Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(936)

Side by Side Diff: public/common/unicode/utext.h

Issue 18836004: Move ICU headers from public/{common,i18n} to source/{common,i18n} (Closed) Base URL: https://chromium.googlesource.com/chromium/deps/icu46.git@master
Patch Set: same as ps #3. retry uploading Created 7 years, 5 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View unified diff | Download patch
« no previous file with comments | « public/common/unicode/usystem.h ('k') | public/common/unicode/utf.h » ('j') | no next file with comments »
Toggle Intra-line Diffs ('i') | Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
OLDNEW
(Empty)
1 /*
2 *******************************************************************************
3 *
4 * Copyright (C) 2004-2010, International Business Machines
5 * Corporation and others. All Rights Reserved.
6 *
7 *******************************************************************************
8 * file name: utext.h
9 * encoding: US-ASCII
10 * tab size: 8 (not used)
11 * indentation:4
12 *
13 * created on: 2004oct06
14 * created by: Markus W. Scherer
15 */
16
17 #ifndef __UTEXT_H__
18 #define __UTEXT_H__
19
20 /**
21 * \file
22 * \brief C API: Abstract Unicode Text API
23 *
24 * The Text Access API provides a means to allow text that is stored in alternat ive
25 * formats to work with ICU services. ICU normally operates on text that is
26 * stored in UTF-16 format, in (UChar *) arrays for the C APIs or as type
27 * UnicodeString for C++ APIs.
28 *
29 * ICU Text Access allows other formats, such as UTF-8 or non-contiguous
30 * UTF-16 strings, to be placed in a UText wrapper and then passed to ICU servic es.
31 *
32 * There are three general classes of usage for UText:
33 *
34 * Application Level Use. This is the simplest usage - applications would
35 * use one of the utext_open() functions on their input text, and pass
36 * the resulting UText to the desired ICU service.
37 *
38 * Second is usage in ICU Services, such as break iteration, that will need to
39 * operate on input presented to them as a UText. These implementations
40 * will need to use the iteration and related UText functions to gain
41 * access to the actual text.
42 *
43 * The third class of UText users are "text providers." These are the
44 * UText implementations for the various text storage formats. An applicati on
45 * or system with a unique text storage format can implement a set of
46 * UText provider functions for that format, which will then allow
47 * ICU services to operate on that format.
48 *
49 *
50 * <em>Iterating over text</em>
51 *
52 * Here is sample code for a forward iteration over the contents of a UText
53 *
54 * \code
55 * UChar32 c;
56 * UText *ut = whatever();
57 *
58 * for (c=utext_next32From(ut, 0); c>=0; c=utext_next32(ut)) {
59 * // do whatever with the codepoint c here.
60 * }
61 * \endcode
62 *
63 * And here is similar code to iterate in the reverse direction, from the end
64 * of the text towards the beginning.
65 *
66 * \code
67 * UChar32 c;
68 * UText *ut = whatever();
69 * int textLength = utext_nativeLength(ut);
70 * for (c=utext_previous32From(ut, textLength); c>=0; c=utext_previous32(ut)) {
71 * // do whatever with the codepoint c here.
72 * }
73 * \endcode
74 *
75 * <em>Characters and Indexing</em>
76 *
77 * Indexing into text by UText functions is nearly always in terms of the native
78 * indexing of the underlying text storage. The storage format could be UTF-8
79 * or UTF-32, for example. When coding to the UText access API, no assumptions
80 * can be made regarding the size of characters, or how far an index
81 * may move when iterating between characters.
82 *
83 * All indices supplied to UText functions are pinned to the length of the
84 * text. An out-of-bounds index is not considered to be an error, but is
85 * adjusted to be in the range 0 <= index <= length of input text.
86 *
87 *
88 * When an index position is returned from a UText function, it will be
89 * a native index to the underlying text. In the case of multi-unit characters,
90 * it will always refer to the first position of the character,
91 * never to the interior. This is essentially the same thing as saying that
92 * a returned index will always point to a boundary between characters.
93 *
94 * When a native index is supplied to a UText function, all indices that
95 * refer to any part of a multi-unit character representation are considered
96 * to be equivalent. In the case of multi-unit characters, an incoming index
97 * will be logically normalized to refer to the start of the character.
98 *
99 * It is possible to test whether a native index is on a code point boundary
100 * by doing a utext_setNativeIndex() followed by a utext_getNativeIndex().
101 * If the index is returned unchanged, it was on a code point boundary. If
102 * an adjusted index is returned, the original index referred to the
103 * interior of a character.
104 *
105 * <em>Conventions for calling UText functions</em>
106 *
107 * Most UText access functions have as their first parameter a (UText *) pointer ,
108 * which specifies the UText to be used. Unless otherwise noted, the
109 * pointer must refer to a valid, open UText. Attempting to
110 * use a closed UText or passing a NULL pointer is a programming error and
111 * will produce undefined results or NULL pointer exceptions.
112 *
113 * The UText_Open family of functions can either open an existing (closed)
114 * UText, or heap allocate a new UText. Here is sample code for creating
115 * a stack-allocated UText.
116 *
117 * \code
118 * char *s = whatever(); // A utf-8 string
119 * U_ErrorCode status = U_ZERO_ERROR;
120 * UText ut = UTEXT_INITIALIZER;
121 * utext_openUTF8(ut, s, -1, &status);
122 * if (U_FAILURE(status)) {
123 * // error handling
124 * } else {
125 * // work with the UText
126 * }
127 * \endcode
128 *
129 * Any existing UText passed to an open function _must_ have been initialized,
130 * either by the UTEXT_INITIALIZER, or by having been originally heap-allocated
131 * by an open function. Passing NULL will cause the open function to
132 * heap-allocate and fully initialize a new UText.
133 *
134 */
135
136
137
138 #include "unicode/utypes.h"
139 #include "unicode/uchar.h"
140 #if U_SHOW_CPLUSPLUS_API
141 #include "unicode/localpointer.h"
142 #include "unicode/rep.h"
143 #include "unicode/unistr.h"
144 #include "unicode/chariter.h"
145 #endif
146
147
148 U_CDECL_BEGIN
149
150 struct UText;
151 typedef struct UText UText; /**< C typedef for struct UText. @stable ICU 3.6 */
152
153
154 /******************************************************************************* ********
155 *
156 * C Functions for creating UText wrappers around various kinds of text string s.
157 *
158 ******************************************************************************* *********/
159
160
161 /**
162 * Close function for UText instances.
163 * Cleans up, releases any resources being held by an open UText.
164 * <p>
165 * If the UText was originally allocated by one of the utext_open functions,
166 * the storage associated with the utext will also be freed.
167 * If the UText storage originated with the application, as it would with
168 * a local or static instance, the storage will not be deleted.
169 *
170 * An open UText can be reset to refer to new string by using one of the utex t_open()
171 * functions without first closing the UText.
172 *
173 * @param ut The UText to be closed.
174 * @return NULL if the UText struct was deleted by the close. If the UText struct
175 * was originally provided by the caller to the open function, it is
176 * returned by this function, and may be safely used again in
177 * a subsequent utext_open.
178 *
179 * @stable ICU 3.4
180 */
181 U_STABLE UText * U_EXPORT2
182 utext_close(UText *ut);
183
184 #if U_SHOW_CPLUSPLUS_API
185
186 U_NAMESPACE_BEGIN
187
188 /**
189 * \class LocalUTextPointer
190 * "Smart pointer" class, closes a UText via utext_close().
191 * For most methods see the LocalPointerBase base class.
192 *
193 * @see LocalPointerBase
194 * @see LocalPointer
195 * @stable ICU 4.4
196 */
197 U_DEFINE_LOCAL_OPEN_POINTER(LocalUTextPointer, UText, utext_close);
198
199 U_NAMESPACE_END
200
201 #endif
202
203 /**
204 * Open a read-only UText implementation for UTF-8 strings.
205 *
206 * \htmlonly
207 * Any invalid UTF-8 in the input will be handled in this way:
208 * a sequence of bytes that has the form of a truncated, but otherwise valid,
209 * UTF-8 sequence will be replaced by a single unicode replacement character, \u FFFD.
210 * Any other illegal bytes will each be replaced by a \uFFFD.
211 * \endhtmlonly
212 *
213 * @param ut Pointer to a UText struct. If NULL, a new UText will be create d.
214 * If non-NULL, must refer to an initialized UText struct, which w ill then
215 * be reset to reference the specified UTF-8 string.
216 * @param s A UTF-8 string. Must not be NULL.
217 * @param length The length of the UTF-8 string in bytes, or -1 if the string is
218 * zero terminated.
219 * @param status Errors are returned here.
220 * @return A pointer to the UText. If a pre-allocated UText was provided, it
221 * will always be used and returned.
222 * @stable ICU 3.4
223 */
224 U_STABLE UText * U_EXPORT2
225 utext_openUTF8(UText *ut, const char *s, int64_t length, UErrorCode *status);
226
227
228 /**
229 * Open a read-only UText for UChar * string.
230 *
231 * @param ut Pointer to a UText struct. If NULL, a new UText will be create d.
232 * If non-NULL, must refer to an initialized UText struct, which w ill then
233 * be reset to reference the specified UChar string.
234 * @param s A UChar (UTF-16) string
235 * @param length The number of UChars in the input string, or -1 if the string i s
236 * zero terminated.
237 * @param status Errors are returned here.
238 * @return A pointer to the UText. If a pre-allocated UText was provided, it
239 * will always be used and returned.
240 * @stable ICU 3.4
241 */
242 U_STABLE UText * U_EXPORT2
243 utext_openUChars(UText *ut, const UChar *s, int64_t length, UErrorCode *status);
244
245
246 #if U_SHOW_CPLUSPLUS_API
247 /**
248 * Open a writable UText for a non-const UnicodeString.
249 *
250 * @param ut Pointer to a UText struct. If NULL, a new UText will be creat ed.
251 * If non-NULL, must refer to an initialized UText struct, which will then
252 * be reset to reference the specified input string.
253 * @param s A UnicodeString.
254 * @param status Errors are returned here.
255 * @return Pointer to the UText. If a UText was supplied as input, this
256 * will always be used and returned.
257 * @stable ICU 3.4
258 */
259 U_STABLE UText * U_EXPORT2
260 utext_openUnicodeString(UText *ut, U_NAMESPACE_QUALIFIER UnicodeString *s, UErro rCode *status);
261
262
263 /**
264 * Open a UText for a const UnicodeString. The resulting UText will not be wri table.
265 *
266 * @param ut Pointer to a UText struct. If NULL, a new UText will be created .
267 * If non-NULL, must refer to an initialized UText struct, which w ill then
268 * be reset to reference the specified input string.
269 * @param s A const UnicodeString to be wrapped.
270 * @param status Errors are returned here.
271 * @return Pointer to the UText. If a UText was supplied as input, this
272 * will always be used and returned.
273 * @stable ICU 3.4
274 */
275 U_STABLE UText * U_EXPORT2
276 utext_openConstUnicodeString(UText *ut, const U_NAMESPACE_QUALIFIER UnicodeStrin g *s, UErrorCode *status);
277
278
279 /**
280 * Open a writable UText implementation for an ICU Replaceable object.
281 * @param ut Pointer to a UText struct. If NULL, a new UText will be created .
282 * If non-NULL, must refer to an already existing UText, which wil l then
283 * be reset to reference the specified replaceable text.
284 * @param rep A Replaceable text object.
285 * @param status Errors are returned here.
286 * @return Pointer to the UText. If a UText was supplied as input, this
287 * will always be used and returned.
288 * @see Replaceable
289 * @stable ICU 3.4
290 */
291 U_STABLE UText * U_EXPORT2
292 utext_openReplaceable(UText *ut, U_NAMESPACE_QUALIFIER Replaceable *rep, UErrorC ode *status);
293
294 /**
295 * Open a UText implementation over an ICU CharacterIterator.
296 * @param ut Pointer to a UText struct. If NULL, a new UText will be created .
297 * If non-NULL, must refer to an already existing UText, which wil l then
298 * be reset to reference the specified replaceable text.
299 * @param ci A Character Iterator.
300 * @param status Errors are returned here.
301 * @return Pointer to the UText. If a UText was supplied as input, this
302 * will always be used and returned.
303 * @see Replaceable
304 * @stable ICU 3.4
305 */
306 U_STABLE UText * U_EXPORT2
307 utext_openCharacterIterator(UText *ut, U_NAMESPACE_QUALIFIER CharacterIterator * ic, UErrorCode *status);
308
309 #endif
310
311
312 /**
313 * Clone a UText. This is much like opening a UText where the source text is itself
314 * another UText.
315 *
316 * A deep clone will copy both the UText data structures and the underlying te xt.
317 * The original and cloned UText will operate completely independently; modifi cations
318 * made to the text in one will not affect the other. Text providers are not
319 * required to support deep clones. The user of clone() must check the status return
320 * and be prepared to handle failures.
321 *
322 * The standard UText implementations for UTF8, UChar *, UnicodeString and
323 * Replaceable all support deep cloning.
324 *
325 * The UText returned from a deep clone will be writable, assuming that the te xt
326 * provider is able to support writing, even if the source UText had been made
327 * non-writable by means of UText_freeze().
328 *
329 * A shallow clone replicates only the UText data structures; it does not make
330 * a copy of the underlying text. Shallow clones can be used as an efficient way to
331 * have multiple iterators active in a single text string that is not being
332 * modified.
333 *
334 * A shallow clone operation will not fail, barring truly exceptional conditio ns such
335 * as memory allocation failures.
336 *
337 * Shallow UText clones should be avoided if the UText functions that modify t he
338 * text are expected to be used, either on the original or the cloned UText.
339 * Any such modifications can cause unpredictable behavior. Read Only
340 * shallow clones provide some protection against errors of this type by
341 * disabling text modification via the cloned UText.
342 *
343 * A shallow clone made with the readOnly parameter == FALSE will preserve the
344 * utext_isWritable() state of the source object. Note, however, that
345 * write operations must be avoided while more than one UText exists that refe r
346 * to the same underlying text.
347 *
348 * A UText and its clone may be safely concurrently accessed by separate threa ds.
349 * This is true for read access only with shallow clones, and for both read an d
350 * write access with deep clones.
351 * It is the responsibility of the Text Provider to ensure that this thread sa fety
352 * constraint is met.
353 *
354 * @param dest A UText struct to be filled in with the result of the clone o peration,
355 * or NULL if the clone function should heap-allocate a new UTex t struct.
356 * If non-NULL, must refer to an already existing UText, which w ill then
357 * be reset to become the clone.
358 * @param src The UText to be cloned.
359 * @param deep TRUE to request a deep clone, FALSE for a shallow clone.
360 * @param readOnly TRUE to request that the cloned UText have read only access to the
361 * underlying text.
362
363 * @param status Errors are returned here. For deep clones, U_UNSUPPORTED_ERR OR
364 * will be returned if the text provider is unable to clone the
365 * original text.
366 * @return The newly created clone, or NULL if the clone operation faile d.
367 * @stable ICU 3.4
368 */
369 U_STABLE UText * U_EXPORT2
370 utext_clone(UText *dest, const UText *src, UBool deep, UBool readOnly, UErrorCod e *status);
371
372
373 /**
374 * Compare two UText objects for equality.
375 * UTexts are equal if they are iterating over the same text, and
376 * have the same iteration position within the text.
377 * If either or both of the parameters are NULL, the comparison is FALSE.
378 *
379 * @param a The first of the two UTexts to compare.
380 * @param b The other UText to be compared.
381 * @return TRUE if the two UTexts are equal.
382 * @stable ICU 3.6
383 */
384 U_STABLE UBool U_EXPORT2
385 utext_equals(const UText *a, const UText *b);
386
387
388 /*****************************************************************************
389 *
390 * Functions to work with the text represeted by a UText wrapper
391 *
392 *****************************************************************************/
393
394 /**
395 * Get the length of the text. Depending on the characteristics
396 * of the underlying text representation, this may be expensive.
397 * @see utext_isLengthExpensive()
398 *
399 *
400 * @param ut the text to be accessed.
401 * @return the length of the text, expressed in native units.
402 *
403 * @stable ICU 3.4
404 */
405 U_STABLE int64_t U_EXPORT2
406 utext_nativeLength(UText *ut);
407
408 /**
409 * Return TRUE if calculating the length of the text could be expensive.
410 * Finding the length of NUL terminated strings is considered to be expensive.
411 *
412 * Note that the value of this function may change
413 * as the result of other operations on a UText.
414 * Once the length of a string has been discovered, it will no longer
415 * be expensive to report it.
416 *
417 * @param ut the text to be accessed.
418 * @return TRUE if determining the length of the text could be time consuming.
419 * @stable ICU 3.4
420 */
421 U_STABLE UBool U_EXPORT2
422 utext_isLengthExpensive(const UText *ut);
423
424 /**
425 * Returns the code point at the requested index,
426 * or U_SENTINEL (-1) if it is out of bounds.
427 *
428 * If the specified index points to the interior of a multi-unit
429 * character - one of the trail bytes of a UTF-8 sequence, for example -
430 * the complete code point will be returned.
431 *
432 * The iteration position will be set to the start of the returned code point.
433 *
434 * This function is roughly equivalent to the the sequence
435 * utext_setNativeIndex(index);
436 * utext_current32();
437 * (There is a subtle difference if the index is out of bounds by being less tha n zero -
438 * utext_setNativeIndex(negative value) sets the index to zero, after which utex t_current()
439 * will return the char at zero. utext_char32At(negative index), on the other h and, will
440 * return the U_SENTINEL value of -1.)
441 *
442 * @param ut the text to be accessed
443 * @param nativeIndex the native index of the character to be accessed. If the index points
444 * to other than the first unit of a multi-unit character, it will be adj usted
445 * to the start of the character.
446 * @return the code point at the specified index.
447 * @stable ICU 3.4
448 */
449 U_STABLE UChar32 U_EXPORT2
450 utext_char32At(UText *ut, int64_t nativeIndex);
451
452
453 /**
454 *
455 * Get the code point at the current iteration position,
456 * or U_SENTINEL (-1) if the iteration has reached the end of
457 * the input text.
458 *
459 * @param ut the text to be accessed.
460 * @return the Unicode code point at the current iterator position.
461 * @stable ICU 3.4
462 */
463 U_STABLE UChar32 U_EXPORT2
464 utext_current32(UText *ut);
465
466
467 /**
468 * Get the code point at the current iteration position of the UText, and
469 * advance the position to the first index following the character.
470 *
471 * If the position is at the end of the text (the index following
472 * the last character, which is also the length of the text),
473 * return U_SENTINEL (-1) and do not advance the index.
474 *
475 * This is a post-increment operation.
476 *
477 * An inline macro version of this function, UTEXT_NEXT32(),
478 * is available for performance critical use.
479 *
480 * @param ut the text to be accessed.
481 * @return the Unicode code point at the iteration position.
482 * @see UTEXT_NEXT32
483 * @stable ICU 3.4
484 */
485 U_STABLE UChar32 U_EXPORT2
486 utext_next32(UText *ut);
487
488
489 /**
490 * Move the iterator position to the character (code point) whose
491 * index precedes the current position, and return that character.
492 * This is a pre-decrement operation.
493 *
494 * If the initial position is at the start of the text (index of 0)
495 * return U_SENTINEL (-1), and leave the position unchanged.
496 *
497 * An inline macro version of this function, UTEXT_PREVIOUS32(),
498 * is available for performance critical use.
499 *
500 * @param ut the text to be accessed.
501 * @return the previous UChar32 code point, or U_SENTINEL (-1)
502 * if the iteration has reached the start of the text.
503 * @see UTEXT_PREVIOUS32
504 * @stable ICU 3.4
505 */
506 U_STABLE UChar32 U_EXPORT2
507 utext_previous32(UText *ut);
508
509
510 /**
511 * Set the iteration index and return the code point at that index.
512 * Leave the iteration index at the start of the following code point.
513 *
514 * This function is the most efficient and convenient way to
515 * begin a forward iteration. The results are identical to the those
516 * from the sequence
517 * \code
518 * utext_setIndex();
519 * utext_next32();
520 * \endcode
521 *
522 * @param ut the text to be accessed.
523 * @param nativeIndex Iteration index, in the native units of the text provide r.
524 * @return Code point which starts at or before index,
525 * or U_SENTINEL (-1) if it is out of bounds.
526 * @stable ICU 3.4
527 */
528 U_STABLE UChar32 U_EXPORT2
529 utext_next32From(UText *ut, int64_t nativeIndex);
530
531
532
533 /**
534 * Set the iteration index, and return the code point preceding the
535 * one specified by the initial index. Leave the iteration position
536 * at the start of the returned code point.
537 *
538 * This function is the most efficient and convenient way to
539 * begin a backwards iteration.
540 *
541 * @param ut the text to be accessed.
542 * @param nativeIndex Iteration index in the native units of the text provider.
543 * @return Code point preceding the one at the initial index,
544 * or U_SENTINEL (-1) if it is out of bounds.
545 *
546 * @stable ICU 3.4
547 */
548 U_STABLE UChar32 U_EXPORT2
549 utext_previous32From(UText *ut, int64_t nativeIndex);
550
551 /**
552 * Get the current iterator position, which can range from 0 to
553 * the length of the text.
554 * The position is a native index into the input text, in whatever format it
555 * may have (possibly UTF-8 for example), and may not always be the same as
556 * the corresponding UChar (UTF-16) index.
557 * The returned position will always be aligned to a code point boundary.
558 *
559 * @param ut the text to be accessed.
560 * @return the current index position, in the native units of the text provider .
561 * @stable ICU 3.4
562 */
563 U_STABLE int64_t U_EXPORT2
564 utext_getNativeIndex(const UText *ut);
565
566 /**
567 * Set the current iteration position to the nearest code point
568 * boundary at or preceding the specified index.
569 * The index is in the native units of the original input text.
570 * If the index is out of range, it will be pinned to be within
571 * the range of the input text.
572 * <p>
573 * It will usually be more efficient to begin an iteration
574 * using the functions utext_next32From() or utext_previous32From()
575 * rather than setIndex().
576 * <p>
577 * Moving the index position to an adjacent character is best done
578 * with utext_next32(), utext_previous32() or utext_moveIndex32().
579 * Attempting to do direct arithmetic on the index position is
580 * complicated by the fact that the size (in native units) of a
581 * character depends on the underlying representation of the character
582 * (UTF-8, UTF-16, UTF-32, arbitrary codepage), and is not
583 * easily knowable.
584 *
585 * @param ut the text to be accessed.
586 * @param nativeIndex the native unit index of the new iteration position.
587 * @stable ICU 3.4
588 */
589 U_STABLE void U_EXPORT2
590 utext_setNativeIndex(UText *ut, int64_t nativeIndex);
591
592 /**
593 * Move the iterator postion by delta code points. The number of code points
594 * is a signed number; a negative delta will move the iterator backwards,
595 * towards the start of the text.
596 * <p>
597 * The index is moved by <code>delta</code> code points
598 * forward or backward, but no further backward than to 0 and
599 * no further forward than to utext_nativeLength().
600 * The resulting index value will be in between 0 and length, inclusive.
601 *
602 * @param ut the text to be accessed.
603 * @param delta the signed number of code points to move the iteration position.
604 * @return TRUE if the position could be moved the requested number of positions while
605 * staying within the range [0 - text length].
606 * @stable ICU 3.4
607 */
608 U_STABLE UBool U_EXPORT2
609 utext_moveIndex32(UText *ut, int32_t delta);
610
611 /**
612 * Get the native index of the character preceeding the current position.
613 * If the iteration position is already at the start of the text, zero
614 * is returned.
615 * The value returned is the same as that obtained from the following sequence,
616 * but without the side effect of changing the iteration position.
617 *
618 * \code
619 * UText *ut = whatever;
620 * ...
621 * utext_previous(ut)
622 * utext_getNativeIndex(ut);
623 * \endcode
624 *
625 * This function is most useful during forwards iteration, where it will get the
626 * native index of the character most recently returned from utext_next().
627 *
628 * @param ut the text to be accessed
629 * @return the native index of the character preceeding the current index positi on,
630 * or zero if the current position is at the start of the text.
631 * @stable ICU 3.6
632 */
633 U_STABLE int64_t U_EXPORT2
634 utext_getPreviousNativeIndex(UText *ut);
635
636
637 /**
638 *
639 * Extract text from a UText into a UChar buffer. The range of text to be extra cted
640 * is specified in the native indices of the UText provider. These may not nece ssarily
641 * be UTF-16 indices.
642 * <p>
643 * The size (number of 16 bit UChars) of the data to be extracted is returned. The
644 * full number of UChars is returned, even when the extracted text is truncated
645 * because the specified buffer size is too small.
646 * <p>
647 * The extracted string will (if you are a user) / must (if you are a text provi der)
648 * be NUL-terminated if there is sufficient space in the destination buffer. Th is
649 * terminating NUL is not included in the returned length.
650 * <p>
651 * The iteration index is left at the position following the last extracted char acter.
652 *
653 * @param ut the UText from which to extract data.
654 * @param nativeStart the native index of the first character to extract.\
655 * If the specified index is out of range,
656 * it will be pinned to to be within 0 <= index <= textLength
657 * @param nativeLimit the native string index of the position following the las t
658 * character to extract. If the specified index is out of range,
659 * it will be pinned to to be within 0 <= index <= textLength.
660 * nativeLimit must be >= nativeStart.
661 * @param dest the UChar (UTF-16) buffer into which the extracted text is plac ed
662 * @param destCapacity The size, in UChars, of the destination buffer. May be zero
663 * for precomputing the required size.
664 * @param status receives any error status.
665 * U_BUFFER_OVERFLOW_ERROR: the extracted text was truncated because the
666 * buffer was too small. Returns number of UChars for preflighting.
667 * @return Number of UChars in the data to be extracted. Does not include a tra iling NUL.
668 *
669 * @stable ICU 3.4
670 */
671 U_STABLE int32_t U_EXPORT2
672 utext_extract(UText *ut,
673 int64_t nativeStart, int64_t nativeLimit,
674 UChar *dest, int32_t destCapacity,
675 UErrorCode *status);
676
677
678 /**
679 * Compare two UTexts (binary order). The comparison begins at each source text' s
680 * iteration position. The iteration position of each UText will be left followi ng
681 * the last character compared.
682 *
683 * The comparison is done in code point order; unlike u_strCompare, you
684 * cannot choose to use code unit order. This is because the characters
685 * in a UText are accessed one code point at a time, and may not be from a UTF-1 6
686 * context.
687 *
688 * This functions works with strings of different explicitly specified lengths
689 * unlike the ANSI C-like u_strcmp() and u_memcmp() etc.
690 * A length argument of -1 signifies that as much of the string should be used a s
691 * is necessary to compare with the other string. If both length arguments are - 1,
692 * the entire remaining portionss of both strings are used.
693 *
694 * @param s1 First source string.
695 * @param length1 Length of first source string in UTF-32 code points.
696 *
697 * @param s2 Second source string.
698 * @param length2 Length of second source string in UTF-32 code points.
699 *
700 * @return <0 or 0 or >0 as usual for string comparisons
701 *
702 * @internal ICU 4.4 technology preview
703 */
704 U_INTERNAL int32_t U_EXPORT2
705 utext_compare(UText *s1, int32_t length1,
706 UText *s2, int32_t length2);
707
708 /**
709 * Compare two UTexts (binary order). The comparison begins at each source text' s
710 * iteration position. The iteration position of each UText will be left followi ng
711 * the last character compared. This method differs from utext_compare in that
712 * it accepts native limits rather than lengths for each string.
713 *
714 * The comparison is done in code point order; unlike u_strCompare, you
715 * cannot choose to use code unit order. This is because the characters
716 * in a UText are accessed one code point at a time, and may not be from a UTF-1 6
717 * context.
718 *
719 * This functions works with strings of different explicitly specified lengths
720 * unlike the ANSI C-like u_strcmp() and u_memcmp() etc.
721 * A limit argument of -1 signifies that as much of the string should be used as
722 * is necessary to compare with the other string. If both limit arguments are -1 ,
723 * the entire remaining portionss of both strings are used.
724 *
725 * @param s1 First source string.
726 * @param limit1 Native index of the last character in the first source string t o be considered.
727 *
728 * @param s2 Second source string.
729 * @param limit2 Native index of the last character in the second source string to be considered.
730 *
731 * @return <0 or 0 or >0 as usual for string comparisons
732 *
733 * @internal ICU 4.4 technology preview
734 */
735 U_INTERNAL int32_t U_EXPORT2
736 utext_compareNativeLimit(UText *s1, int64_t limit1,
737 UText *s2, int64_t limit2);
738
739 /**
740 * Compare two UTexts case-insensitively using full case folding. The comparison
741 * begins at each source text's iteration position. The iteration position of ea ch
742 * UText will be left following the last character compared.
743 *
744 * The comparison is done in code point order; this is because the characters
745 * in a UText are accessed one code point at a time, and may not be from a UTF-1 6
746 * context.
747 *
748 * This functions works with strings of different explicitly specified lengths
749 * unlike the ANSI C-like u_strcmp() and u_memcmp() etc.
750 * A length argument of -1 signifies that as much of the string should be used a s
751 * is necessary to compare with the other string. If both length arguments are - 1,
752 * the entire remaining portionss of both strings are used.
753 *
754 * @param s1 First source string.
755 * @param length1 Length of first source string in UTF-32 code points.
756 *
757 * @param s2 Second source string.
758 * @param length2 Length of second source string in UTF-32 code points.
759 *
760 * @param options A bit set of options:
761 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
762 * Comparison in code point order with default case folding.
763 *
764 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
765 *
766 * @param pErrorCode Must be a valid pointer to an error code value,
767 * which must not indicate a failure before the function call.
768 *
769 * @return <0 or 0 or >0 as usual for string comparisons
770 *
771 * @internal ICU 4.4 technology preview
772 */
773 U_INTERNAL int32_t U_EXPORT2
774 utext_caseCompare(UText *s1, int32_t length1,
775 UText *s2, int32_t length2,
776 uint32_t options, UErrorCode *pErrorCode);
777
778 /**
779 * Compare two UTexts case-insensitively using full case folding. The comparison
780 * begins at each source text's iteration position. The iteration position of ea ch
781 * UText will be left following the last character compared. This method differs from
782 * utext_caseCompare in that it accepts native limits rather than lengths for ea ch
783 * string.
784 *
785 * The comparison is done in code point order; this is because the characters
786 * in a UText are accessed one code point at a time, and may not be from a UTF-1 6
787 * context.
788 *
789 * This functions works with strings of different explicitly specified lengths
790 * unlike the ANSI C-like u_strcmp() and u_memcmp() etc.
791 * A limit argument of -1 signifies that as much of the string should be used as
792 * is necessary to compare with the other string. If both length arguments are - 1,
793 * the entire remaining portionss of both strings are used.
794 *
795 * @param s1 First source string.
796 * @param limit1 Native index of the last character in the first source string t o be considered.
797 *
798 * @param s2 Second source string.
799 * @param limit2 Native index of the last character in the second source string to be considered.
800 *
801 * @param options A bit set of options:
802 * - U_FOLD_CASE_DEFAULT or 0 is used for default options:
803 * Comparison in code point order with default case folding.
804 *
805 * - U_FOLD_CASE_EXCLUDE_SPECIAL_I
806 *
807 * @param pErrorCode Must be a valid pointer to an error code value,
808 * which must not indicate a failure before the function call.
809 *
810 * @return <0 or 0 or >0 as usual for string comparisons
811 *
812 * @internal ICU 4.4 technology preview
813 */
814 U_INTERNAL int32_t U_EXPORT2
815 utext_caseCompareNativeLimit(UText *s1, int64_t limit1,
816 UText *s2, int64_t limit2,
817 uint32_t options, UErrorCode *pErrorCode);
818
819
820 /******************************************************************************* *****
821 *
822 * #define inline versions of selected performance-critical text access functio ns
823 * Caution: do not use auto increment++ or decrement-- expressions
824 * as parameters to these macros.
825 *
826 * For most use, where there is no extreme performance constraint, the
827 * normal, non-inline functions are a better choice. The resulting cod e
828 * will be smaller, and, if the need ever arises, easier to debug.
829 *
830 * These are implemented as #defines rather than real functions
831 * because there is no fully portable way to do inline functions in pla in C.
832 *
833 ******************************************************************************* *****/
834
835 /**
836 * inline version of utext_current32(), for performance-critical situations.
837 *
838 * Get the code point at the current iteration position of the UText.
839 * Returns U_SENTINEL (-1) if the position is at the end of the
840 * text.
841 *
842 * @internal ICU 4.4 technology preview
843 */
844 #define UTEXT_CURRENT32(ut) \
845 ((ut)->chunkOffset < (ut)->chunkLength && ((ut)->chunkContents)[(ut)->chunkO ffset]<0xd800 ? \
846 ((ut)->chunkContents)[((ut)->chunkOffset)] : utext_current32(ut))
847
848 /**
849 * inline version of utext_next32(), for performance-critical situations.
850 *
851 * Get the code point at the current iteration position of the UText, and
852 * advance the position to the first index following the character.
853 * This is a post-increment operation.
854 * Returns U_SENTINEL (-1) if the position is at the end of the
855 * text.
856 *
857 * @stable ICU 3.4
858 */
859 #define UTEXT_NEXT32(ut) \
860 ((ut)->chunkOffset < (ut)->chunkLength && ((ut)->chunkContents)[(ut)->chunkO ffset]<0xd800 ? \
861 ((ut)->chunkContents)[((ut)->chunkOffset)++] : utext_next32(ut))
862
863 /**
864 * inline version of utext_previous32(), for performance-critical situations.
865 *
866 * Move the iterator position to the character (code point) whose
867 * index precedes the current position, and return that character.
868 * This is a pre-decrement operation.
869 * Returns U_SENTINEL (-1) if the position is at the start of the text.
870 *
871 * @stable ICU 3.4
872 */
873 #define UTEXT_PREVIOUS32(ut) \
874 ((ut)->chunkOffset > 0 && \
875 (ut)->chunkContents[(ut)->chunkOffset-1] < 0xd800 ? \
876 (ut)->chunkContents[--((ut)->chunkOffset)] : utext_previous32(ut))
877
878 /**
879 * inline version of utext_getNativeIndex(), for performance-critical situatio ns.
880 *
881 * Get the current iterator position, which can range from 0 to
882 * the length of the text.
883 * The position is a native index into the input text, in whatever format it
884 * may have (possibly UTF-8 for example), and may not always be the same as
885 * the corresponding UChar (UTF-16) index.
886 * The returned position will always be aligned to a code point boundary.
887 *
888 * @stable ICU 3.6
889 */
890 #define UTEXT_GETNATIVEINDEX(ut) \
891 ((ut)->chunkOffset <= (ut)->nativeIndexingLimit? \
892 (ut)->chunkNativeStart+(ut)->chunkOffset : \
893 (ut)->pFuncs->mapOffsetToNative(ut))
894
895 /**
896 * inline version of utext_setNativeIndex(), for performance-critical situatio ns.
897 *
898 * Set the current iteration position to the nearest code point
899 * boundary at or preceding the specified index.
900 * The index is in the native units of the original input text.
901 * If the index is out of range, it will be pinned to be within
902 * the range of the input text.
903 *
904 * @stable ICU 3.8
905 */
906 #define UTEXT_SETNATIVEINDEX(ut, ix) \
907 { int64_t __offset = (ix) - (ut)->chunkNativeStart; \
908 if (__offset>=0 && __offset<=(int64_t)(ut)->nativeIndexingLimit) { \
909 (ut)->chunkOffset=(int32_t)__offset; \
910 } else { \
911 utext_setNativeIndex((ut), (ix)); } }
912
913
914
915 /******************************************************************************* *****
916 *
917 * Functions related to writing or modifying the text.
918 * These will work only with modifiable UTexts. Attempting to
919 * modify a read-only UText will return an error status.
920 *
921 ******************************************************************************* *****/
922
923
924 /**
925 * Return TRUE if the text can be written (modified) with utext_replace() or
926 * utext_copy(). For the text to be writable, the text provider must
927 * be of a type that supports writing and the UText must not be frozen.
928 *
929 * Attempting to modify text when utext_isWriteable() is FALSE will fail -
930 * the text will not be modified, and an error will be returned from the functi on
931 * that attempted the modification.
932 *
933 * @param ut the UText to be tested.
934 * @return TRUE if the text is modifiable.
935 *
936 * @see utext_freeze()
937 * @see utext_replace()
938 * @see utext_copy()
939 * @stable ICU 3.4
940 *
941 */
942 U_STABLE UBool U_EXPORT2
943 utext_isWritable(const UText *ut);
944
945
946 /**
947 * Test whether there is meta data associated with the text.
948 * @see Replaceable::hasMetaData()
949 *
950 * @param ut The UText to be tested
951 * @return TRUE if the underlying text includes meta data.
952 * @stable ICU 3.4
953 */
954 U_STABLE UBool U_EXPORT2
955 utext_hasMetaData(const UText *ut);
956
957
958 /**
959 * Replace a range of the original text with a replacement text.
960 *
961 * Leaves the current iteration position at the position following the
962 * newly inserted replacement text.
963 *
964 * This function is only available on UText types that support writing,
965 * that is, ones where utext_isWritable() returns TRUE.
966 *
967 * When using this function, there should be only a single UText opened onto the
968 * underlying native text string. Behavior after a replace operation
969 * on a UText is undefined for any other additional UTexts that refer to the
970 * modified string.
971 *
972 * @param ut the UText representing the text to be operated on.
973 * @param nativeStart the native index of the start of the region to be rep laced
974 * @param nativeLimit the native index of the character following the regio n to be replaced.
975 * @param replacementText pointer to the replacement text
976 * @param replacementLength length of the replacement text, or -1 if the text is NUL terminated.
977 * @param status receives any error status. Possible errors include
978 * U_NO_WRITE_PERMISSION
979 *
980 * @return The signed number of (native) storage units by which
981 * the length of the text expanded or contracted.
982 *
983 * @stable ICU 3.4
984 */
985 U_STABLE int32_t U_EXPORT2
986 utext_replace(UText *ut,
987 int64_t nativeStart, int64_t nativeLimit,
988 const UChar *replacementText, int32_t replacementLength,
989 UErrorCode *status);
990
991
992
993 /**
994 *
995 * Copy or move a substring from one position to another within the text,
996 * while retaining any metadata associated with the text.
997 * This function is used to duplicate or reorder substrings.
998 * The destination index must not overlap the source range.
999 *
1000 * The text to be copied or moved is inserted at destIndex;
1001 * it does not replace or overwrite any existing text.
1002 *
1003 * The iteration position is left following the newly inserted text
1004 * at the destination position.
1005 *
1006 * This function is only available on UText types that support writing,
1007 * that is, ones where utext_isWritable() returns TRUE.
1008 *
1009 * When using this function, there should be only a single UText opened onto the
1010 * underlying native text string. Behavior after a copy operation
1011 * on a UText is undefined in any other additional UTexts that refer to the
1012 * modified string.
1013 *
1014 * @param ut The UText representing the text to be operated on.
1015 * @param nativeStart The native index of the start of the region to be copied or moved
1016 * @param nativeLimit The native index of the character position following the region
1017 * to be copied.
1018 * @param destIndex The native destination index to which the source substrin g is
1019 * copied or moved.
1020 * @param move If TRUE, then the substring is moved, not copied/duplicat ed.
1021 * @param status receives any error status. Possible errors include U_NO_ WRITE_PERMISSION
1022 *
1023 * @stable ICU 3.4
1024 */
1025 U_STABLE void U_EXPORT2
1026 utext_copy(UText *ut,
1027 int64_t nativeStart, int64_t nativeLimit,
1028 int64_t destIndex,
1029 UBool move,
1030 UErrorCode *status);
1031
1032
1033 /**
1034 * <p>
1035 * Freeze a UText. This prevents any modification to the underlying text itse lf
1036 * by means of functions operating on this UText.
1037 * </p>
1038 * <p>
1039 * Once frozen, a UText can not be unfrozen. The intent is to ensure
1040 * that a the text underlying a frozen UText wrapper cannot be modified via th at UText.
1041 * </p>
1042 * <p>
1043 * Caution: freezing a UText will disable changes made via the specific
1044 * frozen UText wrapper only; it will not have any effect on the ability to
1045 * directly modify the text by bypassing the UText. Any such backdoor modifi cations
1046 * are always an error while UText access is occuring because the underlying
1047 * text can get out of sync with UText's buffering.
1048 * </p>
1049 *
1050 * @param ut The UText to be frozen.
1051 * @see utext_isWritable()
1052 * @stable ICU 3.6
1053 */
1054 U_STABLE void U_EXPORT2
1055 utext_freeze(UText *ut);
1056
1057
1058 /**
1059 * UText provider properties (bit field indexes).
1060 *
1061 * @see UText
1062 * @stable ICU 3.4
1063 */
1064 enum {
1065 /**
1066 * It is potentially time consuming for the provider to determine the length of the text.
1067 * @stable ICU 3.4
1068 */
1069 UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE = 1,
1070 /**
1071 * Text chunks remain valid and usable until the text object is modified or
1072 * deleted, not just until the next time the access() function is called
1073 * (which is the default).
1074 * @stable ICU 3.4
1075 */
1076 UTEXT_PROVIDER_STABLE_CHUNKS = 2,
1077 /**
1078 * The provider supports modifying the text via the replace() and copy()
1079 * functions.
1080 * @see Replaceable
1081 * @stable ICU 3.4
1082 */
1083 UTEXT_PROVIDER_WRITABLE = 3,
1084 /**
1085 * There is meta data associated with the text.
1086 * @see Replaceable::hasMetaData()
1087 * @stable ICU 3.4
1088 */
1089 UTEXT_PROVIDER_HAS_META_DATA = 4,
1090 /**
1091 * Text provider owns the text storage.
1092 * Generally occurs as the result of a deep clone of the UText.
1093 * When closing the UText, the associated text must
1094 * also be closed/deleted/freed/ whatever is appropriate.
1095 * @stable ICU 3.6
1096 */
1097 UTEXT_PROVIDER_OWNS_TEXT = 5
1098 };
1099
1100 /**
1101 * Function type declaration for UText.clone().
1102 *
1103 * clone a UText. Much like opening a UText where the source text is itself
1104 * another UText.
1105 *
1106 * A deep clone will copy both the UText data structures and the underlying te xt.
1107 * The original and cloned UText will operate completely independently; modifi cations
1108 * made to the text in one will not effect the other. Text providers are not
1109 * required to support deep clones. The user of clone() must check the status return
1110 * and be prepared to handle failures.
1111 *
1112 * A shallow clone replicates only the UText data structures; it does not make
1113 * a copy of the underlying text. Shallow clones can be used as an efficient way to
1114 * have multiple iterators active in a single text string that is not being
1115 * modified.
1116 *
1117 * A shallow clone operation must not fail except for truly exceptional condit ions such
1118 * as memory allocation failures.
1119 *
1120 * A UText and its clone may be safely concurrently accessed by separate threa ds.
1121 * This is true for both shallow and deep clones.
1122 * It is the responsibility of the Text Provider to ensure that this thread sa fety
1123 * constraint is met.
1124
1125 *
1126 * @param dest A UText struct to be filled in with the result of the clone o peration,
1127 * or NULL if the clone function should heap-allocate a new UTex t struct.
1128 * @param src The UText to be cloned.
1129 * @param deep TRUE to request a deep clone, FALSE for a shallow clone.
1130 * @param status Errors are returned here. For deep clones, U_UNSUPPORTED_ERR OR
1131 * should be returned if the text provider is unable to clone th e
1132 * original text.
1133 * @return The newly created clone, or NULL if the clone operation faile d.
1134 *
1135 * @stable ICU 3.4
1136 */
1137 typedef UText * U_CALLCONV
1138 UTextClone(UText *dest, const UText *src, UBool deep, UErrorCode *status);
1139
1140
1141 /**
1142 * Function type declaration for UText.nativeLength().
1143 *
1144 * @param ut the UText to get the length of.
1145 * @return the length, in the native units of the original text string.
1146 * @see UText
1147 * @stable ICU 3.4
1148 */
1149 typedef int64_t U_CALLCONV
1150 UTextNativeLength(UText *ut);
1151
1152 /**
1153 * Function type declaration for UText.access(). Get the description of the tex t chunk
1154 * containing the text at a requested native index. The UText's iteration
1155 * position will be left at the requested index. If the index is out
1156 * of bounds, the iteration position will be left at the start or end
1157 * of the string, as appropriate.
1158 *
1159 * Chunks must begin and end on code point boundaries. A single code point
1160 * comprised of multiple storage units must never span a chunk boundary.
1161 *
1162 *
1163 * @param ut the UText being accessed.
1164 * @param nativeIndex Requested index of the text to be accessed.
1165 * @param forward If TRUE, then the returned chunk must contain text
1166 * starting from the index, so that start<=index<limit.
1167 * If FALSE, then the returned chunk must contain text
1168 * before the index, so that start<index<=limit.
1169 * @return True if the requested index could be accessed. The chunk
1170 * will contain the requested text.
1171 * False value if a chunk cannot be accessed
1172 * (the requested index is out of bounds).
1173 *
1174 * @see UText
1175 * @stable ICU 3.4
1176 */
1177 typedef UBool U_CALLCONV
1178 UTextAccess(UText *ut, int64_t nativeIndex, UBool forward);
1179
1180 /**
1181 * Function type declaration for UText.extract().
1182 *
1183 * Extract text from a UText into a UChar buffer. The range of text to be extra cted
1184 * is specified in the native indices of the UText provider. These may not nece ssarily
1185 * be UTF-16 indices.
1186 * <p>
1187 * The size (number of 16 bit UChars) in the data to be extracted is returned. The
1188 * full amount is returned, even when the specified buffer size is smaller.
1189 * <p>
1190 * The extracted string will (if you are a user) / must (if you are a text provi der)
1191 * be NUL-terminated if there is sufficient space in the destination buffer.
1192 *
1193 * @param ut the UText from which to extract data.
1194 * @param nativeStart the native index of the first characer to extract.
1195 * @param nativeLimit the native string index of the position following the l ast
1196 * character to extract.
1197 * @param dest the UChar (UTF-16) buffer into which the extracted text is placed
1198 * @param destCapacity The size, in UChars, of the destination buffer. May be zero
1199 * for precomputing the required size.
1200 * @param status receives any error status.
1201 * If U_BUFFER_OVERFLOW_ERROR: Returns number of UChars fo r
1202 * preflighting.
1203 * @return Number of UChars in the data. Does not include a trailing NUL.
1204 *
1205 * @stable ICU 3.4
1206 */
1207 typedef int32_t U_CALLCONV
1208 UTextExtract(UText *ut,
1209 int64_t nativeStart, int64_t nativeLimit,
1210 UChar *dest, int32_t destCapacity,
1211 UErrorCode *status);
1212
1213 /**
1214 * Function type declaration for UText.replace().
1215 *
1216 * Replace a range of the original text with a replacement text.
1217 *
1218 * Leaves the current iteration position at the position following the
1219 * newly inserted replacement text.
1220 *
1221 * This function need only be implemented on UText types that support writing.
1222 *
1223 * When using this function, there should be only a single UText opened onto the
1224 * underlying native text string. The function is responsible for updating the
1225 * text chunk within the UText to reflect the updated iteration position,
1226 * taking into account any changes to the underlying string's structure caused
1227 * by the replace operation.
1228 *
1229 * @param ut the UText representing the text to be operated on.
1230 * @param nativeStart the index of the start of the region to be replaced
1231 * @param nativeLimit the index of the character following the region to be replaced.
1232 * @param replacementText pointer to the replacement text
1233 * @param replacmentLength length of the replacement text in UChars, or -1 if th e text is NUL terminated.
1234 * @param status receives any error status. Possible errors include
1235 * U_NO_WRITE_PERMISSION
1236 *
1237 * @return The signed number of (native) storage units by which
1238 * the length of the text expanded or contracted.
1239 *
1240 * @stable ICU 3.4
1241 */
1242 typedef int32_t U_CALLCONV
1243 UTextReplace(UText *ut,
1244 int64_t nativeStart, int64_t nativeLimit,
1245 const UChar *replacementText, int32_t replacmentLength,
1246 UErrorCode *status);
1247
1248 /**
1249 * Function type declaration for UText.copy().
1250 *
1251 * Copy or move a substring from one position to another within the text,
1252 * while retaining any metadata associated with the text.
1253 * This function is used to duplicate or reorder substrings.
1254 * The destination index must not overlap the source range.
1255 *
1256 * The text to be copied or moved is inserted at destIndex;
1257 * it does not replace or overwrite any existing text.
1258 *
1259 * This function need only be implemented for UText types that support writing.
1260 *
1261 * When using this function, there should be only a single UText opened onto the
1262 * underlying native text string. The function is responsible for updating the
1263 * text chunk within the UText to reflect the updated iteration position,
1264 * taking into account any changes to the underlying string's structure caused
1265 * by the replace operation.
1266 *
1267 * @param ut The UText representing the text to be operated on.
1268 * @param nativeStart The index of the start of the region to be copied or move d
1269 * @param nativeLimit The index of the character following the region to be rep laced.
1270 * @param nativeDest The destination index to which the source substring is co pied or moved.
1271 * @param move If TRUE, then the substring is moved, not copied/duplicat ed.
1272 * @param status receives any error status. Possible errors include U_NO_ WRITE_PERMISSION
1273 *
1274 * @stable ICU 3.4
1275 */
1276 typedef void U_CALLCONV
1277 UTextCopy(UText *ut,
1278 int64_t nativeStart, int64_t nativeLimit,
1279 int64_t nativeDest,
1280 UBool move,
1281 UErrorCode *status);
1282
1283 /**
1284 * Function type declaration for UText.mapOffsetToNative().
1285 * Map from the current UChar offset within the current text chunk to
1286 * the corresponding native index in the original source text.
1287 *
1288 * This is required only for text providers that do not use native UTF-16 indexe s.
1289 *
1290 * @param ut the UText.
1291 * @return Absolute (native) index corresponding to chunkOffset in the current c hunk.
1292 * The returned native index should always be to a code point boundary.
1293 *
1294 * @stable ICU 3.4
1295 */
1296 typedef int64_t U_CALLCONV
1297 UTextMapOffsetToNative(const UText *ut);
1298
1299 /**
1300 * Function type declaration for UText.mapIndexToUTF16().
1301 * Map from a native index to a UChar offset within a text chunk.
1302 * Behavior is undefined if the native index does not fall within the
1303 * current chunk.
1304 *
1305 * This function is required only for text providers that do not use native UTF- 16 indexes.
1306 *
1307 * @param ut The UText containing the text chunk.
1308 * @param nativeIndex Absolute (native) text index, chunk->start<=index<=chunk-> limit.
1309 * @return Chunk-relative UTF-16 offset corresponding to the specifie d native
1310 * index.
1311 *
1312 * @stable ICU 3.4
1313 */
1314 typedef int32_t U_CALLCONV
1315 UTextMapNativeIndexToUTF16(const UText *ut, int64_t nativeIndex);
1316
1317
1318 /**
1319 * Function type declaration for UText.utextClose().
1320 *
1321 * A Text Provider close function is only required for provider types that make
1322 * allocations in their open function (or other functions) that must be
1323 * cleaned when the UText is closed.
1324 *
1325 * The allocation of the UText struct itself and any "extra" storage
1326 * associated with the UText is handled by the common UText implementation
1327 * and does not require provider specific cleanup in a close function.
1328 *
1329 * Most UText provider implementations do not need to implement this function.
1330 *
1331 * @param ut A UText object to be closed.
1332 *
1333 * @stable ICU 3.4
1334 */
1335 typedef void U_CALLCONV
1336 UTextClose(UText *ut);
1337
1338
1339 /**
1340 * (public) Function dispatch table for UText.
1341 * Conceptually very much like a C++ Virtual Function Table.
1342 * This struct defines the organization of the table.
1343 * Each text provider implementation must provide an
1344 * actual table that is initialized with the appropriate functions
1345 * for the type of text being handled.
1346 * @stable ICU 3.6
1347 */
1348 struct UTextFuncs {
1349 /**
1350 * (public) Function table size, sizeof(UTextFuncs)
1351 * Intended for use should the table grow to accomodate added
1352 * functions in the future, to allow tests for older format
1353 * function tables that do not contain the extensions.
1354 *
1355 * Fields are placed for optimal alignment on
1356 * 32/64/128-bit-pointer machines, by normally grouping together
1357 * 4 32-bit fields,
1358 * 4 pointers,
1359 * 2 64-bit fields
1360 * in sequence.
1361 * @stable ICU 3.6
1362 */
1363 int32_t tableSize;
1364
1365 /**
1366 * (private) Alignment padding.
1367 * Do not use, reserved for use by the UText framework only.
1368 * @internal
1369 */
1370 int32_t reserved1, /** @internal */ reserved2, /** @internal */ reserv ed3;
1371
1372
1373 /**
1374 * (public) Function pointer for UTextClone
1375 *
1376 * @see UTextClone
1377 * @stable ICU 3.6
1378 */
1379 UTextClone *clone;
1380
1381 /**
1382 * (public) function pointer for UTextLength
1383 * May be expensive to compute!
1384 *
1385 * @see UTextLength
1386 * @stable ICU 3.6
1387 */
1388 UTextNativeLength *nativeLength;
1389
1390 /**
1391 * (public) Function pointer for UTextAccess.
1392 *
1393 * @see UTextAccess
1394 * @stable ICU 3.6
1395 */
1396 UTextAccess *access;
1397
1398 /**
1399 * (public) Function pointer for UTextExtract.
1400 *
1401 * @see UTextExtract
1402 * @stable ICU 3.6
1403 */
1404 UTextExtract *extract;
1405
1406 /**
1407 * (public) Function pointer for UTextReplace.
1408 *
1409 * @see UTextReplace
1410 * @stable ICU 3.6
1411 */
1412 UTextReplace *replace;
1413
1414 /**
1415 * (public) Function pointer for UTextCopy.
1416 *
1417 * @see UTextCopy
1418 * @stable ICU 3.6
1419 */
1420 UTextCopy *copy;
1421
1422 /**
1423 * (public) Function pointer for UTextMapOffsetToNative.
1424 *
1425 * @see UTextMapOffsetToNative
1426 * @stable ICU 3.6
1427 */
1428 UTextMapOffsetToNative *mapOffsetToNative;
1429
1430 /**
1431 * (public) Function pointer for UTextMapNativeIndexToUTF16.
1432 *
1433 * @see UTextMapNativeIndexToUTF16
1434 * @stable ICU 3.6
1435 */
1436 UTextMapNativeIndexToUTF16 *mapNativeIndexToUTF16;
1437
1438 /**
1439 * (public) Function pointer for UTextClose.
1440 *
1441 * @see UTextClose
1442 * @stable ICU 3.6
1443 */
1444 UTextClose *close;
1445
1446 /**
1447 * (private) Spare function pointer
1448 * @internal
1449 */
1450 UTextClose *spare1;
1451
1452 /**
1453 * (private) Spare function pointer
1454 * @internal
1455 */
1456 UTextClose *spare2;
1457
1458 /**
1459 * (private) Spare function pointer
1460 * @internal
1461 */
1462 UTextClose *spare3;
1463
1464 };
1465 /**
1466 * Function dispatch table for UText
1467 * @see UTextFuncs
1468 */
1469 typedef struct UTextFuncs UTextFuncs;
1470
1471 /**
1472 * UText struct. Provides the interface between the generic UText access cod e
1473 * and the UText provider code that works on specific kinds of
1474 * text (UTF-8, noncontiguous UTF-16, whatever.)
1475 *
1476 * Applications that are using predefined types of text provid ers
1477 * to pass text data to ICU services will have no need to view the
1478 * internals of the UText structs that they open.
1479 *
1480 * @stable ICU 3.6
1481 */
1482 struct UText {
1483 /**
1484 * (private) Magic. Used to help detect when UText functions are hande d
1485 * invalid or unitialized UText structs.
1486 * utext_openXYZ() functions take an initialized,
1487 * but not necessarily open, UText struct as an
1488 * optional fill-in parameter. This magic field
1489 * is used to check for that initialization.
1490 * Text provider close functions must NOT clear
1491 * the magic field because that would prevent
1492 * reuse of the UText struct.
1493 * @internal
1494 */
1495 uint32_t magic;
1496
1497
1498 /**
1499 * (private) Flags for managing the allocation and freeing of
1500 * memory associated with this UText.
1501 * @internal
1502 */
1503 int32_t flags;
1504
1505
1506 /**
1507 * Text provider properties. This set of flags is maintainted by the
1508 * text provider implementation.
1509 * @stable ICU 3.4
1510 */
1511 int32_t providerProperties;
1512
1513 /**
1514 * (public) sizeOfStruct=sizeof(UText)
1515 * Allows possible backward compatible extension.
1516 *
1517 * @stable ICU 3.4
1518 */
1519 int32_t sizeOfStruct;
1520
1521 /* ------ 16 byte alignment boundary ----------- */
1522
1523
1524 /**
1525 * (protected) Native index of the first character position following
1526 * the current chunk.
1527 * @stable ICU 3.6
1528 */
1529 int64_t chunkNativeLimit;
1530
1531 /**
1532 * (protected) Size in bytes of the extra space (pExtra).
1533 * @stable ICU 3.4
1534 */
1535 int32_t extraSize;
1536
1537 /**
1538 * (protected) The highest chunk offset where native indexing and
1539 * chunk (UTF-16) indexing correspond. For UTF-16 sources, value
1540 * will be equal to chunkLength.
1541 *
1542 * @stable ICU 3.6
1543 */
1544 int32_t nativeIndexingLimit;
1545
1546 /* ---- 16 byte alignment boundary------ */
1547
1548 /**
1549 * (protected) Native index of the first character in the text chunk.
1550 * @stable ICU 3.6
1551 */
1552 int64_t chunkNativeStart;
1553
1554 /**
1555 * (protected) Current iteration position within the text chunk (UTF-16 buf fer).
1556 * This is the index to the character that will be returned by utext_next32 ().
1557 * @stable ICU 3.6
1558 */
1559 int32_t chunkOffset;
1560
1561 /**
1562 * (protected) Length the text chunk (UTF-16 buffer), in UChars.
1563 * @stable ICU 3.6
1564 */
1565 int32_t chunkLength;
1566
1567 /* ---- 16 byte alignment boundary-- */
1568
1569
1570 /**
1571 * (protected) pointer to a chunk of text in UTF-16 format.
1572 * May refer either to original storage of the source of the text, or
1573 * if conversion was required, to a buffer owned by the UText.
1574 * @stable ICU 3.6
1575 */
1576 const UChar *chunkContents;
1577
1578 /**
1579 * (public) Pointer to Dispatch table for accessing functions for this UText.
1580 * @stable ICU 3.6
1581 */
1582 const UTextFuncs *pFuncs;
1583
1584 /**
1585 * (protected) Pointer to additional space requested by the
1586 * text provider during the utext_open operation.
1587 * @stable ICU 3.4
1588 */
1589 void *pExtra;
1590
1591 /**
1592 * (protected) Pointer to string or text-containin object or similar.
1593 * This is the source of the text that this UText is wrapping, in a format
1594 * that is known to the text provider functions.
1595 * @stable ICU 3.4
1596 */
1597 const void *context;
1598
1599 /* --- 16 byte alignment boundary--- */
1600
1601 /**
1602 * (protected) Pointer fields available for use by the text provider.
1603 * Not used by UText common code.
1604 * @stable ICU 3.6
1605 */
1606 const void *p;
1607 /**
1608 * (protected) Pointer fields available for use by the text provider.
1609 * Not used by UText common code.
1610 * @stable ICU 3.6
1611 */
1612 const void *q;
1613 /**
1614 * (protected) Pointer fields available for use by the text provider.
1615 * Not used by UText common code.
1616 * @stable ICU 3.6
1617 */
1618 const void *r;
1619
1620 /**
1621 * Private field reserved for future use by the UText framework
1622 * itself. This is not to be touched by the text providers.
1623 * @internal ICU 3.4
1624 */
1625 void *privP;
1626
1627
1628 /* --- 16 byte alignment boundary--- */
1629
1630
1631 /**
1632 * (protected) Integer field reserved for use by the text provider.
1633 * Not used by the UText framework, or by the client (user) of the UText.
1634 * @stable ICU 3.4
1635 */
1636 int64_t a;
1637
1638 /**
1639 * (protected) Integer field reserved for use by the text provider.
1640 * Not used by the UText framework, or by the client (user) of the UText.
1641 * @stable ICU 3.4
1642 */
1643 int32_t b;
1644
1645 /**
1646 * (protected) Integer field reserved for use by the text provider.
1647 * Not used by the UText framework, or by the client (user) of the UText.
1648 * @stable ICU 3.4
1649 */
1650 int32_t c;
1651
1652 /* ---- 16 byte alignment boundary---- */
1653
1654
1655 /**
1656 * Private field reserved for future use by the UText framework
1657 * itself. This is not to be touched by the text providers.
1658 * @internal ICU 3.4
1659 */
1660 int64_t privA;
1661 /**
1662 * Private field reserved for future use by the UText framework
1663 * itself. This is not to be touched by the text providers.
1664 * @internal ICU 3.4
1665 */
1666 int32_t privB;
1667 /**
1668 * Private field reserved for future use by the UText framework
1669 * itself. This is not to be touched by the text providers.
1670 * @internal ICU 3.4
1671 */
1672 int32_t privC;
1673 };
1674
1675
1676 /**
1677 * Common function for use by Text Provider implementations to allocate and/or initialize
1678 * a new UText struct. To be called in the implementation of utext_open() func tions.
1679 * If the supplied UText parameter is null, a new UText struct will be allocate d on the heap.
1680 * If the supplied UText is already open, the provider's close function will be called
1681 * so that the struct can be reused by the open that is in progress.
1682 *
1683 * @param ut pointer to a UText struct to be re-used, or null if a new UText
1684 * should be allocated.
1685 * @param extraSpace The amount of additional space to be allocated as part
1686 * of this UText, for use by types of providers that require
1687 * additional storage.
1688 * @param status Errors are returned here.
1689 * @return pointer to the UText, allocated if necessary, with extra space set up if requested.
1690 * @stable ICU 3.4
1691 */
1692 U_STABLE UText * U_EXPORT2
1693 utext_setup(UText *ut, int32_t extraSpace, UErrorCode *status);
1694
1695 /**
1696 * @internal
1697 * Value used to help identify correctly initialized UText structs.
1698 * Note: must be publicly visible so that UTEXT_INITIALIZER can access it.
1699 */
1700 enum {
1701 UTEXT_MAGIC = 0x345ad82c
1702 };
1703
1704 /**
1705 * initializer to be used with local (stack) instances of a UText
1706 * struct. UText structs must be initialized before passing
1707 * them to one of the utext_open functions.
1708 *
1709 * @stable ICU 3.6
1710 */
1711 #define UTEXT_INITIALIZER { \
1712 UTEXT_MAGIC, /* magic */ \
1713 0, /* flags */ \
1714 0, /* providerProps */ \
1715 sizeof(UText), /* sizeOfStruct */ \
1716 0, /* chunkNativeLimit */ \
1717 0, /* extraSize */ \
1718 0, /* nativeIndexingLimit */ \
1719 0, /* chunkNativeStart */ \
1720 0, /* chunkOffset */ \
1721 0, /* chunkLength */ \
1722 NULL, /* chunkContents */ \
1723 NULL, /* pFuncs */ \
1724 NULL, /* pExtra */ \
1725 NULL, /* context */ \
1726 NULL, NULL, NULL, /* p, q, r */ \
1727 NULL, /* privP */ \
1728 0, 0, 0, /* a, b, c */ \
1729 0, 0, 0 /* privA,B,C, */ \
1730 }
1731
1732
1733 U_CDECL_END
1734
1735
1736
1737 #endif
OLDNEW
« no previous file with comments | « public/common/unicode/usystem.h ('k') | public/common/unicode/utf.h » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698