- support of [Strings.LanguageID]-sections for inf-files added in setupapi
[reactos.git] / reactos / lib / 3rdparty / icu4ros / icu / source / common / unicode / ubidi.h
1 /*
2 ******************************************************************************
3 *
4 * Copyright (C) 1999-2007, International Business Machines
5 * Corporation and others. All Rights Reserved.
6 *
7 ******************************************************************************
8 * file name: ubidi.h
9 * encoding: US-ASCII
10 * tab size: 8 (not used)
11 * indentation:4
12 *
13 * created on: 1999jul27
14 * created by: Markus W. Scherer, updated by Matitiahu Allouche
15 */
16
17 #ifndef UBIDI_H
18 #define UBIDI_H
19
20 #include "unicode/utypes.h"
21 #include "unicode/uchar.h"
22
23 /**
24 *\file
25 * \brief C API: Bidi algorithm
26 *
27 * <h2>Bidi algorithm for ICU</h2>
28 *
29 * This is an implementation of the Unicode Bidirectional algorithm.
30 * The algorithm is defined in the
31 * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Annex #9</a>,
32 * version 13, also described in The Unicode Standard, Version 4.0 .<p>
33 *
34 * Note: Libraries that perform a bidirectional algorithm and
35 * reorder strings accordingly are sometimes called "Storage Layout Engines".
36 * ICU's Bidi and shaping (u_shapeArabic()) APIs can be used at the core of such
37 * "Storage Layout Engines".
38 *
39 * <h3>General remarks about the API:</h3>
40 *
41 * In functions with an error code parameter,
42 * the <code>pErrorCode</code> pointer must be valid
43 * and the value that it points to must not indicate a failure before
44 * the function call. Otherwise, the function returns immediately.
45 * After the function call, the value indicates success or failure.<p>
46 *
47 * The &quot;limit&quot; of a sequence of characters is the position just after their
48 * last character, i.e., one more than that position.<p>
49 *
50 * Some of the API functions provide access to &quot;runs&quot;.
51 * Such a &quot;run&quot; is defined as a sequence of characters
52 * that are at the same embedding level
53 * after performing the Bidi algorithm.<p>
54 *
55 * @author Markus W. Scherer
56 * @version 1.0
57 *
58 *
59 * <h4> Sample code for the ICU Bidi API </h4>
60 *
61 * <h5>Rendering a paragraph with the ICU Bidi API</h5>
62 *
63 * This is (hypothetical) sample code that illustrates
64 * how the ICU Bidi API could be used to render a paragraph of text.
65 * Rendering code depends highly on the graphics system,
66 * therefore this sample code must make a lot of assumptions,
67 * which may or may not match any existing graphics system's properties.
68 *
69 * <p>The basic assumptions are:</p>
70 * <ul>
71 * <li>Rendering is done from left to right on a horizontal line.</li>
72 * <li>A run of single-style, unidirectional text can be rendered at once.</li>
73 * <li>Such a run of text is passed to the graphics system with
74 * characters (code units) in logical order.</li>
75 * <li>The line-breaking algorithm is very complicated
76 * and Locale-dependent -
77 * and therefore its implementation omitted from this sample code.</li>
78 * </ul>
79 *
80 * <pre>
81 * \code
82 *#include "unicode/ubidi.h"
83 *
84 *typedef enum {
85 * styleNormal=0, styleSelected=1,
86 * styleBold=2, styleItalics=4,
87 * styleSuper=8, styleSub=16
88 *} Style;
89 *
90 *typedef struct { int32_t limit; Style style; } StyleRun;
91 *
92 *int getTextWidth(const UChar *text, int32_t start, int32_t limit,
93 * const StyleRun *styleRuns, int styleRunCount);
94 *
95 * // set *pLimit and *pStyleRunLimit for a line
96 * // from text[start] and from styleRuns[styleRunStart]
97 * // using ubidi_getLogicalRun(para, ...)
98 *void getLineBreak(const UChar *text, int32_t start, int32_t *pLimit,
99 * UBiDi *para,
100 * const StyleRun *styleRuns, int styleRunStart, int *pStyleRunLimit,
101 * int *pLineWidth);
102 *
103 * // render runs on a line sequentially, always from left to right
104 *
105 * // prepare rendering a new line
106 * void startLine(UBiDiDirection textDirection, int lineWidth);
107 *
108 * // render a run of text and advance to the right by the run width
109 * // the text[start..limit-1] is always in logical order
110 * void renderRun(const UChar *text, int32_t start, int32_t limit,
111 * UBiDiDirection textDirection, Style style);
112 *
113 * // We could compute a cross-product
114 * // from the style runs with the directional runs
115 * // and then reorder it.
116 * // Instead, here we iterate over each run type
117 * // and render the intersections -
118 * // with shortcuts in simple (and common) cases.
119 * // renderParagraph() is the main function.
120 *
121 * // render a directional run with
122 * // (possibly) multiple style runs intersecting with it
123 * void renderDirectionalRun(const UChar *text,
124 * int32_t start, int32_t limit,
125 * UBiDiDirection direction,
126 * const StyleRun *styleRuns, int styleRunCount) {
127 * int i;
128 *
129 * // iterate over style runs
130 * if(direction==UBIDI_LTR) {
131 * int styleLimit;
132 *
133 * for(i=0; i<styleRunCount; ++i) {
134 * styleLimit=styleRun[i].limit;
135 * if(start<styleLimit) {
136 * if(styleLimit>limit) { styleLimit=limit; }
137 * renderRun(text, start, styleLimit,
138 * direction, styleRun[i].style);
139 * if(styleLimit==limit) { break; }
140 * start=styleLimit;
141 * }
142 * }
143 * } else {
144 * int styleStart;
145 *
146 * for(i=styleRunCount-1; i>=0; --i) {
147 * if(i>0) {
148 * styleStart=styleRun[i-1].limit;
149 * } else {
150 * styleStart=0;
151 * }
152 * if(limit>=styleStart) {
153 * if(styleStart<start) { styleStart=start; }
154 * renderRun(text, styleStart, limit,
155 * direction, styleRun[i].style);
156 * if(styleStart==start) { break; }
157 * limit=styleStart;
158 * }
159 * }
160 * }
161 * }
162 *
163 * // the line object represents text[start..limit-1]
164 * void renderLine(UBiDi *line, const UChar *text,
165 * int32_t start, int32_t limit,
166 * const StyleRun *styleRuns, int styleRunCount) {
167 * UBiDiDirection direction=ubidi_getDirection(line);
168 * if(direction!=UBIDI_MIXED) {
169 * // unidirectional
170 * if(styleRunCount<=1) {
171 * renderRun(text, start, limit, direction, styleRuns[0].style);
172 * } else {
173 * renderDirectionalRun(text, start, limit,
174 * direction, styleRuns, styleRunCount);
175 * }
176 * } else {
177 * // mixed-directional
178 * int32_t count, i, length;
179 * UBiDiLevel level;
180 *
181 * count=ubidi_countRuns(para, pErrorCode);
182 * if(U_SUCCESS(*pErrorCode)) {
183 * if(styleRunCount<=1) {
184 * Style style=styleRuns[0].style;
185 *
186 * // iterate over directional runs
187 * for(i=0; i<count; ++i) {
188 * direction=ubidi_getVisualRun(para, i, &start, &length);
189 * renderRun(text, start, start+length, direction, style);
190 * }
191 * } else {
192 * int32_t j;
193 *
194 * // iterate over both directional and style runs
195 * for(i=0; i<count; ++i) {
196 * direction=ubidi_getVisualRun(line, i, &start, &length);
197 * renderDirectionalRun(text, start, start+length,
198 * direction, styleRuns, styleRunCount);
199 * }
200 * }
201 * }
202 * }
203 * }
204 *
205 *void renderParagraph(const UChar *text, int32_t length,
206 * UBiDiDirection textDirection,
207 * const StyleRun *styleRuns, int styleRunCount,
208 * int lineWidth,
209 * UErrorCode *pErrorCode) {
210 * UBiDi *para;
211 *
212 * if(pErrorCode==NULL || U_FAILURE(*pErrorCode) || length<=0) {
213 * return;
214 * }
215 *
216 * para=ubidi_openSized(length, 0, pErrorCode);
217 * if(para==NULL) { return; }
218 *
219 * ubidi_setPara(para, text, length,
220 * textDirection ? UBIDI_DEFAULT_RTL : UBIDI_DEFAULT_LTR,
221 * NULL, pErrorCode);
222 * if(U_SUCCESS(*pErrorCode)) {
223 * UBiDiLevel paraLevel=1&ubidi_getParaLevel(para);
224 * StyleRun styleRun={ length, styleNormal };
225 * int width;
226 *
227 * if(styleRuns==NULL || styleRunCount<=0) {
228 * styleRunCount=1;
229 * styleRuns=&styleRun;
230 * }
231 *
232 * // assume styleRuns[styleRunCount-1].limit>=length
233 *
234 * width=getTextWidth(text, 0, length, styleRuns, styleRunCount);
235 * if(width<=lineWidth) {
236 * // everything fits onto one line
237 *
238 * // prepare rendering a new line from either left or right
239 * startLine(paraLevel, width);
240 *
241 * renderLine(para, text, 0, length,
242 * styleRuns, styleRunCount);
243 * } else {
244 * UBiDi *line;
245 *
246 * // we need to render several lines
247 * line=ubidi_openSized(length, 0, pErrorCode);
248 * if(line!=NULL) {
249 * int32_t start=0, limit;
250 * int styleRunStart=0, styleRunLimit;
251 *
252 * for(;;) {
253 * limit=length;
254 * styleRunLimit=styleRunCount;
255 * getLineBreak(text, start, &limit, para,
256 * styleRuns, styleRunStart, &styleRunLimit,
257 * &width);
258 * ubidi_setLine(para, start, limit, line, pErrorCode);
259 * if(U_SUCCESS(*pErrorCode)) {
260 * // prepare rendering a new line
261 * // from either left or right
262 * startLine(paraLevel, width);
263 *
264 * renderLine(line, text, start, limit,
265 * styleRuns+styleRunStart,
266 * styleRunLimit-styleRunStart);
267 * }
268 * if(limit==length) { break; }
269 * start=limit;
270 * styleRunStart=styleRunLimit-1;
271 * if(start>=styleRuns[styleRunStart].limit) {
272 * ++styleRunStart;
273 * }
274 * }
275 *
276 * ubidi_close(line);
277 * }
278 * }
279 * }
280 *
281 * ubidi_close(para);
282 *}
283 *\endcode
284 * </pre>
285 */
286
287 /*DOCXX_TAG*/
288 /*@{*/
289
290 /**
291 * UBiDiLevel is the type of the level values in this
292 * Bidi implementation.
293 * It holds an embedding level and indicates the visual direction
294 * by its bit&nbsp;0 (even/odd value).<p>
295 *
296 * It can also hold non-level values for the
297 * <code>paraLevel</code> and <code>embeddingLevels</code>
298 * arguments of <code>ubidi_setPara()</code>; there:
299 * <ul>
300 * <li>bit&nbsp;7 of an <code>embeddingLevels[]</code>
301 * value indicates whether the using application is
302 * specifying the level of a character to <i>override</i> whatever the
303 * Bidi implementation would resolve it to.</li>
304 * <li><code>paraLevel</code> can be set to the
305 * pseudo-level values <code>UBIDI_DEFAULT_LTR</code>
306 * and <code>UBIDI_DEFAULT_RTL</code>.</li>
307 * </ul>
308 *
309 * @see ubidi_setPara
310 *
311 * <p>The related constants are not real, valid level values.
312 * <code>UBIDI_DEFAULT_XXX</code> can be used to specify
313 * a default for the paragraph level for
314 * when the <code>ubidi_setPara()</code> function
315 * shall determine it but there is no
316 * strongly typed character in the input.<p>
317 *
318 * Note that the value for <code>UBIDI_DEFAULT_LTR</code> is even
319 * and the one for <code>UBIDI_DEFAULT_RTL</code> is odd,
320 * just like with normal LTR and RTL level values -
321 * these special values are designed that way. Also, the implementation
322 * assumes that UBIDI_MAX_EXPLICIT_LEVEL is odd.
323 *
324 * @see UBIDI_DEFAULT_LTR
325 * @see UBIDI_DEFAULT_RTL
326 * @see UBIDI_LEVEL_OVERRIDE
327 * @see UBIDI_MAX_EXPLICIT_LEVEL
328 * @stable ICU 2.0
329 */
330 typedef uint8_t UBiDiLevel;
331
332 /** Paragraph level setting.<p>
333 *
334 * Constant indicating that the base direction depends on the first strong
335 * directional character in the text according to the Unicode Bidirectional
336 * Algorithm. If no strong directional character is present,
337 * then set the paragraph level to 0 (left-to-right).<p>
338 *
339 * If this value is used in conjunction with reordering modes
340 * <code>UBIDI_REORDER_INVERSE_LIKE_DIRECT</code> or
341 * <code>UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the text to reorder
342 * is assumed to be visual LTR, and the text after reordering is required
343 * to be the corresponding logical string with appropriate contextual
344 * direction. The direction of the result string will be RTL if either
345 * the righmost or leftmost strong character of the source text is RTL
346 * or Arabic Letter, the direction will be LTR otherwise.<p>
347 *
348 * If reordering option <code>UBIDI_OPTION_INSERT_MARKS</code> is set, an RLM may
349 * be added at the beginning of the result string to ensure round trip
350 * (that the result string, when reordered back to visual, will produce
351 * the original source text).
352 * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT
353 * @see UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL
354 * @stable ICU 2.0
355 */
356 #define UBIDI_DEFAULT_LTR 0xfe
357
358 /** Paragraph level setting.<p>
359 *
360 * Constant indicating that the base direction depends on the first strong
361 * directional character in the text according to the Unicode Bidirectional
362 * Algorithm. If no strong directional character is present,
363 * then set the paragraph level to 1 (right-to-left).<p>
364 *
365 * If this value is used in conjunction with reordering modes
366 * <code>UBIDI_REORDER_INVERSE_LIKE_DIRECT</code> or
367 * <code>UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the text to reorder
368 * is assumed to be visual LTR, and the text after reordering is required
369 * to be the corresponding logical string with appropriate contextual
370 * direction. The direction of the result string will be RTL if either
371 * the righmost or leftmost strong character of the source text is RTL
372 * or Arabic Letter, or if the text contains no strong character;
373 * the direction will be LTR otherwise.<p>
374 *
375 * If reordering option <code>UBIDI_OPTION_INSERT_MARKS</code> is set, an RLM may
376 * be added at the beginning of the result string to ensure round trip
377 * (that the result string, when reordered back to visual, will produce
378 * the original source text).
379 * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT
380 * @see UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL
381 * @stable ICU 2.0
382 */
383 #define UBIDI_DEFAULT_RTL 0xff
384
385 /**
386 * Maximum explicit embedding level.
387 * (The maximum resolved level can be up to <code>UBIDI_MAX_EXPLICIT_LEVEL+1</code>).
388 * @stable ICU 2.0
389 */
390 #define UBIDI_MAX_EXPLICIT_LEVEL 61
391
392 /** Bit flag for level input.
393 * Overrides directional properties.
394 * @stable ICU 2.0
395 */
396 #define UBIDI_LEVEL_OVERRIDE 0x80
397
398 /**
399 * Special value which can be returned by the mapping functions when a logical
400 * index has no corresponding visual index or vice-versa. This may happen
401 * for the logical-to-visual mapping of a Bidi control when option
402 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is specified. This can also happen
403 * for the visual-to-logical mapping of a Bidi mark (LRM or RLM) inserted
404 * by option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
405 * @see ubidi_getVisualIndex
406 * @see ubidi_getVisualMap
407 * @see ubidi_getLogicalIndex
408 * @see ubidi_getLogicalMap
409 * @stable ICU 3.6
410 */
411 #define UBIDI_MAP_NOWHERE (-1)
412
413 /**
414 * <code>UBiDiDirection</code> values indicate the text direction.
415 * @stable ICU 2.0
416 */
417 enum UBiDiDirection {
418 /** All left-to-right text. This is a 0 value. @stable ICU 2.0 */
419 UBIDI_LTR,
420 /** All right-to-left text. This is a 1 value. @stable ICU 2.0 */
421 UBIDI_RTL,
422 /** Mixed-directional text. @stable ICU 2.0 */
423 UBIDI_MIXED
424 };
425
426 /** @stable ICU 2.0 */
427 typedef enum UBiDiDirection UBiDiDirection;
428
429 /**
430 * Forward declaration of the <code>UBiDi</code> structure for the declaration of
431 * the API functions. Its fields are implementation-specific.<p>
432 * This structure holds information about a paragraph (or multiple paragraphs)
433 * of text with Bidi-algorithm-related details, or about one line of
434 * such a paragraph.<p>
435 * Reordering can be done on a line, or on one or more paragraphs which are
436 * then interpreted each as one single line.
437 * @stable ICU 2.0
438 */
439 struct UBiDi;
440
441 /** @stable ICU 2.0 */
442 typedef struct UBiDi UBiDi;
443
444 /**
445 * Allocate a <code>UBiDi</code> structure.
446 * Such an object is initially empty. It is assigned
447 * the Bidi properties of a piece of text containing one or more paragraphs
448 * by <code>ubidi_setPara()</code>
449 * or the Bidi properties of a line within a paragraph by
450 * <code>ubidi_setLine()</code>.<p>
451 * This object can be reused for as long as it is not deallocated
452 * by calling <code>ubidi_close()</code>.<p>
453 * <code>ubidi_setPara()</code> and <code>ubidi_setLine()</code> will allocate
454 * additional memory for internal structures as necessary.
455 *
456 * @return An empty <code>UBiDi</code> object.
457 * @stable ICU 2.0
458 */
459 U_STABLE UBiDi * U_EXPORT2
460 ubidi_open(void);
461
462 /**
463 * Allocate a <code>UBiDi</code> structure with preallocated memory
464 * for internal structures.
465 * This function provides a <code>UBiDi</code> object like <code>ubidi_open()</code>
466 * with no arguments, but it also preallocates memory for internal structures
467 * according to the sizings supplied by the caller.<p>
468 * Subsequent functions will not allocate any more memory, and are thus
469 * guaranteed not to fail because of lack of memory.<p>
470 * The preallocation can be limited to some of the internal memory
471 * by setting some values to 0 here. That means that if, e.g.,
472 * <code>maxRunCount</code> cannot be reasonably predetermined and should not
473 * be set to <code>maxLength</code> (the only failproof value) to avoid
474 * wasting memory, then <code>maxRunCount</code> could be set to 0 here
475 * and the internal structures that are associated with it will be allocated
476 * on demand, just like with <code>ubidi_open()</code>.
477 *
478 * @param maxLength is the maximum text or line length that internal memory
479 * will be preallocated for. An attempt to associate this object with a
480 * longer text will fail, unless this value is 0, which leaves the allocation
481 * up to the implementation.
482 *
483 * @param maxRunCount is the maximum anticipated number of same-level runs
484 * that internal memory will be preallocated for. An attempt to access
485 * visual runs on an object that was not preallocated for as many runs
486 * as the text was actually resolved to will fail,
487 * unless this value is 0, which leaves the allocation up to the implementation.<br><br>
488 * The number of runs depends on the actual text and maybe anywhere between
489 * 1 and <code>maxLength</code>. It is typically small.
490 *
491 * @param pErrorCode must be a valid pointer to an error code value.
492 *
493 * @return An empty <code>UBiDi</code> object with preallocated memory.
494 * @stable ICU 2.0
495 */
496 U_STABLE UBiDi * U_EXPORT2
497 ubidi_openSized(int32_t maxLength, int32_t maxRunCount, UErrorCode *pErrorCode);
498
499 /**
500 * <code>ubidi_close()</code> must be called to free the memory
501 * associated with a UBiDi object.<p>
502 *
503 * <strong>Important: </strong>
504 * A parent <code>UBiDi</code> object must not be destroyed or reused if
505 * it still has children.
506 * If a <code>UBiDi</code> object has become the <i>child</i>
507 * of another one (its <i>parent</i>) by calling
508 * <code>ubidi_setLine()</code>, then the child object must
509 * be destroyed (closed) or reused (by calling
510 * <code>ubidi_setPara()</code> or <code>ubidi_setLine()</code>)
511 * before the parent object.
512 *
513 * @param pBiDi is a <code>UBiDi</code> object.
514 *
515 * @see ubidi_setPara
516 * @see ubidi_setLine
517 * @stable ICU 2.0
518 */
519 U_STABLE void U_EXPORT2
520 ubidi_close(UBiDi *pBiDi);
521
522 /**
523 * Modify the operation of the Bidi algorithm such that it
524 * approximates an "inverse Bidi" algorithm. This function
525 * must be called before <code>ubidi_setPara()</code>.
526 *
527 * <p>The normal operation of the Bidi algorithm as described
528 * in the Unicode Technical Report is to take text stored in logical
529 * (keyboard, typing) order and to determine the reordering of it for visual
530 * rendering.
531 * Some legacy systems store text in visual order, and for operations
532 * with standard, Unicode-based algorithms, the text needs to be transformed
533 * to logical order. This is effectively the inverse algorithm of the
534 * described Bidi algorithm. Note that there is no standard algorithm for
535 * this "inverse Bidi" and that the current implementation provides only an
536 * approximation of "inverse Bidi".</p>
537 *
538 * <p>With <code>isInverse</code> set to <code>TRUE</code>,
539 * this function changes the behavior of some of the subsequent functions
540 * in a way that they can be used for the inverse Bidi algorithm.
541 * Specifically, runs of text with numeric characters will be treated in a
542 * special way and may need to be surrounded with LRM characters when they are
543 * written in reordered sequence.</p>
544 *
545 * <p>Output runs should be retrieved using <code>ubidi_getVisualRun()</code>.
546 * Since the actual input for "inverse Bidi" is visually ordered text and
547 * <code>ubidi_getVisualRun()</code> gets the reordered runs, these are actually
548 * the runs of the logically ordered output.</p>
549 *
550 * <p>Calling this function with argument <code>isInverse</code> set to
551 * <code>TRUE</code> is equivalent to calling
552 * <code>ubidi_setReorderingMode</code> with argument
553 * <code>reorderingMode</code>
554 * set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br>
555 * Calling this function with argument <code>isInverse</code> set to
556 * <code>FALSE</code> is equivalent to calling
557 * <code>ubidi_setReorderingMode</code> with argument
558 * <code>reorderingMode</code>
559 * set to <code>#UBIDI_REORDER_DEFAULT</code>.
560 *
561 * @param pBiDi is a <code>UBiDi</code> object.
562 *
563 * @param isInverse specifies "forward" or "inverse" Bidi operation.
564 *
565 * @see ubidi_setPara
566 * @see ubidi_writeReordered
567 * @see ubidi_setReorderingMode
568 * @stable ICU 2.0
569 */
570 U_STABLE void U_EXPORT2
571 ubidi_setInverse(UBiDi *pBiDi, UBool isInverse);
572
573 /**
574 * Is this Bidi object set to perform the inverse Bidi algorithm?
575 * <p>Note: calling this function after setting the reordering mode with
576 * <code>ubidi_setReorderingMode</code> will return <code>TRUE</code> if the
577 * reordering mode was set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>,
578 * <code>FALSE</code> for all other values.</p>
579 *
580 * @param pBiDi is a <code>UBiDi</code> object.
581 * @return TRUE if the Bidi object is set to perform the inverse Bidi algorithm
582 * by handling numbers as L.
583 *
584 * @see ubidi_setInverse
585 * @see ubidi_setReorderingMode
586 * @stable ICU 2.0
587 */
588
589 U_STABLE UBool U_EXPORT2
590 ubidi_isInverse(UBiDi *pBiDi);
591
592 /**
593 * Specify whether block separators must be allocated level zero,
594 * so that successive paragraphs will progress from left to right.
595 * This function must be called before <code>ubidi_setPara()</code>.
596 * Paragraph separators (B) may appear in the text. Setting them to level zero
597 * means that all paragraph separators (including one possibly appearing
598 * in the last text position) are kept in the reordered text after the text
599 * that they follow in the source text.
600 * When this feature is not enabled, a paragraph separator at the last
601 * position of the text before reordering will go to the first position
602 * of the reordered text when the paragraph level is odd.
603 *
604 * @param pBiDi is a <code>UBiDi</code> object.
605 *
606 * @param orderParagraphsLTR specifies whether paragraph separators (B) must
607 * receive level 0, so that successive paragraphs progress from left to right.
608 *
609 * @see ubidi_setPara
610 * @stable ICU 3.4
611 */
612 U_STABLE void U_EXPORT2
613 ubidi_orderParagraphsLTR(UBiDi *pBiDi, UBool orderParagraphsLTR);
614
615 /**
616 * Is this Bidi object set to allocate level 0 to block separators so that
617 * successive paragraphs progress from left to right?
618 *
619 * @param pBiDi is a <code>UBiDi</code> object.
620 * @return TRUE if the Bidi object is set to allocate level 0 to block
621 * separators.
622 *
623 * @see ubidi_orderParagraphsLTR
624 * @stable ICU 3.4
625 */
626 U_STABLE UBool U_EXPORT2
627 ubidi_isOrderParagraphsLTR(UBiDi *pBiDi);
628
629 /**
630 * <code>UBiDiReorderingMode</code> values indicate which variant of the Bidi
631 * algorithm to use.
632 *
633 * @see ubidi_setReorderingMode
634 * @stable ICU 3.6
635 */
636 typedef enum UBiDiReorderingMode {
637 /** Regular Logical to Visual Bidi algorithm according to Unicode.
638 * This is a 0 value.
639 * @stable ICU 3.6 */
640 UBIDI_REORDER_DEFAULT = 0,
641 /** Logical to Visual algorithm which handles numbers in a way which
642 * mimicks the behavior of Windows XP.
643 * @stable ICU 3.6 */
644 UBIDI_REORDER_NUMBERS_SPECIAL,
645 /** Logical to Visual algorithm grouping numbers with adjacent R characters
646 * (reversible algorithm).
647 * @stable ICU 3.6 */
648 UBIDI_REORDER_GROUP_NUMBERS_WITH_R,
649 /** Reorder runs only to transform a Logical LTR string to the Logical RTL
650 * string with the same display, or vice-versa.<br>
651 * If this mode is set together with option
652 * <code>#UBIDI_OPTION_INSERT_MARKS</code>, some Bidi controls in the source
653 * text may be removed and other controls may be added to produce the
654 * minimum combination which has the required display.
655 * @stable ICU 3.6 */
656 UBIDI_REORDER_RUNS_ONLY,
657 /** Visual to Logical algorithm which handles numbers like L
658 * (same algorithm as selected by <code>ubidi_setInverse(TRUE)</code>.
659 * @see ubidi_setInverse
660 * @stable ICU 3.6 */
661 UBIDI_REORDER_INVERSE_NUMBERS_AS_L,
662 /** Visual to Logical algorithm equivalent to the regular Logical to Visual
663 * algorithm.
664 * @stable ICU 3.6 */
665 UBIDI_REORDER_INVERSE_LIKE_DIRECT,
666 /** Inverse Bidi (Visual to Logical) algorithm for the
667 * <code>UBIDI_REORDER_NUMBERS_SPECIAL</code> Bidi algorithm.
668 * @stable ICU 3.6 */
669 UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL,
670 /** Number of values for reordering mode.
671 * @stable ICU 3.6 */
672 UBIDI_REORDER_COUNT
673 } UBiDiReorderingMode;
674
675 /**
676 * Modify the operation of the Bidi algorithm such that it implements some
677 * variant to the basic Bidi algorithm or approximates an "inverse Bidi"
678 * algorithm, depending on different values of the "reordering mode".
679 * This function must be called before <code>ubidi_setPara()</code>, and stays
680 * in effect until called again with a different argument.
681 *
682 * <p>The normal operation of the Bidi algorithm as described
683 * in the Unicode Standard Annex #9 is to take text stored in logical
684 * (keyboard, typing) order and to determine how to reorder it for visual
685 * rendering.</p>
686 *
687 * <p>With the reordering mode set to a value other than
688 * <code>#UBIDI_REORDER_DEFAULT</code>, this function changes the behavior of
689 * some of the subsequent functions in a way such that they implement an
690 * inverse Bidi algorithm or some other algorithm variants.</p>
691 *
692 * <p>Some legacy systems store text in visual order, and for operations
693 * with standard, Unicode-based algorithms, the text needs to be transformed
694 * into logical order. This is effectively the inverse algorithm of the
695 * described Bidi algorithm. Note that there is no standard algorithm for
696 * this "inverse Bidi", so a number of variants are implemented here.</p>
697 *
698 * <p>In other cases, it may be desirable to emulate some variant of the
699 * Logical to Visual algorithm (e.g. one used in MS Windows), or perform a
700 * Logical to Logical transformation.</p>
701 *
702 * <ul>
703 * <li>When the reordering mode is set to <code>#UBIDI_REORDER_DEFAULT</code>,
704 * the standard Bidi Logical to Visual algorithm is applied.</li>
705 *
706 * <li>When the reordering mode is set to
707 * <code>#UBIDI_REORDER_NUMBERS_SPECIAL</code>,
708 * the algorithm used to perform Bidi transformations when calling
709 * <code>ubidi_setPara</code> should approximate the algorithm used in
710 * Microsoft Windows XP rather than strictly conform to the Unicode Bidi
711 * algorithm.
712 * <br>
713 * The differences between the basic algorithm and the algorithm addressed
714 * by this option are as follows:
715 * <ul>
716 * <li>Within text at an even embedding level, the sequence "123AB"
717 * (where AB represent R or AL letters) is transformed to "123BA" by the
718 * Unicode algorithm and to "BA123" by the Windows algorithm.</li>
719 * <li>Arabic-Indic numbers (AN) are handled by the Windows algorithm just
720 * like regular numbers (EN).</li>
721 * </ul></li>
722 *
723 * <li>When the reordering mode is set to
724 * <code>#UBIDI_REORDER_GROUP_NUMBERS_WITH_R</code>,
725 * numbers located between LTR text and RTL text are associated with the RTL
726 * text. For instance, an LTR paragraph with content "abc 123 DEF" (where
727 * upper case letters represent RTL characters) will be transformed to
728 * "abc FED 123" (and not "abc 123 FED"), "DEF 123 abc" will be transformed
729 * to "123 FED abc" and "123 FED abc" will be transformed to "DEF 123 abc".
730 * This makes the algorithm reversible and makes it useful when round trip
731 * (from visual to logical and back to visual) must be achieved without
732 * adding LRM characters. However, this is a variation from the standard
733 * Unicode Bidi algorithm.<br>
734 * The source text should not contain Bidi control characters other than LRM
735 * or RLM.</li>
736 *
737 * <li>When the reordering mode is set to
738 * <code>#UBIDI_REORDER_RUNS_ONLY</code>,
739 * a "Logical to Logical" transformation must be performed:
740 * <ul>
741 * <li>If the default text level of the source text (argument <code>paraLevel</code>
742 * in <code>ubidi_setPara</code>) is even, the source text will be handled as
743 * LTR logical text and will be transformed to the RTL logical text which has
744 * the same LTR visual display.</li>
745 * <li>If the default level of the source text is odd, the source text
746 * will be handled as RTL logical text and will be transformed to the
747 * LTR logical text which has the same LTR visual display.</li>
748 * </ul>
749 * This mode may be needed when logical text which is basically Arabic or
750 * Hebrew, with possible included numbers or phrases in English, has to be
751 * displayed as if it had an even embedding level (this can happen if the
752 * displaying application treats all text as if it was basically LTR).
753 * <br>
754 * This mode may also be needed in the reverse case, when logical text which is
755 * basically English, with possible included phrases in Arabic or Hebrew, has to
756 * be displayed as if it had an odd embedding level.
757 * <br>
758 * Both cases could be handled by adding LRE or RLE at the head of the text,
759 * if the display subsystem supports these formatting controls. If it does not,
760 * the problem may be handled by transforming the source text in this mode
761 * before displaying it, so that it will be displayed properly.<br>
762 * The source text should not contain Bidi control characters other than LRM
763 * or RLM.</li>
764 *
765 * <li>When the reordering mode is set to
766 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>, an "inverse Bidi" algorithm
767 * is applied.
768 * Runs of text with numeric characters will be treated like LTR letters and
769 * may need to be surrounded with LRM characters when they are written in
770 * reordered sequence (the option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> can
771 * be used with function <code>ubidi_writeReordered</code> to this end. This
772 * mode is equivalent to calling <code>ubidi_setInverse()</code> with
773 * argument <code>isInverse</code> set to <code>TRUE</code>.</li>
774 *
775 * <li>When the reordering mode is set to
776 * <code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code>, the "direct" Logical to Visual
777 * Bidi algorithm is used as an approximation of an "inverse Bidi" algorithm.
778 * This mode is similar to mode <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>
779 * but is closer to the regular Bidi algorithm.
780 * <br>
781 * For example, an LTR paragraph with the content "FED 123 456 CBA" (where
782 * upper case represents RTL characters) will be transformed to
783 * "ABC 456 123 DEF", as opposed to "DEF 123 456 ABC"
784 * with mode <code>UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br>
785 * When used in conjunction with option
786 * <code>#UBIDI_OPTION_INSERT_MARKS</code>, this mode generally
787 * adds Bidi marks to the output significantly more sparingly than mode
788 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> with option
789 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to
790 * <code>ubidi_writeReordered</code>.</li>
791 *
792 * <li>When the reordering mode is set to
793 * <code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the Logical to Visual
794 * Bidi algorithm used in Windows XP is used as an approximation of an
795 * "inverse Bidi" algorithm.
796 * <br>
797 * For example, an LTR paragraph with the content "abc FED123" (where
798 * upper case represents RTL characters) will be transformed to
799 * "abc 123DEF.</li>
800 * </ul>
801 *
802 * <p>In all the reordering modes specifying an "inverse Bidi" algorithm
803 * (i.e. those with a name starting with <code>UBIDI_REORDER_INVERSE</code>),
804 * output runs should be retrieved using
805 * <code>ubidi_getVisualRun()</code>, and the output text with
806 * <code>ubidi_writeReordered()</code>. The caller should keep in mind that in
807 * "inverse Bidi" modes the input is actually visually ordered text and
808 * reordered output returned by <code>ubidi_getVisualRun()</code> or
809 * <code>ubidi_writeReordered()</code> are actually runs or character string
810 * of logically ordered output.<br>
811 * For all the "inverse Bidi" modes, the source text should not contain
812 * Bidi control characters other than LRM or RLM.</p>
813 *
814 * <p>Note that option <code>#UBIDI_OUTPUT_REVERSE</code> of
815 * <code>ubidi_writeReordered</code> has no useful meaning and should not be
816 * used in conjunction with any value of the reordering mode specifying
817 * "inverse Bidi" or with value <code>UBIDI_REORDER_RUNS_ONLY</code>.
818 *
819 * @param pBiDi is a <code>UBiDi</code> object.
820 * @param reorderingMode specifies the required variant of the Bidi algorithm.
821 *
822 * @see UBiDiReorderingMode
823 * @see ubidi_setInverse
824 * @see ubidi_setPara
825 * @see ubidi_writeReordered
826 * @stable ICU 3.6
827 */
828 U_STABLE void U_EXPORT2
829 ubidi_setReorderingMode(UBiDi *pBiDi, UBiDiReorderingMode reorderingMode);
830
831 /**
832 * What is the requested reordering mode for a given Bidi object?
833 *
834 * @param pBiDi is a <code>UBiDi</code> object.
835 * @return the current reordering mode of the Bidi object
836 * @see ubidi_setReorderingMode
837 * @stable ICU 3.6
838 */
839 U_STABLE UBiDiReorderingMode U_EXPORT2
840 ubidi_getReorderingMode(UBiDi *pBiDi);
841
842 /**
843 * <code>UBiDiReorderingOption</code> values indicate which options are
844 * specified to affect the Bidi algorithm.
845 *
846 * @see ubidi_setReorderingOptions
847 * @stable ICU 3.6
848 */
849 typedef enum UBiDiReorderingOption {
850 /**
851 * option value for <code>ubidi_setReorderingOptions</code>:
852 * disable all the options which can be set with this function
853 * @see ubidi_setReorderingOptions
854 * @stable ICU 3.6
855 */
856 UBIDI_OPTION_DEFAULT = 0,
857
858 /**
859 * option bit for <code>ubidi_setReorderingOptions</code>:
860 * insert Bidi marks (LRM or RLM) when needed to ensure correct result of
861 * a reordering to a Logical order
862 *
863 * <p>This option must be set or reset before calling
864 * <code>ubidi_setPara</code>.</p>
865 *
866 * <p>This option is significant only with reordering modes which generate
867 * a result with Logical order, specifically:</p>
868 * <ul>
869 * <li><code>#UBIDI_REORDER_RUNS_ONLY</code></li>
870 * <li><code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code></li>
871 * <li><code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code></li>
872 * <li><code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code></li>
873 * </ul>
874 *
875 * <p>If this option is set in conjunction with reordering mode
876 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> or with calling
877 * <code>ubidi_setInverse(TRUE)</code>, it implies
878 * option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>
879 * in calls to function <code>ubidi_writeReordered()</code>.</p>
880 *
881 * <p>For other reordering modes, a minimum number of LRM or RLM characters
882 * will be added to the source text after reordering it so as to ensure
883 * round trip, i.e. when applying the inverse reordering mode on the
884 * resulting logical text with removal of Bidi marks
885 * (option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> set before calling
886 * <code>ubidi_setPara()</code> or option <code>#UBIDI_REMOVE_BIDI_CONTROLS</code>
887 * in <code>ubidi_writeReordered</code>), the result will be identical to the
888 * source text in the first transformation.
889 *
890 * <p>This option will be ignored if specified together with option
891 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>. It inhibits option
892 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to function
893 * <code>ubidi_writeReordered()</code> and it implies option
894 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to function
895 * <code>ubidi_writeReordered()</code> if the reordering mode is
896 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.</p>
897 *
898 * @see ubidi_setReorderingMode
899 * @see ubidi_setReorderingOptions
900 * @stable ICU 3.6
901 */
902 UBIDI_OPTION_INSERT_MARKS = 1,
903
904 /**
905 * option bit for <code>ubidi_setReorderingOptions</code>:
906 * remove Bidi control characters
907 *
908 * <p>This option must be set or reset before calling
909 * <code>ubidi_setPara</code>.</p>
910 *
911 * <p>This option nullifies option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
912 * It inhibits option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls
913 * to function <code>ubidi_writeReordered()</code> and it implies option
914 * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to that function.</p>
915 *
916 * @see ubidi_setReorderingMode
917 * @see ubidi_setReorderingOptions
918 * @stable ICU 3.6
919 */
920 UBIDI_OPTION_REMOVE_CONTROLS = 2,
921
922 /**
923 * option bit for <code>ubidi_setReorderingOptions</code>:
924 * process the output as part of a stream to be continued
925 *
926 * <p>This option must be set or reset before calling
927 * <code>ubidi_setPara</code>.</p>
928 *
929 * <p>This option specifies that the caller is interested in processing large
930 * text object in parts.
931 * The results of the successive calls are expected to be concatenated by the
932 * caller. Only the call for the last part will have this option bit off.</p>
933 *
934 * <p>When this option bit is on, <code>ubidi_setPara()</code> may process
935 * less than the full source text in order to truncate the text at a meaningful
936 * boundary. The caller should call <code>ubidi_getProcessedLength()</code>
937 * immediately after calling <code>ubidi_setPara()</code> in order to
938 * determine how much of the source text has been processed.
939 * Source text beyond that length should be resubmitted in following calls to
940 * <code>ubidi_setPara</code>. The processed length may be less than
941 * the length of the source text if a character preceding the last character of
942 * the source text constitutes a reasonable boundary (like a block separator)
943 * for text to be continued.<br>
944 * If the last character of the source text constitutes a reasonable
945 * boundary, the whole text will be processed at once.<br>
946 * If nowhere in the source text there exists
947 * such a reasonable boundary, the processed length will be zero.<br>
948 * The caller should check for such an occurrence and do one of the following:
949 * <ul><li>submit a larger amount of text with a better chance to include
950 * a reasonable boundary.</li>
951 * <li>resubmit the same text after turning off option
952 * <code>UBIDI_OPTION_STREAMING</code>.</li></ul>
953 * In all cases, this option should be turned off before processing the last
954 * part of the text.</p>
955 *
956 * <p>When the <code>UBIDI_OPTION_STREAMING</code> option is used,
957 * it is recommended to call <code>ubidi_orderParagraphsLTR()</code> with
958 * argument <code>orderParagraphsLTR</code> set to <code>TRUE</code> before
959 * calling <code>ubidi_setPara</code> so that later paragraphs may be
960 * concatenated to previous paragraphs on the right.</p>
961 *
962 * @see ubidi_setReorderingMode
963 * @see ubidi_setReorderingOptions
964 * @see ubidi_getProcessedLength
965 * @see ubidi_orderParagraphsLTR
966 * @stable ICU 3.6
967 */
968 UBIDI_OPTION_STREAMING = 4
969 } UBiDiReorderingOption;
970
971 /**
972 * Specify which of the reordering options
973 * should be applied during Bidi transformations.
974 *
975 * @param pBiDi is a <code>UBiDi</code> object.
976 * @param reorderingOptions is a combination of zero or more of the following
977 * options:
978 * <code>#UBIDI_OPTION_DEFAULT</code>, <code>#UBIDI_OPTION_INSERT_MARKS</code>,
979 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>, <code>#UBIDI_OPTION_STREAMING</code>.
980 *
981 * @see ubidi_getReorderingOptions
982 * @stable ICU 3.6
983 */
984 U_STABLE void U_EXPORT2
985 ubidi_setReorderingOptions(UBiDi *pBiDi, uint32_t reorderingOptions);
986
987 /**
988 * What are the reordering options applied to a given Bidi object?
989 *
990 * @param pBiDi is a <code>UBiDi</code> object.
991 * @return the current reordering options of the Bidi object
992 * @see ubidi_setReorderingOptions
993 * @stable ICU 3.6
994 */
995 U_STABLE uint32_t U_EXPORT2
996 ubidi_getReorderingOptions(UBiDi *pBiDi);
997
998 /**
999 * Perform the Unicode Bidi algorithm. It is defined in the
1000 * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Anned #9</a>,
1001 * version 13,
1002 * also described in The Unicode Standard, Version 4.0 .<p>
1003 *
1004 * This function takes a piece of plain text containing one or more paragraphs,
1005 * with or without externally specified embedding levels from <i>styled</i>
1006 * text and computes the left-right-directionality of each character.<p>
1007 *
1008 * If the entire text is all of the same directionality, then
1009 * the function may not perform all the steps described by the algorithm,
1010 * i.e., some levels may not be the same as if all steps were performed.
1011 * This is not relevant for unidirectional text.<br>
1012 * For example, in pure LTR text with numbers the numbers would get
1013 * a resolved level of 2 higher than the surrounding text according to
1014 * the algorithm. This implementation may set all resolved levels to
1015 * the same value in such a case.<p>
1016 *
1017 * The text can be composed of multiple paragraphs. Occurrence of a block
1018 * separator in the text terminates a paragraph, and whatever comes next starts
1019 * a new paragraph. The exception to this rule is when a Carriage Return (CR)
1020 * is followed by a Line Feed (LF). Both CR and LF are block separators, but
1021 * in that case, the pair of characters is considered as terminating the
1022 * preceding paragraph, and a new paragraph will be started by a character
1023 * coming after the LF.
1024 *
1025 * @param pBiDi A <code>UBiDi</code> object allocated with <code>ubidi_open()</code>
1026 * which will be set to contain the reordering information,
1027 * especially the resolved levels for all the characters in <code>text</code>.
1028 *
1029 * @param text is a pointer to the text that the Bidi algorithm will be performed on.
1030 * This pointer is stored in the UBiDi object and can be retrieved
1031 * with <code>ubidi_getText()</code>.<br>
1032 * <strong>Note:</strong> the text must be (at least) <code>length</code> long.
1033 *
1034 * @param length is the length of the text; if <code>length==-1</code> then
1035 * the text must be zero-terminated.
1036 *
1037 * @param paraLevel specifies the default level for the text;
1038 * it is typically 0 (LTR) or 1 (RTL).
1039 * If the function shall determine the paragraph level from the text,
1040 * then <code>paraLevel</code> can be set to
1041 * either <code>#UBIDI_DEFAULT_LTR</code>
1042 * or <code>#UBIDI_DEFAULT_RTL</code>; if the text contains multiple
1043 * paragraphs, the paragraph level shall be determined separately for
1044 * each paragraph; if a paragraph does not include any strongly typed
1045 * character, then the desired default is used (0 for LTR or 1 for RTL).
1046 * Any other value between 0 and <code>#UBIDI_MAX_EXPLICIT_LEVEL</code>
1047 * is also valid, with odd levels indicating RTL.
1048 *
1049 * @param embeddingLevels (in) may be used to preset the embedding and override levels,
1050 * ignoring characters like LRE and PDF in the text.
1051 * A level overrides the directional property of its corresponding
1052 * (same index) character if the level has the
1053 * <code>#UBIDI_LEVEL_OVERRIDE</code> bit set.<br><br>
1054 * Except for that bit, it must be
1055 * <code>paraLevel<=embeddingLevels[]<=UBIDI_MAX_EXPLICIT_LEVEL</code>,
1056 * with one exception: a level of zero may be specified for a paragraph
1057 * separator even if <code>paraLevel>0</code> when multiple paragraphs
1058 * are submitted in the same call to <code>ubidi_setPara()</code>.<br><br>
1059 * <strong>Caution: </strong>A copy of this pointer, not of the levels,
1060 * will be stored in the <code>UBiDi</code> object;
1061 * the <code>embeddingLevels</code> array must not be
1062 * deallocated before the <code>UBiDi</code> structure is destroyed or reused,
1063 * and the <code>embeddingLevels</code>
1064 * should not be modified to avoid unexpected results on subsequent Bidi operations.
1065 * However, the <code>ubidi_setPara()</code> and
1066 * <code>ubidi_setLine()</code> functions may modify some or all of the levels.<br><br>
1067 * After the <code>UBiDi</code> object is reused or destroyed, the caller
1068 * must take care of the deallocation of the <code>embeddingLevels</code> array.<br><br>
1069 * <strong>Note:</strong> the <code>embeddingLevels</code> array must be
1070 * at least <code>length</code> long.
1071 *
1072 * @param pErrorCode must be a valid pointer to an error code value.
1073 * @stable ICU 2.0
1074 */
1075 U_STABLE void U_EXPORT2
1076 ubidi_setPara(UBiDi *pBiDi, const UChar *text, int32_t length,
1077 UBiDiLevel paraLevel, UBiDiLevel *embeddingLevels,
1078 UErrorCode *pErrorCode);
1079
1080 /**
1081 * <code>ubidi_setLine()</code> sets a <code>UBiDi</code> to
1082 * contain the reordering information, especially the resolved levels,
1083 * for all the characters in a line of text. This line of text is
1084 * specified by referring to a <code>UBiDi</code> object representing
1085 * this information for a piece of text containing one or more paragraphs,
1086 * and by specifying a range of indexes in this text.<p>
1087 * In the new line object, the indexes will range from 0 to <code>limit-start-1</code>.<p>
1088 *
1089 * This is used after calling <code>ubidi_setPara()</code>
1090 * for a piece of text, and after line-breaking on that text.
1091 * It is not necessary if each paragraph is treated as a single line.<p>
1092 *
1093 * After line-breaking, rules (L1) and (L2) for the treatment of
1094 * trailing WS and for reordering are performed on
1095 * a <code>UBiDi</code> object that represents a line.<p>
1096 *
1097 * <strong>Important: </strong><code>pLineBiDi</code> shares data with
1098 * <code>pParaBiDi</code>.
1099 * You must destroy or reuse <code>pLineBiDi</code> before <code>pParaBiDi</code>.
1100 * In other words, you must destroy or reuse the <code>UBiDi</code> object for a line
1101 * before the object for its parent paragraph.<p>
1102 *
1103 * The text pointer that was stored in <code>pParaBiDi</code> is also copied,
1104 * and <code>start</code> is added to it so that it points to the beginning of the
1105 * line for this object.
1106 *
1107 * @param pParaBiDi is the parent paragraph object. It must have been set
1108 * by a successful call to ubidi_setPara.
1109 *
1110 * @param start is the line's first index into the text.
1111 *
1112 * @param limit is just behind the line's last index into the text
1113 * (its last index +1).<br>
1114 * It must be <code>0<=start<limit<=</code>containing paragraph limit.
1115 * If the specified line crosses a paragraph boundary, the function
1116 * will terminate with error code U_ILLEGAL_ARGUMENT_ERROR.
1117 *
1118 * @param pLineBiDi is the object that will now represent a line of the text.
1119 *
1120 * @param pErrorCode must be a valid pointer to an error code value.
1121 *
1122 * @see ubidi_setPara
1123 * @see ubidi_getProcessedLength
1124 * @stable ICU 2.0
1125 */
1126 U_STABLE void U_EXPORT2
1127 ubidi_setLine(const UBiDi *pParaBiDi,
1128 int32_t start, int32_t limit,
1129 UBiDi *pLineBiDi,
1130 UErrorCode *pErrorCode);
1131
1132 /**
1133 * Get the directionality of the text.
1134 *
1135 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1136 *
1137 * @return a value of <code>UBIDI_LTR</code>, <code>UBIDI_RTL</code>
1138 * or <code>UBIDI_MIXED</code>
1139 * that indicates if the entire text
1140 * represented by this object is unidirectional,
1141 * and which direction, or if it is mixed-directional.
1142 *
1143 * @see UBiDiDirection
1144 * @stable ICU 2.0
1145 */
1146 U_STABLE UBiDiDirection U_EXPORT2
1147 ubidi_getDirection(const UBiDi *pBiDi);
1148
1149 /**
1150 * Get the pointer to the text.
1151 *
1152 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1153 *
1154 * @return The pointer to the text that the UBiDi object was created for.
1155 *
1156 * @see ubidi_setPara
1157 * @see ubidi_setLine
1158 * @stable ICU 2.0
1159 */
1160 U_STABLE const UChar * U_EXPORT2
1161 ubidi_getText(const UBiDi *pBiDi);
1162
1163 /**
1164 * Get the length of the text.
1165 *
1166 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1167 *
1168 * @return The length of the text that the UBiDi object was created for.
1169 * @stable ICU 2.0
1170 */
1171 U_STABLE int32_t U_EXPORT2
1172 ubidi_getLength(const UBiDi *pBiDi);
1173
1174 /**
1175 * Get the paragraph level of the text.
1176 *
1177 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1178 *
1179 * @return The paragraph level. If there are multiple paragraphs, their
1180 * level may vary if the required paraLevel is UBIDI_DEFAULT_LTR or
1181 * UBIDI_DEFAULT_RTL. In that case, the level of the first paragraph
1182 * is returned.
1183 *
1184 * @see UBiDiLevel
1185 * @see ubidi_getParagraph
1186 * @see ubidi_getParagraphByIndex
1187 * @stable ICU 2.0
1188 */
1189 U_STABLE UBiDiLevel U_EXPORT2
1190 ubidi_getParaLevel(const UBiDi *pBiDi);
1191
1192 /**
1193 * Get the number of paragraphs.
1194 *
1195 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1196 *
1197 * @return The number of paragraphs.
1198 * @stable ICU 3.4
1199 */
1200 U_STABLE int32_t U_EXPORT2
1201 ubidi_countParagraphs(UBiDi *pBiDi);
1202
1203 /**
1204 * Get a paragraph, given a position within the text.
1205 * This function returns information about a paragraph.<br>
1206 * Note: if the paragraph index is known, it is more efficient to
1207 * retrieve the paragraph information using ubidi_getParagraphByIndex().<p>
1208 *
1209 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1210 *
1211 * @param charIndex is the index of a character within the text, in the
1212 * range <code>[0..ubidi_getProcessedLength(pBiDi)-1]</code>.
1213 *
1214 * @param pParaStart will receive the index of the first character of the
1215 * paragraph in the text.
1216 * This pointer can be <code>NULL</code> if this
1217 * value is not necessary.
1218 *
1219 * @param pParaLimit will receive the limit of the paragraph.
1220 * The l-value that you point to here may be the
1221 * same expression (variable) as the one for
1222 * <code>charIndex</code>.
1223 * This pointer can be <code>NULL</code> if this
1224 * value is not necessary.
1225 *
1226 * @param pParaLevel will receive the level of the paragraph.
1227 * This pointer can be <code>NULL</code> if this
1228 * value is not necessary.
1229 *
1230 * @param pErrorCode must be a valid pointer to an error code value.
1231 *
1232 * @return The index of the paragraph containing the specified position.
1233 *
1234 * @see ubidi_getProcessedLength
1235 * @stable ICU 3.4
1236 */
1237 U_STABLE int32_t U_EXPORT2
1238 ubidi_getParagraph(const UBiDi *pBiDi, int32_t charIndex, int32_t *pParaStart,
1239 int32_t *pParaLimit, UBiDiLevel *pParaLevel,
1240 UErrorCode *pErrorCode);
1241
1242 /**
1243 * Get a paragraph, given the index of this paragraph.
1244 *
1245 * This function returns information about a paragraph.<p>
1246 *
1247 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1248 *
1249 * @param paraIndex is the number of the paragraph, in the
1250 * range <code>[0..ubidi_countParagraphs(pBiDi)-1]</code>.
1251 *
1252 * @param pParaStart will receive the index of the first character of the
1253 * paragraph in the text.
1254 * This pointer can be <code>NULL</code> if this
1255 * value is not necessary.
1256 *
1257 * @param pParaLimit will receive the limit of the paragraph.
1258 * This pointer can be <code>NULL</code> if this
1259 * value is not necessary.
1260 *
1261 * @param pParaLevel will receive the level of the paragraph.
1262 * This pointer can be <code>NULL</code> if this
1263 * value is not necessary.
1264 *
1265 * @param pErrorCode must be a valid pointer to an error code value.
1266 *
1267 * @stable ICU 3.4
1268 */
1269 U_STABLE void U_EXPORT2
1270 ubidi_getParagraphByIndex(const UBiDi *pBiDi, int32_t paraIndex,
1271 int32_t *pParaStart, int32_t *pParaLimit,
1272 UBiDiLevel *pParaLevel, UErrorCode *pErrorCode);
1273
1274 /**
1275 * Get the level for one character.
1276 *
1277 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1278 *
1279 * @param charIndex the index of a character. It must be in the range
1280 * [0..ubidi_getProcessedLength(pBiDi)].
1281 *
1282 * @return The level for the character at charIndex (0 if charIndex is not
1283 * in the valid range).
1284 *
1285 * @see UBiDiLevel
1286 * @see ubidi_getProcessedLength
1287 * @stable ICU 2.0
1288 */
1289 U_STABLE UBiDiLevel U_EXPORT2
1290 ubidi_getLevelAt(const UBiDi *pBiDi, int32_t charIndex);
1291
1292 /**
1293 * Get an array of levels for each character.<p>
1294 *
1295 * Note that this function may allocate memory under some
1296 * circumstances, unlike <code>ubidi_getLevelAt()</code>.
1297 *
1298 * @param pBiDi is the paragraph or line <code>UBiDi</code> object, whose
1299 * text length must be strictly positive.
1300 *
1301 * @param pErrorCode must be a valid pointer to an error code value.
1302 *
1303 * @return The levels array for the text,
1304 * or <code>NULL</code> if an error occurs.
1305 *
1306 * @see UBiDiLevel
1307 * @see ubidi_getProcessedLength
1308 * @stable ICU 2.0
1309 */
1310 U_STABLE const UBiDiLevel * U_EXPORT2
1311 ubidi_getLevels(UBiDi *pBiDi, UErrorCode *pErrorCode);
1312
1313 /**
1314 * Get a logical run.
1315 * This function returns information about a run and is used
1316 * to retrieve runs in logical order.<p>
1317 * This is especially useful for line-breaking on a paragraph.
1318 *
1319 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1320 *
1321 * @param logicalPosition is a logical position within the source text.
1322 *
1323 * @param pLogicalLimit will receive the limit of the corresponding run.
1324 * The l-value that you point to here may be the
1325 * same expression (variable) as the one for
1326 * <code>logicalPosition</code>.
1327 * This pointer can be <code>NULL</code> if this
1328 * value is not necessary.
1329 *
1330 * @param pLevel will receive the level of the corresponding run.
1331 * This pointer can be <code>NULL</code> if this
1332 * value is not necessary.
1333 *
1334 * @see ubidi_getProcessedLength
1335 * @stable ICU 2.0
1336 */
1337 U_STABLE void U_EXPORT2
1338 ubidi_getLogicalRun(const UBiDi *pBiDi, int32_t logicalPosition,
1339 int32_t *pLogicalLimit, UBiDiLevel *pLevel);
1340
1341 /**
1342 * Get the number of runs.
1343 * This function may invoke the actual reordering on the
1344 * <code>UBiDi</code> object, after <code>ubidi_setPara()</code>
1345 * may have resolved only the levels of the text. Therefore,
1346 * <code>ubidi_countRuns()</code> may have to allocate memory,
1347 * and may fail doing so.
1348 *
1349 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1350 *
1351 * @param pErrorCode must be a valid pointer to an error code value.
1352 *
1353 * @return The number of runs.
1354 * @stable ICU 2.0
1355 */
1356 U_STABLE int32_t U_EXPORT2
1357 ubidi_countRuns(UBiDi *pBiDi, UErrorCode *pErrorCode);
1358
1359 /**
1360 * Get one run's logical start, length, and directionality,
1361 * which can be 0 for LTR or 1 for RTL.
1362 * In an RTL run, the character at the logical start is
1363 * visually on the right of the displayed run.
1364 * The length is the number of characters in the run.<p>
1365 * <code>ubidi_countRuns()</code> should be called
1366 * before the runs are retrieved.
1367 *
1368 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1369 *
1370 * @param runIndex is the number of the run in visual order, in the
1371 * range <code>[0..ubidi_countRuns(pBiDi)-1]</code>.
1372 *
1373 * @param pLogicalStart is the first logical character index in the text.
1374 * The pointer may be <code>NULL</code> if this index is not needed.
1375 *
1376 * @param pLength is the number of characters (at least one) in the run.
1377 * The pointer may be <code>NULL</code> if this is not needed.
1378 *
1379 * @return the directionality of the run,
1380 * <code>UBIDI_LTR==0</code> or <code>UBIDI_RTL==1</code>,
1381 * never <code>UBIDI_MIXED</code>.
1382 *
1383 * @see ubidi_countRuns
1384 *
1385 * Example:
1386 * <pre>
1387 * \code
1388 * int32_t i, count=ubidi_countRuns(pBiDi),
1389 * logicalStart, visualIndex=0, length;
1390 * for(i=0; i<count; ++i) {
1391 * if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, i, &logicalStart, &length)) {
1392 * do { // LTR
1393 * show_char(text[logicalStart++], visualIndex++);
1394 * } while(--length>0);
1395 * } else {
1396 * logicalStart+=length; // logicalLimit
1397 * do { // RTL
1398 * show_char(text[--logicalStart], visualIndex++);
1399 * } while(--length>0);
1400 * }
1401 * }
1402 *\endcode
1403 * </pre>
1404 *
1405 * Note that in right-to-left runs, code like this places
1406 * modifier letters before base characters and second surrogates
1407 * before first ones.
1408 * @stable ICU 2.0
1409 */
1410 U_STABLE UBiDiDirection U_EXPORT2
1411 ubidi_getVisualRun(UBiDi *pBiDi, int32_t runIndex,
1412 int32_t *pLogicalStart, int32_t *pLength);
1413
1414 /**
1415 * Get the visual position from a logical text position.
1416 * If such a mapping is used many times on the same
1417 * <code>UBiDi</code> object, then calling
1418 * <code>ubidi_getLogicalMap()</code> is more efficient.<p>
1419 *
1420 * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no
1421 * visual position because the corresponding text character is a Bidi control
1422 * removed from output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>.
1423 * <p>
1424 * When the visual output is altered by using options of
1425 * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1426 * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
1427 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the visual position returned may not
1428 * be correct. It is advised to use, when possible, reordering options
1429 * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
1430 * <p>
1431 * Note that in right-to-left runs, this mapping places
1432 * modifier letters before base characters and second surrogates
1433 * before first ones.
1434 *
1435 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1436 *
1437 * @param logicalIndex is the index of a character in the text.
1438 *
1439 * @param pErrorCode must be a valid pointer to an error code value.
1440 *
1441 * @return The visual position of this character.
1442 *
1443 * @see ubidi_getLogicalMap
1444 * @see ubidi_getLogicalIndex
1445 * @see ubidi_getProcessedLength
1446 * @stable ICU 2.0
1447 */
1448 U_STABLE int32_t U_EXPORT2
1449 ubidi_getVisualIndex(UBiDi *pBiDi, int32_t logicalIndex, UErrorCode *pErrorCode);
1450
1451 /**
1452 * Get the logical text position from a visual position.
1453 * If such a mapping is used many times on the same
1454 * <code>UBiDi</code> object, then calling
1455 * <code>ubidi_getVisualMap()</code> is more efficient.<p>
1456 *
1457 * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no
1458 * logical position because the corresponding text character is a Bidi mark
1459 * inserted in the output by option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
1460 * <p>
1461 * This is the inverse function to <code>ubidi_getVisualIndex()</code>.
1462 * <p>
1463 * When the visual output is altered by using options of
1464 * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1465 * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
1466 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the logical position returned may not
1467 * be correct. It is advised to use, when possible, reordering options
1468 * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
1469 *
1470 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1471 *
1472 * @param visualIndex is the visual position of a character.
1473 *
1474 * @param pErrorCode must be a valid pointer to an error code value.
1475 *
1476 * @return The index of this character in the text.
1477 *
1478 * @see ubidi_getVisualMap
1479 * @see ubidi_getVisualIndex
1480 * @see ubidi_getResultLength
1481 * @stable ICU 2.0
1482 */
1483 U_STABLE int32_t U_EXPORT2
1484 ubidi_getLogicalIndex(UBiDi *pBiDi, int32_t visualIndex, UErrorCode *pErrorCode);
1485
1486 /**
1487 * Get a logical-to-visual index map (array) for the characters in the UBiDi
1488 * (paragraph or line) object.
1489 * <p>
1490 * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the
1491 * corresponding text characters are Bidi controls removed from the visual
1492 * output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>.
1493 * <p>
1494 * When the visual output is altered by using options of
1495 * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1496 * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
1497 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the visual positions returned may not
1498 * be correct. It is advised to use, when possible, reordering options
1499 * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
1500 *
1501 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1502 *
1503 * @param indexMap is a pointer to an array of <code>ubidi_getProcessedLength()</code>
1504 * indexes which will reflect the reordering of the characters.
1505 * If option <code>#UBIDI_OPTION_INSERT_MARKS</code> is set, the number
1506 * of elements allocated in <code>indexMap</code> must be no less than
1507 * <code>ubidi_getResultLength()</code>.
1508 * The array does not need to be initialized.<br><br>
1509 * The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>.
1510 *
1511 * @param pErrorCode must be a valid pointer to an error code value.
1512 *
1513 * @see ubidi_getVisualMap
1514 * @see ubidi_getVisualIndex
1515 * @see ubidi_getProcessedLength
1516 * @see ubidi_getResultLength
1517 * @stable ICU 2.0
1518 */
1519 U_STABLE void U_EXPORT2
1520 ubidi_getLogicalMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode);
1521
1522 /**
1523 * Get a visual-to-logical index map (array) for the characters in the UBiDi
1524 * (paragraph or line) object.
1525 * <p>
1526 * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the
1527 * corresponding text characters are Bidi marks inserted in the visual output
1528 * by the option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
1529 * <p>
1530 * When the visual output is altered by using options of
1531 * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1532 * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
1533 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the logical positions returned may not
1534 * be correct. It is advised to use, when possible, reordering options
1535 * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
1536 *
1537 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1538 *
1539 * @param indexMap is a pointer to an array of <code>ubidi_getResultLength()</code>
1540 * indexes which will reflect the reordering of the characters.
1541 * If option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is set, the number
1542 * of elements allocated in <code>indexMap</code> must be no less than
1543 * <code>ubidi_getProcessedLength()</code>.
1544 * The array does not need to be initialized.<br><br>
1545 * The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>.
1546 *
1547 * @param pErrorCode must be a valid pointer to an error code value.
1548 *
1549 * @see ubidi_getLogicalMap
1550 * @see ubidi_getLogicalIndex
1551 * @see ubidi_getProcessedLength
1552 * @see ubidi_getResultLength
1553 * @stable ICU 2.0
1554 */
1555 U_STABLE void U_EXPORT2
1556 ubidi_getVisualMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode);
1557
1558 /**
1559 * This is a convenience function that does not use a UBiDi object.
1560 * It is intended to be used for when an application has determined the levels
1561 * of objects (character sequences) and just needs to have them reordered (L2).
1562 * This is equivalent to using <code>ubidi_getLogicalMap()</code> on a
1563 * <code>UBiDi</code> object.
1564 *
1565 * @param levels is an array with <code>length</code> levels that have been determined by
1566 * the application.
1567 *
1568 * @param length is the number of levels in the array, or, semantically,
1569 * the number of objects to be reordered.
1570 * It must be <code>length>0</code>.
1571 *
1572 * @param indexMap is a pointer to an array of <code>length</code>
1573 * indexes which will reflect the reordering of the characters.
1574 * The array does not need to be initialized.<p>
1575 * The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>.
1576 * @stable ICU 2.0
1577 */
1578 U_STABLE void U_EXPORT2
1579 ubidi_reorderLogical(const UBiDiLevel *levels, int32_t length, int32_t *indexMap);
1580
1581 /**
1582 * This is a convenience function that does not use a UBiDi object.
1583 * It is intended to be used for when an application has determined the levels
1584 * of objects (character sequences) and just needs to have them reordered (L2).
1585 * This is equivalent to using <code>ubidi_getVisualMap()</code> on a
1586 * <code>UBiDi</code> object.
1587 *
1588 * @param levels is an array with <code>length</code> levels that have been determined by
1589 * the application.
1590 *
1591 * @param length is the number of levels in the array, or, semantically,
1592 * the number of objects to be reordered.
1593 * It must be <code>length>0</code>.
1594 *
1595 * @param indexMap is a pointer to an array of <code>length</code>
1596 * indexes which will reflect the reordering of the characters.
1597 * The array does not need to be initialized.<p>
1598 * The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>.
1599 * @stable ICU 2.0
1600 */
1601 U_STABLE void U_EXPORT2
1602 ubidi_reorderVisual(const UBiDiLevel *levels, int32_t length, int32_t *indexMap);
1603
1604 /**
1605 * Invert an index map.
1606 * The index mapping of the first map is inverted and written to
1607 * the second one.
1608 *
1609 * @param srcMap is an array with <code>length</code> elements
1610 * which defines the original mapping from a source array containing
1611 * <code>length</code> elements to a destination array.
1612 * Some elements of the source array may have no mapping in the
1613 * destination array. In that case, their value will be
1614 * the special value <code>UBIDI_MAP_NOWHERE</code>.
1615 * All elements must be >=0 or equal to <code>UBIDI_MAP_NOWHERE</code>.
1616 * Some elements may have a value >= <code>length</code>, if the
1617 * destination array has more elements than the source array.
1618 * There must be no duplicate indexes (two or more elements with the
1619 * same value except <code>UBIDI_MAP_NOWHERE</code>).
1620 *
1621 * @param destMap is an array with a number of elements equal to 1 + the highest
1622 * value in <code>srcMap</code>.
1623 * <code>destMap</code> will be filled with the inverse mapping.
1624 * If element with index i in <code>srcMap</code> has a value k different
1625 * from <code>UBIDI_MAP_NOWHERE</code>, this means that element i of
1626 * the source array maps to element k in the destination array.
1627 * The inverse map will have value i in its k-th element.
1628 * For all elements of the destination array which do not map to
1629 * an element in the source array, the corresponding element in the
1630 * inverse map will have a value equal to <code>UBIDI_MAP_NOWHERE</code>.
1631 *
1632 * @param length is the length of each array.
1633 * @See UBIDI_MAP_NOWHERE
1634 * @stable ICU 2.0
1635 */
1636 U_STABLE void U_EXPORT2
1637 ubidi_invertMap(const int32_t *srcMap, int32_t *destMap, int32_t length);
1638
1639 /** option flags for ubidi_writeReordered() */
1640
1641 /**
1642 * option bit for ubidi_writeReordered():
1643 * keep combining characters after their base characters in RTL runs
1644 *
1645 * @see ubidi_writeReordered
1646 * @stable ICU 2.0
1647 */
1648 #define UBIDI_KEEP_BASE_COMBINING 1
1649
1650 /**
1651 * option bit for ubidi_writeReordered():
1652 * replace characters with the "mirrored" property in RTL runs
1653 * by their mirror-image mappings
1654 *
1655 * @see ubidi_writeReordered
1656 * @stable ICU 2.0
1657 */
1658 #define UBIDI_DO_MIRRORING 2
1659
1660 /**
1661 * option bit for ubidi_writeReordered():
1662 * surround the run with LRMs if necessary;
1663 * this is part of the approximate "inverse Bidi" algorithm
1664 *
1665 * <p>This option does not imply corresponding adjustment of the index
1666 * mappings.</p>
1667 *
1668 * @see ubidi_setInverse
1669 * @see ubidi_writeReordered
1670 * @stable ICU 2.0
1671 */
1672 #define UBIDI_INSERT_LRM_FOR_NUMERIC 4
1673
1674 /**
1675 * option bit for ubidi_writeReordered():
1676 * remove Bidi control characters
1677 * (this does not affect #UBIDI_INSERT_LRM_FOR_NUMERIC)
1678 *
1679 * <p>This option does not imply corresponding adjustment of the index
1680 * mappings.</p>
1681 *
1682 * @see ubidi_writeReordered
1683 * @stable ICU 2.0
1684 */
1685 #define UBIDI_REMOVE_BIDI_CONTROLS 8
1686
1687 /**
1688 * option bit for ubidi_writeReordered():
1689 * write the output in reverse order
1690 *
1691 * <p>This has the same effect as calling <code>ubidi_writeReordered()</code>
1692 * first without this option, and then calling
1693 * <code>ubidi_writeReverse()</code> without mirroring.
1694 * Doing this in the same step is faster and avoids a temporary buffer.
1695 * An example for using this option is output to a character terminal that
1696 * is designed for RTL scripts and stores text in reverse order.</p>
1697 *
1698 * @see ubidi_writeReordered
1699 * @stable ICU 2.0
1700 */
1701 #define UBIDI_OUTPUT_REVERSE 16
1702
1703 /**
1704 * Get the length of the source text processed by the last call to
1705 * <code>ubidi_setPara()</code>. This length may be different from the length
1706 * of the source text if option <code>#UBIDI_OPTION_STREAMING</code>
1707 * has been set.
1708 * <br>
1709 * Note that whenever the length of the text affects the execution or the
1710 * result of a function, it is the processed length which must be considered,
1711 * except for <code>ubidi_setPara</code> (which receives unprocessed source
1712 * text) and <code>ubidi_getLength</code> (which returns the original length
1713 * of the source text).<br>
1714 * In particular, the processed length is the one to consider in the following
1715 * cases:
1716 * <ul>
1717 * <li>maximum value of the <code>limit</code> argument of
1718 * <code>ubidi_setLine</code></li>
1719 * <li>maximum value of the <code>charIndex</code> argument of
1720 * <code>ubidi_getParagraph</code></li>
1721 * <li>maximum value of the <code>charIndex</code> argument of
1722 * <code>ubidi_getLevelAt</code></li>
1723 * <li>number of elements in the array returned by <code>ubidi_getLevels</code></li>
1724 * <li>maximum value of the <code>logicalStart</code> argument of
1725 * <code>ubidi_getLogicalRun</code></li>
1726 * <li>maximum value of the <code>logicalIndex</code> argument of
1727 * <code>ubidi_getVisualIndex</code></li>
1728 * <li>number of elements filled in the <code>*indexMap</code> argument of
1729 * <code>ubidi_getLogicalMap</code></li>
1730 * <li>length of text processed by <code>ubidi_writeReordered</code></li>
1731 * </ul>
1732 *
1733 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1734 *
1735 * @return The length of the part of the source text processed by
1736 * the last call to <code>ubidi_setPara</code>.
1737 * @see ubidi_setPara
1738 * @see UBIDI_OPTION_STREAMING
1739 * @stable ICU 3.6
1740 */
1741 U_STABLE int32_t U_EXPORT2
1742 ubidi_getProcessedLength(const UBiDi *pBiDi);
1743
1744 /**
1745 * Get the length of the reordered text resulting from the last call to
1746 * <code>ubidi_setPara()</code>. This length may be different from the length
1747 * of the source text if option <code>#UBIDI_OPTION_INSERT_MARKS</code>
1748 * or option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> has been set.
1749 * <br>
1750 * This resulting length is the one to consider in the following cases:
1751 * <ul>
1752 * <li>maximum value of the <code>visualIndex</code> argument of
1753 * <code>ubidi_getLogicalIndex</code></li>
1754 * <li>number of elements of the <code>*indexMap</code> argument of
1755 * <code>ubidi_getVisualMap</code></li>
1756 * </ul>
1757 * Note that this length stays identical to the source text length if
1758 * Bidi marks are inserted or removed using option bits of
1759 * <code>ubidi_writeReordered</code>, or if option
1760 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> has been set.
1761 *
1762 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1763 *
1764 * @return The length of the reordered text resulting from
1765 * the last call to <code>ubidi_setPara</code>.
1766 * @see ubidi_setPara
1767 * @see UBIDI_OPTION_INSERT_MARKS
1768 * @see UBIDI_OPTION_REMOVE_CONTROLS
1769 * @stable ICU 3.6
1770 */
1771 U_STABLE int32_t U_EXPORT2
1772 ubidi_getResultLength(const UBiDi *pBiDi);
1773
1774 U_CDECL_BEGIN
1775 /**
1776 * value returned by <code>UBiDiClassCallback</code> callbacks when
1777 * there is no need to override the standard Bidi class for a given code point.
1778 * @see UBiDiClassCallback
1779 * @stable ICU 3.6
1780 */
1781 #define U_BIDI_CLASS_DEFAULT U_CHAR_DIRECTION_COUNT
1782
1783 /**
1784 * Callback type declaration for overriding default Bidi class values with
1785 * custom ones.
1786 * <p>Usually, the function pointer will be propagated to a <code>UBiDi</code>
1787 * object by calling the <code>ubidi_setClassCallback()</code> function;
1788 * then the callback will be invoked by the UBA implementation any time the
1789 * class of a character is to be determined.</p>
1790 *
1791 * @param context is a pointer to the callback private data.
1792 *
1793 * @param c is the code point to get a Bidi class for.
1794 *
1795 * @return The directional property / Bidi class for the given code point
1796 * <code>c</code> if the default class has been overridden, or
1797 * <code>#U_BIDI_CLASS_DEFAULT</code> if the standard Bidi class value
1798 * for <code>c</code> is to be used.
1799 * @see ubidi_setClassCallback
1800 * @see ubidi_getClassCallback
1801 * @stable ICU 3.6
1802 */
1803 typedef UCharDirection U_CALLCONV
1804 UBiDiClassCallback(const void *context, UChar32 c);
1805
1806 U_CDECL_END
1807
1808 /**
1809 * Retrieve the Bidi class for a given code point.
1810 * <p>If a <code>#UBiDiClassCallback</code> callback is defined and returns a
1811 * value other than <code>#U_BIDI_CLASS_DEFAULT</code>, that value is used;
1812 * otherwise the default class determination mechanism is invoked.</p>
1813 *
1814 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1815 *
1816 * @param c is the code point whose Bidi class must be retrieved.
1817 *
1818 * @return The Bidi class for character <code>c</code> based
1819 * on the given <code>pBiDi</code> instance.
1820 * @see UBiDiClassCallback
1821 * @stable ICU 3.6
1822 */
1823 U_STABLE UCharDirection U_EXPORT2
1824 ubidi_getCustomizedClass(UBiDi *pBiDi, UChar32 c);
1825
1826 /**
1827 * Set the callback function and callback data used by the UBA
1828 * implementation for Bidi class determination.
1829 * <p>This may be useful for assigning Bidi classes to PUA characters, or
1830 * for special application needs. For instance, an application may want to
1831 * handle all spaces like L or R characters (according to the base direction)
1832 * when creating the visual ordering of logical lines which are part of a report
1833 * organized in columns: there should not be interaction between adjacent
1834 * cells.<p>
1835 *
1836 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1837 *
1838 * @param newFn is the new callback function pointer.
1839 *
1840 * @param newContext is the new callback context pointer. This can be NULL.
1841 *
1842 * @param oldFn fillin: Returns the old callback function pointer. This can be
1843 * NULL.
1844 *
1845 * @param oldContext fillin: Returns the old callback's context. This can be
1846 * NULL.
1847 *
1848 * @param pErrorCode must be a valid pointer to an error code value.
1849 *
1850 * @see ubidi_getClassCallback
1851 * @stable ICU 3.6
1852 */
1853 U_STABLE void U_EXPORT2
1854 ubidi_setClassCallback(UBiDi *pBiDi, UBiDiClassCallback *newFn,
1855 const void *newContext, UBiDiClassCallback **oldFn,
1856 const void **oldContext, UErrorCode *pErrorCode);
1857
1858 /**
1859 * Get the current callback function used for Bidi class determination.
1860 *
1861 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1862 *
1863 * @param fn fillin: Returns the callback function pointer.
1864 *
1865 * @param context fillin: Returns the callback's private context.
1866 *
1867 * @see ubidi_setClassCallback
1868 * @stable ICU 3.6
1869 */
1870 U_STABLE void U_EXPORT2
1871 ubidi_getClassCallback(UBiDi *pBiDi, UBiDiClassCallback **fn, const void **context);
1872
1873 /**
1874 * Take a <code>UBiDi</code> object containing the reordering
1875 * information for a piece of text (one or more paragraphs) set by
1876 * <code>ubidi_setPara()</code> or for a line of text set by
1877 * <code>ubidi_setLine()</code> and write a reordered string to the
1878 * destination buffer.
1879 *
1880 * This function preserves the integrity of characters with multiple
1881 * code units and (optionally) modifier letters.
1882 * Characters in RTL runs can be replaced by mirror-image characters
1883 * in the destination buffer. Note that "real" mirroring has
1884 * to be done in a rendering engine by glyph selection
1885 * and that for many "mirrored" characters there are no
1886 * Unicode characters as mirror-image equivalents.
1887 * There are also options to insert or remove Bidi control
1888 * characters; see the description of the <code>destSize</code>
1889 * and <code>options</code> parameters and of the option bit flags.
1890 *
1891 * @param pBiDi A pointer to a <code>UBiDi</code> object that
1892 * is set by <code>ubidi_setPara()</code> or
1893 * <code>ubidi_setLine()</code> and contains the reordering
1894 * information for the text that it was defined for,
1895 * as well as a pointer to that text.<br><br>
1896 * The text was aliased (only the pointer was stored
1897 * without copying the contents) and must not have been modified
1898 * since the <code>ubidi_setPara()</code> call.
1899 *
1900 * @param dest A pointer to where the reordered text is to be copied.
1901 * The source text and <code>dest[destSize]</code>
1902 * must not overlap.
1903 *
1904 * @param destSize The size of the <code>dest</code> buffer,
1905 * in number of UChars.
1906 * If the <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>
1907 * option is set, then the destination length could be
1908 * as large as
1909 * <code>ubidi_getLength(pBiDi)+2*ubidi_countRuns(pBiDi)</code>.
1910 * If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option
1911 * is set, then the destination length may be less than
1912 * <code>ubidi_getLength(pBiDi)</code>.
1913 * If none of these options is set, then the destination length
1914 * will be exactly <code>ubidi_getProcessedLength(pBiDi)</code>.
1915 *
1916 * @param options A bit set of options for the reordering that control
1917 * how the reordered text is written.
1918 * The options include mirroring the characters on a code
1919 * point basis and inserting LRM characters, which is used
1920 * especially for transforming visually stored text
1921 * to logically stored text (although this is still an
1922 * imperfect implementation of an "inverse Bidi" algorithm
1923 * because it uses the "forward Bidi" algorithm at its core).
1924 * The available options are:
1925 * <code>#UBIDI_DO_MIRRORING</code>,
1926 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1927 * <code>#UBIDI_KEEP_BASE_COMBINING</code>,
1928 * <code>#UBIDI_OUTPUT_REVERSE</code>,
1929 * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code>
1930 *
1931 * @param pErrorCode must be a valid pointer to an error code value.
1932 *
1933 * @return The length of the output string.
1934 *
1935 * @see ubidi_getProcessedLength
1936 * @stable ICU 2.0
1937 */
1938 U_STABLE int32_t U_EXPORT2
1939 ubidi_writeReordered(UBiDi *pBiDi,
1940 UChar *dest, int32_t destSize,
1941 uint16_t options,
1942 UErrorCode *pErrorCode);
1943
1944 /**
1945 * Reverse a Right-To-Left run of Unicode text.
1946 *
1947 * This function preserves the integrity of characters with multiple
1948 * code units and (optionally) modifier letters.
1949 * Characters can be replaced by mirror-image characters
1950 * in the destination buffer. Note that "real" mirroring has
1951 * to be done in a rendering engine by glyph selection
1952 * and that for many "mirrored" characters there are no
1953 * Unicode characters as mirror-image equivalents.
1954 * There are also options to insert or remove Bidi control
1955 * characters.
1956 *
1957 * This function is the implementation for reversing RTL runs as part
1958 * of <code>ubidi_writeReordered()</code>. For detailed descriptions
1959 * of the parameters, see there.
1960 * Since no Bidi controls are inserted here, the output string length
1961 * will never exceed <code>srcLength</code>.
1962 *
1963 * @see ubidi_writeReordered
1964 *
1965 * @param src A pointer to the RTL run text.
1966 *
1967 * @param srcLength The length of the RTL run.
1968 *
1969 * @param dest A pointer to where the reordered text is to be copied.
1970 * <code>src[srcLength]</code> and <code>dest[destSize]</code>
1971 * must not overlap.
1972 *
1973 * @param destSize The size of the <code>dest</code> buffer,
1974 * in number of UChars.
1975 * If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option
1976 * is set, then the destination length may be less than
1977 * <code>srcLength</code>.
1978 * If this option is not set, then the destination length
1979 * will be exactly <code>srcLength</code>.
1980 *
1981 * @param options A bit set of options for the reordering that control
1982 * how the reordered text is written.
1983 * See the <code>options</code> parameter in <code>ubidi_writeReordered()</code>.
1984 *
1985 * @param pErrorCode must be a valid pointer to an error code value.
1986 *
1987 * @return The length of the output string.
1988 * @stable ICU 2.0
1989 */
1990 U_STABLE int32_t U_EXPORT2
1991 ubidi_writeReverse(const UChar *src, int32_t srcLength,
1992 UChar *dest, int32_t destSize,
1993 uint16_t options,
1994 UErrorCode *pErrorCode);
1995
1996 /*#define BIDI_SAMPLE_CODE*/
1997 /*@}*/
1998
1999 #endif