1/*
2*******************************************************************************
3*
4* Copyright (C) 2004-2012, 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 alternative
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 services.
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 application
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
148U_CDECL_BEGIN
149
150struct UText;
151typedef 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 strings.
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 utext_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 */
181U_STABLE UText * U_EXPORT2
182utext_close(UText *ut);
183
184#if U_SHOW_CPLUSPLUS_API
185
186U_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 */
197U_DEFINE_LOCAL_OPEN_POINTER(LocalUTextPointer, UText, utext_close);
198
199U_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, \uFFFD.
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 created.
214 * If non-NULL, must refer to an initialized UText struct, which will 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 */
224U_STABLE UText * U_EXPORT2
225utext_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 created.
232 * If non-NULL, must refer to an initialized UText struct, which will 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 is
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 */
242U_STABLE UText * U_EXPORT2
243utext_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 created.
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 */
259U_STABLE UText * U_EXPORT2
260utext_openUnicodeString(UText *ut, icu::UnicodeString *s, UErrorCode *status);
261
262
263/**
264 * Open a UText for a const UnicodeString. The resulting UText will not be writable.
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 will 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 */
275U_STABLE UText * U_EXPORT2
276utext_openConstUnicodeString(UText *ut, const icu::UnicodeString *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 will 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 */
291U_STABLE UText * U_EXPORT2
292utext_openReplaceable(UText *ut, icu::Replaceable *rep, UErrorCode *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 will 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 */
306U_STABLE UText * U_EXPORT2
307utext_openCharacterIterator(UText *ut, icu::CharacterIterator *ci, 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 text.
317 * The original and cloned UText will operate completely independently; modifications
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 text
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 conditions such
335 * as memory allocation failures.
336 *
337 * Shallow UText clones should be avoided if the UText functions that modify the
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 refer
346 * to the same underlying text.
347 *
348 * A UText and its clone may be safely concurrently accessed by separate threads.
349 * This is true for read access only with shallow clones, and for both read and
350 * write access with deep clones.
351 * It is the responsibility of the Text Provider to ensure that this thread safety
352 * constraint is met.
353 *
354 * @param dest A UText struct to be filled in with the result of the clone operation,
355 * or NULL if the clone function should heap-allocate a new UText struct.
356 * If non-NULL, must refer to an already existing UText, which will 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_ERROR
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 failed.
367 * @stable ICU 3.4
368 */
369U_STABLE UText * U_EXPORT2
370utext_clone(UText *dest, const UText *src, UBool deep, UBool readOnly, UErrorCode *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 */
384U_STABLE UBool U_EXPORT2
385utext_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 */
405U_STABLE int64_t U_EXPORT2
406utext_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 */
421U_STABLE UBool U_EXPORT2
422utext_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 than zero -
438 * utext_setNativeIndex(negative value) sets the index to zero, after which utext_current()
439 * will return the char at zero. utext_char32At(negative index), on the other hand, 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 adjusted
445 * to the start of the character.
446 * @return the code point at the specified index.
447 * @stable ICU 3.4
448 */
449U_STABLE UChar32 U_EXPORT2
450utext_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 */
463U_STABLE UChar32 U_EXPORT2
464utext_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 */
485U_STABLE UChar32 U_EXPORT2
486utext_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 */
506U_STABLE UChar32 U_EXPORT2
507utext_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 provider.
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 */
528U_STABLE UChar32 U_EXPORT2
529utext_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 */
548U_STABLE UChar32 U_EXPORT2
549utext_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 */
563U_STABLE int64_t U_EXPORT2
564utext_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 */
589U_STABLE void U_EXPORT2
590utext_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 */
608U_STABLE UBool U_EXPORT2
609utext_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 position,
630 * or zero if the current position is at the start of the text.
631 * @stable ICU 3.6
632 */
633U_STABLE int64_t U_EXPORT2
634utext_getPreviousNativeIndex(UText *ut);
635
636
637/**
638 *
639 * Extract text from a UText into a UChar buffer. The range of text to be extracted
640 * is specified in the native indices of the UText provider. These may not necessarily
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 provider)
648 * be NUL-terminated if there is sufficient space in the destination buffer. This
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 character.
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 last
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 placed
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 trailing NUL.
668 *
669 * @stable ICU 3.4
670 */
671U_STABLE int32_t U_EXPORT2
672utext_extract(UText *ut,
673 int64_t nativeStart, int64_t nativeLimit,
674 UChar *dest, int32_t destCapacity,
675 UErrorCode *status);
676
677
678
679/************************************************************************************
680 *
681 * #define inline versions of selected performance-critical text access functions
682 * Caution: do not use auto increment++ or decrement-- expressions
683 * as parameters to these macros.
684 *
685 * For most use, where there is no extreme performance constraint, the
686 * normal, non-inline functions are a better choice. The resulting code
687 * will be smaller, and, if the need ever arises, easier to debug.
688 *
689 * These are implemented as #defines rather than real functions
690 * because there is no fully portable way to do inline functions in plain C.
691 *
692 ************************************************************************************/
693
694#ifndef U_HIDE_INTERNAL_API
695/**
696 * inline version of utext_current32(), for performance-critical situations.
697 *
698 * Get the code point at the current iteration position of the UText.
699 * Returns U_SENTINEL (-1) if the position is at the end of the
700 * text.
701 *
702 * @internal ICU 4.4 technology preview
703 */
704#define UTEXT_CURRENT32(ut) \
705 ((ut)->chunkOffset < (ut)->chunkLength && ((ut)->chunkContents)[(ut)->chunkOffset]<0xd800 ? \
706 ((ut)->chunkContents)[((ut)->chunkOffset)] : utext_current32(ut))
707#endif /* U_HIDE_INTERNAL_API */
708
709/**
710 * inline version of utext_next32(), for performance-critical situations.
711 *
712 * Get the code point at the current iteration position of the UText, and
713 * advance the position to the first index following the character.
714 * This is a post-increment operation.
715 * Returns U_SENTINEL (-1) if the position is at the end of the
716 * text.
717 *
718 * @stable ICU 3.4
719 */
720#define UTEXT_NEXT32(ut) \
721 ((ut)->chunkOffset < (ut)->chunkLength && ((ut)->chunkContents)[(ut)->chunkOffset]<0xd800 ? \
722 ((ut)->chunkContents)[((ut)->chunkOffset)++] : utext_next32(ut))
723
724/**
725 * inline version of utext_previous32(), for performance-critical situations.
726 *
727 * Move the iterator position to the character (code point) whose
728 * index precedes the current position, and return that character.
729 * This is a pre-decrement operation.
730 * Returns U_SENTINEL (-1) if the position is at the start of the text.
731 *
732 * @stable ICU 3.4
733 */
734#define UTEXT_PREVIOUS32(ut) \
735 ((ut)->chunkOffset > 0 && \
736 (ut)->chunkContents[(ut)->chunkOffset-1] < 0xd800 ? \
737 (ut)->chunkContents[--((ut)->chunkOffset)] : utext_previous32(ut))
738
739/**
740 * inline version of utext_getNativeIndex(), for performance-critical situations.
741 *
742 * Get the current iterator position, which can range from 0 to
743 * the length of the text.
744 * The position is a native index into the input text, in whatever format it
745 * may have (possibly UTF-8 for example), and may not always be the same as
746 * the corresponding UChar (UTF-16) index.
747 * The returned position will always be aligned to a code point boundary.
748 *
749 * @stable ICU 3.6
750 */
751#define UTEXT_GETNATIVEINDEX(ut) \
752 ((ut)->chunkOffset <= (ut)->nativeIndexingLimit? \
753 (ut)->chunkNativeStart+(ut)->chunkOffset : \
754 (ut)->pFuncs->mapOffsetToNative(ut))
755
756/**
757 * inline version of utext_setNativeIndex(), for performance-critical situations.
758 *
759 * Set the current iteration position to the nearest code point
760 * boundary at or preceding the specified index.
761 * The index is in the native units of the original input text.
762 * If the index is out of range, it will be pinned to be within
763 * the range of the input text.
764 *
765 * @stable ICU 3.8
766 */
767#define UTEXT_SETNATIVEINDEX(ut, ix) \
768 { int64_t __offset = (ix) - (ut)->chunkNativeStart; \
769 if (__offset>=0 && __offset<=(int64_t)(ut)->nativeIndexingLimit) { \
770 (ut)->chunkOffset=(int32_t)__offset; \
771 } else { \
772 utext_setNativeIndex((ut), (ix)); } }
773
774
775
776/************************************************************************************
777 *
778 * Functions related to writing or modifying the text.
779 * These will work only with modifiable UTexts. Attempting to
780 * modify a read-only UText will return an error status.
781 *
782 ************************************************************************************/
783
784
785/**
786 * Return TRUE if the text can be written (modified) with utext_replace() or
787 * utext_copy(). For the text to be writable, the text provider must
788 * be of a type that supports writing and the UText must not be frozen.
789 *
790 * Attempting to modify text when utext_isWriteable() is FALSE will fail -
791 * the text will not be modified, and an error will be returned from the function
792 * that attempted the modification.
793 *
794 * @param ut the UText to be tested.
795 * @return TRUE if the text is modifiable.
796 *
797 * @see utext_freeze()
798 * @see utext_replace()
799 * @see utext_copy()
800 * @stable ICU 3.4
801 *
802 */
803U_STABLE UBool U_EXPORT2
804utext_isWritable(const UText *ut);
805
806
807/**
808 * Test whether there is meta data associated with the text.
809 * @see Replaceable::hasMetaData()
810 *
811 * @param ut The UText to be tested
812 * @return TRUE if the underlying text includes meta data.
813 * @stable ICU 3.4
814 */
815U_STABLE UBool U_EXPORT2
816utext_hasMetaData(const UText *ut);
817
818
819/**
820 * Replace a range of the original text with a replacement text.
821 *
822 * Leaves the current iteration position at the position following the
823 * newly inserted replacement text.
824 *
825 * This function is only available on UText types that support writing,
826 * that is, ones where utext_isWritable() returns TRUE.
827 *
828 * When using this function, there should be only a single UText opened onto the
829 * underlying native text string. Behavior after a replace operation
830 * on a UText is undefined for any other additional UTexts that refer to the
831 * modified string.
832 *
833 * @param ut the UText representing the text to be operated on.
834 * @param nativeStart the native index of the start of the region to be replaced
835 * @param nativeLimit the native index of the character following the region to be replaced.
836 * @param replacementText pointer to the replacement text
837 * @param replacementLength length of the replacement text, or -1 if the text is NUL terminated.
838 * @param status receives any error status. Possible errors include
839 * U_NO_WRITE_PERMISSION
840 *
841 * @return The signed number of (native) storage units by which
842 * the length of the text expanded or contracted.
843 *
844 * @stable ICU 3.4
845 */
846U_STABLE int32_t U_EXPORT2
847utext_replace(UText *ut,
848 int64_t nativeStart, int64_t nativeLimit,
849 const UChar *replacementText, int32_t replacementLength,
850 UErrorCode *status);
851
852
853
854/**
855 *
856 * Copy or move a substring from one position to another within the text,
857 * while retaining any metadata associated with the text.
858 * This function is used to duplicate or reorder substrings.
859 * The destination index must not overlap the source range.
860 *
861 * The text to be copied or moved is inserted at destIndex;
862 * it does not replace or overwrite any existing text.
863 *
864 * The iteration position is left following the newly inserted text
865 * at the destination position.
866 *
867 * This function is only available on UText types that support writing,
868 * that is, ones where utext_isWritable() returns TRUE.
869 *
870 * When using this function, there should be only a single UText opened onto the
871 * underlying native text string. Behavior after a copy operation
872 * on a UText is undefined in any other additional UTexts that refer to the
873 * modified string.
874 *
875 * @param ut The UText representing the text to be operated on.
876 * @param nativeStart The native index of the start of the region to be copied or moved
877 * @param nativeLimit The native index of the character position following the region
878 * to be copied.
879 * @param destIndex The native destination index to which the source substring is
880 * copied or moved.
881 * @param move If TRUE, then the substring is moved, not copied/duplicated.
882 * @param status receives any error status. Possible errors include U_NO_WRITE_PERMISSION
883 *
884 * @stable ICU 3.4
885 */
886U_STABLE void U_EXPORT2
887utext_copy(UText *ut,
888 int64_t nativeStart, int64_t nativeLimit,
889 int64_t destIndex,
890 UBool move,
891 UErrorCode *status);
892
893
894/**
895 * <p>
896 * Freeze a UText. This prevents any modification to the underlying text itself
897 * by means of functions operating on this UText.
898 * </p>
899 * <p>
900 * Once frozen, a UText can not be unfrozen. The intent is to ensure
901 * that a the text underlying a frozen UText wrapper cannot be modified via that UText.
902 * </p>
903 * <p>
904 * Caution: freezing a UText will disable changes made via the specific
905 * frozen UText wrapper only; it will not have any effect on the ability to
906 * directly modify the text by bypassing the UText. Any such backdoor modifications
907 * are always an error while UText access is occuring because the underlying
908 * text can get out of sync with UText's buffering.
909 * </p>
910 *
911 * @param ut The UText to be frozen.
912 * @see utext_isWritable()
913 * @stable ICU 3.6
914 */
915U_STABLE void U_EXPORT2
916utext_freeze(UText *ut);
917
918
919/**
920 * UText provider properties (bit field indexes).
921 *
922 * @see UText
923 * @stable ICU 3.4
924 */
925enum {
926 /**
927 * It is potentially time consuming for the provider to determine the length of the text.
928 * @stable ICU 3.4
929 */
930 UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE = 1,
931 /**
932 * Text chunks remain valid and usable until the text object is modified or
933 * deleted, not just until the next time the access() function is called
934 * (which is the default).
935 * @stable ICU 3.4
936 */
937 UTEXT_PROVIDER_STABLE_CHUNKS = 2,
938 /**
939 * The provider supports modifying the text via the replace() and copy()
940 * functions.
941 * @see Replaceable
942 * @stable ICU 3.4
943 */
944 UTEXT_PROVIDER_WRITABLE = 3,
945 /**
946 * There is meta data associated with the text.
947 * @see Replaceable::hasMetaData()
948 * @stable ICU 3.4
949 */
950 UTEXT_PROVIDER_HAS_META_DATA = 4,
951 /**
952 * Text provider owns the text storage.
953 * Generally occurs as the result of a deep clone of the UText.
954 * When closing the UText, the associated text must
955 * also be closed/deleted/freed/ whatever is appropriate.
956 * @stable ICU 3.6
957 */
958 UTEXT_PROVIDER_OWNS_TEXT = 5
959};
960
961/**
962 * Function type declaration for UText.clone().
963 *
964 * clone a UText. Much like opening a UText where the source text is itself
965 * another UText.
966 *
967 * A deep clone will copy both the UText data structures and the underlying text.
968 * The original and cloned UText will operate completely independently; modifications
969 * made to the text in one will not effect the other. Text providers are not
970 * required to support deep clones. The user of clone() must check the status return
971 * and be prepared to handle failures.
972 *
973 * A shallow clone replicates only the UText data structures; it does not make
974 * a copy of the underlying text. Shallow clones can be used as an efficient way to
975 * have multiple iterators active in a single text string that is not being
976 * modified.
977 *
978 * A shallow clone operation must not fail except for truly exceptional conditions such
979 * as memory allocation failures.
980 *
981 * A UText and its clone may be safely concurrently accessed by separate threads.
982 * This is true for both shallow and deep clones.
983 * It is the responsibility of the Text Provider to ensure that this thread safety
984 * constraint is met.
985
986 *
987 * @param dest A UText struct to be filled in with the result of the clone operation,
988 * or NULL if the clone function should heap-allocate a new UText struct.
989 * @param src The UText to be cloned.
990 * @param deep TRUE to request a deep clone, FALSE for a shallow clone.
991 * @param status Errors are returned here. For deep clones, U_UNSUPPORTED_ERROR
992 * should be returned if the text provider is unable to clone the
993 * original text.
994 * @return The newly created clone, or NULL if the clone operation failed.
995 *
996 * @stable ICU 3.4
997 */
998typedef UText * U_CALLCONV
999UTextClone(UText *dest, const UText *src, UBool deep, UErrorCode *status);
1000
1001
1002/**
1003 * Function type declaration for UText.nativeLength().
1004 *
1005 * @param ut the UText to get the length of.
1006 * @return the length, in the native units of the original text string.
1007 * @see UText
1008 * @stable ICU 3.4
1009 */
1010typedef int64_t U_CALLCONV
1011UTextNativeLength(UText *ut);
1012
1013/**
1014 * Function type declaration for UText.access(). Get the description of the text chunk
1015 * containing the text at a requested native index. The UText's iteration
1016 * position will be left at the requested index. If the index is out
1017 * of bounds, the iteration position will be left at the start or end
1018 * of the string, as appropriate.
1019 *
1020 * Chunks must begin and end on code point boundaries. A single code point
1021 * comprised of multiple storage units must never span a chunk boundary.
1022 *
1023 *
1024 * @param ut the UText being accessed.
1025 * @param nativeIndex Requested index of the text to be accessed.
1026 * @param forward If TRUE, then the returned chunk must contain text
1027 * starting from the index, so that start<=index<limit.
1028 * If FALSE, then the returned chunk must contain text
1029 * before the index, so that start<index<=limit.
1030 * @return True if the requested index could be accessed. The chunk
1031 * will contain the requested text.
1032 * False value if a chunk cannot be accessed
1033 * (the requested index is out of bounds).
1034 *
1035 * @see UText
1036 * @stable ICU 3.4
1037 */
1038typedef UBool U_CALLCONV
1039UTextAccess(UText *ut, int64_t nativeIndex, UBool forward);
1040
1041/**
1042 * Function type declaration for UText.extract().
1043 *
1044 * Extract text from a UText into a UChar buffer. The range of text to be extracted
1045 * is specified in the native indices of the UText provider. These may not necessarily
1046 * be UTF-16 indices.
1047 * <p>
1048 * The size (number of 16 bit UChars) in the data to be extracted is returned. The
1049 * full amount is returned, even when the specified buffer size is smaller.
1050 * <p>
1051 * The extracted string will (if you are a user) / must (if you are a text provider)
1052 * be NUL-terminated if there is sufficient space in the destination buffer.
1053 *
1054 * @param ut the UText from which to extract data.
1055 * @param nativeStart the native index of the first characer to extract.
1056 * @param nativeLimit the native string index of the position following the last
1057 * character to extract.
1058 * @param dest the UChar (UTF-16) buffer into which the extracted text is placed
1059 * @param destCapacity The size, in UChars, of the destination buffer. May be zero
1060 * for precomputing the required size.
1061 * @param status receives any error status.
1062 * If U_BUFFER_OVERFLOW_ERROR: Returns number of UChars for
1063 * preflighting.
1064 * @return Number of UChars in the data. Does not include a trailing NUL.
1065 *
1066 * @stable ICU 3.4
1067 */
1068typedef int32_t U_CALLCONV
1069UTextExtract(UText *ut,
1070 int64_t nativeStart, int64_t nativeLimit,
1071 UChar *dest, int32_t destCapacity,
1072 UErrorCode *status);
1073
1074/**
1075 * Function type declaration for UText.replace().
1076 *
1077 * Replace a range of the original text with a replacement text.
1078 *
1079 * Leaves the current iteration position at the position following the
1080 * newly inserted replacement text.
1081 *
1082 * This function need only be implemented on UText types that support writing.
1083 *
1084 * When using this function, there should be only a single UText opened onto the
1085 * underlying native text string. The function is responsible for updating the
1086 * text chunk within the UText to reflect the updated iteration position,
1087 * taking into account any changes to the underlying string's structure caused
1088 * by the replace operation.
1089 *
1090 * @param ut the UText representing the text to be operated on.
1091 * @param nativeStart the index of the start of the region to be replaced
1092 * @param nativeLimit the index of the character following the region to be replaced.
1093 * @param replacementText pointer to the replacement text
1094 * @param replacmentLength length of the replacement text in UChars, or -1 if the text is NUL terminated.
1095 * @param status receives any error status. Possible errors include
1096 * U_NO_WRITE_PERMISSION
1097 *
1098 * @return The signed number of (native) storage units by which
1099 * the length of the text expanded or contracted.
1100 *
1101 * @stable ICU 3.4
1102 */
1103typedef int32_t U_CALLCONV
1104UTextReplace(UText *ut,
1105 int64_t nativeStart, int64_t nativeLimit,
1106 const UChar *replacementText, int32_t replacmentLength,
1107 UErrorCode *status);
1108
1109/**
1110 * Function type declaration for UText.copy().
1111 *
1112 * Copy or move a substring from one position to another within the text,
1113 * while retaining any metadata associated with the text.
1114 * This function is used to duplicate or reorder substrings.
1115 * The destination index must not overlap the source range.
1116 *
1117 * The text to be copied or moved is inserted at destIndex;
1118 * it does not replace or overwrite any existing text.
1119 *
1120 * This function need only be implemented for UText types that support writing.
1121 *
1122 * When using this function, there should be only a single UText opened onto the
1123 * underlying native text string. The function is responsible for updating the
1124 * text chunk within the UText to reflect the updated iteration position,
1125 * taking into account any changes to the underlying string's structure caused
1126 * by the replace operation.
1127 *
1128 * @param ut The UText representing the text to be operated on.
1129 * @param nativeStart The index of the start of the region to be copied or moved
1130 * @param nativeLimit The index of the character following the region to be replaced.
1131 * @param nativeDest The destination index to which the source substring is copied or moved.
1132 * @param move If TRUE, then the substring is moved, not copied/duplicated.
1133 * @param status receives any error status. Possible errors include U_NO_WRITE_PERMISSION
1134 *
1135 * @stable ICU 3.4
1136 */
1137typedef void U_CALLCONV
1138UTextCopy(UText *ut,
1139 int64_t nativeStart, int64_t nativeLimit,
1140 int64_t nativeDest,
1141 UBool move,
1142 UErrorCode *status);
1143
1144/**
1145 * Function type declaration for UText.mapOffsetToNative().
1146 * Map from the current UChar offset within the current text chunk to
1147 * the corresponding native index in the original source text.
1148 *
1149 * This is required only for text providers that do not use native UTF-16 indexes.
1150 *
1151 * @param ut the UText.
1152 * @return Absolute (native) index corresponding to chunkOffset in the current chunk.
1153 * The returned native index should always be to a code point boundary.
1154 *
1155 * @stable ICU 3.4
1156 */
1157typedef int64_t U_CALLCONV
1158UTextMapOffsetToNative(const UText *ut);
1159
1160/**
1161 * Function type declaration for UText.mapIndexToUTF16().
1162 * Map from a native index to a UChar offset within a text chunk.
1163 * Behavior is undefined if the native index does not fall within the
1164 * current chunk.
1165 *
1166 * This function is required only for text providers that do not use native UTF-16 indexes.
1167 *
1168 * @param ut The UText containing the text chunk.
1169 * @param nativeIndex Absolute (native) text index, chunk->start<=index<=chunk->limit.
1170 * @return Chunk-relative UTF-16 offset corresponding to the specified native
1171 * index.
1172 *
1173 * @stable ICU 3.4
1174 */
1175typedef int32_t U_CALLCONV
1176UTextMapNativeIndexToUTF16(const UText *ut, int64_t nativeIndex);
1177
1178
1179/**
1180 * Function type declaration for UText.utextClose().
1181 *
1182 * A Text Provider close function is only required for provider types that make
1183 * allocations in their open function (or other functions) that must be
1184 * cleaned when the UText is closed.
1185 *
1186 * The allocation of the UText struct itself and any "extra" storage
1187 * associated with the UText is handled by the common UText implementation
1188 * and does not require provider specific cleanup in a close function.
1189 *
1190 * Most UText provider implementations do not need to implement this function.
1191 *
1192 * @param ut A UText object to be closed.
1193 *
1194 * @stable ICU 3.4
1195 */
1196typedef void U_CALLCONV
1197UTextClose(UText *ut);
1198
1199
1200/**
1201 * (public) Function dispatch table for UText.
1202 * Conceptually very much like a C++ Virtual Function Table.
1203 * This struct defines the organization of the table.
1204 * Each text provider implementation must provide an
1205 * actual table that is initialized with the appropriate functions
1206 * for the type of text being handled.
1207 * @stable ICU 3.6
1208 */
1209struct UTextFuncs {
1210 /**
1211 * (public) Function table size, sizeof(UTextFuncs)
1212 * Intended for use should the table grow to accomodate added
1213 * functions in the future, to allow tests for older format
1214 * function tables that do not contain the extensions.
1215 *
1216 * Fields are placed for optimal alignment on
1217 * 32/64/128-bit-pointer machines, by normally grouping together
1218 * 4 32-bit fields,
1219 * 4 pointers,
1220 * 2 64-bit fields
1221 * in sequence.
1222 * @stable ICU 3.6
1223 */
1224 int32_t tableSize;
1225
1226 /**
1227 * (private) Alignment padding.
1228 * Do not use, reserved for use by the UText framework only.
1229 * @internal
1230 */
1231 int32_t reserved1, /** @internal */ reserved2, /** @internal */ reserved3;
1232
1233
1234 /**
1235 * (public) Function pointer for UTextClone
1236 *
1237 * @see UTextClone
1238 * @stable ICU 3.6
1239 */
1240 UTextClone *clone;
1241
1242 /**
1243 * (public) function pointer for UTextLength
1244 * May be expensive to compute!
1245 *
1246 * @see UTextLength
1247 * @stable ICU 3.6
1248 */
1249 UTextNativeLength *nativeLength;
1250
1251 /**
1252 * (public) Function pointer for UTextAccess.
1253 *
1254 * @see UTextAccess
1255 * @stable ICU 3.6
1256 */
1257 UTextAccess *access;
1258
1259 /**
1260 * (public) Function pointer for UTextExtract.
1261 *
1262 * @see UTextExtract
1263 * @stable ICU 3.6
1264 */
1265 UTextExtract *extract;
1266
1267 /**
1268 * (public) Function pointer for UTextReplace.
1269 *
1270 * @see UTextReplace
1271 * @stable ICU 3.6
1272 */
1273 UTextReplace *replace;
1274
1275 /**
1276 * (public) Function pointer for UTextCopy.
1277 *
1278 * @see UTextCopy
1279 * @stable ICU 3.6
1280 */
1281 UTextCopy *copy;
1282
1283 /**
1284 * (public) Function pointer for UTextMapOffsetToNative.
1285 *
1286 * @see UTextMapOffsetToNative
1287 * @stable ICU 3.6
1288 */
1289 UTextMapOffsetToNative *mapOffsetToNative;
1290
1291 /**
1292 * (public) Function pointer for UTextMapNativeIndexToUTF16.
1293 *
1294 * @see UTextMapNativeIndexToUTF16
1295 * @stable ICU 3.6
1296 */
1297 UTextMapNativeIndexToUTF16 *mapNativeIndexToUTF16;
1298
1299 /**
1300 * (public) Function pointer for UTextClose.
1301 *
1302 * @see UTextClose
1303 * @stable ICU 3.6
1304 */
1305 UTextClose *close;
1306
1307 /**
1308 * (private) Spare function pointer
1309 * @internal
1310 */
1311 UTextClose *spare1;
1312
1313 /**
1314 * (private) Spare function pointer
1315 * @internal
1316 */
1317 UTextClose *spare2;
1318
1319 /**
1320 * (private) Spare function pointer
1321 * @internal
1322 */
1323 UTextClose *spare3;
1324
1325};
1326/**
1327 * Function dispatch table for UText
1328 * @see UTextFuncs
1329 */
1330typedef struct UTextFuncs UTextFuncs;
1331
1332 /**
1333 * UText struct. Provides the interface between the generic UText access code
1334 * and the UText provider code that works on specific kinds of
1335 * text (UTF-8, noncontiguous UTF-16, whatever.)
1336 *
1337 * Applications that are using predefined types of text providers
1338 * to pass text data to ICU services will have no need to view the
1339 * internals of the UText structs that they open.
1340 *
1341 * @stable ICU 3.6
1342 */
1343struct UText {
1344 /**
1345 * (private) Magic. Used to help detect when UText functions are handed
1346 * invalid or unitialized UText structs.
1347 * utext_openXYZ() functions take an initialized,
1348 * but not necessarily open, UText struct as an
1349 * optional fill-in parameter. This magic field
1350 * is used to check for that initialization.
1351 * Text provider close functions must NOT clear
1352 * the magic field because that would prevent
1353 * reuse of the UText struct.
1354 * @internal
1355 */
1356 uint32_t magic;
1357
1358
1359 /**
1360 * (private) Flags for managing the allocation and freeing of
1361 * memory associated with this UText.
1362 * @internal
1363 */
1364 int32_t flags;
1365
1366
1367 /**
1368 * Text provider properties. This set of flags is maintainted by the
1369 * text provider implementation.
1370 * @stable ICU 3.4
1371 */
1372 int32_t providerProperties;
1373
1374 /**
1375 * (public) sizeOfStruct=sizeof(UText)
1376 * Allows possible backward compatible extension.
1377 *
1378 * @stable ICU 3.4
1379 */
1380 int32_t sizeOfStruct;
1381
1382 /* ------ 16 byte alignment boundary ----------- */
1383
1384
1385 /**
1386 * (protected) Native index of the first character position following
1387 * the current chunk.
1388 * @stable ICU 3.6
1389 */
1390 int64_t chunkNativeLimit;
1391
1392 /**
1393 * (protected) Size in bytes of the extra space (pExtra).
1394 * @stable ICU 3.4
1395 */
1396 int32_t extraSize;
1397
1398 /**
1399 * (protected) The highest chunk offset where native indexing and
1400 * chunk (UTF-16) indexing correspond. For UTF-16 sources, value
1401 * will be equal to chunkLength.
1402 *
1403 * @stable ICU 3.6
1404 */
1405 int32_t nativeIndexingLimit;
1406
1407 /* ---- 16 byte alignment boundary------ */
1408
1409 /**
1410 * (protected) Native index of the first character in the text chunk.
1411 * @stable ICU 3.6
1412 */
1413 int64_t chunkNativeStart;
1414
1415 /**
1416 * (protected) Current iteration position within the text chunk (UTF-16 buffer).
1417 * This is the index to the character that will be returned by utext_next32().
1418 * @stable ICU 3.6
1419 */
1420 int32_t chunkOffset;
1421
1422 /**
1423 * (protected) Length the text chunk (UTF-16 buffer), in UChars.
1424 * @stable ICU 3.6
1425 */
1426 int32_t chunkLength;
1427
1428 /* ---- 16 byte alignment boundary-- */
1429
1430
1431 /**
1432 * (protected) pointer to a chunk of text in UTF-16 format.
1433 * May refer either to original storage of the source of the text, or
1434 * if conversion was required, to a buffer owned by the UText.
1435 * @stable ICU 3.6
1436 */
1437 const UChar *chunkContents;
1438
1439 /**
1440 * (public) Pointer to Dispatch table for accessing functions for this UText.
1441 * @stable ICU 3.6
1442 */
1443 const UTextFuncs *pFuncs;
1444
1445 /**
1446 * (protected) Pointer to additional space requested by the
1447 * text provider during the utext_open operation.
1448 * @stable ICU 3.4
1449 */
1450 void *pExtra;
1451
1452 /**
1453 * (protected) Pointer to string or text-containin object or similar.
1454 * This is the source of the text that this UText is wrapping, in a format
1455 * that is known to the text provider functions.
1456 * @stable ICU 3.4
1457 */
1458 const void *context;
1459
1460 /* --- 16 byte alignment boundary--- */
1461
1462 /**
1463 * (protected) Pointer fields available for use by the text provider.
1464 * Not used by UText common code.
1465 * @stable ICU 3.6
1466 */
1467 const void *p;
1468 /**
1469 * (protected) Pointer fields available for use by the text provider.
1470 * Not used by UText common code.
1471 * @stable ICU 3.6
1472 */
1473 const void *q;
1474 /**
1475 * (protected) Pointer fields available for use by the text provider.
1476 * Not used by UText common code.
1477 * @stable ICU 3.6
1478 */
1479 const void *r;
1480
1481 /**
1482 * Private field reserved for future use by the UText framework
1483 * itself. This is not to be touched by the text providers.
1484 * @internal ICU 3.4
1485 */
1486 void *privP;
1487
1488
1489 /* --- 16 byte alignment boundary--- */
1490
1491
1492 /**
1493 * (protected) Integer field reserved for use by the text provider.
1494 * Not used by the UText framework, or by the client (user) of the UText.
1495 * @stable ICU 3.4
1496 */
1497 int64_t a;
1498
1499 /**
1500 * (protected) Integer field reserved for use by the text provider.
1501 * Not used by the UText framework, or by the client (user) of the UText.
1502 * @stable ICU 3.4
1503 */
1504 int32_t b;
1505
1506 /**
1507 * (protected) Integer field reserved for use by the text provider.
1508 * Not used by the UText framework, or by the client (user) of the UText.
1509 * @stable ICU 3.4
1510 */
1511 int32_t c;
1512
1513 /* ---- 16 byte alignment boundary---- */
1514
1515
1516 /**
1517 * Private field reserved for future use by the UText framework
1518 * itself. This is not to be touched by the text providers.
1519 * @internal ICU 3.4
1520 */
1521 int64_t privA;
1522 /**
1523 * Private field reserved for future use by the UText framework
1524 * itself. This is not to be touched by the text providers.
1525 * @internal ICU 3.4
1526 */
1527 int32_t privB;
1528 /**
1529 * Private field reserved for future use by the UText framework
1530 * itself. This is not to be touched by the text providers.
1531 * @internal ICU 3.4
1532 */
1533 int32_t privC;
1534};
1535
1536
1537/**
1538 * Common function for use by Text Provider implementations to allocate and/or initialize
1539 * a new UText struct. To be called in the implementation of utext_open() functions.
1540 * If the supplied UText parameter is null, a new UText struct will be allocated on the heap.
1541 * If the supplied UText is already open, the provider's close function will be called
1542 * so that the struct can be reused by the open that is in progress.
1543 *
1544 * @param ut pointer to a UText struct to be re-used, or null if a new UText
1545 * should be allocated.
1546 * @param extraSpace The amount of additional space to be allocated as part
1547 * of this UText, for use by types of providers that require
1548 * additional storage.
1549 * @param status Errors are returned here.
1550 * @return pointer to the UText, allocated if necessary, with extra space set up if requested.
1551 * @stable ICU 3.4
1552 */
1553U_STABLE UText * U_EXPORT2
1554utext_setup(UText *ut, int32_t extraSpace, UErrorCode *status);
1555
1556#ifndef U_HIDE_INTERNAL_API
1557/**
1558 * @internal
1559 * Value used to help identify correctly initialized UText structs.
1560 * Note: must be publicly visible so that UTEXT_INITIALIZER can access it.
1561 */
1562enum {
1563 UTEXT_MAGIC = 0x345ad82c
1564};
1565#endif /* U_HIDE_INTERNAL_API */
1566
1567/**
1568 * initializer to be used with local (stack) instances of a UText
1569 * struct. UText structs must be initialized before passing
1570 * them to one of the utext_open functions.
1571 *
1572 * @stable ICU 3.6
1573 */
1574#define UTEXT_INITIALIZER { \
1575 UTEXT_MAGIC, /* magic */ \
1576 0, /* flags */ \
1577 0, /* providerProps */ \
1578 sizeof(UText), /* sizeOfStruct */ \
1579 0, /* chunkNativeLimit */ \
1580 0, /* extraSize */ \
1581 0, /* nativeIndexingLimit */ \
1582 0, /* chunkNativeStart */ \
1583 0, /* chunkOffset */ \
1584 0, /* chunkLength */ \
1585 NULL, /* chunkContents */ \
1586 NULL, /* pFuncs */ \
1587 NULL, /* pExtra */ \
1588 NULL, /* context */ \
1589 NULL, NULL, NULL, /* p, q, r */ \
1590 NULL, /* privP */ \
1591 0, 0, 0, /* a, b, c */ \
1592 0, 0, 0 /* privA,B,C, */ \
1593 }
1594
1595
1596U_CDECL_END
1597
1598
1599
1600#endif
1601