]> git.saurik.com Git - apple/icu.git/blame - icuSources/common/unicode/ubidi.h
ICU-8.11.4.tar.gz
[apple/icu.git] / icuSources / common / unicode / ubidi.h
CommitLineData
b75a7d8f
A
1/*
2******************************************************************************
3*
73c04bcf 4* Copyright (C) 1999-2006, International Business Machines
b75a7d8f
A
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
15*/
16
17#ifndef UBIDI_H
18#define UBIDI_H
19
20#include "unicode/utypes.h"
21#include "unicode/uchar.h"
22
b75a7d8f
A
23/**
24 *\file
374ca955 25 * \brief C API: BIDI algorithm
b75a7d8f
A
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
374ca955
A
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>
b75a7d8f
A
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 */
374ca955 286
b75a7d8f
A
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
374ca955 305 * pseudo-level values <code>UBIDI_DEFAULT_LTR</code>
b75a7d8f
A
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 */
330typedef uint8_t UBiDiLevel;
331
332/** Paragraph level setting.
333 * If there is no strong character, then set the paragraph level to 0 (left-to-right).
334 * @stable ICU 2.0
335 */
336#define UBIDI_DEFAULT_LTR 0xfe
337
338/** Paragraph level setting.
339 * If there is no strong character, then set the paragraph level to 1 (right-to-left).
340 * @stable ICU 2.0
341 */
342#define UBIDI_DEFAULT_RTL 0xff
343
344/**
345 * Maximum explicit embedding level.
346 * (The maximum resolved level can be up to <code>UBIDI_MAX_EXPLICIT_LEVEL+1</code>).
347 * @stable ICU 2.0
348 */
349#define UBIDI_MAX_EXPLICIT_LEVEL 61
350
374ca955
A
351/** Bit flag for level input.
352 * Overrides directional properties.
b75a7d8f
A
353 * @stable ICU 2.0
354 */
355#define UBIDI_LEVEL_OVERRIDE 0x80
356
73c04bcf
A
357/**
358 * Special value which can be returned by the mapping functions when a logical
359 * index has no corresponding visual index or vice-versa. This may happen
360 * for the logical-to-visual mapping of a BiDi control when option
361 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is specified. This can also happen
362 * for the visual-to-logical mapping of a BiDi mark (LRM or RLM) inserted
363 * by option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
364 * @see ubidi_getVisualIndex
365 * @see ubidi_getVisualMap
366 * @see ubidi_getLogicalIndex
367 * @see ubidi_getLogicalMap
368 * @draft ICU 3.6
369 */
370#define UBIDI_MAP_NOWHERE (-1)
371
b75a7d8f 372/**
374ca955 373 * <code>UBiDiDirection</code> values indicate the text direction.
b75a7d8f
A
374 * @stable ICU 2.0
375 */
376enum UBiDiDirection {
374ca955 377 /** All left-to-right text. This is a 0 value. @stable ICU 2.0 */
b75a7d8f 378 UBIDI_LTR,
374ca955 379 /** All right-to-left text. This is a 1 value. @stable ICU 2.0 */
b75a7d8f 380 UBIDI_RTL,
374ca955 381 /** Mixed-directional text. @stable ICU 2.0 */
b75a7d8f
A
382 UBIDI_MIXED
383};
384
385/** @stable ICU 2.0 */
386typedef enum UBiDiDirection UBiDiDirection;
387
388/**
389 * Forward declaration of the <code>UBiDi</code> structure for the declaration of
390 * the API functions. Its fields are implementation-specific.<p>
73c04bcf
A
391 * This structure holds information about a paragraph (or multiple paragraphs)
392 * of text with BiDi-algorithm-related details, or about one line of
b75a7d8f 393 * such a paragraph.<p>
73c04bcf
A
394 * Reordering can be done on a line, or on one or more paragraphs which are
395 * then interpreted each as one single line.
b75a7d8f
A
396 * @stable ICU 2.0
397 */
398struct UBiDi;
399
400/** @stable ICU 2.0 */
401typedef struct UBiDi UBiDi;
402
403/**
404 * Allocate a <code>UBiDi</code> structure.
405 * Such an object is initially empty. It is assigned
73c04bcf
A
406 * the BiDi properties of a piece of text containing one or more paragraphs
407 * by <code>ubidi_setPara()</code>
408 * or the BiDi properties of a line within a paragraph by
b75a7d8f
A
409 * <code>ubidi_setLine()</code>.<p>
410 * This object can be reused for as long as it is not deallocated
411 * by calling <code>ubidi_close()</code>.<p>
73c04bcf
A
412 * <code>ubidi_setPara()</code> and <code>ubidi_setLine()</code> will allocate
413 * additional memory for internal structures as necessary.
b75a7d8f
A
414 *
415 * @return An empty <code>UBiDi</code> object.
416 * @stable ICU 2.0
417 */
374ca955 418U_STABLE UBiDi * U_EXPORT2
b75a7d8f
A
419ubidi_open(void);
420
421/**
422 * Allocate a <code>UBiDi</code> structure with preallocated memory
423 * for internal structures.
424 * This function provides a <code>UBiDi</code> object like <code>ubidi_open()</code>
425 * with no arguments, but it also preallocates memory for internal structures
426 * according to the sizings supplied by the caller.<p>
427 * Subsequent functions will not allocate any more memory, and are thus
428 * guaranteed not to fail because of lack of memory.<p>
429 * The preallocation can be limited to some of the internal memory
430 * by setting some values to 0 here. That means that if, e.g.,
431 * <code>maxRunCount</code> cannot be reasonably predetermined and should not
432 * be set to <code>maxLength</code> (the only failproof value) to avoid
433 * wasting memory, then <code>maxRunCount</code> could be set to 0 here
434 * and the internal structures that are associated with it will be allocated
435 * on demand, just like with <code>ubidi_open()</code>.
436 *
73c04bcf 437 * @param maxLength is the maximum text or line length that internal memory
b75a7d8f
A
438 * will be preallocated for. An attempt to associate this object with a
439 * longer text will fail, unless this value is 0, which leaves the allocation
440 * up to the implementation.
441 *
442 * @param maxRunCount is the maximum anticipated number of same-level runs
443 * that internal memory will be preallocated for. An attempt to access
444 * visual runs on an object that was not preallocated for as many runs
445 * as the text was actually resolved to will fail,
73c04bcf 446 * unless this value is 0, which leaves the allocation up to the implementation.<br><br>
b75a7d8f 447 * The number of runs depends on the actual text and maybe anywhere between
73c04bcf 448 * 1 and <code>maxLength</code>. It is typically small.
b75a7d8f 449 *
73c04bcf 450 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
451 *
452 * @return An empty <code>UBiDi</code> object with preallocated memory.
453 * @stable ICU 2.0
454 */
374ca955 455U_STABLE UBiDi * U_EXPORT2
b75a7d8f
A
456ubidi_openSized(int32_t maxLength, int32_t maxRunCount, UErrorCode *pErrorCode);
457
458/**
459 * <code>ubidi_close()</code> must be called to free the memory
460 * associated with a UBiDi object.<p>
461 *
462 * <strong>Important: </strong>
374ca955
A
463 * A parent <code>UBiDi</code> object must not be destroyed or reused if
464 * it still has children.
73c04bcf
A
465 * If a <code>UBiDi</code> object has become the <i>child</i>
466 * of another one (its <i>parent</i>) by calling
b75a7d8f
A
467 * <code>ubidi_setLine()</code>, then the child object must
468 * be destroyed (closed) or reused (by calling
469 * <code>ubidi_setPara()</code> or <code>ubidi_setLine()</code>)
470 * before the parent object.
471 *
472 * @param pBiDi is a <code>UBiDi</code> object.
473 *
474 * @see ubidi_setPara
475 * @see ubidi_setLine
476 * @stable ICU 2.0
477 */
374ca955 478U_STABLE void U_EXPORT2
b75a7d8f
A
479ubidi_close(UBiDi *pBiDi);
480
481/**
482 * Modify the operation of the BiDi algorithm such that it
483 * approximates an "inverse BiDi" algorithm. This function
484 * must be called before <code>ubidi_setPara()</code>.
485 *
486 * <p>The normal operation of the BiDi algorithm as described
487 * in the Unicode Technical Report is to take text stored in logical
488 * (keyboard, typing) order and to determine the reordering of it for visual
489 * rendering.
374ca955 490 * Some legacy systems store text in visual order, and for operations
b75a7d8f
A
491 * with standard, Unicode-based algorithms, the text needs to be transformed
492 * to logical order. This is effectively the inverse algorithm of the
493 * described BiDi algorithm. Note that there is no standard algorithm for
494 * this "inverse BiDi" and that the current implementation provides only an
495 * approximation of "inverse BiDi".</p>
374ca955 496 *
b75a7d8f
A
497 * <p>With <code>isInverse</code> set to <code>TRUE</code>,
498 * this function changes the behavior of some of the subsequent functions
499 * in a way that they can be used for the inverse BiDi algorithm.
500 * Specifically, runs of text with numeric characters will be treated in a
501 * special way and may need to be surrounded with LRM characters when they are
502 * written in reordered sequence.</p>
503 *
504 * <p>Output runs should be retrieved using <code>ubidi_getVisualRun()</code>.
505 * Since the actual input for "inverse BiDi" is visually ordered text and
506 * <code>ubidi_getVisualRun()</code> gets the reordered runs, these are actually
507 * the runs of the logically ordered output.</p>
508 *
73c04bcf
A
509 * <p>Calling this function with argument <code>isInverse</code> set to
510 * <code>TRUE</code> is equivalent to calling
511 * <code>ubidi_setReorderingMode</code> with argument
512 * <code>reorderingMode</code>
513 * set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br>
514 * Calling this function with argument <code>isInverse</code> set to
515 * <code>FALSE</code> is equivalent to calling
516 * <code>ubidi_setReorderingMode</code> with argument
517 * <code>reorderingMode</code>
518 * set to <code>#UBIDI_REORDER_DEFAULT</code>.
519 *
b75a7d8f
A
520 * @param pBiDi is a <code>UBiDi</code> object.
521 *
73c04bcf 522 * @param isInverse specifies "forward" or "inverse" BiDi operation.
b75a7d8f
A
523 *
524 * @see ubidi_setPara
525 * @see ubidi_writeReordered
73c04bcf 526 * @see ubidi_setReorderingMode
b75a7d8f
A
527 * @stable ICU 2.0
528 */
374ca955 529U_STABLE void U_EXPORT2
b75a7d8f
A
530ubidi_setInverse(UBiDi *pBiDi, UBool isInverse);
531
532/**
533 * Is this BiDi object set to perform the inverse BiDi algorithm?
73c04bcf
A
534 * <p>Note: calling this function after setting the reordering mode with
535 * <code>ubidi_setReorderingMode</code> will return <code>TRUE</code> if the
536 * reordering mode was set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>,
537 * <code>FALSE</code> for all other values.</p>
b75a7d8f
A
538 *
539 * @param pBiDi is a <code>UBiDi</code> object.
73c04bcf
A
540 * @return TRUE if the BiDi object is set to perform the inverse BiDi algorithm
541 * by handling numbers as L.
b75a7d8f
A
542 *
543 * @see ubidi_setInverse
73c04bcf 544 * @see ubidi_setReorderingMode
b75a7d8f
A
545 * @stable ICU 2.0
546 */
73c04bcf 547
374ca955 548U_STABLE UBool U_EXPORT2
b75a7d8f
A
549ubidi_isInverse(UBiDi *pBiDi);
550
73c04bcf
A
551/**
552 * Specify whether block separators must be allocated level zero,
553 * so that successive paragraphs will progress from left to right.
554 * This function must be called before <code>ubidi_setPara()</code>.
555 * Paragraph separators (B) may appear in the text. Setting them to level zero
556 * means that all paragraph separators (including one possibly appearing
557 * in the last text position) are kept in the reordered text after the text
558 * that they follow in the source text.
559 * When this feature is not enabled, a paragraph separator at the last
560 * position of the text before reordering will go to the first position
561 * of the reordered text when the paragraph level is odd.
562 *
563 * @param pBiDi is a <code>UBiDi</code> object.
564 *
565 * @param orderParagraphsLTR specifies whether paragraph separators (B) must
566 * receive level 0, so that successive paragraphs progress from left to right.
567 *
568 * @see ubidi_setPara
569 * @stable ICU 3.4
570 */
571U_STABLE void U_EXPORT2
572ubidi_orderParagraphsLTR(UBiDi *pBiDi, UBool orderParagraphsLTR);
573
574/**
575 * Is this BiDi object set to allocate level 0 to block separators so that
576 * successive paragraphs progress from left to right?
577 *
578 * @param pBiDi is a <code>UBiDi</code> object.
579 * @return TRUE if the BiDi object is set to allocate level 0 to block
580 * separators.
581 *
582 * @see ubidi_orderParagraphsLTR
583 * @stable ICU 3.4
584 */
585U_STABLE UBool U_EXPORT2
586ubidi_isOrderParagraphsLTR(UBiDi *pBiDi);
587
588/**
589 * <code>UBiDiReorderingMode</code> values indicate which variant of the BiDi
590 * algorithm to use.
591 *
592 * @see ubidi_setReorderingMode
593 * @draft ICU 3.6
594 */
595typedef enum UBiDiReorderingMode {
596 /** Regular Logical to Visual BiDi algorithm according to Unicode.
597 * This is a 0 value. @draft ICU 3.6 */
598 UBIDI_REORDER_DEFAULT = 0,
599 /** Logical to Visual algorithm which handles numbers in a way which
600 * mimicks the behavior of Windows XP.
601 * @draft ICU 3.6 */
602 UBIDI_REORDER_NUMBERS_SPECIAL,
603 /** Logical to Visual algorithm grouping numbers with adjacent R characters
604 * (reversible algorithm).
605 * @draft ICU 3.6 */
606 UBIDI_REORDER_GROUP_NUMBERS_WITH_R,
607 /** Reorder runs only to transform a Logical LTR string to the Logical RTL
608 * string with the same display, or vice-versa.<br>
609 * If this mode is set together with option
610 * <code>#UBIDI_OPTION_INSERT_MARKS</code>, some BiDi controls in the source
611 * text may be removed and other controls may be added to produce the
612 * minimum combination which has the required display.
613 * @draft ICU 3.6 */
614 UBIDI_REORDER_RUNS_ONLY,
615 /** Visual to Logical algorithm which handles numbers like L
616 * (same algorithm as selected by <code>ubidi_setInverse(TRUE)</code>.
617 * @see ubidi_setInverse
618 * @draft ICU 3.6 */
619 UBIDI_REORDER_INVERSE_NUMBERS_AS_L,
620 /** Visual to Logical algorithm equivalent to the regular Logical to Visual
621 * algorithm. @draft ICU 3.6 */
622 UBIDI_REORDER_INVERSE_LIKE_DIRECT,
623 /** Inverse BiDi (Visual to Logical) algorithm for the
624 * <code>UBIDI_REORDER_NUMBERS_SPECIAL</code> BiDi algorithm.
625 * @draft ICU 3.6 */
626 UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL,
627 /** Number of values for reordering mode.
628 * @draft ICU 3.6 */
629 UBIDI_REORDER_COUNT
630} UBiDiReorderingMode;
631
632/**
633 * Modify the operation of the BiDi algorithm such that it implements some
634 * variant to the basic BiDi algorithm or approximates an "inverse BiDi"
635 * algorithm, depending on different values of the "reordering mode".
636 * This function must be called before <code>ubidi_setPara()</code>, and stays
637 * in effect until called again with a different argument.
638 *
639 * <p>The normal operation of the BiDi algorithm as described
640 * in the Unicode Standard Annex #9 is to take text stored in logical
641 * (keyboard, typing) order and to determine how to reorder it for visual
642 * rendering.</p>
643 *
644 * <p>With the reordering mode set to a value other than
645 * <code>#UBIDI_REORDER_DEFAULT</code>, this function changes the behavior of
646 * some of the subsequent functions in a way such that they implement an
647 * inverse BiDi algorithm or some other algorithm variants.</p>
648 *
649 * <p>Some legacy systems store text in visual order, and for operations
650 * with standard, Unicode-based algorithms, the text needs to be transformed
651 * into logical order. This is effectively the inverse algorithm of the
652 * described BiDi algorithm. Note that there is no standard algorithm for
653 * this "inverse BiDi", so a number of variants are implemented here.</p>
654 *
655 * <p>In other cases, it may be desirable to emulate some variant of the
656 * Logical to Visual algorithm (e.g. one used in MS Windows), or perform a
657 * Logical to Logical transformation.</p>
658 *
659 * <ul>
660 * <li>When the reordering mode is set to <code>#UBIDI_REORDER_DEFAULT</code>,
661 * the standard BiDi Logical to Visual algorithm is applied.</li>
662 *
663 * <li>When the reordering mode is set to
664 * <code>#UBIDI_REORDER_NUMBERS_SPECIAL</code>,
665 * the algorithm used to perform BiDi transformations when calling
666 * <code>ubidi_setPara</code> should approximate the algorithm used in
667 * Microsoft Windows XP rather than strictly conform to the Unicode BiDi
668 * algorithm.
669 * <br>
670 * The differences between the basic algorithm and the algorithm addressed
671 * by this option are as follows:
672 * <ul>
673 * <li>Within text at an even embedding level, the sequence "123AB"
674 * (where AB represent R or AL letters) is transformed to "123BA" by the
675 * Unicode algorithm and to "BA123" by the Windows algorithm.</li>
676 * <li>Arabic-Indic numbers (AN) are handled by the Windows algorithm just
677 * like regular numbers (EN).</li>
678 * </ul></li>
679 *
680 * <li>When the reordering mode is set to
681 * <code>#UBIDI_REORDER_GROUP_NUMBERS_WITH_R</code>,
682 * numbers located between LTR text and RTL text are associated with the RTL
683 * text. For instance, an LTR paragraph with content "abc 123 DEF" (where
684 * upper case letters represent RTL characters) will be transformed to
685 * "abc FED 123" (and not "abc 123 FED"), "DEF 123 abc" will be transformed
686 * to "123 FED abc" and "123 FED abc" will be transformed to "DEF 123 abc".
687 * This makes the algorithm reversible and makes it useful when round trip
688 * (from visual to logical and back to visual) must be achieved without
689 * adding LRM characters. However, this is a variation from the standard
690 * Unicode Bidi algorithm.<br>
691 * The source text should not contain BiDi control characters other than LRM
692 * or RLM.</li>
693 *
694 * <li>When the reordering mode is set to
695 * <code>#UBIDI_REORDER_RUNS_ONLY</code>,
696 * a "Logical to Logical" transformation must be performed:
697 * <ul>
698 * <li>If the default text level of the source text (argument <code>paraLevel</code>
699 * in <code>ubidi_setPara</code>) is even, the source text will be handled as
700 * LTR logical text and will be transformed to the RTL logical text which has
701 * the same LTR visual display.</li>
702 * <li>If the default level of the source text is odd, the source text
703 * will be handled as RTL logical text and will be transformed to the
704 * LTR logical text which has the same LTR visual display.</li>
705 * </ul>
706 * This mode may be needed when logical text which is basically Arabic or
707 * Hebrew, with possible included numbers or phrases in English, has to be
708 * displayed as if it had an even embedding level (this can happen if the
709 * displaying application treats all text as if it was basically LTR.
710 * <br>
711 * This mode may also be needed in the reverse case, when logical text which is
712 * basically English, with possible included phrases in Arabic or Hebrew, has to
713 * be displayed as if it had an odd embedding level.
714 * <br>
715 * Both cases could be handled by adding LRE or RLE at the head of the text,
716 * if the display subsystem supports these formatting controls. If it does not,
717 * the problem may be handled by transforming the source text in this mode
718 * before displaying it, so that it will be displayed properly.<br>
719 * The source text should not contain BiDi control characters other than LRM
720 * or RLM.</li>
721 *
722 * <li>When the reordering mode is set to
723 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>, an "inverse BiDi" algorithm
724 * is applied.
725 * Runs of text with numeric characters will be treated like LTR letters and
726 * may need to be surrounded with LRM characters when they are written in
727 * reordered sequence (the option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> can
728 * be used with function <code>ubidi_writeReordered</code> to this end. This
729 * mode is equivalent to calling <code>ubidi_setInverse()</code> with
730 * argument <code>isInverse</code> set to <code>TRUE</code>.</li>
731 *
732 * <li>When the reordering mode is set to
733 * <code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code>, the "direct" Logical to Visual
734 * BiDi algorithm is used as an approximation of an "inverse BiDi" algorithm.
735 * This mode is similar to mode <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>
736 * but is closer to the regular BiDi algorithm.
737 * <br>
738 * For example, an LTR paragraph with the content "FED 123 456 CBA" (where
739 * upper case represents RTL characters) will be transformed to
740 * "ABC 456 123 DEF", as opposed to "DEF 123 456 ABC"
741 * with mode <code>UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br>
742 * When used in conjunction with option
743 * <code>#UBIDI_OPTION_INSERT_MARKS</code>, this mode generally
744 * adds BiDi marks to the output significantly more sparingly than mode
745 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> with option
746 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to
747 * <code>ubidi_writeReordered</code>.</li>
748 *
749 * <li>When the reordering mode is set to
750 * <code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the Logical to Visual
751 * BiDi algorithm used in Windows XP is used as an approximation of an
752 * "inverse BiDi" algorithm.
753 * <br>
754 * For example, an LTR paragraph with the content "abc FED123" (where
755 * upper case represents RTL characters) will be transformed to
756 * "abc 123DEF.</li>
757 * </ul>
758 *
759 * <p>In all the reordering modes specifying an "inverse BiDi" algorithm
760 * (i.e. those with a name starting with <code>UBIDI_REORDER_INVERSE</code>),
761 * output runs should be retrieved using
762 * <code>ubidi_getVisualRun()</code>, and the output text with
763 * <code>ubidi_writeReordered()</code>. The caller should keep in mind that in
764 * "inverse BiDi" modes the input is actually visually ordered text and
765 * reordered output returned by <code>ubidi_getVisualRun()</code> or
766 * <code>ubidi_writeReordered()</code> are actually runs or character string
767 * of logically ordered output.<br>
768 * For all the "inverse BiDi" modes, the source text should not contain
769 * BiDi control characters other than LRM or RLM.</p>
770 *
771 * <p>Note that option <code>#UBIDI_OUTPUT_REVERSE</code> of
772 * <code>ubidi_writeReordered</code> has no useful meaning and should not be
773 * used in conjunction with any value of the reordering mode specifying
774 * "inverse BiDi" or with value <code>UBIDI_REORDER_RUNS_ONLY</code>.
775 *
776 * @param pBiDi is a <code>UBiDi</code> object.
777 * @param reorderingMode specifies the required variant of the BiDi algorithm.
778 *
779 * @see UBiDiReorderingMode
780 * @see ubidi_setInverse
781 * @see ubidi_setPara
782 * @see ubidi_writeReordered
783 * @draft ICU 3.6
784 */
785U_DRAFT void U_EXPORT2
786ubidi_setReorderingMode(UBiDi *pBiDi, UBiDiReorderingMode reorderingMode);
787
788/**
789 * What is the requested reordering mode for a given BiDi object?
790 *
791 * @param pBiDi is a <code>UBiDi</code> object.
792 * @return the current reordering mode of the BiDi object
793 * @see ubidi_setReorderingMode
794 * @draft ICU 3.6
795 */
796U_DRAFT UBiDiReorderingMode U_EXPORT2
797ubidi_getReorderingMode(UBiDi *pBiDi);
798
799/**
800 * <code>UBiDiReorderingOption</code> values indicate which options are
801 * specified to affect the BiDi algorithm.
802 *
803 * @see ubidi_setReorderingOptions
804 * @draft ICU 3.6
805 */
806typedef enum UBiDiReorderingOption {
807 /**
808 * option value for <code>ubidi_setReorderingOptions</code>:
809 * disable all the options which can be set with this function
810 * @see ubidi_setReorderingOptions
811 * @draft ICU 3.6
812 */
813 UBIDI_OPTION_DEFAULT = 0,
814
815 /**
816 * option bit for <code>ubidi_setReorderingOptions</code>:
817 * insert BiDi marks (LRM or RLM) when needed to ensure correct result of
818 * a reordering to a Logical order
819 *
820 * <p>This option must be set or reset before calling
821 * <code>ubidi_setPara</code>.</p>
822 *
823 * <p>This option is significant only with reordering modes which generate
824 * a result with Logical order, specifically:</p>
825 * <ul>
826 * <li><code>#UBIDI_REORDER_RUNS_ONLY</code></li>
827 * <li><code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code></li>
828 * <li><code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code></li>
829 * <li><code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code></li>
830 * </ul>
831 *
832 * <p>If this option is set in conjunction with reordering mode
833 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> or with calling
834 * <code>ubidi_setInverse(TRUE)</code>, it implies
835 * option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>
836 * in calls to function <code>ubidi_writeReordered()</code>.</p>
837 *
838 * <p>For other reordering modes, a minimum number of LRM or RLM characters
839 * will be added to the source text after reordering it so as to ensure
840 * round trip, i.e. when applying the inverse reordering mode on the
841 * resulting logical text with removal of BiDi marks
842 * (option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> set before calling
843 * <code>ubidi_setPara()</code> or option <code>#UBIDI_REMOVE_BIDI_CONTROLS</code>
844 * in <code>ubidi_writeReordered</code>), the result will be identical to the
845 * source text in the first transformation.
846 *
847 * <p>This option will be ignored if specified together with option
848 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>. It inhibits option
849 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to function
850 * <code>ubidi_writeReordered()</code> and it implies option
851 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to function
852 * <code>ubidi_writeReordered()</code> if the reordering mode is
853 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.</p>
854 *
855 * @see ubidi_setReorderingMode
856 * @see ubidi_setReorderingOptions
857 * @draft ICU 3.6
858 */
859 UBIDI_OPTION_INSERT_MARKS = 1,
860
861 /**
862 * option bit for <code>ubidi_setReorderingOptions</code>:
863 * remove BiDi control characters
864 *
865 * <p>This option must be set or reset before calling
866 * <code>ubidi_setPara</code>.</p>
867 *
868 * <p>This option nullifies option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
869 * It inhibits option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls
870 * to function <code>ubidi_writeReordered()</code> and it implies option
871 * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to that function.</p>
872 *
873 * @see ubidi_setReorderingMode
874 * @see ubidi_setReorderingOptions
875 * @draft ICU 3.6
876 */
877 UBIDI_OPTION_REMOVE_CONTROLS = 2,
878
879 /**
880 * option bit for <code>ubidi_setReorderingOptions</code>:
881 * process the output as part of a stream to be continued
882 *
883 * <p>This option must be set or reset before calling
884 * <code>ubidi_setPara</code>.</p>
885 *
886 * <p>This option specifies that the caller is interested in processing large
887 * text object in parts.
888 * The results of the successive calls are expected to be concatenated by the
889 * caller. Only the call for the last part will have this option bit off.</p>
890 *
891 * <p>When this option bit is on, <code>ubidi_setPara()</code> may process
892 * less than the full source text in order to truncate the text at a meaningful
893 * boundary. The caller should call <code>ubidi_getProcessedLength()</code>
894 * immediately after calling <code>ubidi_setPara()</code> in order to
895 * determine how much of the source text has been processed.
896 * Source text beyond that length should be resubmitted in following calls to
897 * <code>ubidi_setPara</code>. The processed length may be less than
898 * the length of the source text if a character preceding the last character of
899 * the source text constitutes a reasonable boundary (like a block separator)
900 * for text to be continued.<br>
901 * If the last character of the source text constitutes a reasonable
902 * boundary, the whole text will be processed at once.<br>
903 * If nowhere in the source text there exists
904 * such a reasonable boundary, the processed length will be zero.<br>
905 * The caller should check for such an occurrence and do one of the following:
906 * <ul><li>submit a larger amount of text with a better chance to include
907 * a reasonable boundary.</li>
908 * <li>resubmit the same text after turning off option
909 * <code>UBIDI_OPTION_STREAMING</code>.</li></ul>
910 * In all cases, this option should be turned off before processing the last
911 * part of the text.</p>
912 *
913 * <p>When the <code>UBIDI_OPTION_STREAMING</code> option is used,
914 * it is recommended to call <code>ubidi_orderParagraphsLTR()</code> with
915 * argument <code>orderParagraphsLTR</code> set to <code>TRUE</code> before
916 * calling <code>ubidi_setPara</code> so that later paragraphs may be
917 * concatenated to previous paragraphs on the right.</p>
918 *
919 * @see ubidi_setReorderingMode
920 * @see ubidi_setReorderingOptions
921 * @see ubidi_getProcessedLength
922 * @see ubidi_orderParagraphsLTR
923 * @draft ICU 3.6
924 */
925 UBIDI_OPTION_STREAMING = 4
926} UBiDiReorderingOption;
927
928/**
929 * Specify which of the reordering options
930 * should be applied during BiDi transformations.
931 *
932 * @param pBiDi is a <code>UBiDi</code> object.
933 * @param reorderingOptions is a combination of zero or more of the following
934 * options:
935 * <code>#UBIDI_OPTION_DEFAULT</code>, <code>#UBIDI_OPTION_INSERT_MARKS</code>,
936 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>, <code>#UBIDI_OPTION_STREAMING</code>.
937 *
938 * @see ubidi_getReorderingOptions
939 * @draft ICU 3.6
940 */
941U_DRAFT void U_EXPORT2
942ubidi_setReorderingOptions(UBiDi *pBiDi, uint32_t reorderingOptions);
943
944/**
945 * What are the reordering options applied to a given BiDi object?
946 *
947 * @param pBiDi is a <code>UBiDi</code> object.
948 * @return the current reordering options of the BiDi object
949 * @see ubidi_setReorderingOptions
950 * @draft ICU 3.6
951 */
952U_DRAFT uint32_t U_EXPORT2
953ubidi_getReorderingOptions(UBiDi *pBiDi);
954
b75a7d8f
A
955/**
956 * Perform the Unicode BiDi algorithm. It is defined in the
374ca955
A
957 * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Anned #9</a>,
958 * version 13,
959 * also described in The Unicode Standard, Version 4.0 .<p>
b75a7d8f 960 *
73c04bcf
A
961 * This function takes a piece of plain text containing one or more paragraphs,
962 * with or without externally specified embedding levels from <i>styled</i>
963 * text and computes the left-right-directionality of each character.<p>
b75a7d8f 964 *
73c04bcf 965 * If the entire text is all of the same directionality, then
b75a7d8f
A
966 * the function may not perform all the steps described by the algorithm,
967 * i.e., some levels may not be the same as if all steps were performed.
968 * This is not relevant for unidirectional text.<br>
969 * For example, in pure LTR text with numbers the numbers would get
970 * a resolved level of 2 higher than the surrounding text according to
971 * the algorithm. This implementation may set all resolved levels to
972 * the same value in such a case.<p>
973 *
73c04bcf
A
974 * The text can be composed of multiple paragraphs. Occurrence of a block
975 * separator in the text terminates a paragraph, and whatever comes next starts
976 * a new paragraph. The exception to this rule is when a Carriage Return (CR)
977 * is followed by a Line Feed (LF). Both CR and LF are block separators, but
978 * in that case, the pair of characters is considered as terminating the
979 * preceding paragraph, and a new paragraph will be started by a character
980 * coming after the LF.
b75a7d8f
A
981 *
982 * @param pBiDi A <code>UBiDi</code> object allocated with <code>ubidi_open()</code>
983 * which will be set to contain the reordering information,
984 * especially the resolved levels for all the characters in <code>text</code>.
985 *
73c04bcf 986 * @param text is a pointer to the text that the BiDi algorithm will be performed on.
b75a7d8f 987 * This pointer is stored in the UBiDi object and can be retrieved
73c04bcf
A
988 * with <code>ubidi_getText()</code>.<br>
989 * <strong>Note:</strong> the text must be (at least) <code>length</code> long.
b75a7d8f
A
990 *
991 * @param length is the length of the text; if <code>length==-1</code> then
992 * the text must be zero-terminated.
993 *
73c04bcf 994 * @param paraLevel specifies the default level for the text;
b75a7d8f
A
995 * it is typically 0 (LTR) or 1 (RTL).
996 * If the function shall determine the paragraph level from the text,
997 * then <code>paraLevel</code> can be set to
73c04bcf
A
998 * either <code>#UBIDI_DEFAULT_LTR</code>
999 * or <code>#UBIDI_DEFAULT_RTL</code>; if the text contains multiple
1000 * paragraphs, the paragraph level shall be determined separately for
1001 * each paragraph; if a paragraph does not include any strongly typed
1002 * character, then the desired default is used (0 for LTR or 1 for RTL).
1003 * Any other value between 0 and <code>#UBIDI_MAX_EXPLICIT_LEVEL</code>
1004 * is also valid, with odd levels indicating RTL.
b75a7d8f
A
1005 *
1006 * @param embeddingLevels (in) may be used to preset the embedding and override levels,
1007 * ignoring characters like LRE and PDF in the text.
1008 * A level overrides the directional property of its corresponding
1009 * (same index) character if the level has the
73c04bcf 1010 * <code>#UBIDI_LEVEL_OVERRIDE</code> bit set.<br><br>
b75a7d8f 1011 * Except for that bit, it must be
73c04bcf
A
1012 * <code>paraLevel<=embeddingLevels[]<=UBIDI_MAX_EXPLICIT_LEVEL</code>,
1013 * with one exception: a level of zero may be specified for a paragraph
1014 * separator even if <code>paraLevel>0</code> when multiple paragraphs
1015 * are submitted in the same call to <code>ubidi_setPara()</code>.<br><br>
b75a7d8f
A
1016 * <strong>Caution: </strong>A copy of this pointer, not of the levels,
1017 * will be stored in the <code>UBiDi</code> object;
1018 * the <code>embeddingLevels</code> array must not be
1019 * deallocated before the <code>UBiDi</code> structure is destroyed or reused,
1020 * and the <code>embeddingLevels</code>
1021 * should not be modified to avoid unexpected results on subsequent BiDi operations.
1022 * However, the <code>ubidi_setPara()</code> and
73c04bcf 1023 * <code>ubidi_setLine()</code> functions may modify some or all of the levels.<br><br>
b75a7d8f 1024 * After the <code>UBiDi</code> object is reused or destroyed, the caller
73c04bcf
A
1025 * must take care of the deallocation of the <code>embeddingLevels</code> array.<br><br>
1026 * <strong>Note:</strong> the <code>embeddingLevels</code> array must be
1027 * at least <code>length</code> long.
b75a7d8f 1028 *
73c04bcf 1029 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1030 * @stable ICU 2.0
1031 */
374ca955 1032U_STABLE void U_EXPORT2
b75a7d8f
A
1033ubidi_setPara(UBiDi *pBiDi, const UChar *text, int32_t length,
1034 UBiDiLevel paraLevel, UBiDiLevel *embeddingLevels,
1035 UErrorCode *pErrorCode);
1036
1037/**
1038 * <code>ubidi_setLine()</code> sets a <code>UBiDi</code> to
1039 * contain the reordering information, especially the resolved levels,
1040 * for all the characters in a line of text. This line of text is
1041 * specified by referring to a <code>UBiDi</code> object representing
73c04bcf
A
1042 * this information for a piece of text containing one or more paragraphs,
1043 * and by specifying a range of indexes in this text.<p>
374ca955 1044 * In the new line object, the indexes will range from 0 to <code>limit-start-1</code>.<p>
b75a7d8f
A
1045 *
1046 * This is used after calling <code>ubidi_setPara()</code>
73c04bcf
A
1047 * for a piece of text, and after line-breaking on that text.
1048 * It is not necessary if each paragraph is treated as a single line.<p>
b75a7d8f
A
1049 *
1050 * After line-breaking, rules (L1) and (L2) for the treatment of
1051 * trailing WS and for reordering are performed on
1052 * a <code>UBiDi</code> object that represents a line.<p>
1053 *
1054 * <strong>Important: </strong><code>pLineBiDi</code> shares data with
1055 * <code>pParaBiDi</code>.
1056 * You must destroy or reuse <code>pLineBiDi</code> before <code>pParaBiDi</code>.
1057 * In other words, you must destroy or reuse the <code>UBiDi</code> object for a line
1058 * before the object for its parent paragraph.<p>
1059 *
1060 * The text pointer that was stored in <code>pParaBiDi</code> is also copied,
1061 * and <code>start</code> is added to it so that it points to the beginning of the
1062 * line for this object.
1063 *
73c04bcf
A
1064 * @param pParaBiDi is the parent paragraph object. It must have been set
1065 * by a successful call to ubidi_setPara.
b75a7d8f 1066 *
73c04bcf 1067 * @param start is the line's first index into the text.
b75a7d8f 1068 *
73c04bcf 1069 * @param limit is just behind the line's last index into the text
b75a7d8f 1070 * (its last index +1).<br>
73c04bcf
A
1071 * It must be <code>0<=start<=limit<=</code>containing paragraph limit.
1072 * If the specified line crosses a paragraph boundary, the function
1073 * will terminate with error code U_ILLEGAL_ARGUMENT_ERROR.
b75a7d8f 1074 *
73c04bcf 1075 * @param pLineBiDi is the object that will now represent a line of the text.
b75a7d8f 1076 *
73c04bcf 1077 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1078 *
1079 * @see ubidi_setPara
73c04bcf 1080 * @see ubidi_getProcessedLength
b75a7d8f
A
1081 * @stable ICU 2.0
1082 */
374ca955 1083U_STABLE void U_EXPORT2
b75a7d8f
A
1084ubidi_setLine(const UBiDi *pParaBiDi,
1085 int32_t start, int32_t limit,
1086 UBiDi *pLineBiDi,
1087 UErrorCode *pErrorCode);
1088
1089/**
1090 * Get the directionality of the text.
1091 *
1092 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1093 *
1094 * @return A <code>UBIDI_XXX</code> value that indicates if the entire text
1095 * represented by this object is unidirectional,
1096 * and which direction, or if it is mixed-directional.
1097 *
1098 * @see UBiDiDirection
1099 * @stable ICU 2.0
1100 */
374ca955 1101U_STABLE UBiDiDirection U_EXPORT2
b75a7d8f
A
1102ubidi_getDirection(const UBiDi *pBiDi);
1103
1104/**
1105 * Get the pointer to the text.
1106 *
1107 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1108 *
1109 * @return The pointer to the text that the UBiDi object was created for.
1110 *
1111 * @see ubidi_setPara
1112 * @see ubidi_setLine
1113 * @stable ICU 2.0
1114 */
374ca955 1115U_STABLE const UChar * U_EXPORT2
b75a7d8f
A
1116ubidi_getText(const UBiDi *pBiDi);
1117
1118/**
1119 * Get the length of the text.
1120 *
1121 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1122 *
1123 * @return The length of the text that the UBiDi object was created for.
1124 * @stable ICU 2.0
1125 */
374ca955 1126U_STABLE int32_t U_EXPORT2
b75a7d8f
A
1127ubidi_getLength(const UBiDi *pBiDi);
1128
1129/**
1130 * Get the paragraph level of the text.
1131 *
1132 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1133 *
73c04bcf
A
1134 * @return The paragraph level. If there are multiple paragraphs, their
1135 * level may vary if the required paraLevel is UBIDI_DEFAULT_LTR or
1136 * UBIDI_DEFAULT_RTL. In that case, the level of the first paragraph
1137 * is returned.
b75a7d8f
A
1138 *
1139 * @see UBiDiLevel
73c04bcf
A
1140 * @see ubidi_getParagraph
1141 * @see ubidi_getParagraphByIndex
b75a7d8f
A
1142 * @stable ICU 2.0
1143 */
374ca955 1144U_STABLE UBiDiLevel U_EXPORT2
b75a7d8f
A
1145ubidi_getParaLevel(const UBiDi *pBiDi);
1146
73c04bcf
A
1147/**
1148 * Get the number of paragraphs.
1149 *
1150 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1151 *
1152 * @return The number of paragraphs.
1153 * @stable ICU 3.4
1154 */
1155U_STABLE int32_t U_EXPORT2
1156ubidi_countParagraphs(UBiDi *pBiDi);
1157
1158/**
1159 * Get a paragraph, given a position within the text.
1160 * This function returns information about a paragraph.<p>
1161 *
1162 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1163 *
1164 * @param charIndex is the index of a character within the text, in the
1165 * range <code>[0..ubidi_getProcessedLength(pBiDi)-1]</code>.
1166 *
1167 * @param pParaStart will receive the index of the first character of the
1168 * paragraph in the text.
1169 * This pointer can be <code>NULL</code> if this
1170 * value is not necessary.
1171 *
1172 * @param pParaLimit will receive the limit of the paragraph.
1173 * The l-value that you point to here may be the
1174 * same expression (variable) as the one for
1175 * <code>charIndex</code>.
1176 * This pointer can be <code>NULL</code> if this
1177 * value is not necessary.
1178 *
1179 * @param pParaLevel will receive the level of the paragraph.
1180 * This pointer can be <code>NULL</code> if this
1181 * value is not necessary.
1182 *
1183 * @param pErrorCode must be a valid pointer to an error code value.
1184 *
1185 * @return The index of the paragraph containing the specified position.
1186 *
1187 * @see ubidi_getProcessedLength
1188 * @stable ICU 3.4
1189 */
1190U_STABLE int32_t U_EXPORT2
1191ubidi_getParagraph(const UBiDi *pBiDi, int32_t charIndex, int32_t *pParaStart,
1192 int32_t *pParaLimit, UBiDiLevel *pParaLevel,
1193 UErrorCode *pErrorCode);
1194
1195/**
1196 * Get a paragraph, given the index of this paragraph.
1197 *
1198 * This function returns information about a paragraph.<p>
1199 *
1200 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1201 *
1202 * @param paraIndex is the number of the paragraph, in the
1203 * range <code>[0..ubidi_countParagraphs(pBiDi)-1]</code>.
1204 *
1205 * @param pParaStart will receive the index of the first character of the
1206 * paragraph in the text.
1207 * This pointer can be <code>NULL</code> if this
1208 * value is not necessary.
1209 *
1210 * @param pParaLimit will receive the limit of the paragraph.
1211 * This pointer can be <code>NULL</code> if this
1212 * value is not necessary.
1213 *
1214 * @param pParaLevel will receive the level of the paragraph.
1215 * This pointer can be <code>NULL</code> if this
1216 * value is not necessary.
1217 *
1218 * @param pErrorCode must be a valid pointer to an error code value.
1219 *
1220 * @stable ICU 3.4
1221 */
1222U_STABLE void U_EXPORT2
1223ubidi_getParagraphByIndex(const UBiDi *pBiDi, int32_t paraIndex,
1224 int32_t *pParaStart, int32_t *pParaLimit,
1225 UBiDiLevel *pParaLevel, UErrorCode *pErrorCode);
1226
b75a7d8f
A
1227/**
1228 * Get the level for one character.
1229 *
1230 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1231 *
1232 * @param charIndex the index of a character.
1233 *
1234 * @return The level for the character at charIndex.
1235 *
1236 * @see UBiDiLevel
73c04bcf 1237 * @see ubidi_getProcessedLength
b75a7d8f
A
1238 * @stable ICU 2.0
1239 */
374ca955 1240U_STABLE UBiDiLevel U_EXPORT2
b75a7d8f
A
1241ubidi_getLevelAt(const UBiDi *pBiDi, int32_t charIndex);
1242
1243/**
1244 * Get an array of levels for each character.<p>
1245 *
1246 * Note that this function may allocate memory under some
1247 * circumstances, unlike <code>ubidi_getLevelAt()</code>.
1248 *
73c04bcf
A
1249 * @param pBiDi is the paragraph or line <code>UBiDi</code> object, whose
1250 * text length must be strictly positive.
b75a7d8f 1251 *
73c04bcf 1252 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1253 *
1254 * @return The levels array for the text,
1255 * or <code>NULL</code> if an error occurs.
1256 *
1257 * @see UBiDiLevel
73c04bcf 1258 * @see ubidi_getProcessedLength
b75a7d8f
A
1259 * @stable ICU 2.0
1260 */
374ca955 1261U_STABLE const UBiDiLevel * U_EXPORT2
b75a7d8f
A
1262ubidi_getLevels(UBiDi *pBiDi, UErrorCode *pErrorCode);
1263
1264/**
1265 * Get a logical run.
1266 * This function returns information about a run and is used
1267 * to retrieve runs in logical order.<p>
1268 * This is especially useful for line-breaking on a paragraph.
1269 *
1270 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1271 *
1272 * @param logicalStart is the first character of the run.
1273 *
1274 * @param pLogicalLimit will receive the limit of the run.
1275 * The l-value that you point to here may be the
1276 * same expression (variable) as the one for
1277 * <code>logicalStart</code>.
1278 * This pointer can be <code>NULL</code> if this
1279 * value is not necessary.
1280 *
1281 * @param pLevel will receive the level of the run.
1282 * This pointer can be <code>NULL</code> if this
1283 * value is not necessary.
73c04bcf
A
1284 *
1285 * @see ubidi_getProcessedLength
b75a7d8f
A
1286 * @stable ICU 2.0
1287 */
374ca955 1288U_STABLE void U_EXPORT2
b75a7d8f
A
1289ubidi_getLogicalRun(const UBiDi *pBiDi, int32_t logicalStart,
1290 int32_t *pLogicalLimit, UBiDiLevel *pLevel);
1291
1292/**
1293 * Get the number of runs.
1294 * This function may invoke the actual reordering on the
1295 * <code>UBiDi</code> object, after <code>ubidi_setPara()</code>
1296 * may have resolved only the levels of the text. Therefore,
1297 * <code>ubidi_countRuns()</code> may have to allocate memory,
1298 * and may fail doing so.
1299 *
1300 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1301 *
73c04bcf 1302 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1303 *
1304 * @return The number of runs.
1305 * @stable ICU 2.0
1306 */
374ca955 1307U_STABLE int32_t U_EXPORT2
b75a7d8f
A
1308ubidi_countRuns(UBiDi *pBiDi, UErrorCode *pErrorCode);
1309
1310/**
1311 * Get one run's logical start, length, and directionality,
1312 * which can be 0 for LTR or 1 for RTL.
1313 * In an RTL run, the character at the logical start is
1314 * visually on the right of the displayed run.
1315 * The length is the number of characters in the run.<p>
1316 * <code>ubidi_countRuns()</code> should be called
1317 * before the runs are retrieved.
1318 *
1319 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1320 *
1321 * @param runIndex is the number of the run in visual order, in the
1322 * range <code>[0..ubidi_countRuns(pBiDi)-1]</code>.
1323 *
1324 * @param pLogicalStart is the first logical character index in the text.
1325 * The pointer may be <code>NULL</code> if this index is not needed.
1326 *
1327 * @param pLength is the number of characters (at least one) in the run.
1328 * The pointer may be <code>NULL</code> if this is not needed.
1329 *
1330 * @return the directionality of the run,
1331 * <code>UBIDI_LTR==0</code> or <code>UBIDI_RTL==1</code>,
1332 * never <code>UBIDI_MIXED</code>.
1333 *
1334 * @see ubidi_countRuns
1335 *
1336 * Example:
1337 * <pre>
1338 * \code
1339 * int32_t i, count=ubidi_countRuns(pBiDi),
1340 * logicalStart, visualIndex=0, length;
1341 * for(i=0; i<count; ++i) {
1342 * if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, i, &logicalStart, &length)) {
1343 * do { // LTR
1344 * show_char(text[logicalStart++], visualIndex++);
1345 * } while(--length>0);
1346 * } else {
1347 * logicalStart+=length; // logicalLimit
1348 * do { // RTL
1349 * show_char(text[--logicalStart], visualIndex++);
1350 * } while(--length>0);
1351 * }
1352 * }
1353 *\endcode
1354 * </pre>
1355 *
1356 * Note that in right-to-left runs, code like this places
1357 * modifier letters before base characters and second surrogates
1358 * before first ones.
1359 * @stable ICU 2.0
1360 */
374ca955 1361U_STABLE UBiDiDirection U_EXPORT2
b75a7d8f
A
1362ubidi_getVisualRun(UBiDi *pBiDi, int32_t runIndex,
1363 int32_t *pLogicalStart, int32_t *pLength);
1364
1365/**
1366 * Get the visual position from a logical text position.
1367 * If such a mapping is used many times on the same
1368 * <code>UBiDi</code> object, then calling
1369 * <code>ubidi_getLogicalMap()</code> is more efficient.<p>
1370 *
73c04bcf
A
1371 * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no
1372 * visual position because the corresponding text character is a BiDi control
1373 * removed from output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>.
1374 * <p>
b75a7d8f
A
1375 * Note that in right-to-left runs, this mapping places
1376 * modifier letters before base characters and second surrogates
1377 * before first ones.
1378 *
1379 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1380 *
1381 * @param logicalIndex is the index of a character in the text.
1382 *
73c04bcf 1383 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1384 *
1385 * @return The visual position of this character.
1386 *
1387 * @see ubidi_getLogicalMap
1388 * @see ubidi_getLogicalIndex
73c04bcf 1389 * @see ubidi_getProcessedLength
b75a7d8f
A
1390 * @stable ICU 2.0
1391 */
374ca955 1392U_STABLE int32_t U_EXPORT2
b75a7d8f
A
1393ubidi_getVisualIndex(UBiDi *pBiDi, int32_t logicalIndex, UErrorCode *pErrorCode);
1394
1395/**
1396 * Get the logical text position from a visual position.
1397 * If such a mapping is used many times on the same
1398 * <code>UBiDi</code> object, then calling
1399 * <code>ubidi_getVisualMap()</code> is more efficient.<p>
1400 *
73c04bcf
A
1401 * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no
1402 * logical position because the corresponding text character is a BiDi mark
1403 * inserted in the output by option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
1404 * <p>
b75a7d8f
A
1405 * This is the inverse function to <code>ubidi_getVisualIndex()</code>.
1406 *
1407 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1408 *
1409 * @param visualIndex is the visual position of a character.
1410 *
73c04bcf 1411 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1412 *
1413 * @return The index of this character in the text.
1414 *
1415 * @see ubidi_getVisualMap
1416 * @see ubidi_getVisualIndex
73c04bcf 1417 * @see ubidi_getResultLength
b75a7d8f
A
1418 * @stable ICU 2.0
1419 */
374ca955 1420U_STABLE int32_t U_EXPORT2
b75a7d8f
A
1421ubidi_getLogicalIndex(UBiDi *pBiDi, int32_t visualIndex, UErrorCode *pErrorCode);
1422
1423/**
1424 * Get a logical-to-visual index map (array) for the characters in the UBiDi
1425 * (paragraph or line) object.
73c04bcf
A
1426 * <p>
1427 * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the
1428 * corresponding text characters are BiDi controls removed from the visual
1429 * output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>.
b75a7d8f
A
1430 *
1431 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1432 *
73c04bcf 1433 * @param indexMap is a pointer to an array of <code>ubidi_getProcessedLength()</code>
b75a7d8f 1434 * indexes which will reflect the reordering of the characters.
73c04bcf
A
1435 * If option <code>#UBIDI_OPTION_INSERT_MARKS</code> is set, the number
1436 * of elements allocated in <code>indexMap</code> must be no less than
1437 * <code>ubidi_getResultLength()</code>.
1438 * The array does not need to be initialized.<br><br>
1439 * The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>.
b75a7d8f 1440 *
73c04bcf 1441 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1442 *
1443 * @see ubidi_getVisualMap
1444 * @see ubidi_getVisualIndex
73c04bcf
A
1445 * @see ubidi_getProcessedLength
1446 * @see ubidi_getResultLength
b75a7d8f
A
1447 * @stable ICU 2.0
1448 */
374ca955 1449U_STABLE void U_EXPORT2
b75a7d8f
A
1450ubidi_getLogicalMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode);
1451
1452/**
1453 * Get a visual-to-logical index map (array) for the characters in the UBiDi
1454 * (paragraph or line) object.
73c04bcf
A
1455 * <p>
1456 * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the
1457 * corresponding text characters are BiDi marks inserted in the visual output
1458 * by the option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
b75a7d8f
A
1459 *
1460 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1461 *
73c04bcf 1462 * @param indexMap is a pointer to an array of <code>ubidi_getResultLength()</code>
b75a7d8f 1463 * indexes which will reflect the reordering of the characters.
73c04bcf
A
1464 * If option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is set, the number
1465 * of elements allocated in <code>indexMap</code> must be no less than
1466 * <code>ubidi_getProcessedLength()</code>.
1467 * The array does not need to be initialized.<br><br>
1468 * The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>.
b75a7d8f 1469 *
73c04bcf 1470 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1471 *
1472 * @see ubidi_getLogicalMap
1473 * @see ubidi_getLogicalIndex
73c04bcf
A
1474 * @see ubidi_getProcessedLength
1475 * @see ubidi_getResultLength
b75a7d8f
A
1476 * @stable ICU 2.0
1477 */
374ca955 1478U_STABLE void U_EXPORT2
b75a7d8f
A
1479ubidi_getVisualMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode);
1480
1481/**
1482 * This is a convenience function that does not use a UBiDi object.
1483 * It is intended to be used for when an application has determined the levels
1484 * of objects (character sequences) and just needs to have them reordered (L2).
73c04bcf 1485 * This is equivalent to using <code>ubidi_getLogicalMap()</code> on a
b75a7d8f
A
1486 * <code>UBiDi</code> object.
1487 *
1488 * @param levels is an array with <code>length</code> levels that have been determined by
1489 * the application.
1490 *
1491 * @param length is the number of levels in the array, or, semantically,
1492 * the number of objects to be reordered.
1493 * It must be <code>length>0</code>.
1494 *
1495 * @param indexMap is a pointer to an array of <code>length</code>
1496 * indexes which will reflect the reordering of the characters.
1497 * The array does not need to be initialized.<p>
1498 * The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>.
1499 * @stable ICU 2.0
1500 */
374ca955 1501U_STABLE void U_EXPORT2
b75a7d8f
A
1502ubidi_reorderLogical(const UBiDiLevel *levels, int32_t length, int32_t *indexMap);
1503
1504/**
1505 * This is a convenience function that does not use a UBiDi object.
1506 * It is intended to be used for when an application has determined the levels
1507 * of objects (character sequences) and just needs to have them reordered (L2).
73c04bcf 1508 * This is equivalent to using <code>ubidi_getVisualMap()</code> on a
b75a7d8f
A
1509 * <code>UBiDi</code> object.
1510 *
1511 * @param levels is an array with <code>length</code> levels that have been determined by
1512 * the application.
1513 *
1514 * @param length is the number of levels in the array, or, semantically,
1515 * the number of objects to be reordered.
1516 * It must be <code>length>0</code>.
1517 *
1518 * @param indexMap is a pointer to an array of <code>length</code>
1519 * indexes which will reflect the reordering of the characters.
1520 * The array does not need to be initialized.<p>
1521 * The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>.
1522 * @stable ICU 2.0
1523 */
374ca955 1524U_STABLE void U_EXPORT2
b75a7d8f
A
1525ubidi_reorderVisual(const UBiDiLevel *levels, int32_t length, int32_t *indexMap);
1526
1527/**
1528 * Invert an index map.
73c04bcf 1529 * The index mapping of the first map is inverted and written to
b75a7d8f
A
1530 * the second one.
1531 *
1532 * @param srcMap is an array with <code>length</code> indexes
73c04bcf
A
1533 * which defines the original mapping from a source array containing
1534 * <code>length</code> elements to a destination array.
1535 * All indexes must be >=0 or equal to <code>UBIDI_MAP_NOWHERE</code>.
1536 * This special value means that the corresponding elements in the source
1537 * array have no matching element in the destination array.
1538 * Some indexes may have a value >= <code>length</code>, if the
1539 * destination array has more elements than the source array.
1540 * There must be no duplicate indexes (two or more indexes with the
1541 * same value except <code>UBIDI_MAP_NOWHERE</code>).
1542 *
1543 * @param destMap is an array with a number of indexes equal to 1 + the highest
1544 * value in <code>srcMap</code>.
1545 * <code>destMap</code> will be filled with the inverse mapping.
1546 * Elements of <code>destMap</code> which have no matching elements in
1547 * <code>srcMap</code> will receive an index equal to
1548 * <code>UBIDI_MAP_NOWHERE</code>
b75a7d8f
A
1549 *
1550 * @param length is the length of each array.
73c04bcf 1551 * @See UBIDI_MAP_NOWHERE
b75a7d8f
A
1552 * @stable ICU 2.0
1553 */
374ca955 1554U_STABLE void U_EXPORT2
b75a7d8f
A
1555ubidi_invertMap(const int32_t *srcMap, int32_t *destMap, int32_t length);
1556
1557/** option flags for ubidi_writeReordered() */
1558
1559/**
1560 * option bit for ubidi_writeReordered():
1561 * keep combining characters after their base characters in RTL runs
1562 *
1563 * @see ubidi_writeReordered
1564 * @stable ICU 2.0
1565 */
1566#define UBIDI_KEEP_BASE_COMBINING 1
1567
1568/**
1569 * option bit for ubidi_writeReordered():
1570 * replace characters with the "mirrored" property in RTL runs
1571 * by their mirror-image mappings
1572 *
1573 * @see ubidi_writeReordered
1574 * @stable ICU 2.0
1575 */
1576#define UBIDI_DO_MIRRORING 2
1577
1578/**
1579 * option bit for ubidi_writeReordered():
1580 * surround the run with LRMs if necessary;
1581 * this is part of the approximate "inverse BiDi" algorithm
1582 *
73c04bcf
A
1583 * <p>This option does not imply corresponding adjustment of the index
1584 * mappings.</p>
1585 *
b75a7d8f
A
1586 * @see ubidi_setInverse
1587 * @see ubidi_writeReordered
1588 * @stable ICU 2.0
1589 */
1590#define UBIDI_INSERT_LRM_FOR_NUMERIC 4
1591
1592/**
1593 * option bit for ubidi_writeReordered():
1594 * remove BiDi control characters
73c04bcf
A
1595 * (this does not affect #UBIDI_INSERT_LRM_FOR_NUMERIC)
1596 *
1597 * <p>This option does not imply corresponding adjustment of the index
1598 * mappings.</p>
b75a7d8f
A
1599 *
1600 * @see ubidi_writeReordered
1601 * @stable ICU 2.0
1602 */
1603#define UBIDI_REMOVE_BIDI_CONTROLS 8
1604
1605/**
1606 * option bit for ubidi_writeReordered():
1607 * write the output in reverse order
1608 *
1609 * <p>This has the same effect as calling <code>ubidi_writeReordered()</code>
1610 * first without this option, and then calling
1611 * <code>ubidi_writeReverse()</code> without mirroring.
1612 * Doing this in the same step is faster and avoids a temporary buffer.
1613 * An example for using this option is output to a character terminal that
1614 * is designed for RTL scripts and stores text in reverse order.</p>
1615 *
1616 * @see ubidi_writeReordered
1617 * @stable ICU 2.0
1618 */
1619#define UBIDI_OUTPUT_REVERSE 16
1620
73c04bcf
A
1621/**
1622 * Get the length of the source text processed by the last call to
1623 * <code>ubidi_setPara()</code>. This length may be different from the length
1624 * of the source text if option <code>#UBIDI_OPTION_STREAMING</code>
1625 * has been set.
1626 * <br>
1627 * Note that whenever the length of the text affects the execution or the
1628 * result of a function, it is the processed length which must be considered,
1629 * except for <code>ubidi_setPara</code> (which receives unprocessed source
1630 * text) and <code>ubidi_getLength</code> (which returns the original length
1631 * of the source text).<br>
1632 * In particular, the processed length is the one to consider in the following
1633 * cases:
1634 * <ul>
1635 * <li>maximum value of the <code>limit</code> argument of
1636 * <code>ubidi_setLine</code></li>
1637 * <li>maximum value of the <code>charIndex</code> argument of
1638 * <code>ubidi_getParagraph</code></li>
1639 * <li>maximum value of the <code>charIndex</code> argument of
1640 * <code>ubidi_getLevelAt</code></li>
1641 * <li>number of elements in the array returned by <code>ubidi_getLevels</code></li>
1642 * <li>maximum value of the <code>logicalStart</code> argument of
1643 * <code>ubidi_getLogicalRun</code></li>
1644 * <li>maximum value of the <code>logicalIndex</code> argument of
1645 * <code>ubidi_getVisualIndex</code></li>
1646 * <li>number of elements filled in the <code>*indexMap</code> argument of
1647 * <code>ubidi_getLogicalMap</code></li>
1648 * <li>length of text processed by <code>ubidi_writeReordered</code></li>
1649 * </ul>
1650 *
1651 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1652 *
1653 * @return The length of the part of the source text processed by
1654 * the last call to <code>ubidi_setPara</code>.
1655 * @see ubidi_setPara
1656 * @see UBIDI_OPTION_STREAMING
1657 * @draft ICU 3.6
1658 */
1659U_DRAFT int32_t U_EXPORT2
1660ubidi_getProcessedLength(const UBiDi *pBiDi);
1661
1662/**
1663 * Get the length of the reordered text resulting from the last call to
1664 * <code>ubidi_setPara()</code>. This length may be different from the length
1665 * of the source text if option <code>#UBIDI_OPTION_INSERT_MARKS</code>
1666 * or option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> has been set.
1667 * <br>
1668 * This resulting length is the one to consider in the following cases:
1669 * <ul>
1670 * <li>maximum value of the <code>visualIndex</code> argument of
1671 * <code>ubidi_getLogicalIndex</code></li>
1672 * <li>number of elements of the <code>*indexMap</code> argument of
1673 * <code>ubidi_getVisualMap</code></li>
1674 * </ul>
1675 * Note that this length stays identical to the source text length if
1676 * BiDi marks are inserted or removed using option bits of
1677 * <code>ubidi_writeReordered</code>, or if option
1678 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> has been set.
1679 *
1680 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1681 *
1682 * @return The length of the reordered text resulting from
1683 * the last call to <code>ubidi_setPara</code>.
1684 * @see ubidi_setPara
1685 * @see UBIDI_OPTION_INSERT_MARKS
1686 * @see UBIDI_OPTION_REMOVE_CONTROLS
1687 * @draft ICU 3.6
1688 */
1689U_DRAFT int32_t U_EXPORT2
1690ubidi_getResultLength(const UBiDi *pBiDi);
1691
1692U_CDECL_BEGIN
1693/**
1694 * value returned by <code>UBiDiClassCallback</code> callbacks when
1695 * there is no need to override the standard BiDi class for a given code point.
1696 * @see UBiDiClassCallback
1697 * @draft ICU 3.6
1698 */
1699#define U_BIDI_CLASS_DEFAULT U_CHAR_DIRECTION_COUNT
1700
1701/**
1702 * Callback type declaration for overriding default BiDi class values with
1703 * custom ones.
1704 * <p>Usually, the function pointer will be propagated to a <code>UBiDi</code>
1705 * object by calling the <code>ubidi_setClassCallback()</code> function;
1706 * then the callback will be invoked by the UBA implementation any time the
1707 * class of a character is to be determined.</p>
1708 *
1709 * @param context is a pointer to the callback private data.
1710 *
1711 * @param c is the code point to get a BiDi class for.
1712 *
1713 * @return The directional property / BiDi class for the given code point
1714 * <code>c</code> if the default class has been overridden, or
1715 * <code>#U_BIDI_CLASS_DEFAULT</code> if the standard BiDi class value
1716 * for <code>c</code> is to be used.
1717 * @see ubidi_setClassCallback
1718 * @see ubidi_getClassCallback
1719 * @draft ICU 3.6
1720 */
1721typedef UCharDirection U_CALLCONV
1722UBiDiClassCallback(const void *context, UChar32 c);
1723
1724U_CDECL_END
1725
1726/**
1727 * Retrieve the BiDi class for a given code point.
1728 * <p>If a <code>#UBiDiClassCallback</code> callback is defined and returns a
1729 * value other than <code>#U_BIDI_CLASS_DEFAULT</code>, that value is used;
1730 * otherwise the default class determination mechanism is invoked.</p>
1731 *
1732 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1733 *
1734 * @param c is the code point whose BiDi class must be retrieved.
1735 *
1736 * @return The BiDi class for character <code>c</code> based
1737 * on the given <code>pBiDi</code> instance.
1738 * @see UBiDiClassCallback
1739 * @draft ICU 3.6
1740 */
1741U_DRAFT UCharDirection U_EXPORT2
1742ubidi_getCustomizedClass(UBiDi *pBiDi, UChar32 c);
1743
1744/**
1745 * Set the callback function and callback data used by the UBA
1746 * implementation for BiDi class determination.
1747 * <p>This may be useful for assigning BiDi classes to PUA characters, or
1748 * for special application needs. For instance, an application may want to
1749 * handle all spaces like L or R characters (according to the base direction)
1750 * when creating the visual ordering of logical lines which are part of a report
1751 * organized in columns: there should not be interaction between adjacent
1752 * cells.<p>
1753 *
1754 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1755 *
1756 * @param newFn is the new callback function pointer.
1757 *
1758 * @param newContext is the new callback context pointer. This can be NULL.
1759 *
1760 * @param oldFn fillin: Returns the old callback function pointer. This can be
1761 * NULL.
1762 *
1763 * @param oldContext fillin: Returns the old callback's context. This can be
1764 * NULL.
1765 *
1766 * @param pErrorCode must be a valid pointer to an error code value.
1767 *
1768 * @see ubidi_getClassCallback
1769 * @draft ICU 3.6
1770 */
1771U_DRAFT void U_EXPORT2
1772ubidi_setClassCallback(UBiDi *pBiDi, UBiDiClassCallback *newFn,
1773 const void *newContext, UBiDiClassCallback **oldFn,
1774 const void **oldContext, UErrorCode *pErrorCode);
1775
1776/**
1777 * Get the current callback function used for BiDi class determination.
1778 *
1779 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1780 *
1781 * @param fn fillin: Returns the callback function pointer.
1782 *
1783 * @param context fillin: Returns the callback's private context.
1784 *
1785 * @see ubidi_setClassCallback
1786 * @draft ICU 3.6
1787 */
1788U_DRAFT void U_EXPORT2
1789ubidi_getClassCallback(UBiDi *pBiDi, UBiDiClassCallback **fn, const void **context);
1790
b75a7d8f
A
1791/**
1792 * Take a <code>UBiDi</code> object containing the reordering
73c04bcf
A
1793 * information for a piece of text (one or more paragraphs) set by
1794 * <code>ubidi_setPara()</code> or for a line of text set by
1795 * <code>ubidi_setLine()</code> and write a reordered string to the
1796 * destination buffer.
b75a7d8f
A
1797 *
1798 * This function preserves the integrity of characters with multiple
1799 * code units and (optionally) modifier letters.
1800 * Characters in RTL runs can be replaced by mirror-image characters
1801 * in the destination buffer. Note that "real" mirroring has
1802 * to be done in a rendering engine by glyph selection
1803 * and that for many "mirrored" characters there are no
1804 * Unicode characters as mirror-image equivalents.
1805 * There are also options to insert or remove BiDi control
1806 * characters; see the description of the <code>destSize</code>
1807 * and <code>options</code> parameters and of the option bit flags.
1808 *
1809 * @param pBiDi A pointer to a <code>UBiDi</code> object that
1810 * is set by <code>ubidi_setPara()</code> or
1811 * <code>ubidi_setLine()</code> and contains the reordering
1812 * information for the text that it was defined for,
73c04bcf
A
1813 * as well as a pointer to that text.<br><br>
1814 * The text was aliased (only the pointer was stored
b75a7d8f 1815 * without copying the contents) and must not have been modified
73c04bcf 1816 * since the <code>ubidi_setPara()</code> call.
b75a7d8f
A
1817 *
1818 * @param dest A pointer to where the reordered text is to be copied.
1819 * The source text and <code>dest[destSize]</code>
1820 * must not overlap.
1821 *
1822 * @param destSize The size of the <code>dest</code> buffer,
1823 * in number of UChars.
1824 * If the <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>
1825 * option is set, then the destination length could be
1826 * as large as
1827 * <code>ubidi_getLength(pBiDi)+2*ubidi_countRuns(pBiDi)</code>.
1828 * If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option
1829 * is set, then the destination length may be less than
1830 * <code>ubidi_getLength(pBiDi)</code>.
1831 * If none of these options is set, then the destination length
1832 * will be exactly <code>ubidi_getLength(pBiDi)</code>.
1833 *
1834 * @param options A bit set of options for the reordering that control
1835 * how the reordered text is written.
1836 * The options include mirroring the characters on a code
1837 * point basis and inserting LRM characters, which is used
1838 * especially for transforming visually stored text
1839 * to logically stored text (although this is still an
1840 * imperfect implementation of an "inverse BiDi" algorithm
1841 * because it uses the "forward BiDi" algorithm at its core).
374ca955
A
1842 * The available options are:
1843 * <code>#UBIDI_DO_MIRRORING</code>,
1844 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1845 * <code>#UBIDI_KEEP_BASE_COMBINING</code>,
1846 * <code>#UBIDI_OUTPUT_REVERSE</code>,
1847 * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code>
b75a7d8f 1848 *
73c04bcf 1849 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1850 *
1851 * @return The length of the output string.
73c04bcf
A
1852 *
1853 * @see ubidi_getProcessedLength
b75a7d8f
A
1854 * @stable ICU 2.0
1855 */
374ca955 1856U_STABLE int32_t U_EXPORT2
b75a7d8f
A
1857ubidi_writeReordered(UBiDi *pBiDi,
1858 UChar *dest, int32_t destSize,
1859 uint16_t options,
1860 UErrorCode *pErrorCode);
1861
1862/**
1863 * Reverse a Right-To-Left run of Unicode text.
1864 *
1865 * This function preserves the integrity of characters with multiple
1866 * code units and (optionally) modifier letters.
1867 * Characters can be replaced by mirror-image characters
1868 * in the destination buffer. Note that "real" mirroring has
1869 * to be done in a rendering engine by glyph selection
1870 * and that for many "mirrored" characters there are no
1871 * Unicode characters as mirror-image equivalents.
1872 * There are also options to insert or remove BiDi control
1873 * characters.
1874 *
1875 * This function is the implementation for reversing RTL runs as part
1876 * of <code>ubidi_writeReordered()</code>. For detailed descriptions
1877 * of the parameters, see there.
1878 * Since no BiDi controls are inserted here, the output string length
1879 * will never exceed <code>srcLength</code>.
1880 *
1881 * @see ubidi_writeReordered
1882 *
1883 * @param src A pointer to the RTL run text.
1884 *
1885 * @param srcLength The length of the RTL run.
1886 *
1887 * @param dest A pointer to where the reordered text is to be copied.
1888 * <code>src[srcLength]</code> and <code>dest[destSize]</code>
1889 * must not overlap.
1890 *
1891 * @param destSize The size of the <code>dest</code> buffer,
1892 * in number of UChars.
1893 * If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option
1894 * is set, then the destination length may be less than
1895 * <code>srcLength</code>.
1896 * If this option is not set, then the destination length
1897 * will be exactly <code>srcLength</code>.
1898 *
1899 * @param options A bit set of options for the reordering that control
1900 * how the reordered text is written.
374ca955 1901 * See the <code>options</code> parameter in <code>ubidi_writeReordered()</code>.
b75a7d8f 1902 *
73c04bcf 1903 * @param pErrorCode must be a valid pointer to an error code value.
b75a7d8f
A
1904 *
1905 * @return The length of the output string.
1906 * @stable ICU 2.0
1907 */
374ca955 1908U_STABLE int32_t U_EXPORT2
b75a7d8f
A
1909ubidi_writeReverse(const UChar *src, int32_t srcLength,
1910 UChar *dest, int32_t destSize,
1911 uint16_t options,
1912 UErrorCode *pErrorCode);
1913
1914/*#define BIDI_SAMPLE_CODE*/
1915/*@}*/
1916
1917#endif