ | |
| 418 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); \ | |
| 419 (ptr)++; \ | |
| 420 break; \ | |
| 421 } \ | |
| 826 | 422 } while (0) |
| 423 | |
| 867 | 424 /* Given a itext (assumed to point at the beginning of a character or at |
| 826 | 425 the very end of the text), modify that pointer so it points to the |
| 426 beginning of the previous character. | |
| 427 */ | |
| 771 | 428 |
| 800 | 429 #ifdef ERROR_CHECK_TEXT |
| 826 | 430 /* We use a separate definition to avoid warnings about unused dc_ptr1 */ |
| 867 | 431 #define DEC_IBYTEPTR(ptr) do { \ |
| 1333 | 432 const Ibyte *dc_ptr1 = (ptr); \ |
| 826 | 433 do { \ |
| 434 (ptr)--; \ | |
| 867 | 435 } while (!valid_ibyteptr_p (ptr)); \ |
| 826 | 436 text_checking_assert (dc_ptr1 - (ptr) == rep_bytes_by_first_byte (*(ptr))); \ |
| 771 | 437 } while (0) |
| 826 | 438 #else |
| 867 | 439 #define DEC_IBYTEPTR(ptr) do { \ |
| 826 | 440 do { \ |
| 441 (ptr)--; \ | |
| 867 | 442 } while (!valid_ibyteptr_p (ptr)); \ |
| 771 | 443 } while (0) |
| 826 | 444 #endif /* ERROR_CHECK_TEXT */ |
| 445 | |
| 1204 | 446 #define DEC_IBYTEPTR_FMT(ptr, fmt) \ |
| 447 do { \ | |
| 448 Internal_Format __icf_fmt = (fmt); \ | |
| 449 switch (__icf_fmt) \ | |
| 450 { \ | |
| 451 case FORMAT_DEFAULT: \ | |
| 452 DEC_IBYTEPTR (ptr); \ | |
| 453 break; \ | |
| 454 case FORMAT_16_BIT_FIXED: \ | |
| 455 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); \ | |
| 456 (ptr) -= 2; \ | |
| 457 break; \ | |
| 458 case FORMAT_32_BIT_FIXED: \ | |
| 459 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); \ | |
| 460 (ptr) -= 4; \ | |
| 461 break; \ | |
| 462 default: \ | |
| 463 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); \ | |
| 464 (ptr)--; \ | |
| 465 break; \ | |
| 466 } \ | |
| 771 | 467 } while (0) |
| 468 | |
| 469 #ifdef MULE | |
| 470 | |
| 826 | 471 /* Make sure that PTR is pointing to the beginning of a character. If not, |
| 472 back up until this is the case. Note that there are not too many places | |
| 473 where it is legitimate to do this sort of thing. It's an error if | |
| 474 you're passed an "invalid" char * pointer. NOTE: PTR *must* be pointing | |
| 475 to a valid part of the string (i.e. not the very end, unless the string | |
| 476 is zero-terminated or something) in order for this function to not cause | |
| 477 crashes. | |
| 478 */ | |
| 479 | |
| 771 | 480 /* Note that this reads the byte at *PTR! */ |
| 481 | |
| 867 | 482 #define VALIDATE_IBYTEPTR_BACKWARD(ptr) do { \ |
| 483 while (!valid_ibyteptr_p (ptr)) ptr--; \ | |
| 771 | 484 } while (0) |
| 485 | |
| 826 | 486 /* Make sure that PTR is pointing to the beginning of a character. If not, |
| 487 move forward until this is the case. Note that there are not too many | |
| 488 places where it is legitimate to do this sort of thing. It's an error | |
| 489 if you're passed an "invalid" char * pointer. | |
| 490 */ | |
| 771 | 491 |
| 867 | 492 /* This needs to be trickier than VALIDATE_IBYTEPTR_BACKWARD() to avoid the |
| 771 | 493 possibility of running off the end of the string. */ |
| 494 | |
| 867 | 495 #define VALIDATE_IBYTEPTR_FORWARD(ptr) do { \ |
| 496 Ibyte *vcf_ptr = (ptr); \ | |
| 497 VALIDATE_IBYTEPTR_BACKWARD (vcf_ptr); \ | |
| 771 | 498 if (vcf_ptr != (ptr)) \ |
| 499 { \ | |
| 500 (ptr) = vcf_ptr; \ | |
| 867 | 501 INC_IBYTEPTR (ptr); \ |
| 771 | 502 } \ |
| 503 } while (0) | |
| 504 | |
| 505 #else /* not MULE */ | |
| 867 | 506 #define VALIDATE_IBYTEPTR_BACKWARD(ptr) |
| 507 #define VALIDATE_IBYTEPTR_FORWARD(ptr) | |
| 826 | 508 #endif /* not MULE */ |
| 509 | |
| 510 #ifdef MULE | |
| 511 | |
| 867 | 512 /* Given a Ibyte string at PTR of size N, possibly with a partial |
| 826 | 513 character at the end, return the size of the longest substring of |
| 514 complete characters. Does not assume that the byte at *(PTR + N) is | |
| 515 readable. Note that there are not too many places where it is | |
| 516 legitimate to do this sort of thing. It's an error if you're passed an | |
| 517 "invalid" offset. */ | |
| 518 | |
| 519 DECLARE_INLINE_HEADER ( | |
| 520 Bytecount | |
| 867 | 521 validate_ibyte_string_backward (const Ibyte *ptr, Bytecount n) |
| 826 | 522 ) |
| 523 { | |
| 867 | 524 const Ibyte *ptr2; |
| 826 | 525 |
| 526 if (n == 0) | |
| 527 return n; | |
| 528 ptr2 = ptr + n - 1; | |
| 867 | 529 VALIDATE_IBYTEPTR_BACKWARD (ptr2); |
| 826 | 530 if (ptr2 + rep_bytes_by_first_byte (*ptr2) != ptr + n) |
| 531 return ptr2 - ptr; | |
| 532 return n; | |
| 533 } | |
| 534 | |
| 535 #else | |
| 536 | |
| 867 | 537 #define validate_ibyte_string_backward(ptr, n) (n) |
| 826 | 538 |
| 539 #endif /* MULE */ | |
| 771 | 540 |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
541 /* ASSERT_ASCTEXT_ASCII(ptr): Check that an Ascbyte * pointer points to |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
542 purely ASCII text. Useful for checking that putatively ASCII strings |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
543 (i.e. declared as Ascbyte * or const Ascbyte *) are actually ASCII. |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
544 This is important because otherwise we need to worry about what |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
545 encoding they are in -- internal or some external encoding. |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
546 |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
547 ASSERT_ASCTEXT_ASCII_LEN(ptr, len): Same as ASSERT_ASCTEXT_ASCII() |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
548 but where the length has been explicitly given. Useful if the string |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
549 may contain embedded zeroes. |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
550 */ |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
551 |
| 2367 | 552 #ifdef ERROR_CHECK_TEXT |
| 553 #define ASSERT_ASCTEXT_ASCII_LEN(ptr, len) \ | |
| 554 do { \ | |
| 555 int aia2; \ | |
| 556 const Ascbyte *aia2ptr = (ptr); \ | |
| 557 int aia2len = (len); \ | |
| 558 \ | |
| 559 for (aia2 = 0; aia2 < aia2len; aia2++) \ | |
| 560 assert (aia2ptr[aia2] >= 0x00 && aia2ptr[aia2] < 0x7F); \ | |
| 561 } while (0) | |
| 562 #define ASSERT_ASCTEXT_ASCII(ptr) \ | |
| 563 do { \ | |
| 564 const Ascbyte *aiaz2 = (ptr); \ | |
| 565 ASSERT_ASCTEXT_ASCII_LEN (aiaz2, strlen (aiaz2)); \ | |
| 566 } while (0) | |
| 567 #else | |
| 568 #define ASSERT_ASCTEXT_ASCII_LEN(ptr, len) | |
| 569 #define ASSERT_ASCTEXT_ASCII(ptr) | |
| 570 #endif | |
| 571 | |
| 771 | 572 /* -------------------------------------------------------------- */ |
| 826 | 573 /* Working with the length (in bytes and characters) of a */ |
| 574 /* section of internally-formatted text */ | |
| 771 | 575 /* -------------------------------------------------------------- */ |
| 576 | |
| 826 | 577 #ifdef MULE |
| 578 | |
| 1632 | 579 MODULE_API Charcount |
| 580 bytecount_to_charcount_fun (const Ibyte *ptr, Bytecount len); | |
| 581 MODULE_API Bytecount | |
| 582 charcount_to_bytecount_fun (const Ibyte *ptr, Charcount len); | |
| 826 | 583 |
| 584 /* Given a pointer to a text string and a length in bytes, return | |
| 585 the equivalent length in characters. */ | |
| 586 | |
| 587 DECLARE_INLINE_HEADER ( | |
| 588 Charcount | |
| 867 | 589 bytecount_to_charcount (const Ibyte *ptr, Bytecount len) |
| 826 | 590 ) |
| 591 { | |
| 592 if (len < 20) /* Just a random guess, but it should be more or less correct. | |
| 593 If number of bytes is small, just do a simple loop, | |
| 594 which should be more efficient. */ | |
| 595 { | |
| 596 Charcount count = 0; | |
| 867 | 597 const Ibyte *end = ptr + len; |
| 826 | 598 while (ptr < end) |
| 599 { | |
| 867 | 600 INC_IBYTEPTR (ptr); |
| 826 | 601 count++; |
| 602 } | |
| 603 /* Bomb out if the specified substring ends in the middle | |
| 604 of a character. Note that we might have already gotten | |
| 605 a core dump above from an invalid reference, but at least | |
| 606 we will get no farther than here. | |
| 607 | |
| 608 This also catches len < 0. */ | |
| 609 text_checking_assert (ptr == end); | |
| 610 | |
| 611 return count; | |
| 612 } | |
| 613 else | |
| 614 return bytecount_to_charcount_fun (ptr, len); | |
| 615 } | |
| 616 | |
| 617 /* Given a pointer to a text string and a length in characters, return the | |
| 618 equivalent length in bytes. | |
| 619 */ | |
| 620 | |
| 621 DECLARE_INLINE_HEADER ( | |
| 622 Bytecount | |
| 867 | 623 charcount_to_bytecount (const Ibyte *ptr, Charcount len) |
| 826 | 624 ) |
| 625 { | |
| 626 text_checking_assert (len >= 0); | |
| 627 if (len < 20) /* See above */ | |
| 628 { | |
| 867 | 629 const Ibyte *newptr = ptr; |
| 826 | 630 while (len > 0) |
| 631 { | |
| 867 | 632 INC_IBYTEPTR (newptr); |
| 826 | 633 len--; |
| 634 } | |
| 635 return newptr - ptr; | |
| 636 } | |
| 637 else | |
| 638 return charcount_to_bytecount_fun (ptr, len); | |
| 639 } | |
| 640 | |
| 2367 | 641 MODULE_API Bytecount |
| 642 charcount_to_bytecount_down_fun (const Ibyte *ptr, Charcount len); | |
| 643 | |
| 644 /* Given a pointer to a text string and a length in bytes, return | |
| 645 the equivalent length in characters of the stretch [PTR - LEN, PTR). */ | |
| 646 | |
| 647 DECLARE_INLINE_HEADER ( | |
| 648 Charcount | |
| 649 bytecount_to_charcount_down (const Ibyte *ptr, Bytecount len) | |
| 650 ) | |
| 651 { | |
| 652 /* No need to be clever here */ | |
| 653 return bytecount_to_charcount (ptr - len, len); | |
| 654 } | |
| 655 | |
| 656 /* Given a pointer to a text string and a length in characters, return the | |
| 657 equivalent length in bytes of the stretch of characters of that length | |
| 658 BEFORE the pointer. | |
| 659 */ | |
| 660 | |
| 661 DECLARE_INLINE_HEADER ( | |
| 662 Bytecount | |
| 663 charcount_to_bytecount_down (const Ibyte *ptr, Charcount len) | |
| 664 ) | |
| 665 { | |
| 666 #define SLEDGEHAMMER_CHECK_TEXT | |
| 667 #ifdef SLEDGEHAMMER_CHECK_TEXT | |
| 668 Charcount len1 = len; | |
| 669 Bytecount ret1, ret2; | |
| 670 | |
| 671 /* To test the correctness of the function version, always do the | |
| 672 calculation both ways and check that the values are the same. */ | |
| 673 text_checking_assert (len >= 0); | |
| 674 { | |
| 675 const Ibyte *newptr = ptr; | |
| 676 while (len1 > 0) | |
| 677 { | |
| 678 DEC_IBYTEPTR (newptr); | |
| 679 len1--; | |
| 680 } | |
| 681 ret1 = ptr - newptr; | |
| 682 } | |
| 683 ret2 = charcount_to_bytecount_down_fun (ptr, len); | |
| 684 text_checking_assert (ret1 == ret2); | |
| 685 return ret1; | |
| 686 #else | |
| 687 text_checking_assert (len >= 0); | |
| 688 if (len < 20) /* See above */ | |
| 689 { | |
| 690 const Ibyte *newptr = ptr; | |
| 691 while (len > 0) | |
| 692 { | |
| 693 DEC_IBYTEPTR (newptr); | |
| 694 len--; | |
| 695 } | |
| 696 return ptr - newptr; | |
| 697 } | |
| 698 else | |
| 699 return charcount_to_bytecount_down_fun (ptr, len); | |
| 700 #endif /* SLEDGEHAMMER_CHECK_TEXT */ | |
| 701 } | |
| 702 | |
| 826 | 703 /* Given a pointer to a text string in the specified format and a length in |
| 704 bytes, return the equivalent length in characters. | |
| 705 */ | |
| 706 | |
| 707 DECLARE_INLINE_HEADER ( | |
| 708 Charcount | |
| 867 | 709 bytecount_to_charcount_fmt (const Ibyte *ptr, Bytecount len, |
| 826 | 710 Internal_Format fmt) |
| 711 ) | |
| 712 { | |
| 713 switch (fmt) | |
| 714 { | |
| 715 case FORMAT_DEFAULT: | |
| 716 return bytecount_to_charcount (ptr, len); | |
| 717 case FORMAT_16_BIT_FIXED: | |
| 1204 | 718 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); |
| 826 | 719 return (Charcount) (len << 1); |
| 720 case FORMAT_32_BIT_FIXED: | |
| 1204 | 721 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); |
| 826 | 722 return (Charcount) (len << 2); |
| 723 default: | |
| 724 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 725 return (Charcount) len; | |
| 726 } | |
| 727 } | |
| 728 | |
| 729 /* Given a pointer to a text string in the specified format and a length in | |
| 730 characters, return the equivalent length in bytes. | |
| 731 */ | |
| 732 | |
| 733 DECLARE_INLINE_HEADER ( | |
| 734 Bytecount | |
| 867 | 735 charcount_to_bytecount_fmt (const Ibyte *ptr, Charcount len, |
| 826 | 736 Internal_Format fmt) |
| 737 ) | |
| 738 { | |
| 739 switch (fmt) | |
| 740 { | |
| 741 case FORMAT_DEFAULT: | |
| 742 return charcount_to_bytecount (ptr, len); | |
| 743 case FORMAT_16_BIT_FIXED: | |
| 1204 | 744 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); |
| 826 | 745 text_checking_assert (!(len & 1)); |
| 746 return (Bytecount) (len >> 1); | |
| 747 case FORMAT_32_BIT_FIXED: | |
| 748 text_checking_assert (!(len & 3)); | |
| 1204 | 749 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); |
| 826 | 750 return (Bytecount) (len >> 2); |
| 751 default: | |
| 752 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 753 return (Bytecount) len; | |
| 754 } | |
| 755 } | |
| 756 | |
| 757 #else | |
| 758 | |
| 759 #define bytecount_to_charcount(ptr, len) ((Charcount) (len)) | |
| 760 #define bytecount_to_charcount_fmt(ptr, len, fmt) ((Charcount) (len)) | |
| 761 #define charcount_to_bytecount(ptr, len) ((Bytecount) (len)) | |
| 762 #define charcount_to_bytecount_fmt(ptr, len, fmt) ((Bytecount) (len)) | |
| 763 | |
| 764 #endif /* MULE */ | |
| 765 | |
| 766 /* Return the length of the first character at PTR. Equivalent to | |
| 767 charcount_to_bytecount (ptr, 1). | |
| 768 | |
| 769 [Since charcount_to_bytecount() is Written as inline, a smart compiler | |
| 770 should really optimize charcount_to_bytecount (ptr, 1) to the same as | |
| 771 the following, with no error checking. But since this idiom occurs so | |
| 772 often, we'll be helpful and define a special macro for it.] | |
| 773 */ | |
| 774 | |
| 867 | 775 #define itext_ichar_len(ptr) rep_bytes_by_first_byte (*(ptr)) |
| 826 | 776 |
| 777 /* Return the length of the first character at PTR, which is in the | |
| 778 specified internal format. Equivalent to charcount_to_bytecount_fmt | |
| 779 (ptr, 1, fmt). | |
| 780 */ | |
| 781 | |
| 782 DECLARE_INLINE_HEADER ( | |
| 783 Bytecount | |
| 4853 | 784 itext_ichar_len_fmt (const Ibyte *ptr, Internal_Format fmt) |
| 826 | 785 ) |
| 786 { | |
| 787 switch (fmt) | |
| 788 { | |
| 789 case FORMAT_DEFAULT: | |
| 867 | 790 return itext_ichar_len (ptr); |
| 826 | 791 case FORMAT_16_BIT_FIXED: |
| 1204 | 792 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); |
| 826 | 793 return 2; |
| 794 case FORMAT_32_BIT_FIXED: | |
| 1204 | 795 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); |
| 826 | 796 return 4; |
| 797 default: | |
| 798 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 799 return 1; | |
| 800 } | |
| 801 } | |
| 802 | |
| 803 /* Return a pointer to the beginning of the character offset N (in | |
| 804 characters) from PTR. | |
| 805 */ | |
| 806 | |
| 807 DECLARE_INLINE_HEADER ( | |
| 867 | 808 const Ibyte * |
| 809 itext_n_addr (const Ibyte *ptr, Charcount offset) | |
| 826 | 810 ) |
| 771 | 811 { |
| 812 return ptr + charcount_to_bytecount (ptr, offset); | |
| 813 } | |
| 814 | |
| 867 | 815 /* Given a itext and an offset into the text pointed to by the itext, |
| 826 | 816 modify the offset so it points to the beginning of the next character. |
| 817 */ | |
| 818 | |
| 819 #define INC_BYTECOUNT(ptr, pos) do { \ | |
| 867 | 820 assert_valid_ibyteptr (ptr); \ |
| 826 | 821 (pos += rep_bytes_by_first_byte (* ((ptr) + (pos)))); \ |
| 822 } while (0) | |
| 823 | |
| 771 | 824 /* -------------------------------------------------------------------- */ |
| 867 | 825 /* Retrieving or changing the character pointed to by a itext */ |
| 771 | 826 /* -------------------------------------------------------------------- */ |
| 827 | |
| 867 | 828 #define simple_itext_ichar(ptr) ((Ichar) (ptr)[0]) |
| 829 #define simple_set_itext_ichar(ptr, x) \ | |
| 830 ((ptr)[0] = (Ibyte) (x), (Bytecount) 1) | |
| 831 #define simple_itext_copy_ichar(src, dst) \ | |
| 814 | 832 ((dst)[0] = *(src), (Bytecount) 1) |
| 771 | 833 |
| 834 #ifdef MULE | |
| 835 | |
| 1632 | 836 MODULE_API Ichar non_ascii_itext_ichar (const Ibyte *ptr); |
| 837 MODULE_API Bytecount non_ascii_set_itext_ichar (Ibyte *ptr, Ichar c); | |
| 838 MODULE_API Bytecount non_ascii_itext_copy_ichar (const Ibyte *src, Ibyte *dst); | |
| 867 | 839 |
| 840 /* Retrieve the character pointed to by PTR as an Ichar. */ | |
| 826 | 841 |
| 842 DECLARE_INLINE_HEADER ( | |
| 867 | 843 Ichar |
| 844 itext_ichar (const Ibyte *ptr) | |
| 826 | 845 ) |
| 771 | 846 { |
| 826 | 847 return byte_ascii_p (*ptr) ? |
| 867 | 848 simple_itext_ichar (ptr) : |
| 849 non_ascii_itext_ichar (ptr); | |
| 771 | 850 } |
| 851 | |
| 826 | 852 /* Retrieve the character pointed to by PTR (a pointer to text in the |
| 853 format FMT, coming from OBJECT [a buffer, string?, or nil]) as an | |
| 867 | 854 Ichar. |
| 826 | 855 |
| 856 Note: For these and other *_fmt() functions, if you pass in a constant | |
| 857 FMT, the switch will be optimized out of existence. Therefore, there is | |
| 858 no need to create separate versions for the various formats for | |
| 867 | 859 "efficiency reasons". In fact, we don't really need itext_ichar() |
| 826 | 860 and such written separately, but they are used often so it's simpler |
| 861 that way. */ | |
| 862 | |
| 863 DECLARE_INLINE_HEADER ( | |
| 867 | 864 Ichar |
| 865 itext_ichar_fmt (const Ibyte *ptr, Internal_Format fmt, | |
| 2286 | 866 Lisp_Object UNUSED (object)) |
| 826 | 867 ) |
| 868 { | |
| 869 switch (fmt) | |
| 870 { | |
| 871 case FORMAT_DEFAULT: | |
| 867 | 872 return itext_ichar (ptr); |
| 826 | 873 case FORMAT_16_BIT_FIXED: |
| 1204 | 874 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); |
| 867 | 875 return raw_16_bit_fixed_to_ichar (* (UINT_16_BIT *) ptr, object); |
| 826 | 876 case FORMAT_32_BIT_FIXED: |
| 1204 | 877 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); |
| 867 | 878 return raw_32_bit_fixed_to_ichar (* (UINT_32_BIT *) ptr, object); |
| 826 | 879 default: |
| 880 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 867 | 881 return raw_8_bit_fixed_to_ichar (*ptr, object); |
| 826 | 882 } |
| 883 } | |
| 884 | |
| 885 /* Return the character at PTR (which is in format FMT), suitable for | |
| 886 comparison with an ASCII character. This guarantees that if the | |
| 887 character at PTR is ASCII (range 0 - 127), that character will be | |
| 888 returned; otherwise, some character outside of the ASCII range will be | |
| 889 returned, but not necessarily the character actually at PTR. This will | |
| 867 | 890 be faster than itext_ichar_fmt() for some formats -- in particular, |
| 826 | 891 FORMAT_DEFAULT. */ |
| 892 | |
| 893 DECLARE_INLINE_HEADER ( | |
| 867 | 894 Ichar |
| 895 itext_ichar_ascii_fmt (const Ibyte *ptr, Internal_Format fmt, | |
| 2286 | 896 Lisp_Object UNUSED (object)) |
| 826 | 897 ) |
| 898 { | |
| 899 switch (fmt) | |
| 900 { | |
| 901 case FORMAT_DEFAULT: | |
| 867 | 902 return (Ichar) *ptr; |
| 826 | 903 case FORMAT_16_BIT_FIXED: |
| 1204 | 904 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); |
| 867 | 905 return raw_16_bit_fixed_to_ichar (* (UINT_16_BIT *) ptr, object); |
| 826 | 906 case FORMAT_32_BIT_FIXED: |
| 1204 | 907 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); |
| 867 | 908 return raw_32_bit_fixed_to_ichar (* (UINT_32_BIT *) ptr, object); |
| 826 | 909 default: |
| 910 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 867 | 911 return raw_8_bit_fixed_to_ichar (*ptr, object); |
| 826 | 912 } |
| 913 } | |
| 914 | |
| 915 /* Return the "raw value" of the character at PTR, in format FMT. This is | |
| 916 useful when searching for a character; convert the character using | |
| 867 | 917 ichar_to_raw(). */ |
| 826 | 918 |
| 919 DECLARE_INLINE_HEADER ( | |
| 867 | 920 Raw_Ichar |
| 921 itext_ichar_raw_fmt (const Ibyte *ptr, Internal_Format fmt) | |
| 826 | 922 ) |
| 923 { | |
| 924 switch (fmt) | |
| 925 { | |
| 926 case FORMAT_DEFAULT: | |
| 867 | 927 return (Raw_Ichar) itext_ichar (ptr); |
| 826 | 928 case FORMAT_16_BIT_FIXED: |
| 1204 | 929 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); |
| 867 | 930 return (Raw_Ichar) (* (UINT_16_BIT *) ptr); |
| 826 | 931 case FORMAT_32_BIT_FIXED: |
| 1204 | 932 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); |
| 867 | 933 return (Raw_Ichar) (* (UINT_32_BIT *) ptr); |
| 826 | 934 default: |
| 935 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 867 | 936 return (Raw_Ichar) (*ptr); |
| 826 | 937 } |
| 938 } | |
| 939 | |
| 867 | 940 /* Store the character CH (an Ichar) as internally-formatted text starting |
| 826 | 941 at PTR. Return the number of bytes stored. |
| 942 */ | |
| 943 | |
| 944 DECLARE_INLINE_HEADER ( | |
| 945 Bytecount | |
| 867 | 946 set_itext_ichar (Ibyte *ptr, Ichar x) |
| 826 | 947 ) |
| 771 | 948 { |
| 867 | 949 return !ichar_multibyte_p (x) ? |
| 950 simple_set_itext_ichar (ptr, x) : | |
| 951 non_ascii_set_itext_ichar (ptr, x); | |
| 771 | 952 } |
| 953 | |
| 867 | 954 /* Store the character CH (an Ichar) as internally-formatted text of |
| 826 | 955 format FMT starting at PTR, which comes from OBJECT. Return the number |
| 956 of bytes stored. | |
| 957 */ | |
| 958 | |
| 959 DECLARE_INLINE_HEADER ( | |
| 960 Bytecount | |
| 867 | 961 set_itext_ichar_fmt (Ibyte *ptr, Ichar x, Internal_Format fmt, |
| 2286 | 962 Lisp_Object UNUSED (object)) |
| 826 | 963 ) |
| 771 | 964 { |
| 826 | 965 switch (fmt) |
| 966 { | |
| 967 case FORMAT_DEFAULT: | |
| 867 | 968 return set_itext_ichar (ptr, x); |
| 826 | 969 case FORMAT_16_BIT_FIXED: |
| 867 | 970 text_checking_assert (ichar_16_bit_fixed_p (x, object)); |
| 1204 | 971 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); |
| 867 | 972 * (UINT_16_BIT *) ptr = ichar_to_raw_16_bit_fixed (x, object); |
| 826 | 973 return 2; |
| 974 case FORMAT_32_BIT_FIXED: | |
| 1204 | 975 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); |
| 867 | 976 * (UINT_32_BIT *) ptr = ichar_to_raw_32_bit_fixed (x, object); |
| 826 | 977 return 4; |
| 978 default: | |
| 979 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); | |
| 867 | 980 text_checking_assert (ichar_8_bit_fixed_p (x, object)); |
| 981 *ptr = ichar_to_raw_8_bit_fixed (x, object); | |
| 826 | 982 return 1; |
| 983 } | |
| 984 } | |
| 985 | |
| 986 /* Retrieve the character pointed to by SRC and store it as | |
| 987 internally-formatted text in DST. | |
| 988 */ | |
| 989 | |
| 990 DECLARE_INLINE_HEADER ( | |
| 991 Bytecount | |
| 867 | 992 itext_copy_ichar (const Ibyte *src, Ibyte *dst) |
| 826 | 993 ) |
| 994 { | |
| 995 return byte_ascii_p (*src) ? | |
| 867 | 996 simple_itext_copy_ichar (src, dst) : |
| 997 non_ascii_itext_copy_ichar (src, dst); | |
| 771 | 998 } |
| 999 | |
| 1000 #else /* not MULE */ | |
| 1001 | |
| 867 | 1002 # define itext_ichar(ptr) simple_itext_ichar (ptr) |
| 1003 # define itext_ichar_fmt(ptr, fmt, object) itext_ichar (ptr) | |
| 1004 # define itext_ichar_ascii_fmt(ptr, fmt, object) itext_ichar (ptr) | |
| 1005 # define itext_ichar_raw_fmt(ptr, fmt) itext_ichar (ptr) | |
| 1006 # define set_itext_ichar(ptr, x) simple_set_itext_ichar (ptr, x) | |
| 1007 # define set_itext_ichar_fmt(ptr, x, fmt, obj) set_itext_ichar (ptr, x) | |
| 1008 # define itext_copy_ichar(src, dst) simple_itext_copy_ichar (src, dst) | |
| 771 | 1009 |
| 1010 #endif /* not MULE */ | |
| 1011 | |
| 826 | 1012 /* Retrieve the character at offset N (in characters) from PTR, as an |
| 867 | 1013 Ichar. |
| 826 | 1014 */ |
| 1015 | |
| 867 | 1016 #define itext_ichar_n(ptr, offset) \ |
| 1017 itext_ichar (itext_n_addr (ptr, offset)) | |
| 771 | 1018 |
| 1019 | |
| 1020 /************************************************************************/ | |
| 1021 /* */ | |
| 826 | 1022 /* working with Lisp strings */ |
| 1023 /* */ | |
| 1024 /************************************************************************/ | |
| 1025 | |
| 1026 #define string_char_length(s) \ | |
| 1027 string_index_byte_to_char (s, XSTRING_LENGTH (s)) | |
| 1028 #define string_byte(s, i) (XSTRING_DATA (s)[i] + 0) | |
| 1029 /* In case we ever allow strings to be in a different format ... */ | |
| 1030 #define set_string_byte(s, i, c) (XSTRING_DATA (s)[i] = (c)) | |
| 1031 | |
| 1032 #define ASSERT_VALID_CHAR_STRING_INDEX_UNSAFE(s, x) do { \ | |
| 1033 text_checking_assert ((x) >= 0 && x <= string_char_length (s)); \ | |
| 1034 } while (0) | |
| 1035 | |
| 1036 #define ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE(s, x) do { \ | |
| 1037 text_checking_assert ((x) >= 0 && x <= XSTRING_LENGTH (s)); \ | |
| 867 | 1038 text_checking_assert (valid_ibyteptr_p (string_byte_addr (s, x))); \ |
| 826 | 1039 } while (0) |
| 1040 | |
| 1041 /* Convert offset I in string S to a pointer to text there. */ | |
| 1042 #define string_byte_addr(s, i) (&(XSTRING_DATA (s)[i])) | |
| 1043 /* Convert pointer to text in string S into the byte offset to that text. */ | |
| 1044 #define string_addr_to_byte(s, ptr) ((Bytecount) ((ptr) - XSTRING_DATA (s))) | |
| 867 | 1045 /* Return the Ichar at *CHARACTER* offset I. */ |
| 1046 #define string_ichar(s, i) itext_ichar (string_char_addr (s, i)) | |
| 826 | 1047 |
| 1048 #ifdef ERROR_CHECK_TEXT | |
| 1049 #define SLEDGEHAMMER_CHECK_ASCII_BEGIN | |
| 1050 #endif | |
| 1051 | |
| 1052 #ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN | |
| 1053 void sledgehammer_check_ascii_begin (Lisp_Object str); | |
| 1054 #else | |
| 1055 #define sledgehammer_check_ascii_begin(str) | |
| 1056 #endif | |
| 1057 | |
| 1058 /* Make an alloca'd copy of a Lisp string */ | |
| 1059 #define LISP_STRING_TO_ALLOCA(s, lval) \ | |
| 1060 do { \ | |
| 1315 | 1061 Ibyte **_lta_ = (Ibyte **) &(lval); \ |
| 826 | 1062 Lisp_Object _lta_2 = (s); \ |
| 2367 | 1063 *_lta_ = alloca_ibytes (1 + XSTRING_LENGTH (_lta_2)); \ |
| 826 | 1064 memcpy (*_lta_, XSTRING_DATA (_lta_2), 1 + XSTRING_LENGTH (_lta_2)); \ |
| 1065 } while (0) | |
| 1066 | |
| 1067 void resize_string (Lisp_Object s, Bytecount pos, Bytecount delta); | |
| 1068 | |
| 1069 /* Convert a byte index into a string into a char index. */ | |
| 1070 DECLARE_INLINE_HEADER ( | |
| 1071 Charcount | |
| 4853 | 1072 string_index_byte_to_char (Lisp_Object s, Bytecount idx) |
| 826 | 1073 ) |
| 1074 { | |
| 1075 Charcount retval; | |
| 1076 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, idx); | |
| 1077 #ifdef MULE | |
| 1078 if (idx <= (Bytecount) XSTRING_ASCII_BEGIN (s)) | |
| 1079 retval = (Charcount) idx; | |
| 1080 else | |
| 1081 retval = (XSTRING_ASCII_BEGIN (s) + | |
| 1082 bytecount_to_charcount (XSTRING_DATA (s) + | |
| 1083 XSTRING_ASCII_BEGIN (s), | |
| 1084 idx - XSTRING_ASCII_BEGIN (s))); | |
| 1085 # ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN | |
| 1086 assert (retval == bytecount_to_charcount (XSTRING_DATA (s), idx)); | |
| 1087 # endif | |
| 1088 #else | |
| 1089 retval = (Charcount) idx; | |
| 1090 #endif | |
| 1091 /* Don't call ASSERT_VALID_CHAR_STRING_INDEX_UNSAFE() here because it will | |
| 1092 call string_index_byte_to_char(). */ | |
| 1093 return retval; | |
| 1094 } | |
| 1095 | |
| 1096 /* Convert a char index into a string into a byte index. */ | |
| 1097 DECLARE_INLINE_HEADER ( | |
| 1098 Bytecount | |
| 4853 | 1099 string_index_char_to_byte (Lisp_Object s, Charcount idx) |
| 826 | 1100 ) |
| 1101 { | |
| 1102 Bytecount retval; | |
| 1103 ASSERT_VALID_CHAR_STRING_INDEX_UNSAFE (s, idx); | |
| 1104 #ifdef MULE | |
| 1105 if (idx <= (Charcount) XSTRING_ASCII_BEGIN (s)) | |
| 1106 retval = (Bytecount) idx; | |
| 1107 else | |
| 1108 retval = (XSTRING_ASCII_BEGIN (s) + | |
| 1109 charcount_to_bytecount (XSTRING_DATA (s) + | |
| 1110 XSTRING_ASCII_BEGIN (s), | |
| 1111 idx - XSTRING_ASCII_BEGIN (s))); | |
| 1112 # ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN | |
| 1113 assert (retval == charcount_to_bytecount (XSTRING_DATA (s), idx)); | |
| 1114 # endif | |
| 1115 #else | |
| 1116 retval = (Bytecount) idx; | |
| 1117 #endif | |
| 1118 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, retval); | |
| 1119 return retval; | |
| 1120 } | |
| 1121 | |
| 1122 /* Convert a substring length (starting at byte offset OFF) from bytes to | |
| 1123 chars. */ | |
| 1124 DECLARE_INLINE_HEADER ( | |
| 1125 Charcount | |
| 4853 | 1126 string_offset_byte_to_char_len (Lisp_Object s, Bytecount off, Bytecount len) |
| 826 | 1127 ) |
| 1128 { | |
| 1129 Charcount retval; | |
| 1130 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, off); | |
| 1131 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, off + len); | |
| 1132 #ifdef MULE | |
| 1133 if (off + len <= (Bytecount) XSTRING_ASCII_BEGIN (s)) | |
| 1134 retval = (Charcount) len; | |
| 1135 else if (off < (Bytecount) XSTRING_ASCII_BEGIN (s)) | |
| 1136 retval = | |
| 1137 XSTRING_ASCII_BEGIN (s) - (Charcount) off + | |
| 1138 bytecount_to_charcount (XSTRING_DATA (s) + XSTRING_ASCII_BEGIN (s), | |
| 1139 len - (XSTRING_ASCII_BEGIN (s) - off)); | |
| 1140 else | |
| 1141 retval = bytecount_to_charcount (XSTRING_DATA (s) + off, len); | |
| 1142 # ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN | |
| 1143 assert (retval == bytecount_to_charcount (XSTRING_DATA (s) + off, len)); | |
| 1144 # endif | |
| 1145 #else | |
| 1146 retval = (Charcount) len; | |
| 1147 #endif | |
| 1148 return retval; | |
| 1149 } | |
| 1150 | |
| 1151 /* Convert a substring length (starting at byte offset OFF) from chars to | |
| 1152 bytes. */ | |
| 1153 DECLARE_INLINE_HEADER ( | |
| 1154 Bytecount | |
| 4853 | 1155 string_offset_char_to_byte_len (Lisp_Object s, Bytecount off, Charcount len) |
| 826 | 1156 ) |
| 1157 { | |
| 1158 Bytecount retval; | |
| 1159 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, off); | |
| 1160 #ifdef MULE | |
| 1161 /* casts to avoid errors from combining Bytecount/Charcount and warnings | |
| 1162 from signed/unsigned comparisons */ | |
| 1163 if (off + (Bytecount) len <= (Bytecount) XSTRING_ASCII_BEGIN (s)) | |
| 1164 retval = (Bytecount) len; | |
| 1165 else if (off < (Bytecount) XSTRING_ASCII_BEGIN (s)) | |
| 1166 retval = | |
| 1167 XSTRING_ASCII_BEGIN (s) - off + | |
| 1168 charcount_to_bytecount (XSTRING_DATA (s) + XSTRING_ASCII_BEGIN (s), | |
| 1169 len - (XSTRING_ASCII_BEGIN (s) - | |
| 1170 (Charcount) off)); | |
| 1171 else | |
| 1172 retval = charcount_to_bytecount (XSTRING_DATA (s) + off, len); | |
| 1173 # ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN | |
| 1174 assert (retval == charcount_to_bytecount (XSTRING_DATA (s) + off, len)); | |
| 1175 # endif | |
| 1176 #else | |
| 1177 retval = (Bytecount) len; | |
| 1178 #endif | |
| 1179 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, off + retval); | |
| 1180 return retval; | |
| 1181 } | |
| 1182 | |
| 1183 DECLARE_INLINE_HEADER ( | |
| 867 | 1184 const Ibyte * |
| 826 | 1185 string_char_addr (Lisp_Object s, Charcount idx) |
| 1186 ) | |
| 1187 { | |
| 1188 return XSTRING_DATA (s) + string_index_char_to_byte (s, idx); | |
| 1189 } | |
| 1190 | |
| 1191 /* WARNING: If you modify an existing string, you must call | |
| 1192 bump_string_modiff() afterwards. */ | |
| 1193 #ifdef MULE | |
| 867 | 1194 void set_string_char (Lisp_Object s, Charcount i, Ichar c); |
| 826 | 1195 #else |
| 1196 #define set_string_char(s, i, c) set_string_byte (s, i, c) | |
| 1197 #endif /* not MULE */ | |
| 1198 | |
| 1199 /* Return index to character before the one at IDX. */ | |
| 1200 DECLARE_INLINE_HEADER ( | |
| 1201 Bytecount | |
| 1202 prev_string_index (Lisp_Object s, Bytecount idx) | |
| 1203 ) | |
| 1204 { | |
| 867 | 1205 const Ibyte *ptr = string_byte_addr (s, idx); |
| 1206 DEC_IBYTEPTR (ptr); | |
| 826 | 1207 return string_addr_to_byte (s, ptr); |
| 1208 } | |
| 1209 | |
| 1210 /* Return index to character after the one at IDX. */ | |
| 1211 DECLARE_INLINE_HEADER ( | |
| 1212 Bytecount | |
| 1213 next_string_index (Lisp_Object s, Bytecount idx) | |
| 1214 ) | |
| 1215 { | |
| 867 | 1216 const Ibyte *ptr = string_byte_addr (s, idx); |
| 1217 INC_IBYTEPTR (ptr); | |
| 826 | 1218 return string_addr_to_byte (s, ptr); |
| 1219 } | |
| 1220 | |
| 1221 | |
| 1222 /************************************************************************/ | |
| 1223 /* */ | |
| 771 | 1224 /* working with Eistrings */ |
| 1225 /* */ | |
| 1226 /************************************************************************/ | |
| 1227 | |
| 1228 /* | |
| 1229 #### NOTE: This is a work in progress. Neither the API nor especially | |
| 1230 the implementation is finished. | |
| 1231 | |
| 1232 NOTE: An Eistring is a structure that makes it easy to work with | |
| 1233 internally-formatted strings of data. It provides operations similar | |
| 1234 in feel to the standard strcpy(), strcat(), strlen(), etc., but | |
| 1235 | |
| 1236 (a) it is Mule-correct | |
| 1237 (b) it does dynamic allocation so you never have to worry about size | |
| 793 | 1238 restrictions |
| 851 | 1239 (c) it comes in an ALLOCA() variety (all allocation is stack-local, |
| 793 | 1240 so there is no need to explicitly clean up) as well as a malloc() |
| 1241 variety | |
| 1242 (d) it knows its own length, so it does not suffer from standard null | |
| 1243 byte brain-damage -- but it null-terminates the data anyway, so | |
| 1244 it can be passed to standard routines | |
| 1245 (e) it provides a much more powerful set of operations and knows about | |
| 771 | 1246 all the standard places where string data might reside: Lisp_Objects, |
| 867 | 1247 other Eistrings, Ibyte * data with or without an explicit length, |
| 1248 ASCII strings, Ichars, etc. | |
| 793 | 1249 (f) it provides easy operations to convert to/from externally-formatted |
| 1250 data, and is easier to use than the standard TO_INTERNAL_FORMAT | |
| 771 | 1251 and TO_EXTERNAL_FORMAT macros. (An Eistring can store both the internal |
| 1252 and external version of its data, but the external version is only | |
| 1253 initialized or changed when you call eito_external().) | |
| 1254 | |
| 793 | 1255 The idea is to make it as easy to write Mule-correct string manipulation |
| 1256 code as it is to write normal string manipulation code. We also make | |
| 1257 the API sufficiently general that it can handle multiple internal data | |
| 1258 formats (e.g. some fixed-width optimizing formats and a default variable | |
| 1259 width format) and allows for *ANY* data format we might choose in the | |
| 1260 future for the default format, including UCS2. (In other words, we can't | |
| 1261 assume that the internal format is ASCII-compatible and we can't assume | |
| 1262 it doesn't have embedded null bytes. We do assume, however, that any | |
| 1263 chosen format will have the concept of null-termination.) All of this is | |
| 1264 hidden from the user. | |
| 771 | 1265 |
| 1266 #### It is really too bad that we don't have a real object-oriented | |
| 1267 language, or at least a language with polymorphism! | |
| 1268 | |
| 1269 | |
| 1270 ********************************************** | |
| 1271 * Declaration * | |
| 1272 ********************************************** | |
| 1273 | |
| 1274 To declare an Eistring, either put one of the following in the local | |
| 1275 variable section: | |
| 1276 | |
| 1277 DECLARE_EISTRING (name); | |
| 2367 | 1278 Declare a new Eistring and initialize it to the empy string. This |
| 1279 is a standard local variable declaration and can go anywhere in the | |
| 1280 variable declaration section. NAME itself is declared as an | |
| 1281 Eistring *, and its storage declared on the stack. | |
| 771 | 1282 |
| 1283 DECLARE_EISTRING_MALLOC (name); | |
| 2367 | 1284 Declare and initialize a new Eistring, which uses malloc()ed |
| 1285 instead of ALLOCA()ed data. This is a standard local variable | |
| 1286 declaration and can go anywhere in the variable declaration | |
| 1287 section. Once you initialize the Eistring, you will have to free | |
| 1288 it using eifree() to avoid memory leaks. You will need to use this | |
| 1289 form if you are passing an Eistring to any function that modifies | |
| 1290 it (otherwise, the modified data may be in stack space and get | |
| 1291 overwritten when the function returns). | |
| 771 | 1292 |
| 1293 or use | |
| 1294 | |
| 793 | 1295 Eistring ei; |
| 1296 void eiinit (Eistring *ei); | |
| 1297 void eiinit_malloc (Eistring *einame); | |
| 771 | 1298 If you need to put an Eistring elsewhere than in a local variable |
| 1299 declaration (e.g. in a structure), declare it as shown and then | |
| 1300 call one of the init macros. | |
| 1301 | |
| 1302 Also note: | |
| 1303 | |
| 793 | 1304 void eifree (Eistring *ei); |
| 771 | 1305 If you declared an Eistring to use malloc() to hold its data, |
| 1306 or converted it to the heap using eito_malloc(), then this | |
| 1307 releases any data in it and afterwards resets the Eistring | |
| 1308 using eiinit_malloc(). Otherwise, it just resets the Eistring | |
| 1309 using eiinit(). | |
| 1310 | |
| 1311 | |
| 1312 ********************************************** | |
| 1313 * Conventions * | |
| 1314 ********************************************** | |
| 1315 | |
| 1316 - The names of the functions have been chosen, where possible, to | |
| 1317 match the names of str*() functions in the standard C API. | |
| 1318 - | |
| 1319 | |
| 1320 | |
| 1321 ********************************************** | |
| 1322 * Initialization * | |
| 1323 ********************************************** | |
| 1324 | |
| 1325 void eireset (Eistring *eistr); | |
| 1326 Initialize the Eistring to the empty string. | |
| 1327 | |
| 1328 void eicpy_* (Eistring *eistr, ...); | |
| 1329 Initialize the Eistring from somewhere: | |
| 1330 | |
| 1331 void eicpy_ei (Eistring *eistr, Eistring *eistr2); | |
| 1332 ... from another Eistring. | |
| 1333 void eicpy_lstr (Eistring *eistr, Lisp_Object lisp_string); | |
| 1334 ... from a Lisp_Object string. | |
| 867 | 1335 void eicpy_ch (Eistring *eistr, Ichar ch); |
| 1336 ... from an Ichar (this can be a conventional C character). | |
| 771 | 1337 |
| 1338 void eicpy_lstr_off (Eistring *eistr, Lisp_Object lisp_string, | |
| 1339 Bytecount off, Charcount charoff, | |
| 1340 Bytecount len, Charcount charlen); | |
| 1341 ... from a section of a Lisp_Object string. | |
| 1342 void eicpy_lbuf (Eistring *eistr, Lisp_Object lisp_buf, | |
| 1343 Bytecount off, Charcount charoff, | |
| 1344 Bytecount len, Charcount charlen); | |
| 1345 ... from a section of a Lisp_Object buffer. | |
| 867 | 1346 void eicpy_raw (Eistring *eistr, const Ibyte *data, Bytecount len); |
| 771 | 1347 ... from raw internal-format data in the default internal format. |
| 867 | 1348 void eicpy_rawz (Eistring *eistr, const Ibyte *data); |
| 771 | 1349 ... from raw internal-format data in the default internal format |
| 1350 that is "null-terminated" (the meaning of this depends on the nature | |
| 1351 of the default internal format). | |
| 867 | 1352 void eicpy_raw_fmt (Eistring *eistr, const Ibyte *data, Bytecount len, |
| 826 | 1353 Internal_Format intfmt, Lisp_Object object); |
| 771 | 1354 ... from raw internal-format data in the specified format. |
| 867 | 1355 void eicpy_rawz_fmt (Eistring *eistr, const Ibyte *data, |
| 826 | 1356 Internal_Format intfmt, Lisp_Object object); |
| 771 | 1357 ... from raw internal-format data in the specified format that is |
| 1358 "null-terminated" (the meaning of this depends on the nature of | |
| 1359 the specific format). | |
| 2421 | 1360 void eicpy_ascii (Eistring *eistr, const Ascbyte *ascstr); |
| 771 | 1361 ... from an ASCII null-terminated string. Non-ASCII characters in |
| 2500 | 1362 the string are *ILLEGAL* (read ABORT() with error-checking defined). |
| 2421 | 1363 void eicpy_ascii_len (Eistring *eistr, const Ascbyte *ascstr, len); |
| 771 | 1364 ... from an ASCII string, with length specified. Non-ASCII characters |
| 2500 | 1365 in the string are *ILLEGAL* (read ABORT() with error-checking defined). |
| 771 | 1366 void eicpy_ext (Eistring *eistr, const Extbyte *extdata, |
| 1318 | 1367 Lisp_Object codesys); |
| 771 | 1368 ... from external null-terminated data, with coding system specified. |
| 1369 void eicpy_ext_len (Eistring *eistr, const Extbyte *extdata, | |
| 1318 | 1370 Bytecount extlen, Lisp_Object codesys); |
| 771 | 1371 ... from external data, with length and coding system specified. |
| 1372 void eicpy_lstream (Eistring *eistr, Lisp_Object lstream); | |
| 1373 ... from an lstream; reads data till eof. Data must be in default | |
| 1374 internal format; otherwise, interpose a decoding lstream. | |
| 1375 | |
| 1376 | |
| 1377 ********************************************** | |
| 1378 * Getting the data out of the Eistring * | |
| 1379 ********************************************** | |
| 1380 | |
| 867 | 1381 Ibyte *eidata (Eistring *eistr); |
| 771 | 1382 Return a pointer to the raw data in an Eistring. This is NOT |
| 1383 a copy. | |
| 1384 | |
| 1385 Lisp_Object eimake_string (Eistring *eistr); | |
| 1386 Make a Lisp string out of the Eistring. | |
| 1387 | |
| 1388 Lisp_Object eimake_string_off (Eistring *eistr, | |
| 1389 Bytecount off, Charcount charoff, | |
| 1390 Bytecount len, Charcount charlen); | |
| 1391 Make a Lisp string out of a section of the Eistring. | |
| 1392 | |
| 867 | 1393 void eicpyout_alloca (Eistring *eistr, LVALUE: Ibyte *ptr_out, |
| 771 | 1394 LVALUE: Bytecount len_out); |
| 851 | 1395 Make an ALLOCA() copy of the data in the Eistring, using the |
| 1396 default internal format. Due to the nature of ALLOCA(), this | |
| 771 | 1397 must be a macro, with all lvalues passed in as parameters. |
| 793 | 1398 (More specifically, not all compilers correctly handle using |
| 851 | 1399 ALLOCA() as the argument to a function call -- GCC on x86 |
| 1400 didn't used to, for example.) A pointer to the ALLOCA()ed data | |
| 793 | 1401 is stored in PTR_OUT, and the length of the data (not including |
| 1402 the terminating zero) is stored in LEN_OUT. | |
| 771 | 1403 |
| 867 | 1404 void eicpyout_alloca_fmt (Eistring *eistr, LVALUE: Ibyte *ptr_out, |
| 771 | 1405 LVALUE: Bytecount len_out, |
| 826 | 1406 Internal_Format intfmt, Lisp_Object object); |
| 771 | 1407 Like eicpyout_alloca(), but converts to the specified internal |
| 1408 format. (No formats other than FORMAT_DEFAULT are currently | |
| 1409 implemented, and you get an assertion failure if you try.) | |
| 1410 | |
| 867 | 1411 Ibyte *eicpyout_malloc (Eistring *eistr, Bytecount *intlen_out); |
| 771 | 1412 Make a malloc() copy of the data in the Eistring, using the |
| 1413 default internal format. This is a real function. No lvalues | |
| 1414 passed in. Returns the new data, and stores the length (not | |
| 1415 including the terminating zero) using INTLEN_OUT, unless it's | |
| 1416 a NULL pointer. | |
| 1417 | |
| 867 | 1418 Ibyte *eicpyout_malloc_fmt (Eistring *eistr, Internal_Format intfmt, |
| 826 | 1419 Bytecount *intlen_out, Lisp_Object object); |
| 771 | 1420 Like eicpyout_malloc(), but converts to the specified internal |
| 1421 format. (No formats other than FORMAT_DEFAULT are currently | |
| 1422 implemented, and you get an assertion failure if you try.) | |
| 1423 | |
| 1424 | |
| 1425 ********************************************** | |
| 1426 * Moving to the heap * | |
| 1427 ********************************************** | |
| 1428 | |
| 1429 void eito_malloc (Eistring *eistr); | |
| 1430 Move this Eistring to the heap. Its data will be stored in a | |
| 1431 malloc()ed block rather than the stack. Subsequent changes to | |
| 1432 this Eistring will realloc() the block as necessary. Use this | |
| 1433 when you want the Eistring to remain in scope past the end of | |
| 1434 this function call. You will have to manually free the data | |
| 1435 in the Eistring using eifree(). | |
| 1436 | |
| 1437 void eito_alloca (Eistring *eistr); | |
| 1438 Move this Eistring back to the stack, if it was moved to the | |
| 1439 heap with eito_malloc(). This will automatically free any | |
| 1440 heap-allocated data. | |
| 1441 | |
| 1442 | |
| 1443 | |
| 1444 ********************************************** | |
| 1445 * Retrieving the length * | |
| 1446 ********************************************** | |
| 1447 | |
| 1448 Bytecount eilen (Eistring *eistr); | |
| 1449 Return the length of the internal data, in bytes. See also | |
| 1450 eiextlen(), below. | |
| 1451 Charcount eicharlen (Eistring *eistr); | |
| 1452 Return the length of the internal data, in characters. | |
| 1453 | |
| 1454 | |
| 1455 ********************************************** | |
| 1456 * Working with positions * | |
| 1457 ********************************************** | |
| 1458 | |
| 1459 Bytecount eicharpos_to_bytepos (Eistring *eistr, Charcount charpos); | |
| 1460 Convert a char offset to a byte offset. | |
| 1461 Charcount eibytepos_to_charpos (Eistring *eistr, Bytecount bytepos); | |
| 1462 Convert a byte offset to a char offset. | |
| 1463 Bytecount eiincpos (Eistring *eistr, Bytecount bytepos); | |
| 1464 Increment the given position by one character. | |
| 1465 Bytecount eiincpos_n (Eistring *eistr, Bytecount bytepos, Charcount n); | |
| 1466 Increment the given position by N characters. | |
| 1467 Bytecount eidecpos (Eistring *eistr, Bytecount bytepos); | |
| 1468 Decrement the given position by one character. | |
| 1469 Bytecount eidecpos_n (Eistring *eistr, Bytecount bytepos, Charcount n); | |
| 1470 Deccrement the given position by N characters. | |
| 1471 | |
| 1472 | |
| 1473 ********************************************** | |
| 1474 * Getting the character at a position * | |
| 1475 ********************************************** | |
| 1476 | |
| 867 | 1477 Ichar eigetch (Eistring *eistr, Bytecount bytepos); |
| 771 | 1478 Return the character at a particular byte offset. |
| 867 | 1479 Ichar eigetch_char (Eistring *eistr, Charcount charpos); |
| 771 | 1480 Return the character at a particular character offset. |
| 1481 | |
| 1482 | |
| 1483 ********************************************** | |
| 1484 * Setting the character at a position * | |
| 1485 ********************************************** | |
| 1486 | |
| 867 | 1487 Ichar eisetch (Eistring *eistr, Bytecount bytepos, Ichar chr); |
| 771 | 1488 Set the character at a particular byte offset. |
| 867 | 1489 Ichar eisetch_char (Eistring *eistr, Charcount charpos, Ichar chr); |
| 771 | 1490 Set the character at a particular character offset. |
| 1491 | |
| 1492 | |
| 1493 ********************************************** | |
| 1494 * Concatenation * | |
| 1495 ********************************************** | |
| 1496 | |
| 1497 void eicat_* (Eistring *eistr, ...); | |
| 1498 Concatenate onto the end of the Eistring, with data coming from the | |
| 1499 same places as above: | |
| 1500 | |
| 1501 void eicat_ei (Eistring *eistr, Eistring *eistr2); | |
| 1502 ... from another Eistring. | |
| 2421 | 1503 void eicat_ascii (Eistring *eistr, Ascbyte *ascstr); |
| 771 | 1504 ... from an ASCII null-terminated string. Non-ASCII characters in |
| 2500 | 1505 the string are *ILLEGAL* (read ABORT() with error-checking defined). |
| 867 | 1506 void eicat_raw (ei, const Ibyte *data, Bytecount len); |
| 771 | 1507 ... from raw internal-format data in the default internal format. |
| 867 | 1508 void eicat_rawz (ei, const Ibyte *data); |
| 771 | 1509 ... from raw internal-format data in the default internal format |
| 1510 that is "null-terminated" (the meaning of this depends on the nature | |
| 1511 of the default internal format). | |
| 1512 void eicat_lstr (ei, Lisp_Object lisp_string); | |
| 1513 ... from a Lisp_Object string. | |
| 867 | 1514 void eicat_ch (ei, Ichar ch); |
| 1515 ... from an Ichar. | |
| 771 | 1516 |
| 1517 (All except the first variety are convenience functions. | |
| 1518 In the general case, create another Eistring from the source.) | |
| 1519 | |
| 1520 | |
| 1521 ********************************************** | |
| 1522 * Replacement * | |
| 1523 ********************************************** | |
| 1524 | |
| 1525 void eisub_* (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1526 Bytecount len, Charcount charlen, ...); | |
| 1527 Replace a section of the Eistring, specifically: | |
| 1528 | |
| 1529 void eisub_ei (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1530 Bytecount len, Charcount charlen, Eistring *eistr2); | |
| 1531 ... with another Eistring. | |
| 2421 | 1532 void eisub_ascii (Eistring *eistr, Bytecount off, Charcount charoff, |
| 1533 Bytecount len, Charcount charlen, Ascbyte *ascstr); | |
| 771 | 1534 ... with an ASCII null-terminated string. Non-ASCII characters in |
| 2500 | 1535 the string are *ILLEGAL* (read ABORT() with error-checking defined). |
| 771 | 1536 void eisub_ch (Eistring *eistr, Bytecount off, Charcount charoff, |
| 867 | 1537 Bytecount len, Charcount charlen, Ichar ch); |
| 1538 ... with an Ichar. | |
| 771 | 1539 |
| 1540 void eidel (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1541 Bytecount len, Charcount charlen); | |
| 1542 Delete a section of the Eistring. | |
| 1543 | |
| 1544 | |
| 1545 ********************************************** | |
| 1546 * Converting to an external format * | |
| 1547 ********************************************** | |
| 1548 | |
| 1318 | 1549 void eito_external (Eistring *eistr, Lisp_Object codesys); |
| 771 | 1550 Convert the Eistring to an external format and store the result |
| 1551 in the string. NOTE: Further changes to the Eistring will *NOT* | |
| 1552 change the external data stored in the string. You will have to | |
| 1553 call eito_external() again in such a case if you want the external | |
| 1554 data. | |
| 1555 | |
| 1556 Extbyte *eiextdata (Eistring *eistr); | |
| 1557 Return a pointer to the external data stored in the Eistring as | |
| 1558 a result of a prior call to eito_external(). | |
| 1559 | |
| 1560 Bytecount eiextlen (Eistring *eistr); | |
| 1561 Return the length in bytes of the external data stored in the | |
| 1562 Eistring as a result of a prior call to eito_external(). | |
| 1563 | |
| 1564 | |
| 1565 ********************************************** | |
| 1566 * Searching in the Eistring for a character * | |
| 1567 ********************************************** | |
| 1568 | |
| 867 | 1569 Bytecount eichr (Eistring *eistr, Ichar chr); |
| 1570 Charcount eichr_char (Eistring *eistr, Ichar chr); | |
| 1571 Bytecount eichr_off (Eistring *eistr, Ichar chr, Bytecount off, | |
| 771 | 1572 Charcount charoff); |
| 867 | 1573 Charcount eichr_off_char (Eistring *eistr, Ichar chr, Bytecount off, |
| 771 | 1574 Charcount charoff); |
| 867 | 1575 Bytecount eirchr (Eistring *eistr, Ichar chr); |
| 1576 Charcount eirchr_char (Eistring *eistr, Ichar chr); | |
| 1577 Bytecount eirchr_off (Eistring *eistr, Ichar chr, Bytecount off, | |
| 771 | 1578 Charcount charoff); |
| 867 | 1579 Charcount eirchr_off_char (Eistring *eistr, Ichar chr, Bytecount off, |
| 771 | 1580 Charcount charoff); |
| 1581 | |
| 1582 | |
| 1583 ********************************************** | |
| 1584 * Searching in the Eistring for a string * | |
| 1585 ********************************************** | |
| 1586 | |
| 1587 Bytecount eistr_ei (Eistring *eistr, Eistring *eistr2); | |
| 1588 Charcount eistr_ei_char (Eistring *eistr, Eistring *eistr2); | |
| 1589 Bytecount eistr_ei_off (Eistring *eistr, Eistring *eistr2, Bytecount off, | |
| 1590 Charcount charoff); | |
| 1591 Charcount eistr_ei_off_char (Eistring *eistr, Eistring *eistr2, | |
| 1592 Bytecount off, Charcount charoff); | |
| 1593 Bytecount eirstr_ei (Eistring *eistr, Eistring *eistr2); | |
| 1594 Charcount eirstr_ei_char (Eistring *eistr, Eistring *eistr2); | |
| 1595 Bytecount eirstr_ei_off (Eistring *eistr, Eistring *eistr2, Bytecount off, | |
| 1596 Charcount charoff); | |
| 1597 Charcount eirstr_ei_off_char (Eistring *eistr, Eistring *eistr2, | |
| 1598 Bytecount off, Charcount charoff); | |
| 1599 | |
| 2421 | 1600 Bytecount eistr_ascii (Eistring *eistr, Ascbyte *ascstr); |
| 1601 Charcount eistr_ascii_char (Eistring *eistr, Ascbyte *ascstr); | |
| 1602 Bytecount eistr_ascii_off (Eistring *eistr, Ascbyte *ascstr, Bytecount off, | |
| 771 | 1603 Charcount charoff); |
| 2421 | 1604 Charcount eistr_ascii_off_char (Eistring *eistr, Ascbyte *ascstr, |
| 771 | 1605 Bytecount off, Charcount charoff); |
| 2421 | 1606 Bytecount eirstr_ascii (Eistring *eistr, Ascbyte *ascstr); |
| 1607 Charcount eirstr_ascii_char (Eistring *eistr, Ascbyte *ascstr); | |
| 1608 Bytecount eirstr_ascii_off (Eistring *eistr, Ascbyte *ascstr, | |
| 771 | 1609 Bytecount off, Charcount charoff); |
| 2421 | 1610 Charcount eirstr_ascii_off_char (Eistring *eistr, Ascbyte *ascstr, |
| 771 | 1611 Bytecount off, Charcount charoff); |
| 1612 | |
| 1613 | |
| 1614 ********************************************** | |
| 1615 * Comparison * | |
| 1616 ********************************************** | |
| 1617 | |
| 1618 int eicmp_* (Eistring *eistr, ...); | |
| 1619 int eicmp_off_* (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1620 Bytecount len, Charcount charlen, ...); | |
| 1621 int eicasecmp_* (Eistring *eistr, ...); | |
| 1622 int eicasecmp_off_* (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1623 Bytecount len, Charcount charlen, ...); | |
| 1624 int eicasecmp_i18n_* (Eistring *eistr, ...); | |
| 1625 int eicasecmp_i18n_off_* (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1626 Bytecount len, Charcount charlen, ...); | |
| 1627 | |
| 1628 Compare the Eistring with the other data. Return value same as | |
| 1629 from strcmp. The `*' is either `ei' for another Eistring (in | |
| 1630 which case `...' is an Eistring), or `c' for a pure-ASCII string | |
| 1631 (in which case `...' is a pointer to that string). For anything | |
| 1632 more complex, first create an Eistring out of the source. | |
| 1633 Comparison is either simple (`eicmp_...'), ASCII case-folding | |
| 1634 (`eicasecmp_...'), or multilingual case-folding | |
| 1635 (`eicasecmp_i18n_...). | |
| 1636 | |
| 1637 | |
| 1638 More specifically, the prototypes are: | |
| 1639 | |
| 1640 int eicmp_ei (Eistring *eistr, Eistring *eistr2); | |
| 1641 int eicmp_off_ei (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1642 Bytecount len, Charcount charlen, Eistring *eistr2); | |
| 1643 int eicasecmp_ei (Eistring *eistr, Eistring *eistr2); | |
| 1644 int eicasecmp_off_ei (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1645 Bytecount len, Charcount charlen, Eistring *eistr2); | |
| 1646 int eicasecmp_i18n_ei (Eistring *eistr, Eistring *eistr2); | |
| 1647 int eicasecmp_i18n_off_ei (Eistring *eistr, Bytecount off, | |
| 1648 Charcount charoff, Bytecount len, | |
| 1649 Charcount charlen, Eistring *eistr2); | |
| 1650 | |
| 2421 | 1651 int eicmp_ascii (Eistring *eistr, Ascbyte *ascstr); |
| 1652 int eicmp_off_ascii (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 1653 Bytecount len, Charcount charlen, Ascbyte *ascstr); | |
| 1654 int eicasecmp_ascii (Eistring *eistr, Ascbyte *ascstr); | |
| 1655 int eicasecmp_off_ascii (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 771 | 1656 Bytecount len, Charcount charlen, |
| 2421 | 1657 Ascbyte *ascstr); |
| 1658 int eicasecmp_i18n_ascii (Eistring *eistr, Ascbyte *ascstr); | |
| 1659 int eicasecmp_i18n_off_ascii (Eistring *eistr, Bytecount off, Charcount charoff, | |
| 771 | 1660 Bytecount len, Charcount charlen, |
| 2421 | 1661 Ascbyte *ascstr); |
| 771 | 1662 |
| 1663 | |
| 1664 ********************************************** | |
| 1665 * Case-changing the Eistring * | |
| 1666 ********************************************** | |
| 1667 | |
| 1668 void eilwr (Eistring *eistr); | |
| 1669 Convert all characters in the Eistring to lowercase. | |
| 1670 void eiupr (Eistring *eistr); | |
| 1671 Convert all characters in the Eistring to uppercase. | |
| 1672 */ | |
| 1673 | |
| 1674 | |
| 1675 /* Principles for writing Eistring functions: | |
| 1676 | |
| 1677 (1) Unfortunately, we have to write most of the Eistring functions | |
| 851 | 1678 as macros, because of the use of ALLOCA(). The principle used |
| 771 | 1679 below to assure no conflict in local variables is to prefix all |
| 1680 local variables with "ei" plus a number, which should be unique | |
| 1681 among macros. In practice, when finding a new number, find the | |
| 1682 highest so far used, and add 1. | |
| 1683 | |
| 1684 (2) We also suffix the Eistring fields with an _ to avoid problems | |
| 1685 with macro parameters of the same name. (And as the standard | |
| 1686 signal not to access these fields directly.) | |
| 1687 | |
| 1688 (3) We maintain both the length in bytes and chars of the data in | |
| 1689 the Eistring at all times, for convenient retrieval by outside | |
| 1690 functions. That means when writing functions that manipulate | |
| 1691 Eistrings, you too need to keep both lengths up to date for all | |
| 1692 data that you work with. | |
| 1693 | |
| 1694 (4) When writing a new type of operation (e.g. substitution), you | |
| 1695 will often find yourself working with outside data, and thus | |
| 1696 have a series of related API's, for different forms that the | |
| 1697 outside data is in. Generally, you will want to choose a | |
| 1698 subset of the forms supported by eicpy_*, which has to be | |
| 1699 totally general because that's the fundamental way to get data | |
| 1700 into an Eistring, and once the data is into the string, it | |
| 1701 would be to create a whole series of Ei operations that work on | |
| 1702 nothing but Eistrings. Although theoretically nice, in | |
| 1703 practice it's a hassle, so we suggest that you provide | |
| 1704 convenience functions. In particular, there are two paths you | |
| 1705 can take. One is minimalist -- it only allows other Eistrings | |
| 867 | 1706 and ASCII data, and Ichars if the particular operation makes |
| 771 | 1707 sense with a character. The other provides interfaces for the |
| 1708 most commonly-used forms -- Eistring, ASCII data, Lisp string, | |
| 1709 raw internal-format string with length, raw internal-format | |
| 867 | 1710 string without, and possibly Ichar. (In the function names, |
| 771 | 1711 these are designated `ei', `c', `lstr', `raw', `rawz', and |
| 1712 `ch', respectively.) | |
| 1713 | |
| 1714 (5) When coding a new type of operation, such as was discussed in | |
| 1715 previous section, the correct approach is to declare an worker | |
| 1716 function that does the work of everything, and is called by the | |
| 1717 other "container" macros that handle the different outside data | |
| 1718 forms. The data coming into the worker function, which | |
| 1719 typically ends in `_1', is in the form of three parameters: | |
| 1720 DATA, LEN, CHARLEN. (See point [3] about having two lengths and | |
| 1721 keeping them in sync.) | |
| 1722 | |
| 1723 (6) Handling argument evaluation in macros: We take great care | |
| 1724 never to evaluate any argument more than once in any macro, | |
| 1725 except the initial Eistring parameter. This can and will be | |
| 1726 evaluated multiple times, but it should pretty much always just | |
| 1727 be a simple variable. This means, for example, that if an | |
| 1728 Eistring is the second (not first) argument of a macro, it | |
| 1729 doesn't fall under the "initial Eistring" exemption, so it | |
| 1730 needs protection against multi-evaluation. (Take the address of | |
| 1731 the Eistring structure, store in a temporary variable, and use | |
| 1732 temporary variable for all access to the Eistring. | |
| 1733 Essentially, we want it to appear as if these Eistring macros | |
| 1734 are functions -- we would like to declare them as functions but | |
| 851 | 1735 they use ALLOCA(), so we can't (and we can't make them inline |
| 1736 functions either -- ALLOCA() is explicitly disallowed in inline | |
| 771 | 1737 functions.) |
| 1738 | |
| 1739 (7) Note that our rules regarding multiple evaluation are *more* | |
| 1740 strict than the rules listed above under the heading "working | |
| 1741 with raw internal-format data". | |
| 1742 */ | |
| 1743 | |
| 1744 | |
| 1745 /* ----- Declaration ----- */ | |
| 1746 | |
| 1747 typedef struct | |
| 1748 { | |
| 1749 /* Data for the Eistring, stored in the default internal format. | |
| 1750 Always includes terminating null. */ | |
| 867 | 1751 Ibyte *data_; |
| 771 | 1752 /* Total number of bytes allocated in DATA (including null). */ |
| 1753 Bytecount max_size_allocated_; | |
| 1754 Bytecount bytelen_; | |
| 1755 Charcount charlen_; | |
| 1756 int mallocp_; | |
| 1757 | |
| 1758 Extbyte *extdata_; | |
| 1759 Bytecount extlen_; | |
| 1760 } Eistring; | |
| 1761 | |
| 1762 extern Eistring the_eistring_zero_init, the_eistring_malloc_zero_init; | |
| 1763 | |
| 1764 #define DECLARE_EISTRING(name) \ | |
| 1765 Eistring __ ## name ## __storage__ = the_eistring_zero_init; \ | |
| 1766 Eistring *name = & __ ## name ## __storage__ | |
| 1767 #define DECLARE_EISTRING_MALLOC(name) \ | |
| 1768 Eistring __ ## name ## __storage__ = the_eistring_malloc_zero_init; \ | |
| 1769 Eistring *name = & __ ## name ## __storage__ | |
| 1770 | |
| 1771 #define eiinit(ei) \ | |
| 1772 do { \ | |
| 793 | 1773 *(ei) = the_eistring_zero_init; \ |
| 771 | 1774 } while (0) |
| 1775 | |
| 1776 #define eiinit_malloc(ei) \ | |
| 1777 do { \ | |
| 793 | 1778 *(ei) = the_eistring_malloc_zero_init; \ |
| 771 | 1779 } while (0) |
| 1780 | |
| 1781 | |
| 1782 /* ----- Utility ----- */ | |
| 1783 | |
| 1784 /* Make sure both LEN and CHARLEN are specified, in case one is given | |
| 1785 as -1. PTR evaluated at most once, others multiply. */ | |
| 1786 #define eifixup_bytechar(ptr, len, charlen) \ | |
| 1787 do { \ | |
| 1788 if ((len) == -1) \ | |
| 1789 (len) = charcount_to_bytecount (ptr, charlen); \ | |
| 1790 else if ((charlen) == -1) \ | |
| 1791 (charlen) = bytecount_to_charcount (ptr, len); \ | |
| 1792 } while (0) | |
| 1793 | |
| 1794 /* Make sure LEN is specified, in case it's is given as -1. PTR | |
| 1795 evaluated at most once, others multiply. */ | |
| 1796 #define eifixup_byte(ptr, len, charlen) \ | |
| 1797 do { \ | |
| 1798 if ((len) == -1) \ | |
| 1799 (len) = charcount_to_bytecount (ptr, charlen); \ | |
| 1800 } while (0) | |
| 1801 | |
| 1802 /* Make sure CHARLEN is specified, in case it's is given as -1. PTR | |
| 1803 evaluated at most once, others multiply. */ | |
| 1804 #define eifixup_char(ptr, len, charlen) \ | |
| 1805 do { \ | |
| 1806 if ((charlen) == -1) \ | |
| 1807 (charlen) = bytecount_to_charcount (ptr, len); \ | |
| 1808 } while (0) | |
| 1809 | |
| 1810 | |
| 1811 | |
| 1812 /* Make sure we can hold NEWBYTELEN bytes (which is NEWCHARLEN chars) | |
| 1813 plus a zero terminator. Preserve existing data as much as possible, | |
| 1814 including existing zero terminator. Put a new zero terminator where it | |
| 1815 should go if NEWZ if non-zero. All args but EI are evalled only once. */ | |
| 1816 | |
| 1817 #define EI_ALLOC(ei, newbytelen, newcharlen, newz) \ | |
| 1818 do { \ | |
| 1819 int ei1oldeibytelen = (ei)->bytelen_; \ | |
| 1820 \ | |
| 1821 (ei)->charlen_ = (newcharlen); \ | |
| 1822 (ei)->bytelen_ = (newbytelen); \ | |
| 1823 \ | |
| 1824 if (ei1oldeibytelen != (ei)->bytelen_) \ | |
| 1825 { \ | |
| 1826 int ei1newsize = (ei)->max_size_allocated_; \ | |
| 1827 while (ei1newsize < (ei)->bytelen_ + 1) \ | |
| 1828 { \ | |
| 1829 ei1newsize = (int) (ei1newsize * 1.5); \ | |
| 1830 if (ei1newsize < 32) \ | |
| 1831 ei1newsize = 32; \ | |
| 1832 } \ | |
| 1833 if (ei1newsize != (ei)->max_size_allocated_) \ | |
| 1834 { \ | |
| 1835 if ((ei)->mallocp_) \ | |
| 1836 /* xrealloc always preserves existing data as much as possible */ \ | |
| 1333 | 1837 (ei)->data_ = (Ibyte *) xrealloc ((ei)->data_, ei1newsize); \ |
| 771 | 1838 else \ |
| 1839 { \ | |
| 851 | 1840 /* We don't have realloc, so ALLOCA() more space and copy the \ |
| 771 | 1841 data into it. */ \ |
| 867 | 1842 Ibyte *ei1oldeidata = (ei)->data_; \ |
| 2367 | 1843 (ei)->data_ = alloca_ibytes (ei1newsize); \ |
| 771 | 1844 if (ei1oldeidata) \ |
| 1845 memcpy ((ei)->data_, ei1oldeidata, ei1oldeibytelen + 1); \ | |
| 1846 } \ | |
| 1847 (ei)->max_size_allocated_ = ei1newsize; \ | |
| 1848 } \ | |
| 1849 if (newz) \ | |
| 1850 (ei)->data_[(ei)->bytelen_] = '\0'; \ | |
| 1851 } \ | |
| 1852 } while (0) | |
| 1853 | |
| 1854 #define EI_ALLOC_AND_COPY(ei, data, bytelen, charlen) \ | |
| 1855 do { \ | |
| 1856 EI_ALLOC (ei, bytelen, charlen, 1); \ | |
| 1857 memcpy ((ei)->data_, data, (ei)->bytelen_); \ | |
| 1858 } while (0) | |
| 1859 | |
| 1860 /* ----- Initialization ----- */ | |
| 1861 | |
| 1862 #define eicpy_ei(ei, eicpy) \ | |
| 1863 do { \ | |
| 1864 const Eistring *ei2 = (eicpy); \ | |
| 1865 EI_ALLOC_AND_COPY (ei, ei2->data_, ei2->bytelen_, ei2->charlen_); \ | |
| 1866 } while (0) | |
| 1867 | |
| 1868 #define eicpy_lstr(ei, lisp_string) \ | |
| 1869 do { \ | |
| 1870 Lisp_Object ei3 = (lisp_string); \ | |
| 1871 EI_ALLOC_AND_COPY (ei, XSTRING_DATA (ei3), XSTRING_LENGTH (ei3), \ | |
| 1333 | 1872 string_char_length (ei3)); \ |
| 771 | 1873 } while (0) |
| 1874 | |
| 1875 #define eicpy_lstr_off(ei, lisp_string, off, charoff, len, charlen) \ | |
| 1876 do { \ | |
| 1877 Lisp_Object ei23lstr = (lisp_string); \ | |
| 1878 int ei23off = (off); \ | |
| 1879 int ei23charoff = (charoff); \ | |
| 1880 int ei23len = (len); \ | |
| 1881 int ei23charlen = (charlen); \ | |
| 867 | 1882 const Ibyte *ei23data = XSTRING_DATA (ei23lstr); \ |
| 771 | 1883 \ |
| 1884 int ei23oldbytelen = (ei)->bytelen_; \ | |
| 1885 \ | |
| 1886 eifixup_byte (ei23data, ei23off, ei23charoff); \ | |
| 1887 eifixup_bytechar (ei23data + ei23off, ei23len, ei23charlen); \ | |
| 1888 \ | |
| 1889 EI_ALLOC_AND_COPY (ei, ei23data + ei23off, ei23len, ei23charlen); \ | |
| 1890 } while (0) | |
| 1891 | |
| 826 | 1892 #define eicpy_raw_fmt(ei, ptr, len, fmt, object) \ |
| 771 | 1893 do { \ |
| 1333 | 1894 const Ibyte *ei12ptr = (ptr); \ |
| 771 | 1895 Internal_Format ei12fmt = (fmt); \ |
| 1896 int ei12len = (len); \ | |
| 1897 assert (ei12fmt == FORMAT_DEFAULT); \ | |
| 1898 EI_ALLOC_AND_COPY (ei, ei12ptr, ei12len, \ | |
| 1899 bytecount_to_charcount (ei12ptr, ei12len)); \ | |
| 1900 } while (0) | |
| 1901 | |
| 826 | 1902 #define eicpy_raw(ei, ptr, len) \ |
| 1903 eicpy_raw_fmt (ei, ptr, len, FORMAT_DEFAULT, Qnil) | |
| 1904 | |
| 1905 #define eicpy_rawz_fmt(ei, ptr, fmt, object) \ | |
| 1906 do { \ | |
| 867 | 1907 const Ibyte *ei12p1ptr = (ptr); \ |
| 826 | 1908 Internal_Format ei12p1fmt = (fmt); \ |
| 1909 assert (ei12p1fmt == FORMAT_DEFAULT); \ | |
| 1910 eicpy_raw_fmt (ei, ei12p1ptr, qxestrlen (ei12p1ptr), fmt, object); \ | |
| 771 | 1911 } while (0) |
| 1912 | |
| 826 | 1913 #define eicpy_rawz(ei, ptr) eicpy_rawz_fmt (ei, ptr, FORMAT_DEFAULT, Qnil) |
| 771 | 1914 |
| 1333 | 1915 #define eicpy_ch(ei, ch) \ |
| 1916 do { \ | |
| 867 | 1917 Ibyte ei12p2[MAX_ICHAR_LEN]; \ |
| 1918 Bytecount ei12p2len = set_itext_ichar (ei12p2, ch); \ | |
| 1333 | 1919 EI_ALLOC_AND_COPY (ei, ei12p2, ei12p2len, 1); \ |
| 771 | 1920 } while (0) |
| 1921 | |
| 2421 | 1922 #define eicpy_ascii(ei, ascstr) \ |
| 771 | 1923 do { \ |
| 2421 | 1924 const Ascbyte *ei4 = (ascstr); \ |
| 771 | 1925 \ |
| 2367 | 1926 ASSERT_ASCTEXT_ASCII (ei4); \ |
| 771 | 1927 eicpy_ext (ei, ei4, Qbinary); \ |
| 1928 } while (0) | |
| 1929 | |
| 2421 | 1930 #define eicpy_ascii_len(ei, ascstr, c_len) \ |
| 771 | 1931 do { \ |
| 2421 | 1932 const Ascbyte *ei6 = (ascstr); \ |
| 771 | 1933 int ei6len = (c_len); \ |
| 1934 \ | |
| 2367 | 1935 ASSERT_ASCTEXT_ASCII_LEN (ei6, ei6len); \ |
| 771 | 1936 eicpy_ext_len (ei, ei6, ei6len, Qbinary); \ |
| 1937 } while (0) | |
| 1938 | |
|
4981
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1939 #define eicpy_ext_len(ei, extdata, extlen, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1940 do { \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1941 const Extbyte *ei7 = (extdata); \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1942 int ei7len = (extlen); \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1943 \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1944 TO_INTERNAL_FORMAT (DATA, (ei7, ei7len), \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1945 ALLOCA, ((ei)->data_, (ei)->bytelen_), \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1946 codesys); \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
1947 (ei)->max_size_allocated_ = (ei)->bytelen_ + 1; \ |
| 771 | 1948 (ei)->charlen_ = bytecount_to_charcount ((ei)->data_, (ei)->bytelen_); \ |
| 1949 } while (0) | |
| 1950 | |
| 1318 | 1951 #define eicpy_ext(ei, extdata, codesys) \ |
| 1952 do { \ | |
| 1953 const Extbyte *ei8 = (extdata); \ | |
| 1954 \ | |
| 1955 eicpy_ext_len (ei, ei8, dfc_external_data_len (ei8, codesys), \ | |
| 1956 codesys); \ | |
| 771 | 1957 } while (0) |
| 1958 | |
| 1959 #define eicpy_lbuf(eistr, lisp_buf, off, charoff, len, charlen) \ | |
| 1960 NOT YET IMPLEMENTED | |
| 1961 | |
| 1962 #define eicpy_lstream(eistr, lstream) \ | |
| 1963 NOT YET IMPLEMENTED | |
| 1964 | |
| 867 | 1965 #define eireset(eistr) eicpy_rawz (eistr, (Ibyte *) "") |
| 771 | 1966 |
| 1967 /* ----- Getting the data out of the Eistring ----- */ | |
| 1968 | |
| 1969 #define eidata(ei) ((ei)->data_) | |
| 1970 | |
| 1971 #define eimake_string(ei) make_string (eidata (ei), eilen (ei)) | |
| 1972 | |
| 1973 #define eimake_string_off(eistr, off, charoff, len, charlen) \ | |
| 1974 do { \ | |
| 1975 Lisp_Object ei24lstr; \ | |
| 1976 int ei24off = (off); \ | |
| 1977 int ei24charoff = (charoff); \ | |
| 1978 int ei24len = (len); \ | |
| 1979 int ei24charlen = (charlen); \ | |
| 1980 \ | |
| 1981 eifixup_byte ((eistr)->data_, ei24off, ei24charoff); \ | |
| 1982 eifixup_byte ((eistr)->data_ + ei24off, ei24len, ei24charlen); \ | |
| 1983 \ | |
| 1984 return make_string ((eistr)->data_ + ei24off, ei24len); \ | |
| 1985 } while (0) | |
| 1986 | |
| 1987 #define eicpyout_alloca(eistr, ptrout, lenout) \ | |
| 826 | 1988 eicpyout_alloca_fmt (eistr, ptrout, lenout, FORMAT_DEFAULT, Qnil) |
| 771 | 1989 #define eicpyout_malloc(eistr, lenout) \ |
| 826 | 1990 eicpyout_malloc_fmt (eistr, lenout, FORMAT_DEFAULT, Qnil) |
| 867 | 1991 Ibyte *eicpyout_malloc_fmt (Eistring *eistr, Bytecount *len_out, |
| 826 | 1992 Internal_Format fmt, Lisp_Object object); |
| 1993 #define eicpyout_alloca_fmt(eistr, ptrout, lenout, fmt, object) \ | |
| 771 | 1994 do { \ |
| 1995 Internal_Format ei23fmt = (fmt); \ | |
| 867 | 1996 Ibyte *ei23ptrout = &(ptrout); \ |
| 771 | 1997 Bytecount *ei23lenout = &(lenout); \ |
| 1998 \ | |
| 1999 assert (ei23fmt == FORMAT_DEFAULT); \ | |
| 2000 \ | |
| 2001 *ei23lenout = (eistr)->bytelen_; \ | |
| 2367 | 2002 *ei23ptrout = alloca_ibytes ((eistr)->bytelen_ + 1); \ |
| 771 | 2003 memcpy (*ei23ptrout, (eistr)->data_, (eistr)->bytelen_ + 1); \ |
| 2004 } while (0) | |
| 2005 | |
| 2006 /* ----- Moving to the heap ----- */ | |
| 2007 | |
| 2008 #define eifree(ei) \ | |
| 2009 do { \ | |
| 2010 if ((ei)->mallocp_) \ | |
| 2011 { \ | |
| 2012 if ((ei)->data_) \ | |
|
4976
16112448d484
Rename xfree(FOO, TYPE) -> xfree(FOO)
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2013 xfree ((ei)->data_); \ |
| 771 | 2014 if ((ei)->extdata_) \ |
|
4976
16112448d484
Rename xfree(FOO, TYPE) -> xfree(FOO)
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2015 xfree ((ei)->extdata_); \ |
| 771 | 2016 eiinit_malloc (ei); \ |
| 2017 } \ | |
| 2018 else \ | |
| 2019 eiinit (ei); \ | |
| 2020 } while (0) | |
| 2021 | |
| 2022 int eifind_large_enough_buffer (int oldbufsize, int needed_size); | |
| 2023 void eito_malloc_1 (Eistring *ei); | |
| 2024 | |
| 2025 #define eito_malloc(ei) eito_malloc_1 (ei) | |
| 2026 | |
| 2027 #define eito_alloca(ei) \ | |
| 2028 do { \ | |
| 2029 if (!(ei)->mallocp_) \ | |
| 2030 return; \ | |
| 2031 (ei)->mallocp_ = 0; \ | |
| 2032 if ((ei)->data_) \ | |
| 2033 { \ | |
| 867 | 2034 Ibyte *ei13newdata; \ |
| 771 | 2035 \ |
| 2036 (ei)->max_size_allocated_ = \ | |
| 2037 eifind_large_enough_buffer (0, (ei)->bytelen_ + 1); \ | |
| 2367 | 2038 ei13newdata = alloca_ibytes ((ei)->max_size_allocated_); \ |
| 771 | 2039 memcpy (ei13newdata, (ei)->data_, (ei)->bytelen_ + 1); \ |
|
4976
16112448d484
Rename xfree(FOO, TYPE) -> xfree(FOO)
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2040 xfree ((ei)->data_); \ |
| 771 | 2041 (ei)->data_ = ei13newdata; \ |
| 2042 } \ | |
| 2043 \ | |
| 2044 if ((ei)->extdata_) \ | |
| 2045 { \ | |
| 2367 | 2046 Extbyte *ei13newdata = alloca_extbytes ((ei)->extlen_ + 2); \ |
| 771 | 2047 \ |
| 2048 memcpy (ei13newdata, (ei)->extdata_, (ei)->extlen_); \ | |
| 2049 /* Double null-terminate in case of Unicode data */ \ | |
| 2050 ei13newdata[(ei)->extlen_] = '\0'; \ | |
| 2051 ei13newdata[(ei)->extlen_ + 1] = '\0'; \ | |
|
4976
16112448d484
Rename xfree(FOO, TYPE) -> xfree(FOO)
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2052 xfree ((ei)->extdata_); \ |
| 771 | 2053 (ei)->extdata_ = ei13newdata; \ |
| 2054 } \ | |
| 2055 } while (0) | |
| 2056 | |
| 2057 | |
| 2058 /* ----- Retrieving the length ----- */ | |
| 2059 | |
| 2060 #define eilen(ei) ((ei)->bytelen_) | |
| 2061 #define eicharlen(ei) ((ei)->charlen_) | |
| 2062 | |
| 2063 | |
| 2064 /* ----- Working with positions ----- */ | |
| 2065 | |
| 2066 #define eicharpos_to_bytepos(ei, charpos) \ | |
| 2067 charcount_to_bytecount ((ei)->data_, charpos) | |
| 2068 #define eibytepos_to_charpos(ei, bytepos) \ | |
| 2069 bytecount_to_charcount ((ei)->data_, bytepos) | |
| 2070 | |
| 2071 DECLARE_INLINE_HEADER (Bytecount eiincpos_1 (Eistring *eistr, | |
| 2072 Bytecount bytepos, | |
| 2073 Charcount n)) | |
| 2074 { | |
| 867 | 2075 Ibyte *pos = eistr->data_ + bytepos; |
| 814 | 2076 Charcount i; |
| 771 | 2077 |
| 800 | 2078 text_checking_assert (bytepos >= 0 && bytepos <= eistr->bytelen_); |
| 2079 text_checking_assert (n >= 0 && n <= eistr->charlen_); | |
| 771 | 2080 /* We could check N more correctly now, but that would require a |
| 2081 call to bytecount_to_charcount(), which would be needlessly | |
| 2082 expensive (it would convert O(N) algorithms into O(N^2) algorithms | |
| 800 | 2083 with ERROR_CHECK_TEXT, which would be bad). If N is bad, we are |
| 867 | 2084 guaranteed to catch it either inside INC_IBYTEPTR() or in the check |
| 771 | 2085 below. */ |
| 2086 for (i = 0; i < n; i++) | |
| 867 | 2087 INC_IBYTEPTR (pos); |
| 800 | 2088 text_checking_assert (pos - eistr->data_ <= eistr->bytelen_); |
| 771 | 2089 return pos - eistr->data_; |
| 2090 } | |
| 2091 | |
| 2092 #define eiincpos (ei, bytepos) eiincpos_1 (ei, bytepos, 1) | |
| 2093 #define eiincpos_n (ei, bytepos, n) eiincpos_1 (ei, bytepos, n) | |
| 2094 | |
| 2095 DECLARE_INLINE_HEADER (Bytecount eidecpos_1 (Eistring *eistr, | |
| 2096 Bytecount bytepos, | |
| 2097 Charcount n)) | |
| 2098 { | |
| 867 | 2099 Ibyte *pos = eistr->data_ + bytepos; |
| 771 | 2100 int i; |
| 2101 | |
| 800 | 2102 text_checking_assert (bytepos >= 0 && bytepos <= eistr->bytelen_); |
| 2103 text_checking_assert (n >= 0 && n <= eistr->charlen_); | |
| 771 | 2104 /* We could check N more correctly now, but ... see above. */ |
| 2105 for (i = 0; i < n; i++) | |
| 867 | 2106 DEC_IBYTEPTR (pos); |
| 800 | 2107 text_checking_assert (pos - eistr->data_ <= eistr->bytelen_); |
| 771 | 2108 return pos - eistr->data_; |
| 2109 } | |
| 2110 | |
| 2111 #define eidecpos (ei, bytepos) eidecpos_1 (ei, bytepos, 1) | |
| 2112 #define eidecpos_n (ei, bytepos, n) eidecpos_1 (ei, bytepos, n) | |
| 2113 | |
| 2114 | |
| 2115 /* ----- Getting the character at a position ----- */ | |
| 2116 | |
| 2117 #define eigetch(ei, bytepos) \ | |
| 867 | 2118 itext_ichar ((ei)->data_ + (bytepos)) |
| 2119 #define eigetch_char(ei, charpos) itext_ichar_n ((ei)->data_, charpos) | |
| 771 | 2120 |
| 2121 | |
| 2122 /* ----- Setting the character at a position ----- */ | |
| 2123 | |
| 2124 #define eisetch(ei, bytepos, chr) \ | |
| 2125 eisub_ch (ei, bytepos, -1, -1, 1, chr) | |
| 2126 #define eisetch_char(ei, charpos, chr) \ | |
| 2127 eisub_ch (ei, -1, charpos, -1, 1, chr) | |
| 2128 | |
| 2129 | |
| 2130 /* ----- Concatenation ----- */ | |
| 2131 | |
| 2132 #define eicat_1(ei, data, bytelen, charlen) \ | |
| 2133 do { \ | |
| 2134 int ei14oldeibytelen = (ei)->bytelen_; \ | |
| 2135 int ei14bytelen = (bytelen); \ | |
| 2136 EI_ALLOC (ei, (ei)->bytelen_ + ei14bytelen, \ | |
| 2137 (ei)->charlen_ + (charlen), 1); \ | |
| 2138 memcpy ((ei)->data_ + ei14oldeibytelen, (data), \ | |
| 2139 ei14bytelen); \ | |
| 2140 } while (0) | |
| 2141 | |
| 2142 #define eicat_ei(ei, ei2) \ | |
| 2143 do { \ | |
| 2144 const Eistring *ei9 = (ei2); \ | |
| 2145 eicat_1 (ei, ei9->data_, ei9->bytelen_, ei9->charlen_); \ | |
| 2146 } while (0) | |
| 2147 | |
| 2421 | 2148 #define eicat_ascii(ei, ascstr) \ |
| 771 | 2149 do { \ |
| 2421 | 2150 const Ascbyte *ei15 = (ascstr); \ |
| 771 | 2151 int ei15len = strlen (ei15); \ |
| 2152 \ | |
| 2367 | 2153 ASSERT_ASCTEXT_ASCII_LEN (ei15, ei15len); \ |
| 771 | 2154 eicat_1 (ei, ei15, ei15len, \ |
| 867 | 2155 bytecount_to_charcount ((Ibyte *) ei15, ei15len)); \ |
| 771 | 2156 } while (0) |
| 2157 | |
| 2158 #define eicat_raw(ei, data, len) \ | |
| 2159 do { \ | |
| 2160 int ei16len = (len); \ | |
| 867 | 2161 const Ibyte *ei16data = (data); \ |
| 771 | 2162 eicat_1 (ei, ei16data, ei16len, \ |
| 2163 bytecount_to_charcount (ei16data, ei16len)); \ | |
| 2164 } while (0) | |
| 2165 | |
| 2166 #define eicat_rawz(ei, ptr) \ | |
| 2167 do { \ | |
| 867 | 2168 const Ibyte *ei16p5ptr = (ptr); \ |
| 771 | 2169 eicat_raw (ei, ei16p5ptr, qxestrlen (ei16p5ptr)); \ |
| 2170 } while (0) | |
| 2171 | |
| 2172 #define eicat_lstr(ei, lisp_string) \ | |
| 2173 do { \ | |
| 2174 Lisp_Object ei17 = (lisp_string); \ | |
| 2175 eicat_1 (ei, XSTRING_DATA (ei17), XSTRING_LENGTH (ei17), \ | |
| 826 | 2176 string_char_length (ei17)); \ |
| 771 | 2177 } while (0) |
| 2178 | |
| 2179 #define eicat_ch(ei, ch) \ | |
| 2180 do { \ | |
| 1333 | 2181 Ibyte ei22ch[MAX_ICHAR_LEN]; \ |
| 867 | 2182 Bytecount ei22len = set_itext_ichar (ei22ch, ch); \ |
| 771 | 2183 eicat_1 (ei, ei22ch, ei22len, 1); \ |
| 2184 } while (0) | |
| 2185 | |
| 2186 | |
| 2187 /* ----- Replacement ----- */ | |
| 2188 | |
| 2189 /* Replace the section of an Eistring at (OFF, LEN) with the data at | |
| 2190 SRC of length LEN. All positions have corresponding character values, | |
| 2191 and either can be -1 -- it will be computed from the other. */ | |
| 2192 | |
| 2193 #define eisub_1(ei, off, charoff, len, charlen, src, srclen, srccharlen) \ | |
| 2194 do { \ | |
| 2195 int ei18off = (off); \ | |
| 2196 int ei18charoff = (charoff); \ | |
| 2197 int ei18len = (len); \ | |
| 2198 int ei18charlen = (charlen); \ | |
| 867 | 2199 Ibyte *ei18src = (Ibyte *) (src); \ |
| 771 | 2200 int ei18srclen = (srclen); \ |
| 2201 int ei18srccharlen = (srccharlen); \ | |
| 2202 \ | |
| 2203 int ei18oldeibytelen = (ei)->bytelen_; \ | |
| 2204 \ | |
| 2205 eifixup_bytechar ((ei)->data_, ei18off, ei18charoff); \ | |
| 2206 eifixup_bytechar ((ei)->data_ + ei18off, ei18len, ei18charlen); \ | |
| 2207 eifixup_bytechar (ei18src, ei18srclen, ei18srccharlen); \ | |
| 2208 \ | |
| 2209 EI_ALLOC (ei, (ei)->bytelen_ + ei18srclen - ei18len, \ | |
| 2210 (ei)->charlen_ + ei18srccharlen - ei18charlen, 0); \ | |
| 2211 if (ei18len != ei18srclen) \ | |
| 2212 memmove ((ei)->data_ + ei18off + ei18srclen, \ | |
| 2213 (ei)->data_ + ei18off + ei18len, \ | |
| 2214 /* include zero terminator. */ \ | |
| 2215 ei18oldeibytelen - (ei18off + ei18len) + 1); \ | |
| 2216 if (ei18srclen > 0) \ | |
| 2217 memcpy ((ei)->data_ + ei18off, ei18src, ei18srclen); \ | |
| 2218 } while (0) | |
| 2219 | |
| 2220 #define eisub_ei(ei, off, charoff, len, charlen, ei2) \ | |
| 2221 do { \ | |
| 1333 | 2222 const Eistring *ei19 = (ei2); \ |
| 771 | 2223 eisub_1 (ei, off, charoff, len, charlen, ei19->data_, ei19->bytelen_, \ |
| 2224 ei19->charlen_); \ | |
| 2225 } while (0) | |
| 2226 | |
| 2421 | 2227 #define eisub_ascii(ei, off, charoff, len, charlen, ascstr) \ |
| 771 | 2228 do { \ |
| 2421 | 2229 const Ascbyte *ei20 = (ascstr); \ |
| 771 | 2230 int ei20len = strlen (ei20); \ |
| 2367 | 2231 ASSERT_ASCTEXT_ASCII_LEN (ei20, ei20len); \ |
| 771 | 2232 eisub_1 (ei, off, charoff, len, charlen, ei20, ei20len, -1); \ |
| 2233 } while (0) | |
| 2234 | |
| 2235 #define eisub_ch(ei, off, charoff, len, charlen, ch) \ | |
| 2236 do { \ | |
| 1333 | 2237 Ibyte ei21ch[MAX_ICHAR_LEN]; \ |
| 867 | 2238 Bytecount ei21len = set_itext_ichar (ei21ch, ch); \ |
| 771 | 2239 eisub_1 (ei, off, charoff, len, charlen, ei21ch, ei21len, 1); \ |
| 2240 } while (0) | |
| 2241 | |
| 2242 #define eidel(ei, off, charoff, len, charlen) \ | |
| 2243 eisub_1(ei, off, charoff, len, charlen, NULL, 0, 0) | |
| 2244 | |
| 2245 | |
| 2246 /* ----- Converting to an external format ----- */ | |
| 2247 | |
| 1333 | 2248 #define eito_external(ei, codesys) \ |
| 771 | 2249 do { \ |
| 2250 if ((ei)->mallocp_) \ | |
| 2251 { \ | |
| 2252 if ((ei)->extdata_) \ | |
| 2253 { \ | |
|
4976
16112448d484
Rename xfree(FOO, TYPE) -> xfree(FOO)
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2254 xfree ((ei)->extdata_); \ |
| 771 | 2255 (ei)->extdata_ = 0; \ |
| 2256 } \ | |
| 2257 TO_EXTERNAL_FORMAT (DATA, ((ei)->data_, (ei)->bytelen_), \ | |
| 2258 MALLOC, ((ei)->extdata_, (ei)->extlen_), \ | |
| 1333 | 2259 codesys); \ |
| 771 | 2260 } \ |
| 2261 else \ | |
| 2262 TO_EXTERNAL_FORMAT (DATA, ((ei)->data_, (ei)->bytelen_), \ | |
| 2263 ALLOCA, ((ei)->extdata_, (ei)->extlen_), \ | |
| 1318 | 2264 codesys); \ |
| 771 | 2265 } while (0) |
| 2266 | |
| 2267 #define eiextdata(ei) ((ei)->extdata_) | |
| 2268 #define eiextlen(ei) ((ei)->extlen_) | |
| 2269 | |
| 2270 | |
| 2271 /* ----- Searching in the Eistring for a character ----- */ | |
| 2272 | |
| 2273 #define eichr(eistr, chr) \ | |
| 2274 NOT YET IMPLEMENTED | |
| 2275 #define eichr_char(eistr, chr) \ | |
| 2276 NOT YET IMPLEMENTED | |
| 2277 #define eichr_off(eistr, chr, off, charoff) \ | |
| 2278 NOT YET IMPLEMENTED | |
| 2279 #define eichr_off_char(eistr, chr, off, charoff) \ | |
| 2280 NOT YET IMPLEMENTED | |
| 2281 #define eirchr(eistr, chr) \ | |
| 2282 NOT YET IMPLEMENTED | |
| 2283 #define eirchr_char(eistr, chr) \ | |
| 2284 NOT YET IMPLEMENTED | |
| 2285 #define eirchr_off(eistr, chr, off, charoff) \ | |
| 2286 NOT YET IMPLEMENTED | |
| 2287 #define eirchr_off_char(eistr, chr, off, charoff) \ | |
| 2288 NOT YET IMPLEMENTED | |
| 2289 | |
| 2290 | |
| 2291 /* ----- Searching in the Eistring for a string ----- */ | |
| 2292 | |
| 2293 #define eistr_ei(eistr, eistr2) \ | |
| 2294 NOT YET IMPLEMENTED | |
| 2295 #define eistr_ei_char(eistr, eistr2) \ | |
| 2296 NOT YET IMPLEMENTED | |
| 2297 #define eistr_ei_off(eistr, eistr2, off, charoff) \ | |
| 2298 NOT YET IMPLEMENTED | |
| 2299 #define eistr_ei_off_char(eistr, eistr2, off, charoff) \ | |
| 2300 NOT YET IMPLEMENTED | |
| 2301 #define eirstr_ei(eistr, eistr2) \ | |
| 2302 NOT YET IMPLEMENTED | |
| 2303 #define eirstr_ei_char(eistr, eistr2) \ | |
| 2304 NOT YET IMPLEMENTED | |
| 2305 #define eirstr_ei_off(eistr, eistr2, off, charoff) \ | |
| 2306 NOT YET IMPLEMENTED | |
| 2307 #define eirstr_ei_off_char(eistr, eistr2, off, charoff) \ | |
| 2308 NOT YET IMPLEMENTED | |
| 2309 | |
| 2421 | 2310 #define eistr_ascii(eistr, ascstr) \ |
| 771 | 2311 NOT YET IMPLEMENTED |
| 2421 | 2312 #define eistr_ascii_char(eistr, ascstr) \ |
| 771 | 2313 NOT YET IMPLEMENTED |
| 2421 | 2314 #define eistr_ascii_off(eistr, ascstr, off, charoff) \ |
| 771 | 2315 NOT YET IMPLEMENTED |
| 2421 | 2316 #define eistr_ascii_off_char(eistr, ascstr, off, charoff) \ |
| 771 | 2317 NOT YET IMPLEMENTED |
| 2421 | 2318 #define eirstr_ascii(eistr, ascstr) \ |
| 771 | 2319 NOT YET IMPLEMENTED |
| 2421 | 2320 #define eirstr_ascii_char(eistr, ascstr) \ |
| 771 | 2321 NOT YET IMPLEMENTED |
| 2421 | 2322 #define eirstr_ascii_off(eistr, ascstr, off, charoff) \ |
| 771 | 2323 NOT YET IMPLEMENTED |
| 2421 | 2324 #define eirstr_ascii_off_char(eistr, ascstr, off, charoff) \ |
| 771 | 2325 NOT YET IMPLEMENTED |
| 2326 | |
| 2327 | |
| 2328 /* ----- Comparison ----- */ | |
| 2329 | |
| 2330 int eicmp_1 (Eistring *ei, Bytecount off, Charcount charoff, | |
| 867 | 2331 Bytecount len, Charcount charlen, const Ibyte *data, |
| 2526 | 2332 const Eistring *ei2, int is_ascii, int fold_case); |
| 771 | 2333 |
| 2334 #define eicmp_ei(eistr, eistr2) \ | |
| 2335 eicmp_1 (eistr, 0, -1, -1, -1, 0, eistr2, 0, 0) | |
| 2336 #define eicmp_off_ei(eistr, off, charoff, len, charlen, eistr2) \ | |
| 2337 eicmp_1 (eistr, off, charoff, len, charlen, 0, eistr2, 0, 0) | |
| 2338 #define eicasecmp_ei(eistr, eistr2) \ | |
| 2339 eicmp_1 (eistr, 0, -1, -1, -1, 0, eistr2, 0, 1) | |
| 2340 #define eicasecmp_off_ei(eistr, off, charoff, len, charlen, eistr2) \ | |
| 2341 eicmp_1 (eistr, off, charoff, len, charlen, 0, eistr2, 0, 1) | |
| 2342 #define eicasecmp_i18n_ei(eistr, eistr2) \ | |
| 2343 eicmp_1 (eistr, 0, -1, -1, -1, 0, eistr2, 0, 2) | |
| 2344 #define eicasecmp_i18n_off_ei(eistr, off, charoff, len, charlen, eistr2) \ | |
| 2345 eicmp_1 (eistr, off, charoff, len, charlen, 0, eistr2, 0, 2) | |
| 2346 | |
| 2421 | 2347 #define eicmp_ascii(eistr, ascstr) \ |
| 2348 eicmp_1 (eistr, 0, -1, -1, -1, (const Ibyte *) ascstr, 0, 1, 0) | |
| 2349 #define eicmp_off_ascii(eistr, off, charoff, len, charlen, ascstr) \ | |
| 2350 eicmp_1 (eistr, off, charoff, len, charlen, (const Ibyte *) ascstr, 0, 1, 0) | |
| 2351 #define eicasecmp_ascii(eistr, ascstr) \ | |
| 2352 eicmp_1 (eistr, 0, -1, -1, -1, (const Ibyte *) ascstr, 0, 1, 1) | |
| 2353 #define eicasecmp_off_ascii(eistr, off, charoff, len, charlen, ascstr) \ | |
| 2354 eicmp_1 (eistr, off, charoff, len, charlen, (const Ibyte *) ascstr, 0, 1, 1) | |
| 2355 #define eicasecmp_i18n_ascii(eistr, ascstr) \ | |
| 2356 eicmp_1 (eistr, 0, -1, -1, -1, (const Ibyte *) ascstr, 0, 1, 2) | |
| 2357 #define eicasecmp_i18n_off_ascii(eistr, off, charoff, len, charlen, ascstr) \ | |
| 2358 eicmp_1 (eistr, off, charoff, len, charlen, (const Ibyte *) ascstr, 0, 1, 2) | |
| 771 | 2359 |
| 2360 | |
| 2361 /* ----- Case-changing the Eistring ----- */ | |
| 2362 | |
| 867 | 2363 int eistr_casefiddle_1 (Ibyte *olddata, Bytecount len, Ibyte *newdata, |
| 771 | 2364 int downp); |
| 2365 | |
| 2366 #define EI_CASECHANGE(ei, downp) \ | |
| 2367 do { \ | |
| 867 | 2368 int ei11new_allocmax = (ei)->charlen_ * MAX_ICHAR_LEN + 1; \ |
| 1333 | 2369 Ibyte *ei11storage = \ |
| 2367 | 2370 (Ibyte *) alloca_ibytes (ei11new_allocmax); \ |
| 771 | 2371 int ei11newlen = eistr_casefiddle_1 ((ei)->data_, (ei)->bytelen_, \ |
| 2372 ei11storage, downp); \ | |
| 2373 \ | |
| 2374 if (ei11newlen) \ | |
| 2375 { \ | |
| 2376 (ei)->max_size_allocated_ = ei11new_allocmax; \ | |
| 1333 | 2377 (ei)->data_ = ei11storage; \ |
| 771 | 2378 (ei)->bytelen_ = ei11newlen; \ |
| 2379 /* charlen is the same. */ \ | |
| 2380 } \ | |
| 2381 } while (0) | |
| 2382 | |
| 2383 #define eilwr(ei) EI_CASECHANGE (ei, 1) | |
| 2384 #define eiupr(ei) EI_CASECHANGE (ei, 0) | |
| 2385 | |
| 1743 | 2386 END_C_DECLS |
| 1650 | 2387 |
| 771 | 2388 |
| 2389 /************************************************************************/ | |
| 2390 /* */ | |
| 2391 /* Converting between internal and external format */ | |
| 2392 /* */ | |
| 2393 /************************************************************************/ | |
| 2394 /* | |
| 1318 | 2395 The macros below are used for converting data between different formats. |
| 2396 Generally, the data is textual, and the formats are related to | |
| 2397 internationalization (e.g. converting between internal-format text and | |
| 2398 UTF-8) -- but the mechanism is general, and could be used for anything, | |
| 2399 e.g. decoding gzipped data. | |
| 2400 | |
| 2401 In general, conversion involves a source of data, a sink, the existing | |
| 2402 format of the source data, and the desired format of the sink. The | |
| 2403 macros below, however, always require that either the source or sink is | |
| 2404 internal-format text. Therefore, in practice the conversions below | |
| 2405 involve source, sink, an external format (specified by a coding system), | |
| 2406 and the direction of conversion (internal->external or vice-versa). | |
| 2407 | |
| 2408 Sources and sinks can be raw data (sized or unsized -- when unsized, | |
| 2409 input data is assumed to be null-terminated [double null-terminated for | |
| 2410 Unicode-format data], and on output the length is not stored anywhere), | |
| 2411 Lisp strings, Lisp buffers, lstreams, and opaque data objects. When the | |
| 2412 output is raw data, the result can be allocated either with alloca() or | |
| 2413 malloc(). (There is currently no provision for writing into a fixed | |
| 2414 buffer. If you want this, use alloca() output and then copy the data -- | |
| 2415 but be careful with the size! Unless you are very sure of the encoding | |
| 2416 being used, upper bounds for the size are not in general computable.) | |
| 2417 The obvious restrictions on source and sink types apply (e.g. Lisp | |
| 2418 strings are a source and sink only for internal data). | |
| 2419 | |
| 2420 All raw data outputted will contain an extra null byte (two bytes for | |
| 2421 Unicode -- currently, in fact, all output data, whether internal or | |
| 2422 external, is double-null-terminated, but you can't count on this; see | |
| 2423 below). This means that enough space is allocated to contain the extra | |
| 2424 nulls; however, these nulls are not reflected in the returned output | |
| 2425 size. | |
| 2426 | |
| 2427 The most basic macros are TO_EXTERNAL_FORMAT and TO_INTERNAL_FORMAT. | |
| 2428 These can be used to convert between any kinds of sources or sinks. | |
| 2429 However, 99% of conversions involve raw data or Lisp strings as both | |
| 2430 source and sink, and usually data is output as alloca() rather than | |
| 2431 malloc(). For this reason, convenience macros are defined for many types | |
| 2432 of conversions involving raw data and/or Lisp strings, especially when | |
| 2433 the output is an alloca()ed string. (When the destination is a | |
| 2434 Lisp_String, there are other functions that should be used instead -- | |
|
4953
304aebb79cd3
function renamings to track names of char typedefs
Ben Wing <ben@xemacs.org>
parents:
4952
diff
changeset
|
2435 build_extstring() and make_extstring(), for example.) The convenience |
| 1318 | 2436 macros are of two types -- the older kind that store the result into a |
| 2437 specified variable, and the newer kind that return the result. The newer | |
| 2438 kind of macros don't exist when the output is sized data, because that | |
| 2439 would have two return values. NOTE: All convenience macros are | |
| 2440 ultimately defined in terms of TO_EXTERNAL_FORMAT and TO_INTERNAL_FORMAT. | |
| 2441 Thus, any comments below about the workings of these macros also apply to | |
| 2442 all convenience macros. | |
| 2443 | |
| 2444 TO_EXTERNAL_FORMAT (source_type, source, sink_type, sink, codesys) | |
| 2445 TO_INTERNAL_FORMAT (source_type, source, sink_type, sink, codesys) | |
| 771 | 2446 |
| 2447 Typical use is | |
| 2448 | |
| 2367 | 2449 TO_EXTERNAL_FORMAT (LISP_STRING, str, C_STRING_MALLOC, ptr, Qfile_name); |
| 2450 | |
| 2451 which means that the contents of the lisp string `str' are written | |
| 2452 to a malloc'ed memory area which will be pointed to by `ptr', after the | |
| 2453 function returns. The conversion will be done using the `file-name' | |
| 2454 coding system (which will be controlled by the user indirectly by | |
| 2455 setting or binding the variable `file-name-coding-system'). | |
| 2456 | |
| 2457 Some sources and sinks require two C variables to specify. We use | |
| 2458 some preprocessor magic to allow different source and sink types, and | |
| 2459 even different numbers of arguments to specify different types of | |
| 2460 sources and sinks. | |
| 2461 | |
| 2462 So we can have a call that looks like | |
| 2463 | |
| 2464 TO_INTERNAL_FORMAT (DATA, (ptr, len), | |
| 2465 MALLOC, (ptr, len), | |
| 2466 coding_system); | |
| 2467 | |
| 2468 The parenthesized argument pairs are required to make the | |
| 2469 preprocessor magic work. | |
| 771 | 2470 |
| 2471 NOTE: GC is inhibited during the entire operation of these macros. This | |
| 2472 is because frequently the data to be converted comes from strings but | |
| 2473 gets passed in as just DATA, and GC may move around the string data. If | |
| 2474 we didn't inhibit GC, there'd have to be a lot of messy recoding, | |
| 2475 alloca-copying of strings and other annoying stuff. | |
| 2476 | |
| 2477 The source or sink can be specified in one of these ways: | |
| 2478 | |
| 2479 DATA, (ptr, len), // input data is a fixed buffer of size len | |
| 851 | 2480 ALLOCA, (ptr, len), // output data is in a ALLOCA()ed buffer of size len |
| 771 | 2481 MALLOC, (ptr, len), // output data is in a malloc()ed buffer of size len |
| 2482 C_STRING_ALLOCA, ptr, // equivalent to ALLOCA (ptr, len_ignored) on output | |
| 2483 C_STRING_MALLOC, ptr, // equivalent to MALLOC (ptr, len_ignored) on output | |
| 2484 C_STRING, ptr, // equivalent to DATA, (ptr, strlen/wcslen (ptr)) | |
| 2485 // on input (the Unicode version is used when correct) | |
| 2486 LISP_STRING, string, // input or output is a Lisp_Object of type string | |
| 2487 LISP_BUFFER, buffer, // output is written to (point) in lisp buffer | |
| 2488 LISP_LSTREAM, lstream, // input or output is a Lisp_Object of type lstream | |
| 2489 LISP_OPAQUE, object, // input or output is a Lisp_Object of type opaque | |
| 2490 | |
| 2491 When specifying the sink, use lvalues, since the macro will assign to them, | |
| 2492 except when the sink is an lstream or a lisp buffer. | |
| 2493 | |
| 2367 | 2494 For the sink types `ALLOCA' and `C_STRING_ALLOCA', the resulting text is |
| 2495 stored in a stack-allocated buffer, which is automatically freed on | |
| 2496 returning from the function. However, the sink types `MALLOC' and | |
| 2497 `C_STRING_MALLOC' return `xmalloc()'ed memory. The caller is responsible | |
| 2498 for freeing this memory using `xfree()'. | |
| 2499 | |
| 771 | 2500 The macros accept the kinds of sources and sinks appropriate for |
| 2501 internal and external data representation. See the type_checking_assert | |
| 2502 macros below for the actual allowed types. | |
| 2503 | |
| 2504 Since some sources and sinks use one argument (a Lisp_Object) to | |
| 2505 specify them, while others take a (pointer, length) pair, we use | |
| 2506 some C preprocessor trickery to allow pair arguments to be specified | |
| 2507 by parenthesizing them, as in the examples above. | |
| 2508 | |
| 2509 Anything prefixed by dfc_ (`data format conversion') is private. | |
| 2510 They are only used to implement these macros. | |
| 2511 | |
| 2512 [[Using C_STRING* is appropriate for using with external APIs that | |
| 2513 take null-terminated strings. For internal data, we should try to | |
| 2514 be '\0'-clean - i.e. allow arbitrary data to contain embedded '\0'. | |
| 2515 | |
| 2516 Sometime in the future we might allow output to C_STRING_ALLOCA or | |
| 2517 C_STRING_MALLOC _only_ with TO_EXTERNAL_FORMAT(), not | |
| 2518 TO_INTERNAL_FORMAT().]] | |
| 2519 | |
| 2520 The above comments are not true. Frequently (most of the time, in | |
| 2521 fact), external strings come as zero-terminated entities, where the | |
| 2522 zero-termination is the only way to find out the length. Even in | |
| 2523 cases where you can get the length, most of the time the system will | |
| 2524 still use the null to signal the end of the string, and there will | |
| 2525 still be no way to either send in or receive a string with embedded | |
| 2526 nulls. In such situations, it's pointless to track the length | |
| 2527 because null bytes can never be in the string. We have a lot of | |
| 2528 operations that make it easy to operate on zero-terminated strings, | |
| 2529 and forcing the user the deal with the length everywhere would only | |
| 2530 make the code uglier and more complicated, for no gain. --ben | |
| 2531 | |
| 2532 There is no problem using the same lvalue for source and sink. | |
| 2533 | |
| 2534 Also, when pointers are required, the code (currently at least) is | |
| 2535 lax and allows any pointer types, either in the source or the sink. | |
| 2536 This makes it possible, e.g., to deal with internal format data held | |
| 2537 in char *'s or external format data held in WCHAR * (i.e. Unicode). | |
| 2538 | |
| 2539 Finally, whenever storage allocation is called for, extra space is | |
| 2540 allocated for a terminating zero, and such a zero is stored in the | |
| 2541 appropriate place, regardless of whether the source data was | |
| 2542 specified using a length or was specified as zero-terminated. This | |
| 2543 allows you to freely pass the resulting data, no matter how | |
| 2544 obtained, to a routine that expects zero termination (modulo, of | |
| 2545 course, that any embedded zeros in the resulting text will cause | |
| 2546 truncation). In fact, currently two embedded zeros are allocated | |
| 2547 and stored after the data result. This is to allow for the | |
| 2548 possibility of storing a Unicode value on output, which needs the | |
| 2549 two zeros. Currently, however, the two zeros are stored regardless | |
| 2550 of whether the conversion is internal or external and regardless of | |
| 2551 whether the external coding system is in fact Unicode. This | |
| 2552 behavior may change in the future, and you cannot rely on this -- | |
| 2553 the most you can rely on is that sink data in Unicode format will | |
| 2554 have two terminating nulls, which combine to form one Unicode null | |
| 2367 | 2555 character. |
| 2556 | |
| 2557 NOTE: You might ask, why are these not written as functions that | |
| 2558 *RETURN* the converted string, since that would allow them to be used | |
| 2559 much more conveniently, without having to constantly declare temporary | |
| 2560 variables? The answer is that in fact I originally did write the | |
| 2561 routines that way, but that required either | |
| 2562 | |
| 2563 (a) calling alloca() inside of a function call, or | |
| 2564 (b) using expressions separated by commas and a global temporary variable, or | |
| 2565 (c) using the GCC extension ({ ... }). | |
| 2566 | |
| 2567 Turned out that all of the above had bugs, all caused by GCC (hence the | |
| 2568 comments about "those GCC wankers" and "ream gcc up the ass"). As for | |
| 2569 (a), some versions of GCC (especially on Intel platforms), which had | |
| 2570 buggy implementations of alloca() that couldn't handle being called | |
| 2571 inside of a function call -- they just decremented the stack right in the | |
| 2572 middle of pushing args. Oops, crash with stack trashing, very bad. (b) | |
| 2573 was an attempt to fix (a), and that led to further GCC crashes, esp. when | |
| 2574 you had two such calls in a single subexpression, because GCC couldn't be | |
| 2575 counted upon to follow even a minimally reasonable order of execution. | |
| 2576 True, you can't count on one argument being evaluated before another, but | |
| 2577 GCC would actually interleave them so that the temp var got stomped on by | |
| 2578 one while the other was accessing it. So I tried (c), which was | |
| 2579 problematic because that GCC extension has more bugs in it than a | |
| 2580 termite's nest. | |
| 2581 | |
| 2582 So reluctantly I converted to the current way. Now, that was awhile ago | |
| 2583 (c. 1994), and it appears that the bug involving alloca in function calls | |
| 2584 has long since been fixed. More recently, I defined the new-dfc routines | |
| 2585 down below, which DO allow exactly such convenience of returning your | |
| 2586 args rather than store them in temp variables, and I also wrote a | |
| 2587 configure check to see whether alloca() causes crashes inside of function | |
| 2588 calls, and if so use the portable alloca() implementation in alloca.c. | |
| 2589 If you define TEST_NEW_DFC, the old routines get written in terms of the | |
| 2590 new ones, and I've had a beta put out with this on and it appeared to | |
| 2591 this appears to cause no problems -- so we should consider | |
| 2592 switching, and feel no compunctions about writing further such function- | |
| 2593 like alloca() routines in lieu of statement-like ones. --ben */ | |
| 771 | 2594 |
| 2595 #define TO_EXTERNAL_FORMAT(source_type, source, sink_type, sink, codesys) \ | |
| 2596 do { \ | |
| 2597 dfc_conversion_type dfc_simplified_source_type; \ | |
| 2598 dfc_conversion_type dfc_simplified_sink_type; \ | |
| 2599 dfc_conversion_data dfc_source; \ | |
| 2600 dfc_conversion_data dfc_sink; \ | |
| 2601 Lisp_Object dfc_codesys = (codesys); \ | |
| 2602 \ | |
| 2603 type_checking_assert \ | |
| 2604 ((DFC_TYPE_##source_type == DFC_TYPE_DATA || \ | |
| 2605 DFC_TYPE_##source_type == DFC_TYPE_C_STRING || \ | |
| 2606 DFC_TYPE_##source_type == DFC_TYPE_LISP_STRING || \ | |
| 2607 DFC_TYPE_##source_type == DFC_TYPE_LISP_OPAQUE || \ | |
| 2608 DFC_TYPE_##source_type == DFC_TYPE_LISP_LSTREAM) \ | |
| 2609 && \ | |
| 2610 (DFC_TYPE_##sink_type == DFC_TYPE_ALLOCA || \ | |
| 2611 DFC_TYPE_##sink_type == DFC_TYPE_MALLOC || \ | |
| 2612 DFC_TYPE_##sink_type == DFC_TYPE_C_STRING_ALLOCA || \ | |
| 2613 DFC_TYPE_##sink_type == DFC_TYPE_C_STRING_MALLOC || \ | |
| 2614 DFC_TYPE_##sink_type == DFC_TYPE_LISP_LSTREAM || \ | |
| 2615 DFC_TYPE_##sink_type == DFC_TYPE_LISP_OPAQUE)); \ | |
| 2616 \ | |
| 2617 DFC_EXT_SOURCE_##source_type##_TO_ARGS (source, dfc_codesys); \ | |
| 2618 DFC_SINK_##sink_type##_TO_ARGS (sink); \ | |
| 2619 \ | |
| 2620 dfc_convert_to_external_format (dfc_simplified_source_type, &dfc_source, \ | |
| 2621 dfc_codesys, \ | |
| 2622 dfc_simplified_sink_type, &dfc_sink); \ | |
| 2623 \ | |
| 2624 DFC_##sink_type##_USE_CONVERTED_DATA (sink); \ | |
| 2625 } while (0) | |
| 2626 | |
| 2627 #define TO_INTERNAL_FORMAT(source_type, source, sink_type, sink, codesys) \ | |
| 2628 do { \ | |
| 2629 dfc_conversion_type dfc_simplified_source_type; \ | |
| 2630 dfc_conversion_type dfc_simplified_sink_type; \ | |
| 2631 dfc_conversion_data dfc_source; \ | |
| 2632 dfc_conversion_data dfc_sink; \ | |
| 2633 Lisp_Object dfc_codesys = (codesys); \ | |
| 2634 \ | |
| 2635 type_checking_assert \ | |
| 2636 ((DFC_TYPE_##source_type == DFC_TYPE_DATA || \ | |
| 2637 DFC_TYPE_##source_type == DFC_TYPE_C_STRING || \ | |
| 2638 DFC_TYPE_##source_type == DFC_TYPE_LISP_OPAQUE || \ | |
| 2639 DFC_TYPE_##source_type == DFC_TYPE_LISP_LSTREAM) \ | |
| 2640 && \ | |
| 2641 (DFC_TYPE_##sink_type == DFC_TYPE_ALLOCA || \ | |
| 2642 DFC_TYPE_##sink_type == DFC_TYPE_MALLOC || \ | |
| 2643 DFC_TYPE_##sink_type == DFC_TYPE_C_STRING_ALLOCA || \ | |
| 2644 DFC_TYPE_##sink_type == DFC_TYPE_C_STRING_MALLOC || \ | |
| 2645 DFC_TYPE_##sink_type == DFC_TYPE_LISP_STRING || \ | |
| 2646 DFC_TYPE_##sink_type == DFC_TYPE_LISP_LSTREAM || \ | |
| 2647 DFC_TYPE_##sink_type == DFC_TYPE_LISP_BUFFER)); \ | |
| 2648 \ | |
| 2649 DFC_INT_SOURCE_##source_type##_TO_ARGS (source, dfc_codesys); \ | |
| 2650 DFC_SINK_##sink_type##_TO_ARGS (sink); \ | |
| 2651 \ | |
| 2652 dfc_convert_to_internal_format (dfc_simplified_source_type, &dfc_source, \ | |
| 2653 dfc_codesys, \ | |
| 2654 dfc_simplified_sink_type, &dfc_sink); \ | |
| 2655 \ | |
| 2656 DFC_##sink_type##_USE_CONVERTED_DATA (sink); \ | |
| 2657 } while (0) | |
| 2658 | |
| 814 | 2659 #ifdef __cplusplus |
| 771 | 2660 |
| 814 | 2661 /* Error if you try to use a union here: "member `struct {anonymous |
| 2662 union}::{anonymous} {anonymous union}::data' with constructor not allowed | |
| 2663 in union" (Bytecount is a class) */ | |
| 2664 | |
| 2665 typedef struct | |
| 2666 #else | |
| 771 | 2667 typedef union |
| 814 | 2668 #endif |
| 771 | 2669 { |
| 2670 struct { const void *ptr; Bytecount len; } data; | |
| 2671 Lisp_Object lisp_object; | |
| 2672 } dfc_conversion_data; | |
| 2673 | |
| 2674 enum dfc_conversion_type | |
| 2675 { | |
| 2676 DFC_TYPE_DATA, | |
| 2677 DFC_TYPE_ALLOCA, | |
| 2678 DFC_TYPE_MALLOC, | |
| 2679 DFC_TYPE_C_STRING, | |
| 2680 DFC_TYPE_C_STRING_ALLOCA, | |
| 2681 DFC_TYPE_C_STRING_MALLOC, | |
| 2682 DFC_TYPE_LISP_STRING, | |
| 2683 DFC_TYPE_LISP_LSTREAM, | |
| 2684 DFC_TYPE_LISP_OPAQUE, | |
| 2685 DFC_TYPE_LISP_BUFFER | |
| 2686 }; | |
| 2687 typedef enum dfc_conversion_type dfc_conversion_type; | |
| 2688 | |
| 1743 | 2689 BEGIN_C_DECLS |
| 1650 | 2690 |
| 771 | 2691 /* WARNING: These use a static buffer. This can lead to disaster if |
| 2692 these functions are not used *very* carefully. Another reason to only use | |
| 2693 TO_EXTERNAL_FORMAT() and TO_INTERNAL_FORMAT(). */ | |
| 1632 | 2694 MODULE_API void |
| 771 | 2695 dfc_convert_to_external_format (dfc_conversion_type source_type, |
| 2696 dfc_conversion_data *source, | |
| 1318 | 2697 Lisp_Object codesys, |
| 771 | 2698 dfc_conversion_type sink_type, |
| 2699 dfc_conversion_data *sink); | |
| 1632 | 2700 MODULE_API void |
| 771 | 2701 dfc_convert_to_internal_format (dfc_conversion_type source_type, |
| 2702 dfc_conversion_data *source, | |
| 1318 | 2703 Lisp_Object codesys, |
| 771 | 2704 dfc_conversion_type sink_type, |
| 2705 dfc_conversion_data *sink); | |
| 2706 /* CPP Trickery */ | |
| 2707 #define DFC_CPP_CAR(x,y) (x) | |
| 2708 #define DFC_CPP_CDR(x,y) (y) | |
| 2709 | |
| 2710 /* Convert `source' to args for dfc_convert_to_external_format() */ | |
| 2711 #define DFC_EXT_SOURCE_DATA_TO_ARGS(val, codesys) do { \ | |
| 2712 dfc_source.data.ptr = DFC_CPP_CAR val; \ | |
| 2713 dfc_source.data.len = DFC_CPP_CDR val; \ | |
| 2714 dfc_simplified_source_type = DFC_TYPE_DATA; \ | |
| 2715 } while (0) | |
| 2716 #define DFC_EXT_SOURCE_C_STRING_TO_ARGS(val, codesys) do { \ | |
| 2717 dfc_source.data.len = \ | |
| 2718 strlen ((char *) (dfc_source.data.ptr = (val))); \ | |
| 2719 dfc_simplified_source_type = DFC_TYPE_DATA; \ | |
| 2720 } while (0) | |
| 2721 #define DFC_EXT_SOURCE_LISP_STRING_TO_ARGS(val, codesys) do { \ | |
| 2722 Lisp_Object dfc_slsta = (val); \ | |
| 2723 type_checking_assert (STRINGP (dfc_slsta)); \ | |
| 2724 dfc_source.lisp_object = dfc_slsta; \ | |
| 2725 dfc_simplified_source_type = DFC_TYPE_LISP_STRING; \ | |
| 2726 } while (0) | |
| 2727 #define DFC_EXT_SOURCE_LISP_LSTREAM_TO_ARGS(val, codesys) do { \ | |
| 2728 Lisp_Object dfc_sllta = (val); \ | |
| 2729 type_checking_assert (LSTREAMP (dfc_sllta)); \ | |
| 2730 dfc_source.lisp_object = dfc_sllta; \ | |
| 2731 dfc_simplified_source_type = DFC_TYPE_LISP_LSTREAM; \ | |
| 2732 } while (0) | |
| 2733 #define DFC_EXT_SOURCE_LISP_OPAQUE_TO_ARGS(val, codesys) do { \ | |
| 2734 Lisp_Opaque *dfc_slota = XOPAQUE (val); \ | |
| 2735 dfc_source.data.ptr = OPAQUE_DATA (dfc_slota); \ | |
| 2736 dfc_source.data.len = OPAQUE_SIZE (dfc_slota); \ | |
| 2737 dfc_simplified_source_type = DFC_TYPE_DATA; \ | |
| 2738 } while (0) | |
| 2739 | |
| 2740 /* Convert `source' to args for dfc_convert_to_internal_format() */ | |
| 2741 #define DFC_INT_SOURCE_DATA_TO_ARGS(val, codesys) \ | |
| 2742 DFC_EXT_SOURCE_DATA_TO_ARGS (val, codesys) | |
| 2743 #define DFC_INT_SOURCE_C_STRING_TO_ARGS(val, codesys) do { \ | |
| 2744 dfc_source.data.len = dfc_external_data_len (dfc_source.data.ptr = (val), \ | |
| 2745 codesys); \ | |
| 2746 dfc_simplified_source_type = DFC_TYPE_DATA; \ | |
| 2747 } while (0) | |
| 2748 #define DFC_INT_SOURCE_LISP_STRING_TO_ARGS(val, codesys) \ | |
| 2749 DFC_EXT_SOURCE_LISP_STRING_TO_ARGS (val, codesys) | |
| 2750 #define DFC_INT_SOURCE_LISP_LSTREAM_TO_ARGS(val, codesys) \ | |
| 2751 DFC_EXT_SOURCE_LISP_LSTREAM_TO_ARGS (val, codesys) | |
| 2752 #define DFC_INT_SOURCE_LISP_OPAQUE_TO_ARGS(val, codesys) \ | |
| 2753 DFC_EXT_SOURCE_LISP_OPAQUE_TO_ARGS (val, codesys) | |
| 2754 | |
| 2755 /* Convert `sink' to args for dfc_convert_to_*_format() */ | |
| 2756 #define DFC_SINK_ALLOCA_TO_ARGS(val) \ | |
| 2757 dfc_simplified_sink_type = DFC_TYPE_DATA | |
| 2758 #define DFC_SINK_C_STRING_ALLOCA_TO_ARGS(val) \ | |
| 2759 dfc_simplified_sink_type = DFC_TYPE_DATA | |
| 2760 #define DFC_SINK_MALLOC_TO_ARGS(val) \ | |
| 2761 dfc_simplified_sink_type = DFC_TYPE_DATA | |
| 2762 #define DFC_SINK_C_STRING_MALLOC_TO_ARGS(val) \ | |
| 2763 dfc_simplified_sink_type = DFC_TYPE_DATA | |
| 2764 #define DFC_SINK_LISP_STRING_TO_ARGS(val) \ | |
| 2765 dfc_simplified_sink_type = DFC_TYPE_DATA | |
| 2766 #define DFC_SINK_LISP_OPAQUE_TO_ARGS(val) \ | |
| 2767 dfc_simplified_sink_type = DFC_TYPE_DATA | |
| 2768 #define DFC_SINK_LISP_LSTREAM_TO_ARGS(val) do { \ | |
| 2769 Lisp_Object dfc_sllta = (val); \ | |
| 2770 type_checking_assert (LSTREAMP (dfc_sllta)); \ | |
| 2771 dfc_sink.lisp_object = dfc_sllta; \ | |
| 2772 dfc_simplified_sink_type = DFC_TYPE_LISP_LSTREAM; \ | |
| 2773 } while (0) | |
| 2774 #define DFC_SINK_LISP_BUFFER_TO_ARGS(val) do { \ | |
| 2775 struct buffer *dfc_slbta = XBUFFER (val); \ | |
| 2776 dfc_sink.lisp_object = \ | |
| 2777 make_lisp_buffer_output_stream \ | |
| 2778 (dfc_slbta, BUF_PT (dfc_slbta), 0); \ | |
| 2779 dfc_simplified_sink_type = DFC_TYPE_LISP_LSTREAM; \ | |
| 2780 } while (0) | |
| 2781 | |
| 2782 /* Assign to the `sink' lvalue(s) using the converted data. */ | |
| 2783 /* + 2 because we double zero-extended to account for Unicode conversion */ | |
| 2784 typedef union { char c; void *p; } *dfc_aliasing_voidpp; | |
| 2785 #define DFC_ALLOCA_USE_CONVERTED_DATA(sink) do { \ | |
| 851 | 2786 void * dfc_sink_ret = ALLOCA (dfc_sink.data.len + 2); \ |
| 771 | 2787 memcpy (dfc_sink_ret, dfc_sink.data.ptr, dfc_sink.data.len + 2); \ |
| 2367 | 2788 VOIDP_CAST (DFC_CPP_CAR sink) = dfc_sink_ret; \ |
| 771 | 2789 (DFC_CPP_CDR sink) = dfc_sink.data.len; \ |
| 2790 } while (0) | |
| 2791 #define DFC_MALLOC_USE_CONVERTED_DATA(sink) do { \ | |
| 2792 void * dfc_sink_ret = xmalloc (dfc_sink.data.len + 2); \ | |
| 2793 memcpy (dfc_sink_ret, dfc_sink.data.ptr, dfc_sink.data.len + 2); \ | |
| 2367 | 2794 VOIDP_CAST (DFC_CPP_CAR sink) = dfc_sink_ret; \ |
| 771 | 2795 (DFC_CPP_CDR sink) = dfc_sink.data.len; \ |
| 2796 } while (0) | |
| 2797 #define DFC_C_STRING_ALLOCA_USE_CONVERTED_DATA(sink) do { \ | |
| 851 | 2798 void * dfc_sink_ret = ALLOCA (dfc_sink.data.len + 2); \ |
| 771 | 2799 memcpy (dfc_sink_ret, dfc_sink.data.ptr, dfc_sink.data.len + 2); \ |
| 2367 | 2800 VOIDP_CAST (sink) = dfc_sink_ret; \ |
| 771 | 2801 } while (0) |
| 2802 #define DFC_C_STRING_MALLOC_USE_CONVERTED_DATA(sink) do { \ | |
| 2803 void * dfc_sink_ret = xmalloc (dfc_sink.data.len + 2); \ | |
| 2804 memcpy (dfc_sink_ret, dfc_sink.data.ptr, dfc_sink.data.len + 2); \ | |
| 2367 | 2805 VOIDP_CAST (sink) = dfc_sink_ret; \ |
| 771 | 2806 } while (0) |
| 2807 #define DFC_LISP_STRING_USE_CONVERTED_DATA(sink) \ | |
| 867 | 2808 sink = make_string ((Ibyte *) dfc_sink.data.ptr, dfc_sink.data.len) |
| 771 | 2809 #define DFC_LISP_OPAQUE_USE_CONVERTED_DATA(sink) \ |
| 2810 sink = make_opaque (dfc_sink.data.ptr, dfc_sink.data.len) | |
| 2811 #define DFC_LISP_LSTREAM_USE_CONVERTED_DATA(sink) /* data already used */ | |
| 2812 #define DFC_LISP_BUFFER_USE_CONVERTED_DATA(sink) \ | |
| 2813 Lstream_delete (XLSTREAM (dfc_sink.lisp_object)) | |
| 2814 | |
| 1318 | 2815 enum new_dfc_src_type |
| 2816 { | |
| 2817 DFC_EXTERNAL, | |
| 2818 DFC_SIZED_EXTERNAL, | |
| 2819 DFC_INTERNAL, | |
| 2820 DFC_SIZED_INTERNAL, | |
| 2821 DFC_LISP_STRING | |
| 2822 }; | |
| 2823 | |
| 1632 | 2824 MODULE_API void *new_dfc_convert_malloc (const void *src, Bytecount src_size, |
| 2825 enum new_dfc_src_type type, | |
| 2826 Lisp_Object codesys); | |
| 2367 | 2827 MODULE_API Bytecount new_dfc_convert_size (const char *srctext, |
| 2828 const void *src, | |
| 1632 | 2829 Bytecount src_size, |
| 2830 enum new_dfc_src_type type, | |
| 2831 Lisp_Object codesys); | |
| 2367 | 2832 MODULE_API void *new_dfc_convert_copy_data (const char *srctext, |
| 2833 void *alloca_data); | |
| 1318 | 2834 |
| 1743 | 2835 END_C_DECLS |
| 1650 | 2836 |
|
4981
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2837 /* Version of EXTERNAL_TO_ITEXT that *RETURNS* the translated string, |
| 1318 | 2838 still in alloca() space. Requires some trickiness to do this, but gets |
| 2839 it done! */ | |
| 2840 | |
| 2841 /* NOTE: If you make two invocations of the dfc functions below in the same | |
| 2842 subexpression and use the exact same expression for the source in both | |
| 2843 cases, you will lose. In this unlikely case, you will get an abort, and | |
| 2844 need to rewrite the code. | |
| 2845 */ | |
| 2846 | |
| 2847 /* We need to use ALLOCA_FUNCALL_OK here. Some compilers have been known | |
| 2848 to choke when alloca() occurs as a funcall argument, and so we check | |
| 2849 this in configure. Rewriting the expressions below to use a temporary | |
| 2850 variable, so that the call to alloca() is outside of | |
| 2382 | 2851 new_dfc_convert_copy_data(), won't help because the entire NEW_DFC call |
| 1318 | 2852 could be inside of a function call. */ |
| 2853 | |
| 2854 #define NEW_DFC_CONVERT_1_ALLOCA(src, src_size, type, codesys) \ | |
| 2367 | 2855 new_dfc_convert_copy_data \ |
| 1318 | 2856 (#src, ALLOCA_FUNCALL_OK (new_dfc_convert_size (#src, src, src_size, \ |
| 2857 type, codesys))) | |
| 2858 | |
|
4981
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2859 #define EXTERNAL_TO_ITEXT(src, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2860 ((Ibyte *) NEW_DFC_CONVERT_1_ALLOCA (src, -1, DFC_EXTERNAL, codesys)) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2861 #define EXTERNAL_TO_ITEXT_MALLOC(src, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2862 ((Ibyte *) new_dfc_convert_malloc (src, -1, DFC_EXTERNAL, codesys)) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2863 #define SIZED_EXTERNAL_TO_ITEXT(src, len, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2864 ((Ibyte *) NEW_DFC_CONVERT_1_ALLOCA (src, len, DFC_SIZED_EXTERNAL, codesys)) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2865 #define SIZED_EXTERNAL_TO_ITEXT_MALLOC(src, len, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2866 ((Ibyte *) new_dfc_convert_malloc (src, len, DFC_SIZED_EXTERNAL, codesys)) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2867 #define ITEXT_TO_EXTERNAL(src, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2868 ((Extbyte *) NEW_DFC_CONVERT_1_ALLOCA (src, -1, DFC_INTERNAL, codesys)) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2869 #define ITEXT_TO_EXTERNAL_MALLOC(src, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2870 ((Extbyte *) new_dfc_convert_malloc (src, -1, DFC_INTERNAL, codesys)) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2871 #define LISP_STRING_TO_EXTERNAL(src, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2872 ((Extbyte *) NEW_DFC_CONVERT_1_ALLOCA (LISP_TO_VOID (src), -1, \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2873 DFC_LISP_STRING, codesys)) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2874 #define LISP_STRING_TO_EXTERNAL_MALLOC(src, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2875 ((Extbyte *) new_dfc_convert_malloc (LISP_TO_VOID (src), -1, \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2876 DFC_LISP_STRING, codesys)) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2877 /* In place of EXTERNAL_TO_LISP_STRING(), use build_extstring() and/or |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2878 make_extstring(). */ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2879 |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2880 /* The next four have two outputs, so we make both of them be parameters */ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2881 #define ITEXT_TO_SIZED_EXTERNAL(in, out, outlen, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2882 TO_EXTERNAL_FORMAT (C_STRING, in, ALLOCA, (out, outlen), codesys) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2883 #define LISP_STRING_TO_SIZED_EXTERNAL(in, out, outlen, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2884 TO_EXTERNAL_FORMAT (LISP_STRING, in, ALLOCA, (out, outlen), codesys) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2885 #define ITEXT_TO_SIZED_EXTERNAL_MALLOC(in, out, outlen, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2886 TO_EXTERNAL_FORMAT (C_STRING, in, MALLOC, (out, outlen), codesys) |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2887 #define LISP_STRING_TO_SIZED_EXTERNAL_MALLOC(in, out, outlen, codesys) \ |
|
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
2888 TO_EXTERNAL_FORMAT (LISP_STRING, in, MALLOC, (out, outlen), codesys) |
| 771 | 2889 |
| 2367 | 2890 /* Wexttext functions. The type of Wexttext is selected at compile time |
| 2891 and will sometimes be wchar_t, sometimes char. */ | |
| 2892 | |
| 2893 int wcscmp_ascii (const wchar_t *s1, const Ascbyte *s2); | |
| 2894 int wcsncmp_ascii (const wchar_t *s1, const Ascbyte *s2, Charcount len); | |
| 2895 | |
| 2896 #ifdef WEXTTEXT_IS_WIDE /* defined under MS Windows i.e. WIN32_NATIVE */ | |
| 2897 #define WEXTTEXT_ZTERM_SIZE sizeof (wchar_t) | |
| 2898 /* Extra indirection needed in case of manifest constant as arg */ | |
| 2899 #define WEXTSTRING_1(arg) L##arg | |
| 2900 #define WEXTSTRING(arg) WEXTSTRING_1(arg) | |
| 2901 #define wext_strlen wcslen | |
| 2902 #define wext_strcmp wcscmp | |
| 2903 #define wext_strncmp wcsncmp | |
| 2904 #define wext_strcmp_ascii wcscmp_ascii | |
| 2905 #define wext_strncmp_ascii wcsncmp_ascii | |
| 2906 #define wext_strcpy wcscpy | |
| 2907 #define wext_strncpy wcsncpy | |
| 2908 #define wext_strchr wcschr | |
| 2909 #define wext_strrchr wcsrchr | |
| 2910 #define wext_strdup wcsdup | |
| 2911 #define wext_atol(str) wcstol (str, 0, 10) | |
| 2912 #define wext_sprintf wsprintfW /* Huh? both wsprintfA and wsprintfW? */ | |
| 2913 #define wext_getenv _wgetenv | |
|
4953
304aebb79cd3
function renamings to track names of char typedefs
Ben Wing <ben@xemacs.org>
parents:
4952
diff
changeset
|
2914 #define build_wext_string(str, cs) build_extstring ((Extbyte *) str, cs) |
| 2367 | 2915 #define WEXTTEXT_TO_8_BIT(arg) WEXTTEXT_TO_MULTIBYTE(arg) |
| 2916 #ifdef WIN32_NATIVE | |
| 2917 int XCDECL wext_retry_open (const Wexttext *path, int oflag, ...); | |
| 2918 #else | |
| 2919 #error Cannot handle Wexttext yet on this system | |
| 2920 #endif | |
| 2921 #define wext_access _waccess | |
| 2922 #define wext_stat _wstat | |
| 2923 #else | |
| 2924 #define WEXTTEXT_ZTERM_SIZE sizeof (char) | |
| 2925 #define WEXTSTRING(arg) arg | |
| 2926 #define wext_strlen strlen | |
| 2927 #define wext_strcmp strcmp | |
| 2928 #define wext_strncmp strncmp | |
| 2929 #define wext_strcmp_ascii strcmp | |
| 2930 #define wext_strncmp_ascii strncmp | |
| 2931 #define wext_strcpy strcpy | |
| 2932 #define wext_strncpy strncpy | |
| 2933 #define wext_strchr strchr | |
| 2934 #define wext_strrchr strrchr | |
| 2935 #define wext_strdup xstrdup | |
| 2936 #define wext_atol(str) atol (str) | |
| 2937 #define wext_sprintf sprintf | |
| 2938 #define wext_getenv getenv | |
|
4953
304aebb79cd3
function renamings to track names of char typedefs
Ben Wing <ben@xemacs.org>
parents:
4952
diff
changeset
|
2939 #define build_wext_string build_extstring |
| 2367 | 2940 #define wext_retry_open retry_open |
| 2941 #define wext_access access | |
| 2942 #define wext_stat stat | |
| 2943 #define WEXTTEXT_TO_8_BIT(arg) ((Extbyte *) arg) | |
| 2944 #endif | |
| 2945 | |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
2946 /* Standins for various encodings. |
| 1318 | 2947 |
| 2948 About encodings in X: | |
| 2949 | |
| 2950 X works with 5 different encodings: | |
| 2951 | |
| 2952 -- "Host Portable Character Encoding" == printable ASCII + space, tab, | |
| 2953 newline | |
| 2954 | |
| 2955 -- STRING encoding == ASCII + Latin-1 + tab, newline | |
| 2956 | |
| 2957 -- Locale-specific encoding | |
| 2958 | |
| 2959 -- Compound text == STRING encoding + ISO-2022 escape sequences to | |
| 2960 switch between different locale-specific encodings. | |
| 2961 | |
| 2962 -- ANSI C wide-character encoding | |
| 2963 | |
| 2964 The Host Portable Character Encoding (HPCE) is used for atom names, font | |
| 2965 names, color names, keysyms, geometry strings, resource manager quarks, | |
| 2966 display names, locale names, and various other things. When describing | |
| 2967 such strings, the X manual typically says "If the ... is not in the Host | |
| 2968 Portable Character Encoding, the result is implementation dependent." | |
| 2969 | |
| 2970 The wide-character encoding is used only in the Xwc* functions, which | |
| 2971 are provided as equivalents to Xmb* functions. | |
| 2972 | |
| 2973 STRING and compound text are used in the value of string properties and | |
| 2974 selection data, both of which are values with an associated type atom, | |
| 2975 which can be STRING or COMPOUND_TEXT. It can also be a locale name, as | |
| 2976 specified in setlocale() (#### as usual, there is no normalization | |
| 2977 whatsoever of these names). | |
| 2978 | |
| 2979 X also defines a type called "TEXT", which is used only as a requested | |
| 2980 type, and produces data in a type "convenient to the owner". However, | |
| 2981 there is some indication that X expects this to be the locale-specific | |
| 2982 encoding. | |
| 2983 | |
| 2984 According to the glossary, the locale is used in | |
| 2985 | |
| 2986 -- Encoding and processing of input method text | |
| 2987 -- Encoding of resource files and values | |
| 2988 -- Encoding and imaging of text strings | |
| 2989 -- Encoding and decoding for inter-client text communication | |
| 2990 | |
| 2991 The functions XmbTextListToTextProperty and XmbTextPropertyToTextList | |
| 2992 (and Xwc* equivalents) can be used to convert between the | |
| 2993 locale-specific encoding (XTextStyle), STRING (XStringStyle), and | |
| 2994 compound text (XCompoundTextStyle), as well as XStdICCTextStyle, which | |
| 2995 converts to STRING if possible, and if not, COMPOUND_TEXT. This is | |
| 2996 used, for example, in XmbSetWMProperties, in the window_name and | |
| 2997 icon_name properties (WM_NAME and WM_ICON_NAME), which are in the | |
| 2998 locale-specific encoding on input, and are stored as STRING if possible, | |
| 2999 COMPOUND_TEXT otherwise. | |
| 3000 */ | |
| 771 | 3001 |
|
4952
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3002 #ifdef WEXTTEXT_IS_WIDE |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3003 #define Qcommand_argument_encoding Qmswindows_unicode |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3004 #define Qenvironment_variable_encoding Qmswindows_unicode |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3005 #else |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3006 #define Qcommand_argument_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3007 #define Qenvironment_variable_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3008 #endif |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3009 #define Qunix_host_name_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3010 #define Qunix_service_name_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3011 #define Qtime_function_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3012 #define Qtime_zone_encoding Qtime_function_encoding |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3013 #define Qmswindows_host_name_encoding Qmswindows_multibyte |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3014 #define Qmswindows_service_name_encoding Qmswindows_multibyte |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3015 #define Quser_name_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3016 #define Qerror_message_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3017 #define Qjpeg_error_message_encoding Qerror_message_encoding |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3018 #define Qtooltalk_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3019 #define Qgtk_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3020 |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3021 #define Qdll_symbol_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3022 #define Qdll_function_name_encoding Qdll_symbol_encoding |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3023 #define Qdll_variable_name_encoding Qdll_symbol_encoding |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3024 #define Qdll_filename_encoding Qfile_name |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3025 #define Qemodule_string_encoding Qnative |
|
19a72041c5ed
Mule-izing, various fixes related to char * arguments
Ben Wing <ben@xemacs.org>
parents:
4853
diff
changeset
|
3026 |
| 771 | 3027 /* !!#### Need to verify the encoding used in lwlib -- Qnative or Qctext? |
| 3028 Almost certainly the former. Use a standin for now. */ | |
| 3029 #define Qlwlib_encoding Qnative | |
| 3030 | |
| 1318 | 3031 /* The Host Portable Character Encoding. */ |
| 3032 #define Qx_hpc_encoding Qnative | |
| 3033 | |
| 3034 #define Qx_atom_name_encoding Qx_hpc_encoding | |
| 3035 #define Qx_font_name_encoding Qx_hpc_encoding | |
| 3036 #define Qx_color_name_encoding Qx_hpc_encoding | |
| 3037 #define Qx_keysym_encoding Qx_hpc_encoding | |
| 3038 #define Qx_geometry_encoding Qx_hpc_encoding | |
| 3039 #define Qx_resource_name_encoding Qx_hpc_encoding | |
| 3040 #define Qx_application_class_encoding Qx_hpc_encoding | |
| 771 | 3041 /* the following probably must agree with Qcommand_argument_encoding and |
| 3042 Qenvironment_variable_encoding */ | |
| 1318 | 3043 #define Qx_display_name_encoding Qx_hpc_encoding |
| 3044 #define Qx_xpm_data_encoding Qx_hpc_encoding | |
|
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:
4790
diff
changeset
|
3045 #define Qx_error_message_encoding Qx_hpc_encoding |
| 1318 | 3046 |
| 2367 | 3047 /* !!#### Verify these! */ |
| 3048 #define Qxt_widget_arg_encoding Qnative | |
| 3049 #define Qdt_dnd_encoding Qnative | |
| 3050 | |
| 1318 | 3051 /* RedHat 6.2 contains a locale called "Francais" with the C-cedilla |
| 3052 encoded in ISO2022! */ | |
| 3053 #define Qlocale_name_encoding Qctext | |
| 771 | 3054 |
| 3055 #define Qstrerror_encoding Qnative | |
| 3056 | |
| 1318 | 3057 /* !!#### This exists to remind us that our hexify routine is totally |
| 3058 un-Muleized. */ | |
| 3059 #define Qdnd_hexify_encoding Qascii | |
| 3060 | |
| 771 | 3061 #define GET_STRERROR(var, num) \ |
| 3062 do { \ | |
| 3063 int __gsnum__ = (num); \ | |
| 3064 Extbyte * __gserr__ = strerror (__gsnum__); \ | |
| 3065 \ | |
| 3066 if (!__gserr__) \ | |
| 3067 { \ | |
|
4981
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
3068 var = alloca_ibytes (99); \ |
| 771 | 3069 qxesprintf (var, "Unknown error %d", __gsnum__); \ |
| 3070 } \ | |
| 3071 else \ | |
|
4981
4aebb0131297
Cleanups/renaming of EXTERNAL_TO_C_STRING and friends
Ben Wing <ben@xemacs.org>
parents:
4953
diff
changeset
|
3072 var = EXTERNAL_TO_ITEXT (__gserr__, Qstrerror_encoding); \ |
| 771 | 3073 } while (0) |
| 3074 | |
| 3075 #endif /* INCLUDED_text_h_ */ |