|
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.
|
|
788
|
4 Copyright (C) 2000, 2001, 2002 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
|
|
|
48 /* ---------------------------------------------------------------------- */
|
|
|
49 /* Super-basic character properties */
|
|
|
50 /* ---------------------------------------------------------------------- */
|
|
|
51
|
|
|
52 /* These properties define the specifics of how our current encoding fits
|
|
|
53 in the basic model used for the encoding. Because this model is the same
|
|
|
54 as is used for UTF-8, all these properties could be defined for it, too.
|
|
|
55 This would instantly make the rest of this file work with UTF-8 (with
|
|
|
56 the exception of a few called functions that would need to be redefined).
|
|
|
57
|
|
|
58 (UTF-2000 implementers, take note!)
|
|
|
59 */
|
|
|
60
|
|
|
61 /* If you want more than this, you need to include charset.h */
|
|
|
62
|
|
|
63 #ifndef MULE
|
|
|
64
|
|
826
|
65 #define rep_bytes_by_first_byte(fb) 1
|
|
|
66 #define byte_ascii_p(byte) 1
|
|
867
|
67 #define MAX_ICHAR_LEN 1
|
|
771
|
68
|
|
|
69 #else /* MULE */
|
|
|
70
|
|
|
71 /* These are carefully designed to work if BYTE is signed or unsigned. */
|
|
|
72 /* Note that SPC and DEL are considered ASCII, not control. */
|
|
|
73
|
|
826
|
74 #define byte_ascii_p(byte) (((byte) & ~0x7f) == 0)
|
|
|
75 #define byte_c0_p(byte) (((byte) & ~0x1f) == 0)
|
|
|
76 #define byte_c1_p(byte) (((byte) & ~0x1f) == 0x80)
|
|
771
|
77
|
|
|
78 /* Does BYTE represent the first byte of a character? */
|
|
|
79
|
|
826
|
80 #ifdef ERROR_CHECK_TEXT
|
|
|
81
|
|
|
82 DECLARE_INLINE_HEADER (
|
|
|
83 int
|
|
867
|
84 ibyte_first_byte_p_1 (int byte, const char *file, int line)
|
|
826
|
85 )
|
|
|
86 {
|
|
|
87 assert_at_line (byte >= 0 && byte < 256, file, line);
|
|
|
88 return byte < 0xA0;
|
|
|
89 }
|
|
|
90
|
|
867
|
91 #define ibyte_first_byte_p(byte) \
|
|
|
92 ibyte_first_byte_p_1 (byte, __FILE__, __LINE__)
|
|
826
|
93
|
|
|
94 #else
|
|
|
95
|
|
867
|
96 #define ibyte_first_byte_p(byte) ((byte) < 0xA0)
|
|
826
|
97
|
|
|
98 #endif
|
|
|
99
|
|
|
100 #ifdef ERROR_CHECK_TEXT
|
|
771
|
101
|
|
|
102 /* Does BYTE represent the first byte of a multi-byte character? */
|
|
|
103
|
|
826
|
104 DECLARE_INLINE_HEADER (
|
|
|
105 int
|
|
867
|
106 ibyte_leading_byte_p_1 (int byte, const char *file, int line)
|
|
826
|
107 )
|
|
|
108 {
|
|
|
109 assert_at_line (byte >= 0 && byte < 256, file, line);
|
|
|
110 return byte_c1_p (byte);
|
|
|
111 }
|
|
|
112
|
|
867
|
113 #define ibyte_leading_byte_p(byte) \
|
|
|
114 ibyte_leading_byte_p_1 (byte, __FILE__, __LINE__)
|
|
826
|
115
|
|
|
116 #else
|
|
|
117
|
|
867
|
118 #define ibyte_leading_byte_p(byte) byte_c1_p (byte)
|
|
826
|
119
|
|
|
120 #endif
|
|
771
|
121
|
|
|
122 /* Table of number of bytes in the string representation of a character
|
|
|
123 indexed by the first byte of that representation.
|
|
|
124
|
|
|
125 This value can be derived in other ways -- e.g. something like
|
|
826
|
126 XCHARSET_REP_BYTES (charset_by_leading_byte (first_byte))
|
|
771
|
127 but it's faster this way. */
|
|
|
128 extern const Bytecount rep_bytes_by_first_byte[0xA0];
|
|
|
129
|
|
|
130 /* Number of bytes in the string representation of a character. */
|
|
788
|
131
|
|
800
|
132 #ifdef ERROR_CHECK_TEXT
|
|
788
|
133
|
|
826
|
134 DECLARE_INLINE_HEADER (
|
|
|
135 Bytecount
|
|
|
136 rep_bytes_by_first_byte_1 (int fb, const char *file, int line)
|
|
|
137 )
|
|
771
|
138 {
|
|
826
|
139 assert_at_line (fb >= 0 && fb < 0xA0, file, line);
|
|
771
|
140 return rep_bytes_by_first_byte[fb];
|
|
|
141 }
|
|
|
142
|
|
826
|
143 #define rep_bytes_by_first_byte(fb) \
|
|
|
144 rep_bytes_by_first_byte_1 (fb, __FILE__, __LINE__)
|
|
788
|
145
|
|
800
|
146 #else /* ERROR_CHECK_TEXT */
|
|
788
|
147
|
|
826
|
148 #define rep_bytes_by_first_byte(fb) (rep_bytes_by_first_byte[fb])
|
|
788
|
149
|
|
800
|
150 #endif /* ERROR_CHECK_TEXT */
|
|
788
|
151
|
|
826
|
152 /* Is this character represented by more than one byte in a string in the
|
|
|
153 default format? */
|
|
|
154
|
|
867
|
155 #define ichar_multibyte_p(c) ((c) >= 0x80)
|
|
|
156
|
|
|
157 #define ichar_ascii_p(c) (!ichar_multibyte_p (c))
|
|
826
|
158
|
|
|
159 /* Maximum number of bytes per Emacs character when represented as text, in
|
|
|
160 any format.
|
|
|
161 */
|
|
771
|
162
|
|
867
|
163 #define MAX_ICHAR_LEN 4
|
|
771
|
164
|
|
826
|
165 #endif /* not MULE */
|
|
|
166
|
|
|
167 /* ---------------- Handling non-default formats ----------------- */
|
|
|
168
|
|
|
169 /* We support, at least to some extent, formats other than the default
|
|
|
170 variable-width format, for speed; all of these alternative formats are
|
|
|
171 fixed-width. Currently we only handle these non-default formats in
|
|
|
172 buffers, because access to their text is strictly controlled and thus
|
|
|
173 the details of the format mostly compartmentalized. The only really
|
|
|
174 tricky part is the search code -- the regex, Boyer-Moore, and
|
|
|
175 simple-search algorithms in search.c and regex.c. All other code that
|
|
|
176 knows directly about the buffer representation is the basic code to
|
|
|
177 modify or retrieve the buffer text.
|
|
|
178
|
|
|
179 Supporting fixed-width formats in Lisp strings is harder, but possible
|
|
|
180 -- FSF currently does this, for example. In this case, however,
|
|
|
181 probably only 8-bit-fixed is reasonable for Lisp strings -- getting
|
|
|
182 non-ASCII-compatible fixed-width formats to work is much, much harder
|
|
|
183 because a lot of code assumes that strings are ASCII-compatible
|
|
|
184 (i.e. ASCII + other characters represented exclusively using high-bit
|
|
|
185 bytes) and a lot of code mixes Lisp strings and non-Lisp strings freely.
|
|
|
186
|
|
|
187 The different possible fixed-width formats are 8-bit fixed, 16-bit
|
|
|
188 fixed, and 32-bit fixed. The latter can represent all possible
|
|
|
189 characters, but at a substantial memory penalty. The other two can
|
|
|
190 represent only a subset of the possible characters. How these subsets
|
|
|
191 are defined can be simple or very tricky.
|
|
|
192
|
|
|
193 Currently we support only the default format and the 8-bit fixed format,
|
|
|
194 and in the latter, we only allow these to be the first 256 characters in
|
|
867
|
195 an Ichar (ASCII and Latin 1).
|
|
826
|
196
|
|
|
197 One reasonable approach for 8-bit fixed is to allow the upper half to
|
|
|
198 represent any 1-byte charset, which is specified on a per-buffer basis.
|
|
|
199 This should work fairly well in practice since most documents are in
|
|
|
200 only one foreign language (possibly with some English mixed in). I
|
|
|
201 think FSF does something like this; or at least, they have something
|
|
|
202 called nonascii-translation-table and use it when converting from
|
|
|
203 8-bit-fixed text ("unibyte text") to default text ("multibyte text").
|
|
|
204 With 16-bit fixed, you could do something like assign chunks of the 64K
|
|
|
205 worth of characters to charsets as they're encountered in documents.
|
|
|
206 This should work well with most Asian documents.
|
|
|
207
|
|
|
208 If/when we switch to using Unicode internally, we might have formats more
|
|
|
209 like this:
|
|
|
210
|
|
|
211 -- UTF-8 or some extension as the default format. Perl uses an
|
|
|
212 extension that handles 64-bit chars and requires as much as 13 bytes per
|
|
|
213 char, vs. the standard of 31-bit chars and 6 bytes max. UTF-8 has the
|
|
|
214 same basic properties as our own variable-width format (see text.c,
|
|
|
215 Internal String Encoding) and so most code would not need to be changed.
|
|
|
216
|
|
|
217 -- UTF-16 as a "pseudo-fixed" format (i.e. 16-bit fixed plus surrogates
|
|
|
218 for representing characters not in the BMP, aka >= 65536). The vast
|
|
|
219 majority of documents will have no surrogates in them so byte/char
|
|
|
220 conversion will be very fast.
|
|
|
221
|
|
|
222 -- an 8-bit fixed format, like currently.
|
|
|
223
|
|
|
224 -- possibly, UCS-4 as a 32-bit fixed format.
|
|
|
225
|
|
|
226 The fixed-width formats essentially treat the buffer as an array of
|
|
|
227 8-bit, 16-bit or 32-bit integers. This means that how they are stored
|
|
|
228 in memory (in particular, big-endian or little-endian) depends on the
|
|
|
229 native format of the machine's processor. It also means we have to
|
|
|
230 worry a bit about alignment (basically, we just need to keep the gap an
|
|
|
231 integral size of the character size, and get things aligned properly
|
|
|
232 when converting the buffer between formats).
|
|
|
233 */
|
|
|
234 typedef enum internal_format
|
|
|
235 {
|
|
|
236 FORMAT_DEFAULT,
|
|
|
237 FORMAT_8_BIT_FIXED,
|
|
|
238 FORMAT_16_BIT_FIXED, /* not implemented */
|
|
|
239 FORMAT_32_BIT_FIXED /* not implemented */
|
|
|
240 } Internal_Format;
|
|
|
241
|
|
|
242 #ifdef MULE
|
|
|
243 /* "OBJECT" below will usually be a buffer, string, or nil. This needs to
|
|
|
244 be passed in because the interpretation of 8-bit-fixed and 16-bit-fixed
|
|
|
245 values may depend on the buffer, e.g. depending on what language the
|
|
|
246 text in the buffer is in. */
|
|
|
247
|
|
867
|
248 /* True if Ichar CH can be represented in 8-bit-fixed format. */
|
|
|
249 #define ichar_8_bit_fixed_p(ch, object) (((ch) & ~0xff) == 0)
|
|
|
250 /* Convert Ichar CH to an 8-bit int, as will be stored in the buffer. */
|
|
|
251 #define ichar_to_raw_8_bit_fixed(ch, object) ((Ibyte) (ch))
|
|
826
|
252 /* Convert the other way. */
|
|
867
|
253 #define raw_8_bit_fixed_to_ichar(ch, object) ((Ichar) (ch))
|
|
|
254
|
|
|
255 #define ichar_16_bit_fixed_p(ch, object) (((ch) & ~0xffff) == 0)
|
|
|
256 /* Convert Ichar CH to a 16-bit int, as will be stored in the buffer. */
|
|
|
257 #define ichar_to_raw_16_bit_fixed(ch, object) ((UINT_16_BIT) (ch))
|
|
826
|
258 /* Convert the other way. */
|
|
867
|
259 #define raw_16_bit_fixed_to_ichar(ch, object) ((Ichar) (ch))
|
|
|
260
|
|
|
261 /* Convert Ichar CH to a 32-bit int, as will be stored in the buffer. */
|
|
|
262 #define ichar_to_raw_32_bit_fixed(ch, object) ((UINT_32_BIT) (ch))
|
|
826
|
263 /* Convert the other way. */
|
|
867
|
264 #define raw_32_bit_fixed_to_ichar(ch, object) ((Ichar) (ch))
|
|
826
|
265
|
|
|
266 /* Return the "raw value" of a character as stored in the buffer. In the
|
|
|
267 default format, this is just the same as the character. In fixed-width
|
|
|
268 formats, this is the actual value in the buffer, which will be limited
|
|
|
269 to the range as established by the format. This is used when searching
|
|
|
270 for a character in a buffer -- it's faster to convert the character to
|
|
|
271 the raw value and look for that, than repeatedly convert each raw value
|
|
|
272 in the buffer into a character. */
|
|
|
273
|
|
|
274 DECLARE_INLINE_HEADER (
|
|
867
|
275 Raw_Ichar
|
|
|
276 ichar_to_raw (Ichar ch, Internal_Format fmt, Lisp_Object object)
|
|
826
|
277 )
|
|
|
278 {
|
|
|
279 switch (fmt)
|
|
|
280 {
|
|
|
281 case FORMAT_DEFAULT:
|
|
867
|
282 return (Raw_Ichar) ch;
|
|
826
|
283 case FORMAT_16_BIT_FIXED:
|
|
867
|
284 text_checking_assert (ichar_16_bit_fixed_p (ch, object));
|
|
|
285 return (Raw_Ichar) ichar_to_raw_16_bit_fixed (ch, object);
|
|
826
|
286 case FORMAT_32_BIT_FIXED:
|
|
867
|
287 return (Raw_Ichar) ichar_to_raw_32_bit_fixed (ch, object);
|
|
826
|
288 default:
|
|
|
289 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
867
|
290 text_checking_assert (ichar_8_bit_fixed_p (ch, object));
|
|
|
291 return (Raw_Ichar) ichar_to_raw_8_bit_fixed (ch, object);
|
|
826
|
292 }
|
|
|
293 }
|
|
|
294
|
|
|
295 /* Return whether CH is representable in the given format in the given
|
|
|
296 object. */
|
|
|
297
|
|
|
298 DECLARE_INLINE_HEADER (
|
|
|
299 int
|
|
867
|
300 ichar_fits_in_format (Ichar ch, Internal_Format fmt, Lisp_Object object)
|
|
826
|
301 )
|
|
|
302 {
|
|
|
303 switch (fmt)
|
|
|
304 {
|
|
|
305 case FORMAT_DEFAULT:
|
|
|
306 return 1;
|
|
|
307 case FORMAT_16_BIT_FIXED:
|
|
867
|
308 return ichar_16_bit_fixed_p (ch, object);
|
|
826
|
309 case FORMAT_32_BIT_FIXED:
|
|
|
310 return 1;
|
|
|
311 default:
|
|
|
312 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
867
|
313 return ichar_8_bit_fixed_p (ch, object);
|
|
826
|
314 }
|
|
|
315 }
|
|
|
316
|
|
|
317 /* Assuming the formats are the same, return whether the two objects
|
|
|
318 represent text in exactly the same way. */
|
|
|
319
|
|
|
320 DECLARE_INLINE_HEADER (
|
|
|
321 int
|
|
|
322 objects_have_same_internal_representation (Lisp_Object srcobj,
|
|
|
323 Lisp_Object dstobj)
|
|
|
324 )
|
|
|
325 {
|
|
|
326 /* &&#### implement this properly when we allow per-object format
|
|
|
327 differences */
|
|
|
328 return 1;
|
|
|
329 }
|
|
|
330
|
|
|
331 #else
|
|
|
332
|
|
867
|
333 #define ichar_to_raw(ch, fmt, object) ((Raw_Ichar) (ch))
|
|
|
334 #define ichar_fits_in_format(ch, fmt, object) 1
|
|
826
|
335 #define objects_have_same_internal_representation(srcobj, dstobj) 1
|
|
|
336
|
|
771
|
337 #endif /* MULE */
|
|
|
338
|
|
|
339 int dfc_coding_system_is_unicode (Lisp_Object coding_system);
|
|
|
340
|
|
|
341 DECLARE_INLINE_HEADER (
|
|
|
342 Bytecount dfc_external_data_len (const void *ptr, Lisp_Object codesys)
|
|
|
343 )
|
|
|
344 {
|
|
|
345 if (dfc_coding_system_is_unicode (codesys))
|
|
|
346 return sizeof (wchar_t) * wcslen ((wchar_t *) ptr);
|
|
|
347 else
|
|
|
348 return strlen ((char *) ptr);
|
|
|
349 }
|
|
|
350
|
|
|
351
|
|
|
352 /************************************************************************/
|
|
|
353 /* */
|
|
|
354 /* working with raw internal-format data */
|
|
|
355 /* */
|
|
|
356 /************************************************************************/
|
|
|
357
|
|
826
|
358 /*
|
|
|
359 Use the following functions/macros on contiguous text in any of the
|
|
|
360 internal formats. Those that take a format arg work on all internal
|
|
|
361 formats; the others work only on the default (variable-width under Mule)
|
|
|
362 format. If the text you're operating on is known to come from a buffer,
|
|
|
363 use the buffer-level functions in buffer.h, which automatically know the
|
|
|
364 correct format and handle the gap.
|
|
|
365
|
|
|
366 Some terminology:
|
|
|
367
|
|
867
|
368 "itext" appearing in the macros means "internal-format text" -- type
|
|
|
369 `Ibyte *'. Operations on such pointers themselves, rather than on the
|
|
|
370 text being pointed to, have "itext" instead of "itext" in the macro
|
|
|
371 name. "ichar" in the macro names means an Ichar -- the representation
|
|
826
|
372 of a character as a single integer rather than a series of bytes, as part
|
|
867
|
373 of "itext". Many of the macros below are for converting between the
|
|
826
|
374 two representations of characters.
|
|
|
375
|
|
867
|
376 Note also that we try to consistently distinguish between an "Ichar" and
|
|
826
|
377 a Lisp character. Stuff working with Lisp characters often just says
|
|
867
|
378 "char", so we consistently use "Ichar" when that's what we're working
|
|
826
|
379 with. */
|
|
|
380
|
|
|
381 /* The three golden rules of macros:
|
|
771
|
382
|
|
|
383 1) Anything that's an lvalue can be evaluated more than once.
|
|
826
|
384
|
|
|
385 2) Macros where anything else can be evaluated more than once should
|
|
|
386 have the word "unsafe" in their name (exceptions may be made for
|
|
|
387 large sets of macros that evaluate arguments of certain types more
|
|
|
388 than once, e.g. struct buffer * arguments, when clearly indicated in
|
|
|
389 the macro documentation). These macros are generally meant to be
|
|
|
390 called only by other macros that have already stored the calling
|
|
|
391 values in temporary variables.
|
|
|
392
|
|
|
393 3) Nothing else can be evaluated more than once. Use inline
|
|
771
|
394 functions, if necessary, to prevent multiple evaluation.
|
|
826
|
395
|
|
|
396 NOTE: The functions and macros below are given full prototypes in their
|
|
|
397 docs, even when the implementation is a macro. In such cases, passing
|
|
|
398 an argument of a type other than expected will produce undefined
|
|
|
399 results. Also, given that macros can do things functions can't (in
|
|
|
400 particular, directly modify arguments as if they were passed by
|
|
|
401 reference), the declaration syntax has been extended to include the
|
|
|
402 call-by-reference syntax from C++, where an & after a type indicates
|
|
|
403 that the argument is an lvalue and is passed by reference, i.e. the
|
|
|
404 function can modify its value. (This is equivalent in C to passing a
|
|
|
405 pointer to the argument, but without the need to explicitly worry about
|
|
|
406 pointers.)
|
|
|
407
|
|
|
408 When to capitalize macros:
|
|
|
409
|
|
|
410 -- Capitalize macros doing stuff obviously impossible with (C)
|
|
|
411 functions, e.g. directly modifying arguments as if they were passed by
|
|
|
412 reference.
|
|
|
413
|
|
|
414 -- Capitalize macros that evaluate *any* argument more than once regardless
|
|
|
415 of whether that's "allowed" (e.g. buffer arguments).
|
|
|
416
|
|
|
417 -- Capitalize macros that directly access a field in a Lisp_Object or
|
|
|
418 its equivalent underlying structure. In such cases, access through the
|
|
|
419 Lisp_Object precedes the macro with an X, and access through the underlying
|
|
|
420 structure doesn't.
|
|
|
421
|
|
|
422 -- Capitalize certain other basic macros relating to Lisp_Objects; e.g.
|
|
|
423 FRAMEP, CHECK_FRAME, etc.
|
|
|
424
|
|
|
425 -- Try to avoid capitalizing any other macros.
|
|
771
|
426 */
|
|
|
427
|
|
|
428 /* ---------------------------------------------------------------------- */
|
|
867
|
429 /* Working with itext's (pointers to internally-formatted text) */
|
|
771
|
430 /* ---------------------------------------------------------------------- */
|
|
|
431
|
|
867
|
432 /* Given an itext, does it point to the beginning of a character?
|
|
826
|
433 */
|
|
|
434
|
|
771
|
435 #ifdef MULE
|
|
867
|
436 # define valid_ibyteptr_p(ptr) ibyte_first_byte_p (* (ptr))
|
|
771
|
437 #else
|
|
867
|
438 # define valid_ibyteptr_p(ptr) 1
|
|
771
|
439 #endif
|
|
|
440
|
|
867
|
441 /* If error-checking is enabled, assert that the given itext points to
|
|
826
|
442 the beginning of a character. Otherwise, do nothing.
|
|
|
443 */
|
|
|
444
|
|
867
|
445 #define assert_valid_ibyteptr(ptr) text_checking_assert (valid_ibyteptr_p (ptr))
|
|
|
446
|
|
|
447 /* Given a itext (assumed to point at the beginning of a character),
|
|
826
|
448 modify that pointer so it points to the beginning of the next character.
|
|
|
449
|
|
867
|
450 Note that INC_IBYTEPTR() and DEC_IBYTEPTR() have to be written in
|
|
|
451 completely separate ways. INC_IBYTEPTR() cannot use the DEC_IBYTEPTR()
|
|
771
|
452 trick of looking for a valid first byte because it might run off
|
|
867
|
453 the end of the string. DEC_IBYTEPTR() can't use the INC_IBYTEPTR()
|
|
771
|
454 method because it doesn't have easy access to the first byte of
|
|
|
455 the character it's moving over. */
|
|
|
456
|
|
867
|
457 #define INC_IBYTEPTR(ptr) do { \
|
|
|
458 assert_valid_ibyteptr (ptr); \
|
|
826
|
459 (ptr) += rep_bytes_by_first_byte (* (ptr)); \
|
|
|
460 } while (0)
|
|
|
461
|
|
1204
|
462 #define INC_IBYTEPTR_FMT(ptr, fmt) \
|
|
|
463 do { \
|
|
|
464 Internal_Format __icf_fmt = (fmt); \
|
|
|
465 switch (__icf_fmt) \
|
|
|
466 { \
|
|
|
467 case FORMAT_DEFAULT: \
|
|
|
468 INC_IBYTEPTR (ptr); \
|
|
|
469 break; \
|
|
|
470 case FORMAT_16_BIT_FIXED: \
|
|
|
471 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); \
|
|
|
472 (ptr) += 2; \
|
|
|
473 break; \
|
|
|
474 case FORMAT_32_BIT_FIXED: \
|
|
|
475 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); \
|
|
|
476 (ptr) += 4; \
|
|
|
477 break; \
|
|
|
478 default: \
|
|
|
479 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); \
|
|
|
480 (ptr)++; \
|
|
|
481 break; \
|
|
|
482 } \
|
|
826
|
483 } while (0)
|
|
|
484
|
|
867
|
485 /* Given a itext (assumed to point at the beginning of a character or at
|
|
826
|
486 the very end of the text), modify that pointer so it points to the
|
|
|
487 beginning of the previous character.
|
|
|
488 */
|
|
771
|
489
|
|
800
|
490 #ifdef ERROR_CHECK_TEXT
|
|
826
|
491 /* We use a separate definition to avoid warnings about unused dc_ptr1 */
|
|
867
|
492 #define DEC_IBYTEPTR(ptr) do { \
|
|
|
493 const Ibyte *dc_ptr1 = (ptr); \
|
|
826
|
494 do { \
|
|
|
495 (ptr)--; \
|
|
867
|
496 } while (!valid_ibyteptr_p (ptr)); \
|
|
826
|
497 text_checking_assert (dc_ptr1 - (ptr) == rep_bytes_by_first_byte (*(ptr))); \
|
|
771
|
498 } while (0)
|
|
826
|
499 #else
|
|
867
|
500 #define DEC_IBYTEPTR(ptr) do { \
|
|
826
|
501 do { \
|
|
|
502 (ptr)--; \
|
|
867
|
503 } while (!valid_ibyteptr_p (ptr)); \
|
|
771
|
504 } while (0)
|
|
826
|
505 #endif /* ERROR_CHECK_TEXT */
|
|
|
506
|
|
1204
|
507 #define DEC_IBYTEPTR_FMT(ptr, fmt) \
|
|
|
508 do { \
|
|
|
509 Internal_Format __icf_fmt = (fmt); \
|
|
|
510 switch (__icf_fmt) \
|
|
|
511 { \
|
|
|
512 case FORMAT_DEFAULT: \
|
|
|
513 DEC_IBYTEPTR (ptr); \
|
|
|
514 break; \
|
|
|
515 case FORMAT_16_BIT_FIXED: \
|
|
|
516 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT)); \
|
|
|
517 (ptr) -= 2; \
|
|
|
518 break; \
|
|
|
519 case FORMAT_32_BIT_FIXED: \
|
|
|
520 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT)); \
|
|
|
521 (ptr) -= 4; \
|
|
|
522 break; \
|
|
|
523 default: \
|
|
|
524 text_checking_assert (fmt == FORMAT_8_BIT_FIXED); \
|
|
|
525 (ptr)--; \
|
|
|
526 break; \
|
|
|
527 } \
|
|
771
|
528 } while (0)
|
|
|
529
|
|
|
530 #ifdef MULE
|
|
|
531
|
|
826
|
532 /* Make sure that PTR is pointing to the beginning of a character. If not,
|
|
|
533 back up until this is the case. Note that there are not too many places
|
|
|
534 where it is legitimate to do this sort of thing. It's an error if
|
|
|
535 you're passed an "invalid" char * pointer. NOTE: PTR *must* be pointing
|
|
|
536 to a valid part of the string (i.e. not the very end, unless the string
|
|
|
537 is zero-terminated or something) in order for this function to not cause
|
|
|
538 crashes.
|
|
|
539 */
|
|
|
540
|
|
771
|
541 /* Note that this reads the byte at *PTR! */
|
|
|
542
|
|
867
|
543 #define VALIDATE_IBYTEPTR_BACKWARD(ptr) do { \
|
|
|
544 while (!valid_ibyteptr_p (ptr)) ptr--; \
|
|
771
|
545 } while (0)
|
|
|
546
|
|
826
|
547 /* Make sure that PTR is pointing to the beginning of a character. If not,
|
|
|
548 move forward until this is the case. Note that there are not too many
|
|
|
549 places where it is legitimate to do this sort of thing. It's an error
|
|
|
550 if you're passed an "invalid" char * pointer.
|
|
|
551 */
|
|
771
|
552
|
|
867
|
553 /* This needs to be trickier than VALIDATE_IBYTEPTR_BACKWARD() to avoid the
|
|
771
|
554 possibility of running off the end of the string. */
|
|
|
555
|
|
867
|
556 #define VALIDATE_IBYTEPTR_FORWARD(ptr) do { \
|
|
|
557 Ibyte *vcf_ptr = (ptr); \
|
|
|
558 VALIDATE_IBYTEPTR_BACKWARD (vcf_ptr); \
|
|
771
|
559 if (vcf_ptr != (ptr)) \
|
|
|
560 { \
|
|
|
561 (ptr) = vcf_ptr; \
|
|
867
|
562 INC_IBYTEPTR (ptr); \
|
|
771
|
563 } \
|
|
|
564 } while (0)
|
|
|
565
|
|
|
566 #else /* not MULE */
|
|
867
|
567 #define VALIDATE_IBYTEPTR_BACKWARD(ptr)
|
|
|
568 #define VALIDATE_IBYTEPTR_FORWARD(ptr)
|
|
826
|
569 #endif /* not MULE */
|
|
|
570
|
|
|
571 #ifdef MULE
|
|
|
572
|
|
867
|
573 /* Given a Ibyte string at PTR of size N, possibly with a partial
|
|
826
|
574 character at the end, return the size of the longest substring of
|
|
|
575 complete characters. Does not assume that the byte at *(PTR + N) is
|
|
|
576 readable. Note that there are not too many places where it is
|
|
|
577 legitimate to do this sort of thing. It's an error if you're passed an
|
|
|
578 "invalid" offset. */
|
|
|
579
|
|
|
580 DECLARE_INLINE_HEADER (
|
|
|
581 Bytecount
|
|
867
|
582 validate_ibyte_string_backward (const Ibyte *ptr, Bytecount n)
|
|
826
|
583 )
|
|
|
584 {
|
|
867
|
585 const Ibyte *ptr2;
|
|
826
|
586
|
|
|
587 if (n == 0)
|
|
|
588 return n;
|
|
|
589 ptr2 = ptr + n - 1;
|
|
867
|
590 VALIDATE_IBYTEPTR_BACKWARD (ptr2);
|
|
826
|
591 if (ptr2 + rep_bytes_by_first_byte (*ptr2) != ptr + n)
|
|
|
592 return ptr2 - ptr;
|
|
|
593 return n;
|
|
|
594 }
|
|
|
595
|
|
|
596 #else
|
|
|
597
|
|
867
|
598 #define validate_ibyte_string_backward(ptr, n) (n)
|
|
826
|
599
|
|
|
600 #endif /* MULE */
|
|
771
|
601
|
|
|
602 /* -------------------------------------------------------------- */
|
|
826
|
603 /* Working with the length (in bytes and characters) of a */
|
|
|
604 /* section of internally-formatted text */
|
|
771
|
605 /* -------------------------------------------------------------- */
|
|
|
606
|
|
826
|
607 #ifdef MULE
|
|
|
608
|
|
867
|
609 Charcount bytecount_to_charcount_fun (const Ibyte *ptr, Bytecount len);
|
|
|
610 Bytecount charcount_to_bytecount_fun (const Ibyte *ptr, Charcount len);
|
|
826
|
611
|
|
|
612 /* Given a pointer to a text string and a length in bytes, return
|
|
|
613 the equivalent length in characters. */
|
|
|
614
|
|
|
615 DECLARE_INLINE_HEADER (
|
|
|
616 Charcount
|
|
867
|
617 bytecount_to_charcount (const Ibyte *ptr, Bytecount len)
|
|
826
|
618 )
|
|
|
619 {
|
|
|
620 if (len < 20) /* Just a random guess, but it should be more or less correct.
|
|
|
621 If number of bytes is small, just do a simple loop,
|
|
|
622 which should be more efficient. */
|
|
|
623 {
|
|
|
624 Charcount count = 0;
|
|
867
|
625 const Ibyte *end = ptr + len;
|
|
826
|
626 while (ptr < end)
|
|
|
627 {
|
|
867
|
628 INC_IBYTEPTR (ptr);
|
|
826
|
629 count++;
|
|
|
630 }
|
|
|
631 /* Bomb out if the specified substring ends in the middle
|
|
|
632 of a character. Note that we might have already gotten
|
|
|
633 a core dump above from an invalid reference, but at least
|
|
|
634 we will get no farther than here.
|
|
|
635
|
|
|
636 This also catches len < 0. */
|
|
|
637 text_checking_assert (ptr == end);
|
|
|
638
|
|
|
639 return count;
|
|
|
640 }
|
|
|
641 else
|
|
|
642 return bytecount_to_charcount_fun (ptr, len);
|
|
|
643 }
|
|
|
644
|
|
|
645 /* Given a pointer to a text string and a length in characters, return the
|
|
|
646 equivalent length in bytes.
|
|
|
647 */
|
|
|
648
|
|
|
649 DECLARE_INLINE_HEADER (
|
|
|
650 Bytecount
|
|
867
|
651 charcount_to_bytecount (const Ibyte *ptr, Charcount len)
|
|
826
|
652 )
|
|
|
653 {
|
|
|
654 text_checking_assert (len >= 0);
|
|
|
655 if (len < 20) /* See above */
|
|
|
656 {
|
|
867
|
657 const Ibyte *newptr = ptr;
|
|
826
|
658 while (len > 0)
|
|
|
659 {
|
|
867
|
660 INC_IBYTEPTR (newptr);
|
|
826
|
661 len--;
|
|
|
662 }
|
|
|
663 return newptr - ptr;
|
|
|
664 }
|
|
|
665 else
|
|
|
666 return charcount_to_bytecount_fun (ptr, len);
|
|
|
667 }
|
|
|
668
|
|
|
669 /* Given a pointer to a text string in the specified format and a length in
|
|
|
670 bytes, return the equivalent length in characters.
|
|
|
671 */
|
|
|
672
|
|
|
673 DECLARE_INLINE_HEADER (
|
|
|
674 Charcount
|
|
867
|
675 bytecount_to_charcount_fmt (const Ibyte *ptr, Bytecount len,
|
|
826
|
676 Internal_Format fmt)
|
|
|
677 )
|
|
|
678 {
|
|
|
679 switch (fmt)
|
|
|
680 {
|
|
|
681 case FORMAT_DEFAULT:
|
|
|
682 return bytecount_to_charcount (ptr, len);
|
|
|
683 case FORMAT_16_BIT_FIXED:
|
|
1204
|
684 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT));
|
|
826
|
685 return (Charcount) (len << 1);
|
|
|
686 case FORMAT_32_BIT_FIXED:
|
|
1204
|
687 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT));
|
|
826
|
688 return (Charcount) (len << 2);
|
|
|
689 default:
|
|
|
690 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
|
691 return (Charcount) len;
|
|
|
692 }
|
|
|
693 }
|
|
|
694
|
|
|
695 /* Given a pointer to a text string in the specified format and a length in
|
|
|
696 characters, return the equivalent length in bytes.
|
|
|
697 */
|
|
|
698
|
|
|
699 DECLARE_INLINE_HEADER (
|
|
|
700 Bytecount
|
|
867
|
701 charcount_to_bytecount_fmt (const Ibyte *ptr, Charcount len,
|
|
826
|
702 Internal_Format fmt)
|
|
|
703 )
|
|
|
704 {
|
|
|
705 switch (fmt)
|
|
|
706 {
|
|
|
707 case FORMAT_DEFAULT:
|
|
|
708 return charcount_to_bytecount (ptr, len);
|
|
|
709 case FORMAT_16_BIT_FIXED:
|
|
1204
|
710 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT));
|
|
826
|
711 text_checking_assert (!(len & 1));
|
|
|
712 return (Bytecount) (len >> 1);
|
|
|
713 case FORMAT_32_BIT_FIXED:
|
|
|
714 text_checking_assert (!(len & 3));
|
|
1204
|
715 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT));
|
|
826
|
716 return (Bytecount) (len >> 2);
|
|
|
717 default:
|
|
|
718 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
|
719 return (Bytecount) len;
|
|
|
720 }
|
|
|
721 }
|
|
|
722
|
|
|
723 #else
|
|
|
724
|
|
|
725 #define bytecount_to_charcount(ptr, len) ((Charcount) (len))
|
|
|
726 #define bytecount_to_charcount_fmt(ptr, len, fmt) ((Charcount) (len))
|
|
|
727 #define charcount_to_bytecount(ptr, len) ((Bytecount) (len))
|
|
|
728 #define charcount_to_bytecount_fmt(ptr, len, fmt) ((Bytecount) (len))
|
|
|
729
|
|
|
730 #endif /* MULE */
|
|
|
731
|
|
|
732 /* Return the length of the first character at PTR. Equivalent to
|
|
|
733 charcount_to_bytecount (ptr, 1).
|
|
|
734
|
|
|
735 [Since charcount_to_bytecount() is Written as inline, a smart compiler
|
|
|
736 should really optimize charcount_to_bytecount (ptr, 1) to the same as
|
|
|
737 the following, with no error checking. But since this idiom occurs so
|
|
|
738 often, we'll be helpful and define a special macro for it.]
|
|
|
739 */
|
|
|
740
|
|
867
|
741 #define itext_ichar_len(ptr) rep_bytes_by_first_byte (*(ptr))
|
|
826
|
742
|
|
|
743 /* Return the length of the first character at PTR, which is in the
|
|
|
744 specified internal format. Equivalent to charcount_to_bytecount_fmt
|
|
|
745 (ptr, 1, fmt).
|
|
|
746 */
|
|
|
747
|
|
|
748 DECLARE_INLINE_HEADER (
|
|
|
749 Bytecount
|
|
867
|
750 itext_ichar_len_fmt (const Ibyte *ptr, Internal_Format fmt)
|
|
826
|
751 )
|
|
|
752 {
|
|
|
753 switch (fmt)
|
|
|
754 {
|
|
|
755 case FORMAT_DEFAULT:
|
|
867
|
756 return itext_ichar_len (ptr);
|
|
826
|
757 case FORMAT_16_BIT_FIXED:
|
|
1204
|
758 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT));
|
|
826
|
759 return 2;
|
|
|
760 case FORMAT_32_BIT_FIXED:
|
|
1204
|
761 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT));
|
|
826
|
762 return 4;
|
|
|
763 default:
|
|
|
764 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
|
765 return 1;
|
|
|
766 }
|
|
|
767 }
|
|
|
768
|
|
|
769 /* Return a pointer to the beginning of the character offset N (in
|
|
|
770 characters) from PTR.
|
|
|
771 */
|
|
|
772
|
|
|
773 DECLARE_INLINE_HEADER (
|
|
867
|
774 const Ibyte *
|
|
|
775 itext_n_addr (const Ibyte *ptr, Charcount offset)
|
|
826
|
776 )
|
|
771
|
777 {
|
|
|
778 return ptr + charcount_to_bytecount (ptr, offset);
|
|
|
779 }
|
|
|
780
|
|
867
|
781 /* Given a itext and an offset into the text pointed to by the itext,
|
|
826
|
782 modify the offset so it points to the beginning of the next character.
|
|
|
783 */
|
|
|
784
|
|
|
785 #define INC_BYTECOUNT(ptr, pos) do { \
|
|
867
|
786 assert_valid_ibyteptr (ptr); \
|
|
826
|
787 (pos += rep_bytes_by_first_byte (* ((ptr) + (pos)))); \
|
|
|
788 } while (0)
|
|
|
789
|
|
771
|
790 /* -------------------------------------------------------------------- */
|
|
867
|
791 /* Retrieving or changing the character pointed to by a itext */
|
|
771
|
792 /* -------------------------------------------------------------------- */
|
|
|
793
|
|
867
|
794 #define simple_itext_ichar(ptr) ((Ichar) (ptr)[0])
|
|
|
795 #define simple_set_itext_ichar(ptr, x) \
|
|
|
796 ((ptr)[0] = (Ibyte) (x), (Bytecount) 1)
|
|
|
797 #define simple_itext_copy_ichar(src, dst) \
|
|
814
|
798 ((dst)[0] = *(src), (Bytecount) 1)
|
|
771
|
799
|
|
|
800 #ifdef MULE
|
|
|
801
|
|
867
|
802 Ichar non_ascii_itext_ichar (const Ibyte *ptr);
|
|
|
803 Bytecount non_ascii_set_itext_ichar (Ibyte *ptr, Ichar c);
|
|
|
804 Bytecount non_ascii_itext_copy_ichar (const Ibyte *src, Ibyte *dst);
|
|
|
805
|
|
|
806 /* Retrieve the character pointed to by PTR as an Ichar. */
|
|
826
|
807
|
|
|
808 DECLARE_INLINE_HEADER (
|
|
867
|
809 Ichar
|
|
|
810 itext_ichar (const Ibyte *ptr)
|
|
826
|
811 )
|
|
771
|
812 {
|
|
826
|
813 return byte_ascii_p (*ptr) ?
|
|
867
|
814 simple_itext_ichar (ptr) :
|
|
|
815 non_ascii_itext_ichar (ptr);
|
|
771
|
816 }
|
|
|
817
|
|
826
|
818 /* Retrieve the character pointed to by PTR (a pointer to text in the
|
|
|
819 format FMT, coming from OBJECT [a buffer, string?, or nil]) as an
|
|
867
|
820 Ichar.
|
|
826
|
821
|
|
|
822 Note: For these and other *_fmt() functions, if you pass in a constant
|
|
|
823 FMT, the switch will be optimized out of existence. Therefore, there is
|
|
|
824 no need to create separate versions for the various formats for
|
|
867
|
825 "efficiency reasons". In fact, we don't really need itext_ichar()
|
|
826
|
826 and such written separately, but they are used often so it's simpler
|
|
|
827 that way. */
|
|
|
828
|
|
|
829 DECLARE_INLINE_HEADER (
|
|
867
|
830 Ichar
|
|
|
831 itext_ichar_fmt (const Ibyte *ptr, Internal_Format fmt,
|
|
826
|
832 Lisp_Object object)
|
|
|
833 )
|
|
|
834 {
|
|
|
835 switch (fmt)
|
|
|
836 {
|
|
|
837 case FORMAT_DEFAULT:
|
|
867
|
838 return itext_ichar (ptr);
|
|
826
|
839 case FORMAT_16_BIT_FIXED:
|
|
1204
|
840 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT));
|
|
867
|
841 return raw_16_bit_fixed_to_ichar (* (UINT_16_BIT *) ptr, object);
|
|
826
|
842 case FORMAT_32_BIT_FIXED:
|
|
1204
|
843 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT));
|
|
867
|
844 return raw_32_bit_fixed_to_ichar (* (UINT_32_BIT *) ptr, object);
|
|
826
|
845 default:
|
|
|
846 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
867
|
847 return raw_8_bit_fixed_to_ichar (*ptr, object);
|
|
826
|
848 }
|
|
|
849 }
|
|
|
850
|
|
|
851 /* Return the character at PTR (which is in format FMT), suitable for
|
|
|
852 comparison with an ASCII character. This guarantees that if the
|
|
|
853 character at PTR is ASCII (range 0 - 127), that character will be
|
|
|
854 returned; otherwise, some character outside of the ASCII range will be
|
|
|
855 returned, but not necessarily the character actually at PTR. This will
|
|
867
|
856 be faster than itext_ichar_fmt() for some formats -- in particular,
|
|
826
|
857 FORMAT_DEFAULT. */
|
|
|
858
|
|
|
859 DECLARE_INLINE_HEADER (
|
|
867
|
860 Ichar
|
|
|
861 itext_ichar_ascii_fmt (const Ibyte *ptr, Internal_Format fmt,
|
|
826
|
862 Lisp_Object object)
|
|
|
863 )
|
|
|
864 {
|
|
|
865 switch (fmt)
|
|
|
866 {
|
|
|
867 case FORMAT_DEFAULT:
|
|
867
|
868 return (Ichar) *ptr;
|
|
826
|
869 case FORMAT_16_BIT_FIXED:
|
|
1204
|
870 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT));
|
|
867
|
871 return raw_16_bit_fixed_to_ichar (* (UINT_16_BIT *) ptr, object);
|
|
826
|
872 case FORMAT_32_BIT_FIXED:
|
|
1204
|
873 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT));
|
|
867
|
874 return raw_32_bit_fixed_to_ichar (* (UINT_32_BIT *) ptr, object);
|
|
826
|
875 default:
|
|
|
876 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
867
|
877 return raw_8_bit_fixed_to_ichar (*ptr, object);
|
|
826
|
878 }
|
|
|
879 }
|
|
|
880
|
|
|
881 /* Return the "raw value" of the character at PTR, in format FMT. This is
|
|
|
882 useful when searching for a character; convert the character using
|
|
867
|
883 ichar_to_raw(). */
|
|
826
|
884
|
|
|
885 DECLARE_INLINE_HEADER (
|
|
867
|
886 Raw_Ichar
|
|
|
887 itext_ichar_raw_fmt (const Ibyte *ptr, Internal_Format fmt)
|
|
826
|
888 )
|
|
|
889 {
|
|
|
890 switch (fmt)
|
|
|
891 {
|
|
|
892 case FORMAT_DEFAULT:
|
|
867
|
893 return (Raw_Ichar) itext_ichar (ptr);
|
|
826
|
894 case FORMAT_16_BIT_FIXED:
|
|
1204
|
895 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT));
|
|
867
|
896 return (Raw_Ichar) (* (UINT_16_BIT *) ptr);
|
|
826
|
897 case FORMAT_32_BIT_FIXED:
|
|
1204
|
898 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT));
|
|
867
|
899 return (Raw_Ichar) (* (UINT_32_BIT *) ptr);
|
|
826
|
900 default:
|
|
|
901 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
867
|
902 return (Raw_Ichar) (*ptr);
|
|
826
|
903 }
|
|
|
904 }
|
|
|
905
|
|
867
|
906 /* Store the character CH (an Ichar) as internally-formatted text starting
|
|
826
|
907 at PTR. Return the number of bytes stored.
|
|
|
908 */
|
|
|
909
|
|
|
910 DECLARE_INLINE_HEADER (
|
|
|
911 Bytecount
|
|
867
|
912 set_itext_ichar (Ibyte *ptr, Ichar x)
|
|
826
|
913 )
|
|
771
|
914 {
|
|
867
|
915 return !ichar_multibyte_p (x) ?
|
|
|
916 simple_set_itext_ichar (ptr, x) :
|
|
|
917 non_ascii_set_itext_ichar (ptr, x);
|
|
771
|
918 }
|
|
|
919
|
|
867
|
920 /* Store the character CH (an Ichar) as internally-formatted text of
|
|
826
|
921 format FMT starting at PTR, which comes from OBJECT. Return the number
|
|
|
922 of bytes stored.
|
|
|
923 */
|
|
|
924
|
|
|
925 DECLARE_INLINE_HEADER (
|
|
|
926 Bytecount
|
|
867
|
927 set_itext_ichar_fmt (Ibyte *ptr, Ichar x, Internal_Format fmt,
|
|
826
|
928 Lisp_Object object)
|
|
|
929 )
|
|
771
|
930 {
|
|
826
|
931 switch (fmt)
|
|
|
932 {
|
|
|
933 case FORMAT_DEFAULT:
|
|
867
|
934 return set_itext_ichar (ptr, x);
|
|
826
|
935 case FORMAT_16_BIT_FIXED:
|
|
867
|
936 text_checking_assert (ichar_16_bit_fixed_p (x, object));
|
|
1204
|
937 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_16_BIT));
|
|
867
|
938 * (UINT_16_BIT *) ptr = ichar_to_raw_16_bit_fixed (x, object);
|
|
826
|
939 return 2;
|
|
|
940 case FORMAT_32_BIT_FIXED:
|
|
1204
|
941 text_checking_assert ((void *) ptr == ALIGN_PTR (ptr, UINT_32_BIT));
|
|
867
|
942 * (UINT_32_BIT *) ptr = ichar_to_raw_32_bit_fixed (x, object);
|
|
826
|
943 return 4;
|
|
|
944 default:
|
|
|
945 text_checking_assert (fmt == FORMAT_8_BIT_FIXED);
|
|
867
|
946 text_checking_assert (ichar_8_bit_fixed_p (x, object));
|
|
|
947 *ptr = ichar_to_raw_8_bit_fixed (x, object);
|
|
826
|
948 return 1;
|
|
|
949 }
|
|
|
950 }
|
|
|
951
|
|
|
952 /* Retrieve the character pointed to by SRC and store it as
|
|
|
953 internally-formatted text in DST.
|
|
|
954 */
|
|
|
955
|
|
|
956 DECLARE_INLINE_HEADER (
|
|
|
957 Bytecount
|
|
867
|
958 itext_copy_ichar (const Ibyte *src, Ibyte *dst)
|
|
826
|
959 )
|
|
|
960 {
|
|
|
961 return byte_ascii_p (*src) ?
|
|
867
|
962 simple_itext_copy_ichar (src, dst) :
|
|
|
963 non_ascii_itext_copy_ichar (src, dst);
|
|
771
|
964 }
|
|
|
965
|
|
|
966 #else /* not MULE */
|
|
|
967
|
|
867
|
968 # define itext_ichar(ptr) simple_itext_ichar (ptr)
|
|
|
969 # define itext_ichar_fmt(ptr, fmt, object) itext_ichar (ptr)
|
|
|
970 # define itext_ichar_ascii_fmt(ptr, fmt, object) itext_ichar (ptr)
|
|
|
971 # define itext_ichar_raw_fmt(ptr, fmt) itext_ichar (ptr)
|
|
|
972 # define set_itext_ichar(ptr, x) simple_set_itext_ichar (ptr, x)
|
|
|
973 # define set_itext_ichar_fmt(ptr, x, fmt, obj) set_itext_ichar (ptr, x)
|
|
|
974 # define itext_copy_ichar(src, dst) simple_itext_copy_ichar (src, dst)
|
|
771
|
975
|
|
|
976 #endif /* not MULE */
|
|
|
977
|
|
826
|
978 /* Retrieve the character at offset N (in characters) from PTR, as an
|
|
867
|
979 Ichar.
|
|
826
|
980 */
|
|
|
981
|
|
867
|
982 #define itext_ichar_n(ptr, offset) \
|
|
|
983 itext_ichar (itext_n_addr (ptr, offset))
|
|
771
|
984
|
|
|
985
|
|
|
986 /* ---------------------------- */
|
|
867
|
987 /* Working with Ichars */
|
|
771
|
988 /* ---------------------------- */
|
|
|
989
|
|
867
|
990 /* NOTE: There are other functions/macros for working with Ichars in
|
|
|
991 charset.h, for retrieving the charset of an Ichar, the length of an
|
|
|
992 Ichar when converted to text, etc.
|
|
826
|
993 */
|
|
|
994
|
|
771
|
995 #ifdef MULE
|
|
|
996
|
|
867
|
997 int non_ascii_valid_ichar_p (Ichar ch);
|
|
|
998
|
|
|
999 /* Return whether the given Ichar is valid.
|
|
826
|
1000 */
|
|
|
1001
|
|
|
1002 DECLARE_INLINE_HEADER (
|
|
|
1003 int
|
|
867
|
1004 valid_ichar_p (Ichar ch)
|
|
826
|
1005 )
|
|
771
|
1006 {
|
|
867
|
1007 return (! (ch & ~0xFF)) || non_ascii_valid_ichar_p (ch);
|
|
771
|
1008 }
|
|
|
1009
|
|
|
1010 #else /* not MULE */
|
|
|
1011
|
|
867
|
1012 #define valid_ichar_p(ch) (! (ch & ~0xFF))
|
|
771
|
1013
|
|
|
1014 #endif /* not MULE */
|
|
|
1015
|
|
831
|
1016 DECLARE_INLINE_HEADER (
|
|
|
1017 Lisp_Object
|
|
867
|
1018 make_char (Ichar val)
|
|
831
|
1019 )
|
|
|
1020 {
|
|
867
|
1021 type_checking_assert (valid_ichar_p (val));
|
|
831
|
1022 return make_char_1 (val);
|
|
|
1023 }
|
|
|
1024
|
|
867
|
1025 #define CHAR_INTP(x) (INTP (x) && valid_ichar_p (XINT (x)))
|
|
771
|
1026
|
|
|
1027 #define CHAR_OR_CHAR_INTP(x) (CHARP (x) || CHAR_INTP (x))
|
|
|
1028
|
|
826
|
1029 DECLARE_INLINE_HEADER (
|
|
867
|
1030 Ichar
|
|
771
|
1031 XCHAR_OR_CHAR_INT (Lisp_Object obj)
|
|
826
|
1032 )
|
|
771
|
1033 {
|
|
|
1034 return CHARP (obj) ? XCHAR (obj) : XINT (obj);
|
|
|
1035 }
|
|
|
1036
|
|
826
|
1037 /* Signal an error if CH is not a valid character or integer Lisp_Object.
|
|
|
1038 If CH is an integer Lisp_Object, convert it to a character Lisp_Object,
|
|
|
1039 but merely by repackaging, without performing tests for char validity.
|
|
|
1040 */
|
|
|
1041
|
|
771
|
1042 #define CHECK_CHAR_COERCE_INT(x) do { \
|
|
|
1043 if (CHARP (x)) \
|
|
|
1044 ; \
|
|
|
1045 else if (CHAR_INTP (x)) \
|
|
|
1046 x = make_char (XINT (x)); \
|
|
|
1047 else \
|
|
|
1048 x = wrong_type_argument (Qcharacterp, x); \
|
|
|
1049 } while (0)
|
|
|
1050
|
|
|
1051
|
|
|
1052
|
|
|
1053 /************************************************************************/
|
|
|
1054 /* */
|
|
826
|
1055 /* working with Lisp strings */
|
|
|
1056 /* */
|
|
|
1057 /************************************************************************/
|
|
|
1058
|
|
|
1059 #define string_char_length(s) \
|
|
|
1060 string_index_byte_to_char (s, XSTRING_LENGTH (s))
|
|
|
1061 #define string_byte(s, i) (XSTRING_DATA (s)[i] + 0)
|
|
|
1062 /* In case we ever allow strings to be in a different format ... */
|
|
|
1063 #define set_string_byte(s, i, c) (XSTRING_DATA (s)[i] = (c))
|
|
|
1064
|
|
|
1065 #define ASSERT_VALID_CHAR_STRING_INDEX_UNSAFE(s, x) do { \
|
|
|
1066 text_checking_assert ((x) >= 0 && x <= string_char_length (s)); \
|
|
|
1067 } while (0)
|
|
|
1068
|
|
|
1069 #define ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE(s, x) do { \
|
|
|
1070 text_checking_assert ((x) >= 0 && x <= XSTRING_LENGTH (s)); \
|
|
867
|
1071 text_checking_assert (valid_ibyteptr_p (string_byte_addr (s, x))); \
|
|
826
|
1072 } while (0)
|
|
|
1073
|
|
|
1074 /* Convert offset I in string S to a pointer to text there. */
|
|
|
1075 #define string_byte_addr(s, i) (&(XSTRING_DATA (s)[i]))
|
|
|
1076 /* Convert pointer to text in string S into the byte offset to that text. */
|
|
|
1077 #define string_addr_to_byte(s, ptr) ((Bytecount) ((ptr) - XSTRING_DATA (s)))
|
|
867
|
1078 /* Return the Ichar at *CHARACTER* offset I. */
|
|
|
1079 #define string_ichar(s, i) itext_ichar (string_char_addr (s, i))
|
|
826
|
1080
|
|
|
1081 #ifdef ERROR_CHECK_TEXT
|
|
|
1082 #define SLEDGEHAMMER_CHECK_ASCII_BEGIN
|
|
|
1083 #endif
|
|
|
1084
|
|
|
1085 #ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN
|
|
|
1086 void sledgehammer_check_ascii_begin (Lisp_Object str);
|
|
|
1087 #else
|
|
|
1088 #define sledgehammer_check_ascii_begin(str)
|
|
|
1089 #endif
|
|
|
1090
|
|
|
1091 /* Make an alloca'd copy of a Lisp string */
|
|
|
1092 #define LISP_STRING_TO_ALLOCA(s, lval) \
|
|
|
1093 do { \
|
|
1315
|
1094 Ibyte **_lta_ = (Ibyte **) &(lval); \
|
|
826
|
1095 Lisp_Object _lta_2 = (s); \
|
|
867
|
1096 *_lta_ = alloca_array (Ibyte, 1 + XSTRING_LENGTH (_lta_2)); \
|
|
826
|
1097 memcpy (*_lta_, XSTRING_DATA (_lta_2), 1 + XSTRING_LENGTH (_lta_2)); \
|
|
|
1098 } while (0)
|
|
|
1099
|
|
1315
|
1100 /* Make an alloca'd copy of a Ibyte * */ \
|
|
867
|
1101 #define IBYTE_STRING_TO_ALLOCA(p, lval) \
|
|
1315
|
1102 do { \
|
|
867
|
1103 Ibyte **_bsta_ = (Ibyte **) &(lval); \
|
|
1315
|
1104 const Ibyte *_bsta_2 = (p); \
|
|
|
1105 Bytecount _bsta_3 = qxestrlen (_bsta_2); \
|
|
867
|
1106 *_bsta_ = alloca_array (Ibyte, 1 + _bsta_3); \
|
|
1315
|
1107 memcpy (*_bsta_, _bsta_2, 1 + _bsta_3); \
|
|
826
|
1108 } while (0)
|
|
|
1109
|
|
851
|
1110
|
|
867
|
1111 #define alloca_ibytes(num) alloca_array (Ibyte, num)
|
|
826
|
1112 #define alloca_extbytes(num) alloca_array (Extbyte, num)
|
|
|
1113
|
|
|
1114 void resize_string (Lisp_Object s, Bytecount pos, Bytecount delta);
|
|
|
1115
|
|
|
1116 /* Convert a byte index into a string into a char index. */
|
|
|
1117 DECLARE_INLINE_HEADER (
|
|
|
1118 Charcount
|
|
|
1119 string_index_byte_to_char (Lisp_Object s, Bytecount idx)
|
|
|
1120 )
|
|
|
1121 {
|
|
|
1122 Charcount retval;
|
|
|
1123 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, idx);
|
|
|
1124 #ifdef MULE
|
|
|
1125 if (idx <= (Bytecount) XSTRING_ASCII_BEGIN (s))
|
|
|
1126 retval = (Charcount) idx;
|
|
|
1127 else
|
|
|
1128 retval = (XSTRING_ASCII_BEGIN (s) +
|
|
|
1129 bytecount_to_charcount (XSTRING_DATA (s) +
|
|
|
1130 XSTRING_ASCII_BEGIN (s),
|
|
|
1131 idx - XSTRING_ASCII_BEGIN (s)));
|
|
|
1132 # ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN
|
|
|
1133 assert (retval == bytecount_to_charcount (XSTRING_DATA (s), idx));
|
|
|
1134 # endif
|
|
|
1135 #else
|
|
|
1136 retval = (Charcount) idx;
|
|
|
1137 #endif
|
|
|
1138 /* Don't call ASSERT_VALID_CHAR_STRING_INDEX_UNSAFE() here because it will
|
|
|
1139 call string_index_byte_to_char(). */
|
|
|
1140 return retval;
|
|
|
1141 }
|
|
|
1142
|
|
|
1143 /* Convert a char index into a string into a byte index. */
|
|
|
1144 DECLARE_INLINE_HEADER (
|
|
|
1145 Bytecount
|
|
|
1146 string_index_char_to_byte (Lisp_Object s, Charcount idx)
|
|
|
1147 )
|
|
|
1148 {
|
|
|
1149 Bytecount retval;
|
|
|
1150 ASSERT_VALID_CHAR_STRING_INDEX_UNSAFE (s, idx);
|
|
|
1151 #ifdef MULE
|
|
|
1152 if (idx <= (Charcount) XSTRING_ASCII_BEGIN (s))
|
|
|
1153 retval = (Bytecount) idx;
|
|
|
1154 else
|
|
|
1155 retval = (XSTRING_ASCII_BEGIN (s) +
|
|
|
1156 charcount_to_bytecount (XSTRING_DATA (s) +
|
|
|
1157 XSTRING_ASCII_BEGIN (s),
|
|
|
1158 idx - XSTRING_ASCII_BEGIN (s)));
|
|
|
1159 # ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN
|
|
|
1160 assert (retval == charcount_to_bytecount (XSTRING_DATA (s), idx));
|
|
|
1161 # endif
|
|
|
1162 #else
|
|
|
1163 retval = (Bytecount) idx;
|
|
|
1164 #endif
|
|
|
1165 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, retval);
|
|
|
1166 return retval;
|
|
|
1167 }
|
|
|
1168
|
|
|
1169 /* Convert a substring length (starting at byte offset OFF) from bytes to
|
|
|
1170 chars. */
|
|
|
1171 DECLARE_INLINE_HEADER (
|
|
|
1172 Charcount
|
|
|
1173 string_offset_byte_to_char_len (Lisp_Object s, Bytecount off, Bytecount len)
|
|
|
1174 )
|
|
|
1175 {
|
|
|
1176 Charcount retval;
|
|
|
1177 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, off);
|
|
|
1178 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, off + len);
|
|
|
1179 #ifdef MULE
|
|
|
1180 if (off + len <= (Bytecount) XSTRING_ASCII_BEGIN (s))
|
|
|
1181 retval = (Charcount) len;
|
|
|
1182 else if (off < (Bytecount) XSTRING_ASCII_BEGIN (s))
|
|
|
1183 retval =
|
|
|
1184 XSTRING_ASCII_BEGIN (s) - (Charcount) off +
|
|
|
1185 bytecount_to_charcount (XSTRING_DATA (s) + XSTRING_ASCII_BEGIN (s),
|
|
|
1186 len - (XSTRING_ASCII_BEGIN (s) - off));
|
|
|
1187 else
|
|
|
1188 retval = bytecount_to_charcount (XSTRING_DATA (s) + off, len);
|
|
|
1189 # ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN
|
|
|
1190 assert (retval == bytecount_to_charcount (XSTRING_DATA (s) + off, len));
|
|
|
1191 # endif
|
|
|
1192 #else
|
|
|
1193 retval = (Charcount) len;
|
|
|
1194 #endif
|
|
|
1195 return retval;
|
|
|
1196 }
|
|
|
1197
|
|
|
1198 /* Convert a substring length (starting at byte offset OFF) from chars to
|
|
|
1199 bytes. */
|
|
|
1200 DECLARE_INLINE_HEADER (
|
|
|
1201 Bytecount
|
|
|
1202 string_offset_char_to_byte_len (Lisp_Object s, Bytecount off, Charcount len)
|
|
|
1203 )
|
|
|
1204 {
|
|
|
1205 Bytecount retval;
|
|
|
1206 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, off);
|
|
|
1207 #ifdef MULE
|
|
|
1208 /* casts to avoid errors from combining Bytecount/Charcount and warnings
|
|
|
1209 from signed/unsigned comparisons */
|
|
|
1210 if (off + (Bytecount) len <= (Bytecount) XSTRING_ASCII_BEGIN (s))
|
|
|
1211 retval = (Bytecount) len;
|
|
|
1212 else if (off < (Bytecount) XSTRING_ASCII_BEGIN (s))
|
|
|
1213 retval =
|
|
|
1214 XSTRING_ASCII_BEGIN (s) - off +
|
|
|
1215 charcount_to_bytecount (XSTRING_DATA (s) + XSTRING_ASCII_BEGIN (s),
|
|
|
1216 len - (XSTRING_ASCII_BEGIN (s) -
|
|
|
1217 (Charcount) off));
|
|
|
1218 else
|
|
|
1219 retval = charcount_to_bytecount (XSTRING_DATA (s) + off, len);
|
|
|
1220 # ifdef SLEDGEHAMMER_CHECK_ASCII_BEGIN
|
|
|
1221 assert (retval == charcount_to_bytecount (XSTRING_DATA (s) + off, len));
|
|
|
1222 # endif
|
|
|
1223 #else
|
|
|
1224 retval = (Bytecount) len;
|
|
|
1225 #endif
|
|
|
1226 ASSERT_VALID_BYTE_STRING_INDEX_UNSAFE (s, off + retval);
|
|
|
1227 return retval;
|
|
|
1228 }
|
|
|
1229
|
|
|
1230 DECLARE_INLINE_HEADER (
|
|
867
|
1231 const Ibyte *
|
|
826
|
1232 string_char_addr (Lisp_Object s, Charcount idx)
|
|
|
1233 )
|
|
|
1234 {
|
|
|
1235 return XSTRING_DATA (s) + string_index_char_to_byte (s, idx);
|
|
|
1236 }
|
|
|
1237
|
|
|
1238 /* WARNING: If you modify an existing string, you must call
|
|
|
1239 bump_string_modiff() afterwards. */
|
|
|
1240 #ifdef MULE
|
|
867
|
1241 void set_string_char (Lisp_Object s, Charcount i, Ichar c);
|
|
826
|
1242 #else
|
|
|
1243 #define set_string_char(s, i, c) set_string_byte (s, i, c)
|
|
|
1244 #endif /* not MULE */
|
|
|
1245
|
|
|
1246 /* Return index to character before the one at IDX. */
|
|
|
1247 DECLARE_INLINE_HEADER (
|
|
|
1248 Bytecount
|
|
|
1249 prev_string_index (Lisp_Object s, Bytecount idx)
|
|
|
1250 )
|
|
|
1251 {
|
|
867
|
1252 const Ibyte *ptr = string_byte_addr (s, idx);
|
|
|
1253 DEC_IBYTEPTR (ptr);
|
|
826
|
1254 return string_addr_to_byte (s, ptr);
|
|
|
1255 }
|
|
|
1256
|
|
|
1257 /* Return index to character after the one at IDX. */
|
|
|
1258 DECLARE_INLINE_HEADER (
|
|
|
1259 Bytecount
|
|
|
1260 next_string_index (Lisp_Object s, Bytecount idx)
|
|
|
1261 )
|
|
|
1262 {
|
|
867
|
1263 const Ibyte *ptr = string_byte_addr (s, idx);
|
|
|
1264 INC_IBYTEPTR (ptr);
|
|
826
|
1265 return string_addr_to_byte (s, ptr);
|
|
|
1266 }
|
|
|
1267
|
|
|
1268
|
|
|
1269 /************************************************************************/
|
|
|
1270 /* */
|
|
771
|
1271 /* working with Eistrings */
|
|
|
1272 /* */
|
|
|
1273 /************************************************************************/
|
|
|
1274
|
|
|
1275 /*
|
|
|
1276 #### NOTE: This is a work in progress. Neither the API nor especially
|
|
|
1277 the implementation is finished.
|
|
|
1278
|
|
|
1279 NOTE: An Eistring is a structure that makes it easy to work with
|
|
|
1280 internally-formatted strings of data. It provides operations similar
|
|
|
1281 in feel to the standard strcpy(), strcat(), strlen(), etc., but
|
|
|
1282
|
|
|
1283 (a) it is Mule-correct
|
|
|
1284 (b) it does dynamic allocation so you never have to worry about size
|
|
793
|
1285 restrictions
|
|
851
|
1286 (c) it comes in an ALLOCA() variety (all allocation is stack-local,
|
|
793
|
1287 so there is no need to explicitly clean up) as well as a malloc()
|
|
|
1288 variety
|
|
|
1289 (d) it knows its own length, so it does not suffer from standard null
|
|
|
1290 byte brain-damage -- but it null-terminates the data anyway, so
|
|
|
1291 it can be passed to standard routines
|
|
|
1292 (e) it provides a much more powerful set of operations and knows about
|
|
771
|
1293 all the standard places where string data might reside: Lisp_Objects,
|
|
867
|
1294 other Eistrings, Ibyte * data with or without an explicit length,
|
|
|
1295 ASCII strings, Ichars, etc.
|
|
793
|
1296 (f) it provides easy operations to convert to/from externally-formatted
|
|
|
1297 data, and is easier to use than the standard TO_INTERNAL_FORMAT
|
|
771
|
1298 and TO_EXTERNAL_FORMAT macros. (An Eistring can store both the internal
|
|
|
1299 and external version of its data, but the external version is only
|
|
|
1300 initialized or changed when you call eito_external().)
|
|
|
1301
|
|
793
|
1302 The idea is to make it as easy to write Mule-correct string manipulation
|
|
|
1303 code as it is to write normal string manipulation code. We also make
|
|
|
1304 the API sufficiently general that it can handle multiple internal data
|
|
|
1305 formats (e.g. some fixed-width optimizing formats and a default variable
|
|
|
1306 width format) and allows for *ANY* data format we might choose in the
|
|
|
1307 future for the default format, including UCS2. (In other words, we can't
|
|
|
1308 assume that the internal format is ASCII-compatible and we can't assume
|
|
|
1309 it doesn't have embedded null bytes. We do assume, however, that any
|
|
|
1310 chosen format will have the concept of null-termination.) All of this is
|
|
|
1311 hidden from the user.
|
|
771
|
1312
|
|
|
1313 #### It is really too bad that we don't have a real object-oriented
|
|
|
1314 language, or at least a language with polymorphism!
|
|
|
1315
|
|
|
1316
|
|
|
1317 **********************************************
|
|
|
1318 * Declaration *
|
|
|
1319 **********************************************
|
|
|
1320
|
|
|
1321 To declare an Eistring, either put one of the following in the local
|
|
|
1322 variable section:
|
|
|
1323
|
|
|
1324 DECLARE_EISTRING (name);
|
|
|
1325 Declare a new Eistring. This is a standard local variable declaration
|
|
|
1326 and can go anywhere in the variable declaration section. NAME itself
|
|
|
1327 is declared as an Eistring *, and its storage declared on the stack.
|
|
|
1328
|
|
|
1329 DECLARE_EISTRING_MALLOC (name);
|
|
851
|
1330 Declare a new Eistring, which uses malloc()ed instead of ALLOCA()ed
|
|
771
|
1331 data. This is a standard local variable declaration and can go
|
|
|
1332 anywhere in the variable declaration section. Once you initialize
|
|
|
1333 the Eistring, you will have to free it using eifree() to avoid
|
|
793
|
1334 memory leaks. You will need to use this form if you are passing
|
|
|
1335 an Eistring to any function that modifies it (otherwise, the
|
|
|
1336 modified data may be in stack space and get overwritten when the
|
|
|
1337 function returns).
|
|
771
|
1338
|
|
|
1339 or use
|
|
|
1340
|
|
793
|
1341 Eistring ei;
|
|
|
1342 void eiinit (Eistring *ei);
|
|
|
1343 void eiinit_malloc (Eistring *einame);
|
|
771
|
1344 If you need to put an Eistring elsewhere than in a local variable
|
|
|
1345 declaration (e.g. in a structure), declare it as shown and then
|
|
|
1346 call one of the init macros.
|
|
|
1347
|
|
|
1348 Also note:
|
|
|
1349
|
|
793
|
1350 void eifree (Eistring *ei);
|
|
771
|
1351 If you declared an Eistring to use malloc() to hold its data,
|
|
|
1352 or converted it to the heap using eito_malloc(), then this
|
|
|
1353 releases any data in it and afterwards resets the Eistring
|
|
|
1354 using eiinit_malloc(). Otherwise, it just resets the Eistring
|
|
|
1355 using eiinit().
|
|
|
1356
|
|
|
1357
|
|
|
1358 **********************************************
|
|
|
1359 * Conventions *
|
|
|
1360 **********************************************
|
|
|
1361
|
|
|
1362 - The names of the functions have been chosen, where possible, to
|
|
|
1363 match the names of str*() functions in the standard C API.
|
|
|
1364 -
|
|
|
1365
|
|
|
1366
|
|
|
1367 **********************************************
|
|
|
1368 * Initialization *
|
|
|
1369 **********************************************
|
|
|
1370
|
|
|
1371 void eireset (Eistring *eistr);
|
|
|
1372 Initialize the Eistring to the empty string.
|
|
|
1373
|
|
|
1374 void eicpy_* (Eistring *eistr, ...);
|
|
|
1375 Initialize the Eistring from somewhere:
|
|
|
1376
|
|
|
1377 void eicpy_ei (Eistring *eistr, Eistring *eistr2);
|
|
|
1378 ... from another Eistring.
|
|
|
1379 void eicpy_lstr (Eistring *eistr, Lisp_Object lisp_string);
|
|
|
1380 ... from a Lisp_Object string.
|
|
867
|
1381 void eicpy_ch (Eistring *eistr, Ichar ch);
|
|
|
1382 ... from an Ichar (this can be a conventional C character).
|
|
771
|
1383
|
|
|
1384 void eicpy_lstr_off (Eistring *eistr, Lisp_Object lisp_string,
|
|
|
1385 Bytecount off, Charcount charoff,
|
|
|
1386 Bytecount len, Charcount charlen);
|
|
|
1387 ... from a section of a Lisp_Object string.
|
|
|
1388 void eicpy_lbuf (Eistring *eistr, Lisp_Object lisp_buf,
|
|
|
1389 Bytecount off, Charcount charoff,
|
|
|
1390 Bytecount len, Charcount charlen);
|
|
|
1391 ... from a section of a Lisp_Object buffer.
|
|
867
|
1392 void eicpy_raw (Eistring *eistr, const Ibyte *data, Bytecount len);
|
|
771
|
1393 ... from raw internal-format data in the default internal format.
|
|
867
|
1394 void eicpy_rawz (Eistring *eistr, const Ibyte *data);
|
|
771
|
1395 ... from raw internal-format data in the default internal format
|
|
|
1396 that is "null-terminated" (the meaning of this depends on the nature
|
|
|
1397 of the default internal format).
|
|
867
|
1398 void eicpy_raw_fmt (Eistring *eistr, const Ibyte *data, Bytecount len,
|
|
826
|
1399 Internal_Format intfmt, Lisp_Object object);
|
|
771
|
1400 ... from raw internal-format data in the specified format.
|
|
867
|
1401 void eicpy_rawz_fmt (Eistring *eistr, const Ibyte *data,
|
|
826
|
1402 Internal_Format intfmt, Lisp_Object object);
|
|
771
|
1403 ... from raw internal-format data in the specified format that is
|
|
|
1404 "null-terminated" (the meaning of this depends on the nature of
|
|
|
1405 the specific format).
|
|
|
1406 void eicpy_c (Eistring *eistr, const Char_ASCII *c_string);
|
|
|
1407 ... from an ASCII null-terminated string. Non-ASCII characters in
|
|
|
1408 the string are *ILLEGAL* (read abort() with error-checking defined).
|
|
|
1409 void eicpy_c_len (Eistring *eistr, const Char_ASCII *c_string, len);
|
|
|
1410 ... from an ASCII string, with length specified. Non-ASCII characters
|
|
|
1411 in the string are *ILLEGAL* (read abort() with error-checking defined).
|
|
|
1412 void eicpy_ext (Eistring *eistr, const Extbyte *extdata,
|
|
|
1413 Lisp_Object coding_system);
|
|
|
1414 ... from external null-terminated data, with coding system specified.
|
|
|
1415 void eicpy_ext_len (Eistring *eistr, const Extbyte *extdata,
|
|
|
1416 Bytecount extlen, Lisp_Object coding_system);
|
|
|
1417 ... from external data, with length and coding system specified.
|
|
|
1418 void eicpy_lstream (Eistring *eistr, Lisp_Object lstream);
|
|
|
1419 ... from an lstream; reads data till eof. Data must be in default
|
|
|
1420 internal format; otherwise, interpose a decoding lstream.
|
|
|
1421
|
|
|
1422
|
|
|
1423 **********************************************
|
|
|
1424 * Getting the data out of the Eistring *
|
|
|
1425 **********************************************
|
|
|
1426
|
|
867
|
1427 Ibyte *eidata (Eistring *eistr);
|
|
771
|
1428 Return a pointer to the raw data in an Eistring. This is NOT
|
|
|
1429 a copy.
|
|
|
1430
|
|
|
1431 Lisp_Object eimake_string (Eistring *eistr);
|
|
|
1432 Make a Lisp string out of the Eistring.
|
|
|
1433
|
|
|
1434 Lisp_Object eimake_string_off (Eistring *eistr,
|
|
|
1435 Bytecount off, Charcount charoff,
|
|
|
1436 Bytecount len, Charcount charlen);
|
|
|
1437 Make a Lisp string out of a section of the Eistring.
|
|
|
1438
|
|
867
|
1439 void eicpyout_alloca (Eistring *eistr, LVALUE: Ibyte *ptr_out,
|
|
771
|
1440 LVALUE: Bytecount len_out);
|
|
851
|
1441 Make an ALLOCA() copy of the data in the Eistring, using the
|
|
|
1442 default internal format. Due to the nature of ALLOCA(), this
|
|
771
|
1443 must be a macro, with all lvalues passed in as parameters.
|
|
793
|
1444 (More specifically, not all compilers correctly handle using
|
|
851
|
1445 ALLOCA() as the argument to a function call -- GCC on x86
|
|
|
1446 didn't used to, for example.) A pointer to the ALLOCA()ed data
|
|
793
|
1447 is stored in PTR_OUT, and the length of the data (not including
|
|
|
1448 the terminating zero) is stored in LEN_OUT.
|
|
771
|
1449
|
|
867
|
1450 void eicpyout_alloca_fmt (Eistring *eistr, LVALUE: Ibyte *ptr_out,
|
|
771
|
1451 LVALUE: Bytecount len_out,
|
|
826
|
1452 Internal_Format intfmt, Lisp_Object object);
|
|
771
|
1453 Like eicpyout_alloca(), but converts to the specified internal
|
|
|
1454 format. (No formats other than FORMAT_DEFAULT are currently
|
|
|
1455 implemented, and you get an assertion failure if you try.)
|
|
|
1456
|
|
867
|
1457 Ibyte *eicpyout_malloc (Eistring *eistr, Bytecount *intlen_out);
|
|
771
|
1458 Make a malloc() copy of the data in the Eistring, using the
|
|
|
1459 default internal format. This is a real function. No lvalues
|
|
|
1460 passed in. Returns the new data, and stores the length (not
|
|
|
1461 including the terminating zero) using INTLEN_OUT, unless it's
|
|
|
1462 a NULL pointer.
|
|
|
1463
|
|
867
|
1464 Ibyte *eicpyout_malloc_fmt (Eistring *eistr, Internal_Format intfmt,
|
|
826
|
1465 Bytecount *intlen_out, Lisp_Object object);
|
|
771
|
1466 Like eicpyout_malloc(), but converts to the specified internal
|
|
|
1467 format. (No formats other than FORMAT_DEFAULT are currently
|
|
|
1468 implemented, and you get an assertion failure if you try.)
|
|
|
1469
|
|
|
1470
|
|
|
1471 **********************************************
|
|
|
1472 * Moving to the heap *
|
|
|
1473 **********************************************
|
|
|
1474
|
|
|
1475 void eito_malloc (Eistring *eistr);
|
|
|
1476 Move this Eistring to the heap. Its data will be stored in a
|
|
|
1477 malloc()ed block rather than the stack. Subsequent changes to
|
|
|
1478 this Eistring will realloc() the block as necessary. Use this
|
|
|
1479 when you want the Eistring to remain in scope past the end of
|
|
|
1480 this function call. You will have to manually free the data
|
|
|
1481 in the Eistring using eifree().
|
|
|
1482
|
|
|
1483 void eito_alloca (Eistring *eistr);
|
|
|
1484 Move this Eistring back to the stack, if it was moved to the
|
|
|
1485 heap with eito_malloc(). This will automatically free any
|
|
|
1486 heap-allocated data.
|
|
|
1487
|
|
|
1488
|
|
|
1489
|
|
|
1490 **********************************************
|
|
|
1491 * Retrieving the length *
|
|
|
1492 **********************************************
|
|
|
1493
|
|
|
1494 Bytecount eilen (Eistring *eistr);
|
|
|
1495 Return the length of the internal data, in bytes. See also
|
|
|
1496 eiextlen(), below.
|
|
|
1497 Charcount eicharlen (Eistring *eistr);
|
|
|
1498 Return the length of the internal data, in characters.
|
|
|
1499
|
|
|
1500
|
|
|
1501 **********************************************
|
|
|
1502 * Working with positions *
|
|
|
1503 **********************************************
|
|
|
1504
|
|
|
1505 Bytecount eicharpos_to_bytepos (Eistring *eistr, Charcount charpos);
|
|
|
1506 Convert a char offset to a byte offset.
|
|
|
1507 Charcount eibytepos_to_charpos (Eistring *eistr, Bytecount bytepos);
|
|
|
1508 Convert a byte offset to a char offset.
|
|
|
1509 Bytecount eiincpos (Eistring *eistr, Bytecount bytepos);
|
|
|
1510 Increment the given position by one character.
|
|
|
1511 Bytecount eiincpos_n (Eistring *eistr, Bytecount bytepos, Charcount n);
|
|
|
1512 Increment the given position by N characters.
|
|
|
1513 Bytecount eidecpos (Eistring *eistr, Bytecount bytepos);
|
|
|
1514 Decrement the given position by one character.
|
|
|
1515 Bytecount eidecpos_n (Eistring *eistr, Bytecount bytepos, Charcount n);
|
|
|
1516 Deccrement the given position by N characters.
|
|
|
1517
|
|
|
1518
|
|
|
1519 **********************************************
|
|
|
1520 * Getting the character at a position *
|
|
|
1521 **********************************************
|
|
|
1522
|
|
867
|
1523 Ichar eigetch (Eistring *eistr, Bytecount bytepos);
|
|
771
|
1524 Return the character at a particular byte offset.
|
|
867
|
1525 Ichar eigetch_char (Eistring *eistr, Charcount charpos);
|
|
771
|
1526 Return the character at a particular character offset.
|
|
|
1527
|
|
|
1528
|
|
|
1529 **********************************************
|
|
|
1530 * Setting the character at a position *
|
|
|
1531 **********************************************
|
|
|
1532
|
|
867
|
1533 Ichar eisetch (Eistring *eistr, Bytecount bytepos, Ichar chr);
|
|
771
|
1534 Set the character at a particular byte offset.
|
|
867
|
1535 Ichar eisetch_char (Eistring *eistr, Charcount charpos, Ichar chr);
|
|
771
|
1536 Set the character at a particular character offset.
|
|
|
1537
|
|
|
1538
|
|
|
1539 **********************************************
|
|
|
1540 * Concatenation *
|
|
|
1541 **********************************************
|
|
|
1542
|
|
|
1543 void eicat_* (Eistring *eistr, ...);
|
|
|
1544 Concatenate onto the end of the Eistring, with data coming from the
|
|
|
1545 same places as above:
|
|
|
1546
|
|
|
1547 void eicat_ei (Eistring *eistr, Eistring *eistr2);
|
|
|
1548 ... from another Eistring.
|
|
|
1549 void eicat_c (Eistring *eistr, Char_ASCII *c_string);
|
|
|
1550 ... from an ASCII null-terminated string. Non-ASCII characters in
|
|
|
1551 the string are *ILLEGAL* (read abort() with error-checking defined).
|
|
867
|
1552 void eicat_raw (ei, const Ibyte *data, Bytecount len);
|
|
771
|
1553 ... from raw internal-format data in the default internal format.
|
|
867
|
1554 void eicat_rawz (ei, const Ibyte *data);
|
|
771
|
1555 ... from raw internal-format data in the default internal format
|
|
|
1556 that is "null-terminated" (the meaning of this depends on the nature
|
|
|
1557 of the default internal format).
|
|
|
1558 void eicat_lstr (ei, Lisp_Object lisp_string);
|
|
|
1559 ... from a Lisp_Object string.
|
|
867
|
1560 void eicat_ch (ei, Ichar ch);
|
|
|
1561 ... from an Ichar.
|
|
771
|
1562
|
|
|
1563 (All except the first variety are convenience functions.
|
|
|
1564 In the general case, create another Eistring from the source.)
|
|
|
1565
|
|
|
1566
|
|
|
1567 **********************************************
|
|
|
1568 * Replacement *
|
|
|
1569 **********************************************
|
|
|
1570
|
|
|
1571 void eisub_* (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1572 Bytecount len, Charcount charlen, ...);
|
|
|
1573 Replace a section of the Eistring, specifically:
|
|
|
1574
|
|
|
1575 void eisub_ei (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1576 Bytecount len, Charcount charlen, Eistring *eistr2);
|
|
|
1577 ... with another Eistring.
|
|
|
1578 void eisub_c (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1579 Bytecount len, Charcount charlen, Char_ASCII *c_string);
|
|
|
1580 ... with an ASCII null-terminated string. Non-ASCII characters in
|
|
|
1581 the string are *ILLEGAL* (read abort() with error-checking defined).
|
|
|
1582 void eisub_ch (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
867
|
1583 Bytecount len, Charcount charlen, Ichar ch);
|
|
|
1584 ... with an Ichar.
|
|
771
|
1585
|
|
|
1586 void eidel (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1587 Bytecount len, Charcount charlen);
|
|
|
1588 Delete a section of the Eistring.
|
|
|
1589
|
|
|
1590
|
|
|
1591 **********************************************
|
|
|
1592 * Converting to an external format *
|
|
|
1593 **********************************************
|
|
|
1594
|
|
|
1595 void eito_external (Eistring *eistr, Lisp_Object coding_system);
|
|
|
1596 Convert the Eistring to an external format and store the result
|
|
|
1597 in the string. NOTE: Further changes to the Eistring will *NOT*
|
|
|
1598 change the external data stored in the string. You will have to
|
|
|
1599 call eito_external() again in such a case if you want the external
|
|
|
1600 data.
|
|
|
1601
|
|
|
1602 Extbyte *eiextdata (Eistring *eistr);
|
|
|
1603 Return a pointer to the external data stored in the Eistring as
|
|
|
1604 a result of a prior call to eito_external().
|
|
|
1605
|
|
|
1606 Bytecount eiextlen (Eistring *eistr);
|
|
|
1607 Return the length in bytes of the external data stored in the
|
|
|
1608 Eistring as a result of a prior call to eito_external().
|
|
|
1609
|
|
|
1610
|
|
|
1611 **********************************************
|
|
|
1612 * Searching in the Eistring for a character *
|
|
|
1613 **********************************************
|
|
|
1614
|
|
867
|
1615 Bytecount eichr (Eistring *eistr, Ichar chr);
|
|
|
1616 Charcount eichr_char (Eistring *eistr, Ichar chr);
|
|
|
1617 Bytecount eichr_off (Eistring *eistr, Ichar chr, Bytecount off,
|
|
771
|
1618 Charcount charoff);
|
|
867
|
1619 Charcount eichr_off_char (Eistring *eistr, Ichar chr, Bytecount off,
|
|
771
|
1620 Charcount charoff);
|
|
867
|
1621 Bytecount eirchr (Eistring *eistr, Ichar chr);
|
|
|
1622 Charcount eirchr_char (Eistring *eistr, Ichar chr);
|
|
|
1623 Bytecount eirchr_off (Eistring *eistr, Ichar chr, Bytecount off,
|
|
771
|
1624 Charcount charoff);
|
|
867
|
1625 Charcount eirchr_off_char (Eistring *eistr, Ichar chr, Bytecount off,
|
|
771
|
1626 Charcount charoff);
|
|
|
1627
|
|
|
1628
|
|
|
1629 **********************************************
|
|
|
1630 * Searching in the Eistring for a string *
|
|
|
1631 **********************************************
|
|
|
1632
|
|
|
1633 Bytecount eistr_ei (Eistring *eistr, Eistring *eistr2);
|
|
|
1634 Charcount eistr_ei_char (Eistring *eistr, Eistring *eistr2);
|
|
|
1635 Bytecount eistr_ei_off (Eistring *eistr, Eistring *eistr2, Bytecount off,
|
|
|
1636 Charcount charoff);
|
|
|
1637 Charcount eistr_ei_off_char (Eistring *eistr, Eistring *eistr2,
|
|
|
1638 Bytecount off, Charcount charoff);
|
|
|
1639 Bytecount eirstr_ei (Eistring *eistr, Eistring *eistr2);
|
|
|
1640 Charcount eirstr_ei_char (Eistring *eistr, Eistring *eistr2);
|
|
|
1641 Bytecount eirstr_ei_off (Eistring *eistr, Eistring *eistr2, Bytecount off,
|
|
|
1642 Charcount charoff);
|
|
|
1643 Charcount eirstr_ei_off_char (Eistring *eistr, Eistring *eistr2,
|
|
|
1644 Bytecount off, Charcount charoff);
|
|
|
1645
|
|
|
1646 Bytecount eistr_c (Eistring *eistr, Char_ASCII *c_string);
|
|
|
1647 Charcount eistr_c_char (Eistring *eistr, Char_ASCII *c_string);
|
|
|
1648 Bytecount eistr_c_off (Eistring *eistr, Char_ASCII *c_string, Bytecount off,
|
|
|
1649 Charcount charoff);
|
|
|
1650 Charcount eistr_c_off_char (Eistring *eistr, Char_ASCII *c_string,
|
|
|
1651 Bytecount off, Charcount charoff);
|
|
|
1652 Bytecount eirstr_c (Eistring *eistr, Char_ASCII *c_string);
|
|
|
1653 Charcount eirstr_c_char (Eistring *eistr, Char_ASCII *c_string);
|
|
|
1654 Bytecount eirstr_c_off (Eistring *eistr, Char_ASCII *c_string,
|
|
|
1655 Bytecount off, Charcount charoff);
|
|
|
1656 Charcount eirstr_c_off_char (Eistring *eistr, Char_ASCII *c_string,
|
|
|
1657 Bytecount off, Charcount charoff);
|
|
|
1658
|
|
|
1659
|
|
|
1660 **********************************************
|
|
|
1661 * Comparison *
|
|
|
1662 **********************************************
|
|
|
1663
|
|
|
1664 int eicmp_* (Eistring *eistr, ...);
|
|
|
1665 int eicmp_off_* (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1666 Bytecount len, Charcount charlen, ...);
|
|
|
1667 int eicasecmp_* (Eistring *eistr, ...);
|
|
|
1668 int eicasecmp_off_* (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1669 Bytecount len, Charcount charlen, ...);
|
|
|
1670 int eicasecmp_i18n_* (Eistring *eistr, ...);
|
|
|
1671 int eicasecmp_i18n_off_* (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1672 Bytecount len, Charcount charlen, ...);
|
|
|
1673
|
|
|
1674 Compare the Eistring with the other data. Return value same as
|
|
|
1675 from strcmp. The `*' is either `ei' for another Eistring (in
|
|
|
1676 which case `...' is an Eistring), or `c' for a pure-ASCII string
|
|
|
1677 (in which case `...' is a pointer to that string). For anything
|
|
|
1678 more complex, first create an Eistring out of the source.
|
|
|
1679 Comparison is either simple (`eicmp_...'), ASCII case-folding
|
|
|
1680 (`eicasecmp_...'), or multilingual case-folding
|
|
|
1681 (`eicasecmp_i18n_...).
|
|
|
1682
|
|
|
1683
|
|
|
1684 More specifically, the prototypes are:
|
|
|
1685
|
|
|
1686 int eicmp_ei (Eistring *eistr, Eistring *eistr2);
|
|
|
1687 int eicmp_off_ei (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1688 Bytecount len, Charcount charlen, Eistring *eistr2);
|
|
|
1689 int eicasecmp_ei (Eistring *eistr, Eistring *eistr2);
|
|
|
1690 int eicasecmp_off_ei (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1691 Bytecount len, Charcount charlen, Eistring *eistr2);
|
|
|
1692 int eicasecmp_i18n_ei (Eistring *eistr, Eistring *eistr2);
|
|
|
1693 int eicasecmp_i18n_off_ei (Eistring *eistr, Bytecount off,
|
|
|
1694 Charcount charoff, Bytecount len,
|
|
|
1695 Charcount charlen, Eistring *eistr2);
|
|
|
1696
|
|
|
1697 int eicmp_c (Eistring *eistr, Char_ASCII *c_string);
|
|
|
1698 int eicmp_off_c (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1699 Bytecount len, Charcount charlen, Char_ASCII *c_string);
|
|
|
1700 int eicasecmp_c (Eistring *eistr, Char_ASCII *c_string);
|
|
|
1701 int eicasecmp_off_c (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1702 Bytecount len, Charcount charlen,
|
|
|
1703 Char_ASCII *c_string);
|
|
|
1704 int eicasecmp_i18n_c (Eistring *eistr, Char_ASCII *c_string);
|
|
|
1705 int eicasecmp_i18n_off_c (Eistring *eistr, Bytecount off, Charcount charoff,
|
|
|
1706 Bytecount len, Charcount charlen,
|
|
|
1707 Char_ASCII *c_string);
|
|
|
1708
|
|
|
1709
|
|
|
1710 **********************************************
|
|
|
1711 * Case-changing the Eistring *
|
|
|
1712 **********************************************
|
|
|
1713
|
|
|
1714 void eilwr (Eistring *eistr);
|
|
|
1715 Convert all characters in the Eistring to lowercase.
|
|
|
1716 void eiupr (Eistring *eistr);
|
|
|
1717 Convert all characters in the Eistring to uppercase.
|
|
|
1718 */
|
|
|
1719
|
|
|
1720
|
|
|
1721 /* Principles for writing Eistring functions:
|
|
|
1722
|
|
|
1723 (1) Unfortunately, we have to write most of the Eistring functions
|
|
851
|
1724 as macros, because of the use of ALLOCA(). The principle used
|
|
771
|
1725 below to assure no conflict in local variables is to prefix all
|
|
|
1726 local variables with "ei" plus a number, which should be unique
|
|
|
1727 among macros. In practice, when finding a new number, find the
|
|
|
1728 highest so far used, and add 1.
|
|
|
1729
|
|
|
1730 (2) We also suffix the Eistring fields with an _ to avoid problems
|
|
|
1731 with macro parameters of the same name. (And as the standard
|
|
|
1732 signal not to access these fields directly.)
|
|
|
1733
|
|
|
1734 (3) We maintain both the length in bytes and chars of the data in
|
|
|
1735 the Eistring at all times, for convenient retrieval by outside
|
|
|
1736 functions. That means when writing functions that manipulate
|
|
|
1737 Eistrings, you too need to keep both lengths up to date for all
|
|
|
1738 data that you work with.
|
|
|
1739
|
|
|
1740 (4) When writing a new type of operation (e.g. substitution), you
|
|
|
1741 will often find yourself working with outside data, and thus
|
|
|
1742 have a series of related API's, for different forms that the
|
|
|
1743 outside data is in. Generally, you will want to choose a
|
|
|
1744 subset of the forms supported by eicpy_*, which has to be
|
|
|
1745 totally general because that's the fundamental way to get data
|
|
|
1746 into an Eistring, and once the data is into the string, it
|
|
|
1747 would be to create a whole series of Ei operations that work on
|
|
|
1748 nothing but Eistrings. Although theoretically nice, in
|
|
|
1749 practice it's a hassle, so we suggest that you provide
|
|
|
1750 convenience functions. In particular, there are two paths you
|
|
|
1751 can take. One is minimalist -- it only allows other Eistrings
|
|
867
|
1752 and ASCII data, and Ichars if the particular operation makes
|
|
771
|
1753 sense with a character. The other provides interfaces for the
|
|
|
1754 most commonly-used forms -- Eistring, ASCII data, Lisp string,
|
|
|
1755 raw internal-format string with length, raw internal-format
|
|
867
|
1756 string without, and possibly Ichar. (In the function names,
|
|
771
|
1757 these are designated `ei', `c', `lstr', `raw', `rawz', and
|
|
|
1758 `ch', respectively.)
|
|
|
1759
|
|
|
1760 (5) When coding a new type of operation, such as was discussed in
|
|
|
1761 previous section, the correct approach is to declare an worker
|
|
|
1762 function that does the work of everything, and is called by the
|
|
|
1763 other "container" macros that handle the different outside data
|
|
|
1764 forms. The data coming into the worker function, which
|
|
|
1765 typically ends in `_1', is in the form of three parameters:
|
|
|
1766 DATA, LEN, CHARLEN. (See point [3] about having two lengths and
|
|
|
1767 keeping them in sync.)
|
|
|
1768
|
|
|
1769 (6) Handling argument evaluation in macros: We take great care
|
|
|
1770 never to evaluate any argument more than once in any macro,
|
|
|
1771 except the initial Eistring parameter. This can and will be
|
|
|
1772 evaluated multiple times, but it should pretty much always just
|
|
|
1773 be a simple variable. This means, for example, that if an
|
|
|
1774 Eistring is the second (not first) argument of a macro, it
|
|
|
1775 doesn't fall under the "initial Eistring" exemption, so it
|
|
|
1776 needs protection against multi-evaluation. (Take the address of
|
|
|
1777 the Eistring structure, store in a temporary variable, and use
|
|
|
1778 temporary variable for all access to the Eistring.
|
|
|
1779 Essentially, we want it to appear as if these Eistring macros
|
|
|
1780 are functions -- we would like to declare them as functions but
|
|
851
|
1781 they use ALLOCA(), so we can't (and we can't make them inline
|
|
|
1782 functions either -- ALLOCA() is explicitly disallowed in inline
|
|
771
|
1783 functions.)
|
|
|
1784
|
|
|
1785 (7) Note that our rules regarding multiple evaluation are *more*
|
|
|
1786 strict than the rules listed above under the heading "working
|
|
|
1787 with raw internal-format data".
|
|
|
1788 */
|
|
|
1789
|
|
|
1790
|
|
|
1791 /* ----- Declaration ----- */
|
|
|
1792
|
|
|
1793 typedef struct
|
|
|
1794 {
|
|
|
1795 /* Data for the Eistring, stored in the default internal format.
|
|
|
1796 Always includes terminating null. */
|
|
867
|
1797 Ibyte *data_;
|
|
771
|
1798 /* Total number of bytes allocated in DATA (including null). */
|
|
|
1799 Bytecount max_size_allocated_;
|
|
|
1800 Bytecount bytelen_;
|
|
|
1801 Charcount charlen_;
|
|
|
1802 int mallocp_;
|
|
|
1803
|
|
|
1804 Extbyte *extdata_;
|
|
|
1805 Bytecount extlen_;
|
|
|
1806 } Eistring;
|
|
|
1807
|
|
|
1808 extern Eistring the_eistring_zero_init, the_eistring_malloc_zero_init;
|
|
|
1809
|
|
|
1810 #define DECLARE_EISTRING(name) \
|
|
|
1811 Eistring __ ## name ## __storage__ = the_eistring_zero_init; \
|
|
|
1812 Eistring *name = & __ ## name ## __storage__
|
|
|
1813 #define DECLARE_EISTRING_MALLOC(name) \
|
|
|
1814 Eistring __ ## name ## __storage__ = the_eistring_malloc_zero_init; \
|
|
|
1815 Eistring *name = & __ ## name ## __storage__
|
|
|
1816
|
|
|
1817 #define eiinit(ei) \
|
|
|
1818 do { \
|
|
793
|
1819 *(ei) = the_eistring_zero_init; \
|
|
771
|
1820 } while (0)
|
|
|
1821
|
|
|
1822 #define eiinit_malloc(ei) \
|
|
|
1823 do { \
|
|
793
|
1824 *(ei) = the_eistring_malloc_zero_init; \
|
|
771
|
1825 } while (0)
|
|
|
1826
|
|
|
1827
|
|
|
1828 /* ----- Utility ----- */
|
|
|
1829
|
|
|
1830 /* Make sure both LEN and CHARLEN are specified, in case one is given
|
|
|
1831 as -1. PTR evaluated at most once, others multiply. */
|
|
|
1832 #define eifixup_bytechar(ptr, len, charlen) \
|
|
|
1833 do { \
|
|
|
1834 if ((len) == -1) \
|
|
|
1835 (len) = charcount_to_bytecount (ptr, charlen); \
|
|
|
1836 else if ((charlen) == -1) \
|
|
|
1837 (charlen) = bytecount_to_charcount (ptr, len); \
|
|
|
1838 } while (0)
|
|
|
1839
|
|
|
1840 /* Make sure LEN is specified, in case it's is given as -1. PTR
|
|
|
1841 evaluated at most once, others multiply. */
|
|
|
1842 #define eifixup_byte(ptr, len, charlen) \
|
|
|
1843 do { \
|
|
|
1844 if ((len) == -1) \
|
|
|
1845 (len) = charcount_to_bytecount (ptr, charlen); \
|
|
|
1846 } while (0)
|
|
|
1847
|
|
|
1848 /* Make sure CHARLEN is specified, in case it's is given as -1. PTR
|
|
|
1849 evaluated at most once, others multiply. */
|
|
|
1850 #define eifixup_char(ptr, len, charlen) \
|
|
|
1851 do { \
|
|
|
1852 if ((charlen) == -1) \
|
|
|
1853 (charlen) = bytecount_to_charcount (ptr, len); \
|
|
|
1854 } while (0)
|
|
|
1855
|
|
|
1856
|
|
|
1857
|
|
|
1858 /* Make sure we can hold NEWBYTELEN bytes (which is NEWCHARLEN chars)
|
|
|
1859 plus a zero terminator. Preserve existing data as much as possible,
|
|
|
1860 including existing zero terminator. Put a new zero terminator where it
|
|
|
1861 should go if NEWZ if non-zero. All args but EI are evalled only once. */
|
|
|
1862
|
|
|
1863 #define EI_ALLOC(ei, newbytelen, newcharlen, newz) \
|
|
|
1864 do { \
|
|
|
1865 int ei1oldeibytelen = (ei)->bytelen_; \
|
|
|
1866 \
|
|
|
1867 (ei)->charlen_ = (newcharlen); \
|
|
|
1868 (ei)->bytelen_ = (newbytelen); \
|
|
|
1869 \
|
|
|
1870 if (ei1oldeibytelen != (ei)->bytelen_) \
|
|
|
1871 { \
|
|
|
1872 int ei1newsize = (ei)->max_size_allocated_; \
|
|
|
1873 while (ei1newsize < (ei)->bytelen_ + 1) \
|
|
|
1874 { \
|
|
|
1875 ei1newsize = (int) (ei1newsize * 1.5); \
|
|
|
1876 if (ei1newsize < 32) \
|
|
|
1877 ei1newsize = 32; \
|
|
|
1878 } \
|
|
|
1879 if (ei1newsize != (ei)->max_size_allocated_) \
|
|
|
1880 { \
|
|
|
1881 if ((ei)->mallocp_) \
|
|
|
1882 /* xrealloc always preserves existing data as much as possible */ \
|
|
867
|
1883 (ei)->data_ = (Ibyte *) xrealloc ((ei)->data_, ei1newsize); \
|
|
771
|
1884 else \
|
|
|
1885 { \
|
|
851
|
1886 /* We don't have realloc, so ALLOCA() more space and copy the \
|
|
771
|
1887 data into it. */ \
|
|
867
|
1888 Ibyte *ei1oldeidata = (ei)->data_; \
|
|
|
1889 (ei)->data_ = (Ibyte *) ALLOCA (ei1newsize); \
|
|
771
|
1890 if (ei1oldeidata) \
|
|
|
1891 memcpy ((ei)->data_, ei1oldeidata, ei1oldeibytelen + 1); \
|
|
|
1892 } \
|
|
|
1893 (ei)->max_size_allocated_ = ei1newsize; \
|
|
|
1894 } \
|
|
|
1895 if (newz) \
|
|
|
1896 (ei)->data_[(ei)->bytelen_] = '\0'; \
|
|
|
1897 } \
|
|
|
1898 } while (0)
|
|
|
1899
|
|
|
1900 #define EI_ALLOC_AND_COPY(ei, data, bytelen, charlen) \
|
|
|
1901 do { \
|
|
|
1902 EI_ALLOC (ei, bytelen, charlen, 1); \
|
|
|
1903 memcpy ((ei)->data_, data, (ei)->bytelen_); \
|
|
|
1904 } while (0)
|
|
|
1905
|
|
800
|
1906 #ifdef ERROR_CHECK_TEXT
|
|
771
|
1907 #define EI_ASSERT_ASCII(ptr, len) \
|
|
|
1908 do { \
|
|
|
1909 int ei5; \
|
|
|
1910 const Char_ASCII *ei5ptr = (ptr); \
|
|
|
1911 int ei5len = (len); \
|
|
|
1912 \
|
|
|
1913 for (ei5 = 0; ei5 < ei5len; ei5++) \
|
|
|
1914 assert (ei5ptr[ei5] >= 0x00 && ei5ptr[ei5] < 0x7F); \
|
|
|
1915 } while (0)
|
|
|
1916 #define EI_ASSERT_ASCIIZ(ptr) \
|
|
|
1917 do { \
|
|
|
1918 const Char_ASCII *ei5p1 = (ptr); \
|
|
|
1919 EI_ASSERT_ASCII (ei5p1, strlen (ei5p1)); \
|
|
|
1920 } while (0)
|
|
|
1921 #else
|
|
|
1922 #define EI_ASSERT_ASCII(ptr, len)
|
|
|
1923 #define EI_ASSERT_ASCIIZ(ptr)
|
|
|
1924 #endif
|
|
|
1925
|
|
|
1926
|
|
|
1927 /* ----- Initialization ----- */
|
|
|
1928
|
|
|
1929 #define eicpy_ei(ei, eicpy) \
|
|
|
1930 do { \
|
|
|
1931 const Eistring *ei2 = (eicpy); \
|
|
|
1932 EI_ALLOC_AND_COPY (ei, ei2->data_, ei2->bytelen_, ei2->charlen_); \
|
|
|
1933 } while (0)
|
|
|
1934
|
|
|
1935 #define eicpy_lstr(ei, lisp_string) \
|
|
|
1936 do { \
|
|
|
1937 Lisp_Object ei3 = (lisp_string); \
|
|
|
1938 EI_ALLOC_AND_COPY (ei, XSTRING_DATA (ei3), XSTRING_LENGTH (ei3), \
|
|
826
|
1939 string_char_length (ei3)); \
|
|
771
|
1940 } while (0)
|
|
|
1941
|
|
|
1942 #define eicpy_lstr_off(ei, lisp_string, off, charoff, len, charlen) \
|
|
|
1943 do { \
|
|
|
1944 Lisp_Object ei23lstr = (lisp_string); \
|
|
|
1945 int ei23off = (off); \
|
|
|
1946 int ei23charoff = (charoff); \
|
|
|
1947 int ei23len = (len); \
|
|
|
1948 int ei23charlen = (charlen); \
|
|
867
|
1949 const Ibyte *ei23data = XSTRING_DATA (ei23lstr); \
|
|
771
|
1950 \
|
|
|
1951 int ei23oldbytelen = (ei)->bytelen_; \
|
|
|
1952 \
|
|
|
1953 eifixup_byte (ei23data, ei23off, ei23charoff); \
|
|
|
1954 eifixup_bytechar (ei23data + ei23off, ei23len, ei23charlen); \
|
|
|
1955 \
|
|
|
1956 EI_ALLOC_AND_COPY (ei, ei23data + ei23off, ei23len, ei23charlen); \
|
|
|
1957 } while (0)
|
|
|
1958
|
|
826
|
1959 #define eicpy_raw_fmt(ei, ptr, len, fmt, object) \
|
|
771
|
1960 do { \
|
|
867
|
1961 const Ibyte *ei12ptr = (ptr); \
|
|
771
|
1962 Internal_Format ei12fmt = (fmt); \
|
|
|
1963 int ei12len = (len); \
|
|
|
1964 assert (ei12fmt == FORMAT_DEFAULT); \
|
|
|
1965 EI_ALLOC_AND_COPY (ei, ei12ptr, ei12len, \
|
|
|
1966 bytecount_to_charcount (ei12ptr, ei12len)); \
|
|
|
1967 } while (0)
|
|
|
1968
|
|
826
|
1969 #define eicpy_raw(ei, ptr, len) \
|
|
|
1970 eicpy_raw_fmt (ei, ptr, len, FORMAT_DEFAULT, Qnil)
|
|
|
1971
|
|
|
1972 #define eicpy_rawz_fmt(ei, ptr, fmt, object) \
|
|
|
1973 do { \
|
|
867
|
1974 const Ibyte *ei12p1ptr = (ptr); \
|
|
826
|
1975 Internal_Format ei12p1fmt = (fmt); \
|
|
|
1976 assert (ei12p1fmt == FORMAT_DEFAULT); \
|
|
|
1977 eicpy_raw_fmt (ei, ei12p1ptr, qxestrlen (ei12p1ptr), fmt, object); \
|
|
771
|
1978 } while (0)
|
|
|
1979
|
|
826
|
1980 #define eicpy_rawz(ei, ptr) eicpy_rawz_fmt (ei, ptr, FORMAT_DEFAULT, Qnil)
|
|
771
|
1981
|
|
|
1982 #define eicpy_ch(ei, ch) \
|
|
|
1983 do { \
|
|
867
|
1984 Ibyte ei12p2[MAX_ICHAR_LEN]; \
|
|
|
1985 Bytecount ei12p2len = set_itext_ichar (ei12p2, ch); \
|
|
771
|
1986 EI_ALLOC_AND_COPY (ei, ei12p2, ei12p2len, 1); \
|
|
|
1987 } while (0)
|
|
|
1988
|
|
|
1989 #define eicpy_c(ei, c_string) \
|
|
|
1990 do { \
|
|
|
1991 const Char_ASCII *ei4 = (c_string); \
|
|
|
1992 \
|
|
|
1993 EI_ASSERT_ASCIIZ (ei4); \
|
|
|
1994 eicpy_ext (ei, ei4, Qbinary); \
|
|
|
1995 } while (0)
|
|
|
1996
|
|
|
1997 #define eicpy_c_len(ei, c_string, c_len) \
|
|
|
1998 do { \
|
|
|
1999 const Char_ASCII *ei6 = (c_string); \
|
|
|
2000 int ei6len = (c_len); \
|
|
|
2001 \
|
|
|
2002 EI_ASSERT_ASCII (ei6, ei6len); \
|
|
|
2003 eicpy_ext_len (ei, ei6, ei6len, Qbinary); \
|
|
|
2004 } while (0)
|
|
|
2005
|
|
|
2006 #define eicpy_ext_len(ei, extdata, extlen, coding_system) \
|
|
|
2007 do { \
|
|
|
2008 const Extbyte *ei7 = (extdata); \
|
|
|
2009 int ei7len = (extlen); \
|
|
|
2010 \
|
|
|
2011 TO_INTERNAL_FORMAT (DATA, (ei7, ei7len), \
|
|
|
2012 ALLOCA, ((ei)->data_, (ei)->bytelen_), \
|
|
|
2013 coding_system); \
|
|
|
2014 (ei)->max_size_allocated_ = (ei)->bytelen_ + 1; \
|
|
|
2015 (ei)->charlen_ = bytecount_to_charcount ((ei)->data_, (ei)->bytelen_); \
|
|
|
2016 } while (0)
|
|
|
2017
|
|
|
2018 #define eicpy_ext(ei, extdata, coding_system) \
|
|
|
2019 do { \
|
|
|
2020 const Extbyte *ei8 = (extdata); \
|
|
|
2021 \
|
|
|
2022 eicpy_ext_len (ei, ei8, dfc_external_data_len (ei8, coding_system), \
|
|
|
2023 coding_system); \
|
|
|
2024 } while (0)
|
|
|
2025
|
|
|
2026 #define eicpy_lbuf(eistr, lisp_buf, off, charoff, len, charlen) \
|
|
|
2027 NOT YET IMPLEMENTED
|
|
|
2028
|
|
|
2029 #define eicpy_lstream(eistr, lstream) \
|
|
|
2030 NOT YET IMPLEMENTED
|
|
|
2031
|
|
867
|
2032 #define eireset(eistr) eicpy_rawz (eistr, (Ibyte *) "")
|
|
771
|
2033
|
|
|
2034 /* ----- Getting the data out of the Eistring ----- */
|
|
|
2035
|
|
|
2036 #define eidata(ei) ((ei)->data_)
|
|
|
2037
|
|
|
2038 #define eimake_string(ei) make_string (eidata (ei), eilen (ei))
|
|
|
2039
|
|
|
2040 #define eimake_string_off(eistr, off, charoff, len, charlen) \
|
|
|
2041 do { \
|
|
|
2042 Lisp_Object ei24lstr; \
|
|
|
2043 int ei24off = (off); \
|
|
|
2044 int ei24charoff = (charoff); \
|
|
|
2045 int ei24len = (len); \
|
|
|
2046 int ei24charlen = (charlen); \
|
|
|
2047 \
|
|
|
2048 eifixup_byte ((eistr)->data_, ei24off, ei24charoff); \
|
|
|
2049 eifixup_byte ((eistr)->data_ + ei24off, ei24len, ei24charlen); \
|
|
|
2050 \
|
|
|
2051 return make_string ((eistr)->data_ + ei24off, ei24len); \
|
|
|
2052 } while (0)
|
|
|
2053
|
|
|
2054 #define eicpyout_alloca(eistr, ptrout, lenout) \
|
|
826
|
2055 eicpyout_alloca_fmt (eistr, ptrout, lenout, FORMAT_DEFAULT, Qnil)
|
|
771
|
2056 #define eicpyout_malloc(eistr, lenout) \
|
|
826
|
2057 eicpyout_malloc_fmt (eistr, lenout, FORMAT_DEFAULT, Qnil)
|
|
867
|
2058 Ibyte *eicpyout_malloc_fmt (Eistring *eistr, Bytecount *len_out,
|
|
826
|
2059 Internal_Format fmt, Lisp_Object object);
|
|
|
2060 #define eicpyout_alloca_fmt(eistr, ptrout, lenout, fmt, object) \
|
|
771
|
2061 do { \
|
|
|
2062 Internal_Format ei23fmt = (fmt); \
|
|
867
|
2063 Ibyte *ei23ptrout = &(ptrout); \
|
|
771
|
2064 Bytecount *ei23lenout = &(lenout); \
|
|
|
2065 \
|
|
|
2066 assert (ei23fmt == FORMAT_DEFAULT); \
|
|
|
2067 \
|
|
|
2068 *ei23lenout = (eistr)->bytelen_; \
|
|
867
|
2069 *ei23ptrout = alloca_array (Ibyte, (eistr)->bytelen_ + 1); \
|
|
771
|
2070 memcpy (*ei23ptrout, (eistr)->data_, (eistr)->bytelen_ + 1); \
|
|
|
2071 } while (0)
|
|
|
2072
|
|
|
2073 /* ----- Moving to the heap ----- */
|
|
|
2074
|
|
|
2075 #define eifree(ei) \
|
|
|
2076 do { \
|
|
|
2077 if ((ei)->mallocp_) \
|
|
|
2078 { \
|
|
|
2079 if ((ei)->data_) \
|
|
|
2080 xfree ((ei)->data_); \
|
|
|
2081 if ((ei)->extdata_) \
|
|
|
2082 xfree ((ei)->extdata_); \
|
|
|
2083 eiinit_malloc (ei); \
|
|
|
2084 } \
|
|
|
2085 else \
|
|
|
2086 eiinit (ei); \
|
|
|
2087 } while (0)
|
|
|
2088
|
|
|
2089 int eifind_large_enough_buffer (int oldbufsize, int needed_size);
|
|
|
2090 void eito_malloc_1 (Eistring *ei);
|
|
|
2091
|
|
|
2092 #define eito_malloc(ei) eito_malloc_1 (ei)
|
|
|
2093
|
|
|
2094 #define eito_alloca(ei) \
|
|
|
2095 do { \
|
|
|
2096 if (!(ei)->mallocp_) \
|
|
|
2097 return; \
|
|
|
2098 (ei)->mallocp_ = 0; \
|
|
|
2099 if ((ei)->data_) \
|
|
|
2100 { \
|
|
867
|
2101 Ibyte *ei13newdata; \
|
|
771
|
2102 \
|
|
|
2103 (ei)->max_size_allocated_ = \
|
|
|
2104 eifind_large_enough_buffer (0, (ei)->bytelen_ + 1); \
|
|
867
|
2105 ei13newdata = (Ibyte *) ALLOCA ((ei)->max_size_allocated_); \
|
|
771
|
2106 memcpy (ei13newdata, (ei)->data_, (ei)->bytelen_ + 1); \
|
|
|
2107 xfree ((ei)->data_); \
|
|
|
2108 (ei)->data_ = ei13newdata; \
|
|
|
2109 } \
|
|
|
2110 \
|
|
|
2111 if ((ei)->extdata_) \
|
|
|
2112 { \
|
|
851
|
2113 Extbyte *ei13newdata = (Extbyte *) ALLOCA ((ei)->extlen_ + 2); \
|
|
771
|
2114 \
|
|
|
2115 memcpy (ei13newdata, (ei)->extdata_, (ei)->extlen_); \
|
|
|
2116 /* Double null-terminate in case of Unicode data */ \
|
|
|
2117 ei13newdata[(ei)->extlen_] = '\0'; \
|
|
|
2118 ei13newdata[(ei)->extlen_ + 1] = '\0'; \
|
|
|
2119 xfree ((ei)->extdata_); \
|
|
|
2120 (ei)->extdata_ = ei13newdata; \
|
|
|
2121 } \
|
|
|
2122 } while (0)
|
|
|
2123
|
|
|
2124
|
|
|
2125 /* ----- Retrieving the length ----- */
|
|
|
2126
|
|
|
2127 #define eilen(ei) ((ei)->bytelen_)
|
|
|
2128 #define eicharlen(ei) ((ei)->charlen_)
|
|
|
2129
|
|
|
2130
|
|
|
2131 /* ----- Working with positions ----- */
|
|
|
2132
|
|
|
2133 #define eicharpos_to_bytepos(ei, charpos) \
|
|
|
2134 charcount_to_bytecount ((ei)->data_, charpos)
|
|
|
2135 #define eibytepos_to_charpos(ei, bytepos) \
|
|
|
2136 bytecount_to_charcount ((ei)->data_, bytepos)
|
|
|
2137
|
|
|
2138 DECLARE_INLINE_HEADER (Bytecount eiincpos_1 (Eistring *eistr,
|
|
|
2139 Bytecount bytepos,
|
|
|
2140 Charcount n))
|
|
|
2141 {
|
|
867
|
2142 Ibyte *pos = eistr->data_ + bytepos;
|
|
814
|
2143 Charcount i;
|
|
771
|
2144
|
|
800
|
2145 text_checking_assert (bytepos >= 0 && bytepos <= eistr->bytelen_);
|
|
|
2146 text_checking_assert (n >= 0 && n <= eistr->charlen_);
|
|
771
|
2147 /* We could check N more correctly now, but that would require a
|
|
|
2148 call to bytecount_to_charcount(), which would be needlessly
|
|
|
2149 expensive (it would convert O(N) algorithms into O(N^2) algorithms
|
|
800
|
2150 with ERROR_CHECK_TEXT, which would be bad). If N is bad, we are
|
|
867
|
2151 guaranteed to catch it either inside INC_IBYTEPTR() or in the check
|
|
771
|
2152 below. */
|
|
|
2153 for (i = 0; i < n; i++)
|
|
867
|
2154 INC_IBYTEPTR (pos);
|
|
800
|
2155 text_checking_assert (pos - eistr->data_ <= eistr->bytelen_);
|
|
771
|
2156 return pos - eistr->data_;
|
|
|
2157 }
|
|
|
2158
|
|
|
2159 #define eiincpos (ei, bytepos) eiincpos_1 (ei, bytepos, 1)
|
|
|
2160 #define eiincpos_n (ei, bytepos, n) eiincpos_1 (ei, bytepos, n)
|
|
|
2161
|
|
|
2162 DECLARE_INLINE_HEADER (Bytecount eidecpos_1 (Eistring *eistr,
|
|
|
2163 Bytecount bytepos,
|
|
|
2164 Charcount n))
|
|
|
2165 {
|
|
867
|
2166 Ibyte *pos = eistr->data_ + bytepos;
|
|
771
|
2167 int i;
|
|
|
2168
|
|
800
|
2169 text_checking_assert (bytepos >= 0 && bytepos <= eistr->bytelen_);
|
|
|
2170 text_checking_assert (n >= 0 && n <= eistr->charlen_);
|
|
771
|
2171 /* We could check N more correctly now, but ... see above. */
|
|
|
2172 for (i = 0; i < n; i++)
|
|
867
|
2173 DEC_IBYTEPTR (pos);
|
|
800
|
2174 text_checking_assert (pos - eistr->data_ <= eistr->bytelen_);
|
|
771
|
2175 return pos - eistr->data_;
|
|
|
2176 }
|
|
|
2177
|
|
|
2178 #define eidecpos (ei, bytepos) eidecpos_1 (ei, bytepos, 1)
|
|
|
2179 #define eidecpos_n (ei, bytepos, n) eidecpos_1 (ei, bytepos, n)
|
|
|
2180
|
|
|
2181
|
|
|
2182 /* ----- Getting the character at a position ----- */
|
|
|
2183
|
|
|
2184 #define eigetch(ei, bytepos) \
|
|
867
|
2185 itext_ichar ((ei)->data_ + (bytepos))
|
|
|
2186 #define eigetch_char(ei, charpos) itext_ichar_n ((ei)->data_, charpos)
|
|
771
|
2187
|
|
|
2188
|
|
|
2189 /* ----- Setting the character at a position ----- */
|
|
|
2190
|
|
|
2191 #define eisetch(ei, bytepos, chr) \
|
|
|
2192 eisub_ch (ei, bytepos, -1, -1, 1, chr)
|
|
|
2193 #define eisetch_char(ei, charpos, chr) \
|
|
|
2194 eisub_ch (ei, -1, charpos, -1, 1, chr)
|
|
|
2195
|
|
|
2196
|
|
|
2197 /* ----- Concatenation ----- */
|
|
|
2198
|
|
|
2199 #define eicat_1(ei, data, bytelen, charlen) \
|
|
|
2200 do { \
|
|
|
2201 int ei14oldeibytelen = (ei)->bytelen_; \
|
|
|
2202 int ei14bytelen = (bytelen); \
|
|
|
2203 EI_ALLOC (ei, (ei)->bytelen_ + ei14bytelen, \
|
|
|
2204 (ei)->charlen_ + (charlen), 1); \
|
|
|
2205 memcpy ((ei)->data_ + ei14oldeibytelen, (data), \
|
|
|
2206 ei14bytelen); \
|
|
|
2207 } while (0)
|
|
|
2208
|
|
|
2209 #define eicat_ei(ei, ei2) \
|
|
|
2210 do { \
|
|
|
2211 const Eistring *ei9 = (ei2); \
|
|
|
2212 eicat_1 (ei, ei9->data_, ei9->bytelen_, ei9->charlen_); \
|
|
|
2213 } while (0)
|
|
|
2214
|
|
|
2215 #define eicat_c(ei, c_string) \
|
|
|
2216 do { \
|
|
|
2217 const Char_ASCII *ei15 = (c_string); \
|
|
|
2218 int ei15len = strlen (ei15); \
|
|
|
2219 \
|
|
|
2220 EI_ASSERT_ASCII (ei15, ei15len); \
|
|
|
2221 eicat_1 (ei, ei15, ei15len, \
|
|
867
|
2222 bytecount_to_charcount ((Ibyte *) ei15, ei15len)); \
|
|
771
|
2223 } while (0)
|
|
|
2224
|
|
|
2225 #define eicat_raw(ei, data, len) \
|
|
|
2226 do { \
|
|
|
2227 int ei16len = (len); \
|
|
867
|
2228 const Ibyte *ei16data = (data); \
|
|
771
|
2229 eicat_1 (ei, ei16data, ei16len, \
|
|
|
2230 bytecount_to_charcount (ei16data, ei16len)); \
|
|
|
2231 } while (0)
|
|
|
2232
|
|
|
2233 #define eicat_rawz(ei, ptr) \
|
|
|
2234 do { \
|
|
867
|
2235 const Ibyte *ei16p5ptr = (ptr); \
|
|
771
|
2236 eicat_raw (ei, ei16p5ptr, qxestrlen (ei16p5ptr)); \
|
|
|
2237 } while (0)
|
|
|
2238
|
|
|
2239 #define eicat_lstr(ei, lisp_string) \
|
|
|
2240 do { \
|
|
|
2241 Lisp_Object ei17 = (lisp_string); \
|
|
|
2242 eicat_1 (ei, XSTRING_DATA (ei17), XSTRING_LENGTH (ei17), \
|
|
826
|
2243 string_char_length (ei17)); \
|
|
771
|
2244 } while (0)
|
|
|
2245
|
|
|
2246 #define eicat_ch(ei, ch) \
|
|
|
2247 do { \
|
|
867
|
2248 Ibyte ei22ch[MAX_ICHAR_LEN]; \
|
|
|
2249 Bytecount ei22len = set_itext_ichar (ei22ch, ch); \
|
|
771
|
2250 eicat_1 (ei, ei22ch, ei22len, 1); \
|
|
|
2251 } while (0)
|
|
|
2252
|
|
|
2253
|
|
|
2254 /* ----- Replacement ----- */
|
|
|
2255
|
|
|
2256 /* Replace the section of an Eistring at (OFF, LEN) with the data at
|
|
|
2257 SRC of length LEN. All positions have corresponding character values,
|
|
|
2258 and either can be -1 -- it will be computed from the other. */
|
|
|
2259
|
|
|
2260 #define eisub_1(ei, off, charoff, len, charlen, src, srclen, srccharlen) \
|
|
|
2261 do { \
|
|
|
2262 int ei18off = (off); \
|
|
|
2263 int ei18charoff = (charoff); \
|
|
|
2264 int ei18len = (len); \
|
|
|
2265 int ei18charlen = (charlen); \
|
|
867
|
2266 Ibyte *ei18src = (Ibyte *) (src); \
|
|
771
|
2267 int ei18srclen = (srclen); \
|
|
|
2268 int ei18srccharlen = (srccharlen); \
|
|
|
2269 \
|
|
|
2270 int ei18oldeibytelen = (ei)->bytelen_; \
|
|
|
2271 \
|
|
|
2272 eifixup_bytechar ((ei)->data_, ei18off, ei18charoff); \
|
|
|
2273 eifixup_bytechar ((ei)->data_ + ei18off, ei18len, ei18charlen); \
|
|
|
2274 eifixup_bytechar (ei18src, ei18srclen, ei18srccharlen); \
|
|
|
2275 \
|
|
|
2276 EI_ALLOC (ei, (ei)->bytelen_ + ei18srclen - ei18len, \
|
|
|
2277 (ei)->charlen_ + ei18srccharlen - ei18charlen, 0); \
|
|
|
2278 if (ei18len != ei18srclen) \
|
|
|
2279 memmove ((ei)->data_ + ei18off + ei18srclen, \
|
|
|
2280 (ei)->data_ + ei18off + ei18len, \
|
|
|
2281 /* include zero terminator. */ \
|
|
|
2282 ei18oldeibytelen - (ei18off + ei18len) + 1); \
|
|
|
2283 if (ei18srclen > 0) \
|
|
|
2284 memcpy ((ei)->data_ + ei18off, ei18src, ei18srclen); \
|
|
|
2285 } while (0)
|
|
|
2286
|
|
|
2287 #define eisub_ei(ei, off, charoff, len, charlen, ei2) \
|
|
|
2288 do { \
|
|
|
2289 const Eistring *ei19 = (ei2); \
|
|
|
2290 eisub_1 (ei, off, charoff, len, charlen, ei19->data_, ei19->bytelen_, \
|
|
|
2291 ei19->charlen_); \
|
|
|
2292 } while (0)
|
|
|
2293
|
|
|
2294 #define eisub_c(ei, off, charoff, len, charlen, c_string) \
|
|
|
2295 do { \
|
|
|
2296 const Char_ASCII *ei20 = (c_string); \
|
|
|
2297 int ei20len = strlen (ei20); \
|
|
|
2298 EI_ASSERT_ASCII (ei20, ei20len); \
|
|
|
2299 eisub_1 (ei, off, charoff, len, charlen, ei20, ei20len, -1); \
|
|
|
2300 } while (0)
|
|
|
2301
|
|
|
2302 #define eisub_ch(ei, off, charoff, len, charlen, ch) \
|
|
|
2303 do { \
|
|
867
|
2304 Ibyte ei21ch[MAX_ICHAR_LEN]; \
|
|
|
2305 Bytecount ei21len = set_itext_ichar (ei21ch, ch); \
|
|
771
|
2306 eisub_1 (ei, off, charoff, len, charlen, ei21ch, ei21len, 1); \
|
|
|
2307 } while (0)
|
|
|
2308
|
|
|
2309 #define eidel(ei, off, charoff, len, charlen) \
|
|
|
2310 eisub_1(ei, off, charoff, len, charlen, NULL, 0, 0)
|
|
|
2311
|
|
|
2312
|
|
|
2313 /* ----- Converting to an external format ----- */
|
|
|
2314
|
|
|
2315 #define eito_external(ei, coding_system) \
|
|
|
2316 do { \
|
|
|
2317 if ((ei)->mallocp_) \
|
|
|
2318 { \
|
|
|
2319 if ((ei)->extdata_) \
|
|
|
2320 { \
|
|
|
2321 xfree ((ei)->extdata_); \
|
|
|
2322 (ei)->extdata_ = 0; \
|
|
|
2323 } \
|
|
|
2324 TO_EXTERNAL_FORMAT (DATA, ((ei)->data_, (ei)->bytelen_), \
|
|
|
2325 MALLOC, ((ei)->extdata_, (ei)->extlen_), \
|
|
|
2326 coding_system); \
|
|
|
2327 } \
|
|
|
2328 else \
|
|
|
2329 TO_EXTERNAL_FORMAT (DATA, ((ei)->data_, (ei)->bytelen_), \
|
|
|
2330 ALLOCA, ((ei)->extdata_, (ei)->extlen_), \
|
|
|
2331 coding_system); \
|
|
|
2332 } while (0)
|
|
|
2333
|
|
|
2334 #define eiextdata(ei) ((ei)->extdata_)
|
|
|
2335 #define eiextlen(ei) ((ei)->extlen_)
|
|
|
2336
|
|
|
2337
|
|
|
2338 /* ----- Searching in the Eistring for a character ----- */
|
|
|
2339
|
|
|
2340 #define eichr(eistr, chr) \
|
|
|
2341 NOT YET IMPLEMENTED
|
|
|
2342 #define eichr_char(eistr, chr) \
|
|
|
2343 NOT YET IMPLEMENTED
|
|
|
2344 #define eichr_off(eistr, chr, off, charoff) \
|
|
|
2345 NOT YET IMPLEMENTED
|
|
|
2346 #define eichr_off_char(eistr, chr, off, charoff) \
|
|
|
2347 NOT YET IMPLEMENTED
|
|
|
2348 #define eirchr(eistr, chr) \
|
|
|
2349 NOT YET IMPLEMENTED
|
|
|
2350 #define eirchr_char(eistr, chr) \
|
|
|
2351 NOT YET IMPLEMENTED
|
|
|
2352 #define eirchr_off(eistr, chr, off, charoff) \
|
|
|
2353 NOT YET IMPLEMENTED
|
|
|
2354 #define eirchr_off_char(eistr, chr, off, charoff) \
|
|
|
2355 NOT YET IMPLEMENTED
|
|
|
2356
|
|
|
2357
|
|
|
2358 /* ----- Searching in the Eistring for a string ----- */
|
|
|
2359
|
|
|
2360 #define eistr_ei(eistr, eistr2) \
|
|
|
2361 NOT YET IMPLEMENTED
|
|
|
2362 #define eistr_ei_char(eistr, eistr2) \
|
|
|
2363 NOT YET IMPLEMENTED
|
|
|
2364 #define eistr_ei_off(eistr, eistr2, off, charoff) \
|
|
|
2365 NOT YET IMPLEMENTED
|
|
|
2366 #define eistr_ei_off_char(eistr, eistr2, off, charoff) \
|
|
|
2367 NOT YET IMPLEMENTED
|
|
|
2368 #define eirstr_ei(eistr, eistr2) \
|
|
|
2369 NOT YET IMPLEMENTED
|
|
|
2370 #define eirstr_ei_char(eistr, eistr2) \
|
|
|
2371 NOT YET IMPLEMENTED
|
|
|
2372 #define eirstr_ei_off(eistr, eistr2, off, charoff) \
|
|
|
2373 NOT YET IMPLEMENTED
|
|
|
2374 #define eirstr_ei_off_char(eistr, eistr2, off, charoff) \
|
|
|
2375 NOT YET IMPLEMENTED
|
|
|
2376
|
|
|
2377 #define eistr_c(eistr, c_string) \
|
|
|
2378 NOT YET IMPLEMENTED
|
|
|
2379 #define eistr_c_char(eistr, c_string) \
|
|
|
2380 NOT YET IMPLEMENTED
|
|
|
2381 #define eistr_c_off(eistr, c_string, off, charoff) \
|
|
|
2382 NOT YET IMPLEMENTED
|
|
|
2383 #define eistr_c_off_char(eistr, c_string, off, charoff) \
|
|
|
2384 NOT YET IMPLEMENTED
|
|
|
2385 #define eirstr_c(eistr, c_string) \
|
|
|
2386 NOT YET IMPLEMENTED
|
|
|
2387 #define eirstr_c_char(eistr, c_string) \
|
|
|
2388 NOT YET IMPLEMENTED
|
|
|
2389 #define eirstr_c_off(eistr, c_string, off, charoff) \
|
|
|
2390 NOT YET IMPLEMENTED
|
|
|
2391 #define eirstr_c_off_char(eistr, c_string, off, charoff) \
|
|
|
2392 NOT YET IMPLEMENTED
|
|
|
2393
|
|
|
2394
|
|
|
2395 /* ----- Comparison ----- */
|
|
|
2396
|
|
|
2397 int eicmp_1 (Eistring *ei, Bytecount off, Charcount charoff,
|
|
867
|
2398 Bytecount len, Charcount charlen, const Ibyte *data,
|
|
771
|
2399 const Eistring *ei2, int is_c, int fold_case);
|
|
|
2400
|
|
|
2401 #define eicmp_ei(eistr, eistr2) \
|
|
|
2402 eicmp_1 (eistr, 0, -1, -1, -1, 0, eistr2, 0, 0)
|
|
|
2403 #define eicmp_off_ei(eistr, off, charoff, len, charlen, eistr2) \
|
|
|
2404 eicmp_1 (eistr, off, charoff, len, charlen, 0, eistr2, 0, 0)
|
|
|
2405 #define eicasecmp_ei(eistr, eistr2) \
|
|
|
2406 eicmp_1 (eistr, 0, -1, -1, -1, 0, eistr2, 0, 1)
|
|
|
2407 #define eicasecmp_off_ei(eistr, off, charoff, len, charlen, eistr2) \
|
|
|
2408 eicmp_1 (eistr, off, charoff, len, charlen, 0, eistr2, 0, 1)
|
|
|
2409 #define eicasecmp_i18n_ei(eistr, eistr2) \
|
|
|
2410 eicmp_1 (eistr, 0, -1, -1, -1, 0, eistr2, 0, 2)
|
|
|
2411 #define eicasecmp_i18n_off_ei(eistr, off, charoff, len, charlen, eistr2) \
|
|
|
2412 eicmp_1 (eistr, off, charoff, len, charlen, 0, eistr2, 0, 2)
|
|
|
2413
|
|
|
2414 #define eicmp_c(eistr, c_string) \
|
|
|
2415 eicmp_1 (eistr, 0, -1, -1, -1, c_string, 0, 1, 0)
|
|
|
2416 #define eicmp_off_c(eistr, off, charoff, len, charlen, c_string) \
|
|
|
2417 eicmp_1 (eistr, off, charoff, len, charlen, c_string, 0, 1, 0)
|
|
|
2418 #define eicasecmp_c(eistr, c_string) \
|
|
|
2419 eicmp_1 (eistr, 0, -1, -1, -1, c_string, 0, 1, 1)
|
|
|
2420 #define eicasecmp_off_c(eistr, off, charoff, len, charlen, c_string) \
|
|
|
2421 eicmp_1 (eistr, off, charoff, len, charlen, c_string, 0, 1, 1)
|
|
|
2422 #define eicasecmp_i18n_c(eistr, c_string) \
|
|
|
2423 eicmp_1 (eistr, 0, -1, -1, -1, c_string, 0, 1, 2)
|
|
|
2424 #define eicasecmp_i18n_off_c(eistr, off, charoff, len, charlen, c_string) \
|
|
|
2425 eicmp_1 (eistr, off, charoff, len, charlen, c_string, 0, 1, 2)
|
|
|
2426
|
|
|
2427
|
|
|
2428 /* ----- Case-changing the Eistring ----- */
|
|
|
2429
|
|
867
|
2430 int eistr_casefiddle_1 (Ibyte *olddata, Bytecount len, Ibyte *newdata,
|
|
771
|
2431 int downp);
|
|
|
2432
|
|
|
2433 #define EI_CASECHANGE(ei, downp) \
|
|
|
2434 do { \
|
|
867
|
2435 int ei11new_allocmax = (ei)->charlen_ * MAX_ICHAR_LEN + 1; \
|
|
|
2436 Ibyte *ei11storage = \
|
|
|
2437 (Ibyte *) alloca_array (Ibyte, ei11new_allocmax); \
|
|
771
|
2438 int ei11newlen = eistr_casefiddle_1 ((ei)->data_, (ei)->bytelen_, \
|
|
|
2439 ei11storage, downp); \
|
|
|
2440 \
|
|
|
2441 if (ei11newlen) \
|
|
|
2442 { \
|
|
|
2443 (ei)->max_size_allocated_ = ei11new_allocmax; \
|
|
|
2444 (ei)->data_ = ei11storage; \
|
|
|
2445 (ei)->bytelen_ = ei11newlen; \
|
|
|
2446 /* charlen is the same. */ \
|
|
|
2447 } \
|
|
|
2448 } while (0)
|
|
|
2449
|
|
|
2450 #define eilwr(ei) EI_CASECHANGE (ei, 1)
|
|
|
2451 #define eiupr(ei) EI_CASECHANGE (ei, 0)
|
|
|
2452
|
|
|
2453
|
|
|
2454 /************************************************************************/
|
|
|
2455 /* */
|
|
|
2456 /* Converting between internal and external format */
|
|
|
2457 /* */
|
|
|
2458 /************************************************************************/
|
|
|
2459 /*
|
|
|
2460 All client code should use only the two macros
|
|
|
2461
|
|
|
2462 TO_EXTERNAL_FORMAT (source_type, source, sink_type, sink, coding_system)
|
|
|
2463 TO_INTERNAL_FORMAT (source_type, source, sink_type, sink, coding_system)
|
|
|
2464
|
|
|
2465 Typical use is
|
|
|
2466
|
|
|
2467 TO_EXTERNAL_FORMAT (DATA, (ptr, len),
|
|
|
2468 LISP_BUFFER, buffer,
|
|
|
2469 Qfile_name);
|
|
|
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
|
|
|
2494 The macros accept the kinds of sources and sinks appropriate for
|
|
|
2495 internal and external data representation. See the type_checking_assert
|
|
|
2496 macros below for the actual allowed types.
|
|
|
2497
|
|
|
2498 Since some sources and sinks use one argument (a Lisp_Object) to
|
|
|
2499 specify them, while others take a (pointer, length) pair, we use
|
|
|
2500 some C preprocessor trickery to allow pair arguments to be specified
|
|
|
2501 by parenthesizing them, as in the examples above.
|
|
|
2502
|
|
|
2503 Anything prefixed by dfc_ (`data format conversion') is private.
|
|
|
2504 They are only used to implement these macros.
|
|
|
2505
|
|
|
2506 [[Using C_STRING* is appropriate for using with external APIs that
|
|
|
2507 take null-terminated strings. For internal data, we should try to
|
|
|
2508 be '\0'-clean - i.e. allow arbitrary data to contain embedded '\0'.
|
|
|
2509
|
|
|
2510 Sometime in the future we might allow output to C_STRING_ALLOCA or
|
|
|
2511 C_STRING_MALLOC _only_ with TO_EXTERNAL_FORMAT(), not
|
|
|
2512 TO_INTERNAL_FORMAT().]]
|
|
|
2513
|
|
|
2514 The above comments are not true. Frequently (most of the time, in
|
|
|
2515 fact), external strings come as zero-terminated entities, where the
|
|
|
2516 zero-termination is the only way to find out the length. Even in
|
|
|
2517 cases where you can get the length, most of the time the system will
|
|
|
2518 still use the null to signal the end of the string, and there will
|
|
|
2519 still be no way to either send in or receive a string with embedded
|
|
|
2520 nulls. In such situations, it's pointless to track the length
|
|
|
2521 because null bytes can never be in the string. We have a lot of
|
|
|
2522 operations that make it easy to operate on zero-terminated strings,
|
|
|
2523 and forcing the user the deal with the length everywhere would only
|
|
|
2524 make the code uglier and more complicated, for no gain. --ben
|
|
|
2525
|
|
|
2526 There is no problem using the same lvalue for source and sink.
|
|
|
2527
|
|
|
2528 Also, when pointers are required, the code (currently at least) is
|
|
|
2529 lax and allows any pointer types, either in the source or the sink.
|
|
|
2530 This makes it possible, e.g., to deal with internal format data held
|
|
|
2531 in char *'s or external format data held in WCHAR * (i.e. Unicode).
|
|
|
2532
|
|
|
2533 Finally, whenever storage allocation is called for, extra space is
|
|
|
2534 allocated for a terminating zero, and such a zero is stored in the
|
|
|
2535 appropriate place, regardless of whether the source data was
|
|
|
2536 specified using a length or was specified as zero-terminated. This
|
|
|
2537 allows you to freely pass the resulting data, no matter how
|
|
|
2538 obtained, to a routine that expects zero termination (modulo, of
|
|
|
2539 course, that any embedded zeros in the resulting text will cause
|
|
|
2540 truncation). In fact, currently two embedded zeros are allocated
|
|
|
2541 and stored after the data result. This is to allow for the
|
|
|
2542 possibility of storing a Unicode value on output, which needs the
|
|
|
2543 two zeros. Currently, however, the two zeros are stored regardless
|
|
|
2544 of whether the conversion is internal or external and regardless of
|
|
|
2545 whether the external coding system is in fact Unicode. This
|
|
|
2546 behavior may change in the future, and you cannot rely on this --
|
|
|
2547 the most you can rely on is that sink data in Unicode format will
|
|
|
2548 have two terminating nulls, which combine to form one Unicode null
|
|
|
2549 character. */
|
|
|
2550
|
|
|
2551 #define TO_EXTERNAL_FORMAT(source_type, source, sink_type, sink, codesys) \
|
|
|
2552 do { \
|
|
|
2553 dfc_conversion_type dfc_simplified_source_type; \
|
|
|
2554 dfc_conversion_type dfc_simplified_sink_type; \
|
|
|
2555 dfc_conversion_data dfc_source; \
|
|
|
2556 dfc_conversion_data dfc_sink; \
|
|
|
2557 Lisp_Object dfc_codesys = (codesys); \
|
|
|
2558 \
|
|
|
2559 type_checking_assert \
|
|
|
2560 ((DFC_TYPE_##source_type == DFC_TYPE_DATA || \
|
|
|
2561 DFC_TYPE_##source_type == DFC_TYPE_C_STRING || \
|
|
|
2562 DFC_TYPE_##source_type == DFC_TYPE_LISP_STRING || \
|
|
|
2563 DFC_TYPE_##source_type == DFC_TYPE_LISP_OPAQUE || \
|
|
|
2564 DFC_TYPE_##source_type == DFC_TYPE_LISP_LSTREAM) \
|
|
|
2565 && \
|
|
|
2566 (DFC_TYPE_##sink_type == DFC_TYPE_ALLOCA || \
|
|
|
2567 DFC_TYPE_##sink_type == DFC_TYPE_MALLOC || \
|
|
|
2568 DFC_TYPE_##sink_type == DFC_TYPE_C_STRING_ALLOCA || \
|
|
|
2569 DFC_TYPE_##sink_type == DFC_TYPE_C_STRING_MALLOC || \
|
|
|
2570 DFC_TYPE_##sink_type == DFC_TYPE_LISP_LSTREAM || \
|
|
|
2571 DFC_TYPE_##sink_type == DFC_TYPE_LISP_OPAQUE)); \
|
|
|
2572 \
|
|
|
2573 DFC_EXT_SOURCE_##source_type##_TO_ARGS (source, dfc_codesys); \
|
|
|
2574 DFC_SINK_##sink_type##_TO_ARGS (sink); \
|
|
|
2575 \
|
|
|
2576 dfc_convert_to_external_format (dfc_simplified_source_type, &dfc_source, \
|
|
|
2577 dfc_codesys, \
|
|
|
2578 dfc_simplified_sink_type, &dfc_sink); \
|
|
|
2579 \
|
|
|
2580 DFC_##sink_type##_USE_CONVERTED_DATA (sink); \
|
|
|
2581 } while (0)
|
|
|
2582
|
|
|
2583 #define TO_INTERNAL_FORMAT(source_type, source, sink_type, sink, codesys) \
|
|
|
2584 do { \
|
|
|
2585 dfc_conversion_type dfc_simplified_source_type; \
|
|
|
2586 dfc_conversion_type dfc_simplified_sink_type; \
|
|
|
2587 dfc_conversion_data dfc_source; \
|
|
|
2588 dfc_conversion_data dfc_sink; \
|
|
|
2589 Lisp_Object dfc_codesys = (codesys); \
|
|
|
2590 \
|
|
|
2591 type_checking_assert \
|
|
|
2592 ((DFC_TYPE_##source_type == DFC_TYPE_DATA || \
|
|
|
2593 DFC_TYPE_##source_type == DFC_TYPE_C_STRING || \
|
|
|
2594 DFC_TYPE_##source_type == DFC_TYPE_LISP_OPAQUE || \
|
|
|
2595 DFC_TYPE_##source_type == DFC_TYPE_LISP_LSTREAM) \
|
|
|
2596 && \
|
|
|
2597 (DFC_TYPE_##sink_type == DFC_TYPE_ALLOCA || \
|
|
|
2598 DFC_TYPE_##sink_type == DFC_TYPE_MALLOC || \
|
|
|
2599 DFC_TYPE_##sink_type == DFC_TYPE_C_STRING_ALLOCA || \
|
|
|
2600 DFC_TYPE_##sink_type == DFC_TYPE_C_STRING_MALLOC || \
|
|
|
2601 DFC_TYPE_##sink_type == DFC_TYPE_LISP_STRING || \
|
|
|
2602 DFC_TYPE_##sink_type == DFC_TYPE_LISP_LSTREAM || \
|
|
|
2603 DFC_TYPE_##sink_type == DFC_TYPE_LISP_BUFFER)); \
|
|
|
2604 \
|
|
|
2605 DFC_INT_SOURCE_##source_type##_TO_ARGS (source, dfc_codesys); \
|
|
|
2606 DFC_SINK_##sink_type##_TO_ARGS (sink); \
|
|
|
2607 \
|
|
|
2608 dfc_convert_to_internal_format (dfc_simplified_source_type, &dfc_source, \
|
|
|
2609 dfc_codesys, \
|
|
|
2610 dfc_simplified_sink_type, &dfc_sink); \
|
|
|
2611 \
|
|
|
2612 DFC_##sink_type##_USE_CONVERTED_DATA (sink); \
|
|
|
2613 } while (0)
|
|
|
2614
|
|
814
|
2615 #ifdef __cplusplus
|
|
771
|
2616
|
|
814
|
2617 /* Error if you try to use a union here: "member `struct {anonymous
|
|
|
2618 union}::{anonymous} {anonymous union}::data' with constructor not allowed
|
|
|
2619 in union" (Bytecount is a class) */
|
|
|
2620
|
|
|
2621 typedef struct
|
|
|
2622 #else
|
|
771
|
2623 typedef union
|
|
814
|
2624 #endif
|
|
771
|
2625 {
|
|
|
2626 struct { const void *ptr; Bytecount len; } data;
|
|
|
2627 Lisp_Object lisp_object;
|
|
|
2628 } dfc_conversion_data;
|
|
|
2629
|
|
|
2630 enum dfc_conversion_type
|
|
|
2631 {
|
|
|
2632 DFC_TYPE_DATA,
|
|
|
2633 DFC_TYPE_ALLOCA,
|
|
|
2634 DFC_TYPE_MALLOC,
|
|
|
2635 DFC_TYPE_C_STRING,
|
|
|
2636 DFC_TYPE_C_STRING_ALLOCA,
|
|
|
2637 DFC_TYPE_C_STRING_MALLOC,
|
|
|
2638 DFC_TYPE_LISP_STRING,
|
|
|
2639 DFC_TYPE_LISP_LSTREAM,
|
|
|
2640 DFC_TYPE_LISP_OPAQUE,
|
|
|
2641 DFC_TYPE_LISP_BUFFER
|
|
|
2642 };
|
|
|
2643 typedef enum dfc_conversion_type dfc_conversion_type;
|
|
|
2644
|
|
|
2645 /* WARNING: These use a static buffer. This can lead to disaster if
|
|
|
2646 these functions are not used *very* carefully. Another reason to only use
|
|
|
2647 TO_EXTERNAL_FORMAT() and TO_INTERNAL_FORMAT(). */
|
|
|
2648 void
|
|
|
2649 dfc_convert_to_external_format (dfc_conversion_type source_type,
|
|
|
2650 dfc_conversion_data *source,
|
|
|
2651 Lisp_Object coding_system,
|
|
|
2652 dfc_conversion_type sink_type,
|
|
|
2653 dfc_conversion_data *sink);
|
|
|
2654 void
|
|
|
2655 dfc_convert_to_internal_format (dfc_conversion_type source_type,
|
|
|
2656 dfc_conversion_data *source,
|
|
|
2657 Lisp_Object coding_system,
|
|
|
2658 dfc_conversion_type sink_type,
|
|
|
2659 dfc_conversion_data *sink);
|
|
|
2660 /* CPP Trickery */
|
|
|
2661 #define DFC_CPP_CAR(x,y) (x)
|
|
|
2662 #define DFC_CPP_CDR(x,y) (y)
|
|
|
2663
|
|
|
2664 /* Convert `source' to args for dfc_convert_to_external_format() */
|
|
|
2665 #define DFC_EXT_SOURCE_DATA_TO_ARGS(val, codesys) do { \
|
|
|
2666 dfc_source.data.ptr = DFC_CPP_CAR val; \
|
|
|
2667 dfc_source.data.len = DFC_CPP_CDR val; \
|
|
|
2668 dfc_simplified_source_type = DFC_TYPE_DATA; \
|
|
|
2669 } while (0)
|
|
|
2670 #define DFC_EXT_SOURCE_C_STRING_TO_ARGS(val, codesys) do { \
|
|
|
2671 dfc_source.data.len = \
|
|
|
2672 strlen ((char *) (dfc_source.data.ptr = (val))); \
|
|
|
2673 dfc_simplified_source_type = DFC_TYPE_DATA; \
|
|
|
2674 } while (0)
|
|
|
2675 #define DFC_EXT_SOURCE_LISP_STRING_TO_ARGS(val, codesys) do { \
|
|
|
2676 Lisp_Object dfc_slsta = (val); \
|
|
|
2677 type_checking_assert (STRINGP (dfc_slsta)); \
|
|
|
2678 dfc_source.lisp_object = dfc_slsta; \
|
|
|
2679 dfc_simplified_source_type = DFC_TYPE_LISP_STRING; \
|
|
|
2680 } while (0)
|
|
|
2681 #define DFC_EXT_SOURCE_LISP_LSTREAM_TO_ARGS(val, codesys) do { \
|
|
|
2682 Lisp_Object dfc_sllta = (val); \
|
|
|
2683 type_checking_assert (LSTREAMP (dfc_sllta)); \
|
|
|
2684 dfc_source.lisp_object = dfc_sllta; \
|
|
|
2685 dfc_simplified_source_type = DFC_TYPE_LISP_LSTREAM; \
|
|
|
2686 } while (0)
|
|
|
2687 #define DFC_EXT_SOURCE_LISP_OPAQUE_TO_ARGS(val, codesys) do { \
|
|
|
2688 Lisp_Opaque *dfc_slota = XOPAQUE (val); \
|
|
|
2689 dfc_source.data.ptr = OPAQUE_DATA (dfc_slota); \
|
|
|
2690 dfc_source.data.len = OPAQUE_SIZE (dfc_slota); \
|
|
|
2691 dfc_simplified_source_type = DFC_TYPE_DATA; \
|
|
|
2692 } while (0)
|
|
|
2693
|
|
|
2694 /* Convert `source' to args for dfc_convert_to_internal_format() */
|
|
|
2695 #define DFC_INT_SOURCE_DATA_TO_ARGS(val, codesys) \
|
|
|
2696 DFC_EXT_SOURCE_DATA_TO_ARGS (val, codesys)
|
|
|
2697 #define DFC_INT_SOURCE_C_STRING_TO_ARGS(val, codesys) do { \
|
|
|
2698 dfc_source.data.len = dfc_external_data_len (dfc_source.data.ptr = (val), \
|
|
|
2699 codesys); \
|
|
|
2700 dfc_simplified_source_type = DFC_TYPE_DATA; \
|
|
|
2701 } while (0)
|
|
|
2702 #define DFC_INT_SOURCE_LISP_STRING_TO_ARGS(val, codesys) \
|
|
|
2703 DFC_EXT_SOURCE_LISP_STRING_TO_ARGS (val, codesys)
|
|
|
2704 #define DFC_INT_SOURCE_LISP_LSTREAM_TO_ARGS(val, codesys) \
|
|
|
2705 DFC_EXT_SOURCE_LISP_LSTREAM_TO_ARGS (val, codesys)
|
|
|
2706 #define DFC_INT_SOURCE_LISP_OPAQUE_TO_ARGS(val, codesys) \
|
|
|
2707 DFC_EXT_SOURCE_LISP_OPAQUE_TO_ARGS (val, codesys)
|
|
|
2708
|
|
|
2709 /* Convert `sink' to args for dfc_convert_to_*_format() */
|
|
|
2710 #define DFC_SINK_ALLOCA_TO_ARGS(val) \
|
|
|
2711 dfc_simplified_sink_type = DFC_TYPE_DATA
|
|
|
2712 #define DFC_SINK_C_STRING_ALLOCA_TO_ARGS(val) \
|
|
|
2713 dfc_simplified_sink_type = DFC_TYPE_DATA
|
|
|
2714 #define DFC_SINK_MALLOC_TO_ARGS(val) \
|
|
|
2715 dfc_simplified_sink_type = DFC_TYPE_DATA
|
|
|
2716 #define DFC_SINK_C_STRING_MALLOC_TO_ARGS(val) \
|
|
|
2717 dfc_simplified_sink_type = DFC_TYPE_DATA
|
|
|
2718 #define DFC_SINK_LISP_STRING_TO_ARGS(val) \
|
|
|
2719 dfc_simplified_sink_type = DFC_TYPE_DATA
|
|
|
2720 #define DFC_SINK_LISP_OPAQUE_TO_ARGS(val) \
|
|
|
2721 dfc_simplified_sink_type = DFC_TYPE_DATA
|
|
|
2722 #define DFC_SINK_LISP_LSTREAM_TO_ARGS(val) do { \
|
|
|
2723 Lisp_Object dfc_sllta = (val); \
|
|
|
2724 type_checking_assert (LSTREAMP (dfc_sllta)); \
|
|
|
2725 dfc_sink.lisp_object = dfc_sllta; \
|
|
|
2726 dfc_simplified_sink_type = DFC_TYPE_LISP_LSTREAM; \
|
|
|
2727 } while (0)
|
|
|
2728 #define DFC_SINK_LISP_BUFFER_TO_ARGS(val) do { \
|
|
|
2729 struct buffer *dfc_slbta = XBUFFER (val); \
|
|
|
2730 dfc_sink.lisp_object = \
|
|
|
2731 make_lisp_buffer_output_stream \
|
|
|
2732 (dfc_slbta, BUF_PT (dfc_slbta), 0); \
|
|
|
2733 dfc_simplified_sink_type = DFC_TYPE_LISP_LSTREAM; \
|
|
|
2734 } while (0)
|
|
|
2735
|
|
|
2736 /* Assign to the `sink' lvalue(s) using the converted data. */
|
|
|
2737 /* + 2 because we double zero-extended to account for Unicode conversion */
|
|
|
2738 typedef union { char c; void *p; } *dfc_aliasing_voidpp;
|
|
|
2739 #define DFC_ALLOCA_USE_CONVERTED_DATA(sink) do { \
|
|
851
|
2740 void * dfc_sink_ret = ALLOCA (dfc_sink.data.len + 2); \
|
|
771
|
2741 memcpy (dfc_sink_ret, dfc_sink.data.ptr, dfc_sink.data.len + 2); \
|
|
|
2742 ((dfc_aliasing_voidpp) &(DFC_CPP_CAR sink))->p = dfc_sink_ret; \
|
|
|
2743 (DFC_CPP_CDR sink) = dfc_sink.data.len; \
|
|
|
2744 } while (0)
|
|
|
2745 #define DFC_MALLOC_USE_CONVERTED_DATA(sink) do { \
|
|
|
2746 void * dfc_sink_ret = xmalloc (dfc_sink.data.len + 2); \
|
|
|
2747 memcpy (dfc_sink_ret, dfc_sink.data.ptr, dfc_sink.data.len + 2); \
|
|
|
2748 ((dfc_aliasing_voidpp) &(DFC_CPP_CAR sink))->p = dfc_sink_ret; \
|
|
|
2749 (DFC_CPP_CDR sink) = dfc_sink.data.len; \
|
|
|
2750 } while (0)
|
|
|
2751 #define DFC_C_STRING_ALLOCA_USE_CONVERTED_DATA(sink) do { \
|
|
851
|
2752 void * dfc_sink_ret = ALLOCA (dfc_sink.data.len + 2); \
|
|
771
|
2753 memcpy (dfc_sink_ret, dfc_sink.data.ptr, dfc_sink.data.len + 2); \
|
|
|
2754 ((dfc_aliasing_voidpp) &(sink))->p = dfc_sink_ret; \
|
|
|
2755 } while (0)
|
|
|
2756 #define DFC_C_STRING_MALLOC_USE_CONVERTED_DATA(sink) do { \
|
|
|
2757 void * dfc_sink_ret = xmalloc (dfc_sink.data.len + 2); \
|
|
|
2758 memcpy (dfc_sink_ret, dfc_sink.data.ptr, dfc_sink.data.len + 2); \
|
|
|
2759 ((dfc_aliasing_voidpp) &(sink))->p = dfc_sink_ret; \
|
|
|
2760 } while (0)
|
|
|
2761 #define DFC_LISP_STRING_USE_CONVERTED_DATA(sink) \
|
|
867
|
2762 sink = make_string ((Ibyte *) dfc_sink.data.ptr, dfc_sink.data.len)
|
|
771
|
2763 #define DFC_LISP_OPAQUE_USE_CONVERTED_DATA(sink) \
|
|
|
2764 sink = make_opaque (dfc_sink.data.ptr, dfc_sink.data.len)
|
|
|
2765 #define DFC_LISP_LSTREAM_USE_CONVERTED_DATA(sink) /* data already used */
|
|
|
2766 #define DFC_LISP_BUFFER_USE_CONVERTED_DATA(sink) \
|
|
|
2767 Lstream_delete (XLSTREAM (dfc_sink.lisp_object))
|
|
|
2768
|
|
|
2769 /* Convenience macros for extremely common invocations */
|
|
|
2770 #define C_STRING_TO_EXTERNAL(in, out, coding_system) \
|
|
|
2771 TO_EXTERNAL_FORMAT (C_STRING, in, C_STRING_ALLOCA, out, coding_system)
|
|
|
2772 #define C_STRING_TO_EXTERNAL_MALLOC(in, out, coding_system) \
|
|
|
2773 TO_EXTERNAL_FORMAT (C_STRING, in, C_STRING_MALLOC, out, coding_system)
|
|
|
2774 #define EXTERNAL_TO_C_STRING(in, out, coding_system) \
|
|
|
2775 TO_INTERNAL_FORMAT (C_STRING, in, C_STRING_ALLOCA, out, coding_system)
|
|
|
2776 #define EXTERNAL_TO_C_STRING_MALLOC(in, out, coding_system) \
|
|
|
2777 TO_INTERNAL_FORMAT (C_STRING, in, C_STRING_MALLOC, out, coding_system)
|
|
|
2778 #define LISP_STRING_TO_EXTERNAL(in, out, coding_system) \
|
|
|
2779 TO_EXTERNAL_FORMAT (LISP_STRING, in, C_STRING_ALLOCA, out, coding_system)
|
|
|
2780 #define LISP_STRING_TO_EXTERNAL_MALLOC(in, out, coding_system) \
|
|
|
2781 TO_EXTERNAL_FORMAT (LISP_STRING, in, C_STRING_MALLOC, out, coding_system)
|
|
|
2782
|
|
|
2783 /* Standins for various encodings, until we know them better */
|
|
|
2784 #define Qcommand_argument_encoding Qnative
|
|
|
2785 #define Qenvironment_variable_encoding Qnative
|
|
|
2786 #define Qunix_host_name_encoding Qnative
|
|
|
2787 #define Qunix_service_name_encoding Qnative
|
|
|
2788 #define Qmswindows_host_name_encoding Qmswindows_multibyte
|
|
|
2789 #define Qmswindows_service_name_encoding Qmswindows_multibyte
|
|
|
2790
|
|
|
2791 /* Standins for various X encodings, until we know them better */
|
|
|
2792
|
|
|
2793 /* !!#### Need to verify the encoding used in lwlib -- Qnative or Qctext?
|
|
|
2794 Almost certainly the former. Use a standin for now. */
|
|
|
2795 #define Qlwlib_encoding Qnative
|
|
|
2796
|
|
|
2797 #define Qx_atom_name_encoding Qctext
|
|
|
2798 /* font names are often stored in atoms, so it gets sticky if we set this
|
|
|
2799 to something different from atom-name encoding */
|
|
|
2800 #define Qx_font_name_encoding Qctext
|
|
|
2801
|
|
|
2802 #define Qx_color_name_encoding Qctext
|
|
|
2803
|
|
|
2804 /* the following probably must agree with Qcommand_argument_encoding and
|
|
|
2805 Qenvironment_variable_encoding */
|
|
|
2806 #define Qx_display_name_encoding Qnative
|
|
|
2807
|
|
|
2808 #define Qstrerror_encoding Qnative
|
|
|
2809
|
|
|
2810 #define GET_STRERROR(var, num) \
|
|
|
2811 do { \
|
|
|
2812 int __gsnum__ = (num); \
|
|
|
2813 Extbyte * __gserr__ = strerror (__gsnum__); \
|
|
|
2814 \
|
|
|
2815 if (!__gserr__) \
|
|
|
2816 { \
|
|
867
|
2817 var = alloca_ibytes (99); \
|
|
771
|
2818 qxesprintf (var, "Unknown error %d", __gsnum__); \
|
|
|
2819 } \
|
|
|
2820 else \
|
|
|
2821 EXTERNAL_TO_C_STRING (__gserr__, var, Qstrerror_encoding); \
|
|
|
2822 } while (0)
|
|
|
2823
|
|
|
2824 #endif /* INCLUDED_text_h_ */
|
