1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
4 ******************************************************************************
6 * Copyright (C) 1999-2015, International Business Machines
7 * Corporation and others. All Rights Reserved.
9 ******************************************************************************
12 * tab size: 8 (not used)
15 * created on: 1999jul27
16 * created by: Markus W. Scherer, updated by Matitiahu Allouche
22 #include "unicode/utypes.h"
23 #include "unicode/uchar.h"
24 #include "unicode/localpointer.h"
28 * \brief C API: Bidi algorithm
30 * <h2>Bidi algorithm for ICU</h2>
32 * This is an implementation of the Unicode Bidirectional Algorithm.
33 * The algorithm is defined in the
34 * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Annex #9</a>.<p>
36 * Note: Libraries that perform a bidirectional algorithm and
37 * reorder strings accordingly are sometimes called "Storage Layout Engines".
38 * ICU's Bidi and shaping (u_shapeArabic()) APIs can be used at the core of such
39 * "Storage Layout Engines".
41 * <h3>General remarks about the API:</h3>
43 * In functions with an error code parameter,
44 * the <code>pErrorCode</code> pointer must be valid
45 * and the value that it points to must not indicate a failure before
46 * the function call. Otherwise, the function returns immediately.
47 * After the function call, the value indicates success or failure.<p>
49 * The "limit" of a sequence of characters is the position just after their
50 * last character, i.e., one more than that position.<p>
52 * Some of the API functions provide access to "runs".
53 * Such a "run" is defined as a sequence of characters
54 * that are at the same embedding level
55 * after performing the Bidi algorithm.<p>
57 * @author Markus W. Scherer
61 * <h4> Sample code for the ICU Bidi API </h4>
63 * <h5>Rendering a paragraph with the ICU Bidi API</h5>
65 * This is (hypothetical) sample code that illustrates
66 * how the ICU Bidi API could be used to render a paragraph of text.
67 * Rendering code depends highly on the graphics system,
68 * therefore this sample code must make a lot of assumptions,
69 * which may or may not match any existing graphics system's properties.
71 * <p>The basic assumptions are:</p>
73 * <li>Rendering is done from left to right on a horizontal line.</li>
74 * <li>A run of single-style, unidirectional text can be rendered at once.</li>
75 * <li>Such a run of text is passed to the graphics system with
76 * characters (code units) in logical order.</li>
77 * <li>The line-breaking algorithm is very complicated
78 * and Locale-dependent -
79 * and therefore its implementation omitted from this sample code.</li>
84 *#include "unicode/ubidi.h"
87 * styleNormal=0, styleSelected=1,
88 * styleBold=2, styleItalics=4,
89 * styleSuper=8, styleSub=16
92 *typedef struct { int32_t limit; Style style; } StyleRun;
94 *int getTextWidth(const UChar *text, int32_t start, int32_t limit,
95 * const StyleRun *styleRuns, int styleRunCount);
97 * // set *pLimit and *pStyleRunLimit for a line
98 * // from text[start] and from styleRuns[styleRunStart]
99 * // using ubidi_getLogicalRun(para, ...)
100 *void getLineBreak(const UChar *text, int32_t start, int32_t *pLimit,
102 * const StyleRun *styleRuns, int styleRunStart, int *pStyleRunLimit,
105 * // render runs on a line sequentially, always from left to right
107 * // prepare rendering a new line
108 * void startLine(UBiDiDirection textDirection, int lineWidth);
110 * // render a run of text and advance to the right by the run width
111 * // the text[start..limit-1] is always in logical order
112 * void renderRun(const UChar *text, int32_t start, int32_t limit,
113 * UBiDiDirection textDirection, Style style);
115 * // We could compute a cross-product
116 * // from the style runs with the directional runs
117 * // and then reorder it.
118 * // Instead, here we iterate over each run type
119 * // and render the intersections -
120 * // with shortcuts in simple (and common) cases.
121 * // renderParagraph() is the main function.
123 * // render a directional run with
124 * // (possibly) multiple style runs intersecting with it
125 * void renderDirectionalRun(const UChar *text,
126 * int32_t start, int32_t limit,
127 * UBiDiDirection direction,
128 * const StyleRun *styleRuns, int styleRunCount) {
131 * // iterate over style runs
132 * if(direction==UBIDI_LTR) {
135 * for(i=0; i<styleRunCount; ++i) {
136 * styleLimit=styleRun[i].limit;
137 * if(start<styleLimit) {
138 * if(styleLimit>limit) { styleLimit=limit; }
139 * renderRun(text, start, styleLimit,
140 * direction, styleRun[i].style);
141 * if(styleLimit==limit) { break; }
148 * for(i=styleRunCount-1; i>=0; --i) {
150 * styleStart=styleRun[i-1].limit;
154 * if(limit>=styleStart) {
155 * if(styleStart<start) { styleStart=start; }
156 * renderRun(text, styleStart, limit,
157 * direction, styleRun[i].style);
158 * if(styleStart==start) { break; }
165 * // the line object represents text[start..limit-1]
166 * void renderLine(UBiDi *line, const UChar *text,
167 * int32_t start, int32_t limit,
168 * const StyleRun *styleRuns, int styleRunCount) {
169 * UBiDiDirection direction=ubidi_getDirection(line);
170 * if(direction!=UBIDI_MIXED) {
172 * if(styleRunCount<=1) {
173 * renderRun(text, start, limit, direction, styleRuns[0].style);
175 * renderDirectionalRun(text, start, limit,
176 * direction, styleRuns, styleRunCount);
179 * // mixed-directional
180 * int32_t count, i, length;
183 * count=ubidi_countRuns(para, pErrorCode);
184 * if(U_SUCCESS(*pErrorCode)) {
185 * if(styleRunCount<=1) {
186 * Style style=styleRuns[0].style;
188 * // iterate over directional runs
189 * for(i=0; i<count; ++i) {
190 * direction=ubidi_getVisualRun(para, i, &start, &length);
191 * renderRun(text, start, start+length, direction, style);
196 * // iterate over both directional and style runs
197 * for(i=0; i<count; ++i) {
198 * direction=ubidi_getVisualRun(line, i, &start, &length);
199 * renderDirectionalRun(text, start, start+length,
200 * direction, styleRuns, styleRunCount);
207 *void renderParagraph(const UChar *text, int32_t length,
208 * UBiDiDirection textDirection,
209 * const StyleRun *styleRuns, int styleRunCount,
211 * UErrorCode *pErrorCode) {
214 * if(pErrorCode==NULL || U_FAILURE(*pErrorCode) || length<=0) {
218 * para=ubidi_openSized(length, 0, pErrorCode);
219 * if(para==NULL) { return; }
221 * ubidi_setPara(para, text, length,
222 * textDirection ? UBIDI_DEFAULT_RTL : UBIDI_DEFAULT_LTR,
224 * if(U_SUCCESS(*pErrorCode)) {
225 * UBiDiLevel paraLevel=1&ubidi_getParaLevel(para);
226 * StyleRun styleRun={ length, styleNormal };
229 * if(styleRuns==NULL || styleRunCount<=0) {
231 * styleRuns=&styleRun;
234 * // assume styleRuns[styleRunCount-1].limit>=length
236 * width=getTextWidth(text, 0, length, styleRuns, styleRunCount);
237 * if(width<=lineWidth) {
238 * // everything fits onto one line
240 * // prepare rendering a new line from either left or right
241 * startLine(paraLevel, width);
243 * renderLine(para, text, 0, length,
244 * styleRuns, styleRunCount);
248 * // we need to render several lines
249 * line=ubidi_openSized(length, 0, pErrorCode);
251 * int32_t start=0, limit;
252 * int styleRunStart=0, styleRunLimit;
256 * styleRunLimit=styleRunCount;
257 * getLineBreak(text, start, &limit, para,
258 * styleRuns, styleRunStart, &styleRunLimit,
260 * ubidi_setLine(para, start, limit, line, pErrorCode);
261 * if(U_SUCCESS(*pErrorCode)) {
262 * // prepare rendering a new line
263 * // from either left or right
264 * startLine(paraLevel, width);
266 * renderLine(line, text, start, limit,
267 * styleRuns+styleRunStart,
268 * styleRunLimit-styleRunStart);
270 * if(limit==length) { break; }
272 * styleRunStart=styleRunLimit-1;
273 * if(start>=styleRuns[styleRunStart].limit) {
293 * UBiDiLevel is the type of the level values in this
294 * Bidi implementation.
295 * It holds an embedding level and indicates the visual direction
296 * by its bit 0 (even/odd value).<p>
298 * It can also hold non-level values for the
299 * <code>paraLevel</code> and <code>embeddingLevels</code>
300 * arguments of <code>ubidi_setPara()</code>; there:
302 * <li>bit 7 of an <code>embeddingLevels[]</code>
303 * value indicates whether the using application is
304 * specifying the level of a character to <i>override</i> whatever the
305 * Bidi implementation would resolve it to.</li>
306 * <li><code>paraLevel</code> can be set to the
307 * pseudo-level values <code>UBIDI_DEFAULT_LTR</code>
308 * and <code>UBIDI_DEFAULT_RTL</code>.</li>
313 * <p>The related constants are not real, valid level values.
314 * <code>UBIDI_DEFAULT_XXX</code> can be used to specify
315 * a default for the paragraph level for
316 * when the <code>ubidi_setPara()</code> function
317 * shall determine it but there is no
318 * strongly typed character in the input.<p>
320 * Note that the value for <code>UBIDI_DEFAULT_LTR</code> is even
321 * and the one for <code>UBIDI_DEFAULT_RTL</code> is odd,
322 * just like with normal LTR and RTL level values -
323 * these special values are designed that way. Also, the implementation
324 * assumes that UBIDI_MAX_EXPLICIT_LEVEL is odd.
326 * Note: The numeric values of the related constants will not change:
327 * They are tied to the use of 7-bit byte values (plus the override bit)
328 * and of the UBiDiLevel=uint8_t data type in this API.
330 * @see UBIDI_DEFAULT_LTR
331 * @see UBIDI_DEFAULT_RTL
332 * @see UBIDI_LEVEL_OVERRIDE
333 * @see UBIDI_MAX_EXPLICIT_LEVEL
336 typedef uint8_t UBiDiLevel
;
338 /** Paragraph level setting.<p>
340 * Constant indicating that the base direction depends on the first strong
341 * directional character in the text according to the Unicode Bidirectional
342 * Algorithm. If no strong directional character is present,
343 * then set the paragraph level to 0 (left-to-right).<p>
345 * If this value is used in conjunction with reordering modes
346 * <code>UBIDI_REORDER_INVERSE_LIKE_DIRECT</code> or
347 * <code>UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the text to reorder
348 * is assumed to be visual LTR, and the text after reordering is required
349 * to be the corresponding logical string with appropriate contextual
350 * direction. The direction of the result string will be RTL if either
351 * the righmost or leftmost strong character of the source text is RTL
352 * or Arabic Letter, the direction will be LTR otherwise.<p>
354 * If reordering option <code>UBIDI_OPTION_INSERT_MARKS</code> is set, an RLM may
355 * be added at the beginning of the result string to ensure round trip
356 * (that the result string, when reordered back to visual, will produce
357 * the original source text).
358 * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT
359 * @see UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL
362 #define UBIDI_DEFAULT_LTR 0xfe
364 /** Paragraph level setting.<p>
366 * Constant indicating that the base direction depends on the first strong
367 * directional character in the text according to the Unicode Bidirectional
368 * Algorithm. If no strong directional character is present,
369 * then set the paragraph level to 1 (right-to-left).<p>
371 * If this value is used in conjunction with reordering modes
372 * <code>UBIDI_REORDER_INVERSE_LIKE_DIRECT</code> or
373 * <code>UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the text to reorder
374 * is assumed to be visual LTR, and the text after reordering is required
375 * to be the corresponding logical string with appropriate contextual
376 * direction. The direction of the result string will be RTL if either
377 * the righmost or leftmost strong character of the source text is RTL
378 * or Arabic Letter, or if the text contains no strong character;
379 * the direction will be LTR otherwise.<p>
381 * If reordering option <code>UBIDI_OPTION_INSERT_MARKS</code> is set, an RLM may
382 * be added at the beginning of the result string to ensure round trip
383 * (that the result string, when reordered back to visual, will produce
384 * the original source text).
385 * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT
386 * @see UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL
389 #define UBIDI_DEFAULT_RTL 0xff
392 * Maximum explicit embedding level.
393 * Same as the max_depth value in the
394 * <a href="http://www.unicode.org/reports/tr9/#BD2">Unicode Bidirectional Algorithm</a>.
395 * (The maximum resolved level can be up to <code>UBIDI_MAX_EXPLICIT_LEVEL+1</code>).
398 #define UBIDI_MAX_EXPLICIT_LEVEL 125
400 /** Bit flag for level input.
401 * Overrides directional properties.
404 #define UBIDI_LEVEL_OVERRIDE 0x80
407 * Special value which can be returned by the mapping functions when a logical
408 * index has no corresponding visual index or vice-versa. This may happen
409 * for the logical-to-visual mapping of a Bidi control when option
410 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is specified. This can also happen
411 * for the visual-to-logical mapping of a Bidi mark (LRM or RLM) inserted
412 * by option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
413 * @see ubidi_getVisualIndex
414 * @see ubidi_getVisualMap
415 * @see ubidi_getLogicalIndex
416 * @see ubidi_getLogicalMap
419 #define UBIDI_MAP_NOWHERE (-1)
422 * <code>UBiDiDirection</code> values indicate the text direction.
425 enum UBiDiDirection
{
426 /** Left-to-right text. This is a 0 value.
428 * <li>As return value for <code>ubidi_getDirection()</code>, it means
429 * that the source string contains no right-to-left characters, or
430 * that the source string is empty and the paragraph level is even.
431 * <li> As return value for <code>ubidi_getBaseDirection()</code>, it
432 * means that the first strong character of the source string has
433 * a left-to-right direction.
438 /** Right-to-left text. This is a 1 value.
440 * <li>As return value for <code>ubidi_getDirection()</code>, it means
441 * that the source string contains no left-to-right characters, or
442 * that the source string is empty and the paragraph level is odd.
443 * <li> As return value for <code>ubidi_getBaseDirection()</code>, it
444 * means that the first strong character of the source string has
445 * a right-to-left direction.
450 /** Mixed-directional text.
451 * <p>As return value for <code>ubidi_getDirection()</code>, it means
452 * that the source string contains both left-to-right and
453 * right-to-left characters.
457 /** No strongly directional text.
458 * <p>As return value for <code>ubidi_getBaseDirection()</code>, it means
459 * that the source string is missing or empty, or contains neither left-to-right
460 * nor right-to-left characters.
466 /** @stable ICU 2.0 */
467 typedef enum UBiDiDirection UBiDiDirection
;
470 * Forward declaration of the <code>UBiDi</code> structure for the declaration of
471 * the API functions. Its fields are implementation-specific.<p>
472 * This structure holds information about a paragraph (or multiple paragraphs)
473 * of text with Bidi-algorithm-related details, or about one line of
474 * such a paragraph.<p>
475 * Reordering can be done on a line, or on one or more paragraphs which are
476 * then interpreted each as one single line.
481 /** @stable ICU 2.0 */
482 typedef struct UBiDi UBiDi
;
485 * Allocate a <code>UBiDi</code> structure.
486 * Such an object is initially empty. It is assigned
487 * the Bidi properties of a piece of text containing one or more paragraphs
488 * by <code>ubidi_setPara()</code>
489 * or the Bidi properties of a line within a paragraph by
490 * <code>ubidi_setLine()</code>.<p>
491 * This object can be reused for as long as it is not deallocated
492 * by calling <code>ubidi_close()</code>.<p>
493 * <code>ubidi_setPara()</code> and <code>ubidi_setLine()</code> will allocate
494 * additional memory for internal structures as necessary.
496 * @return An empty <code>UBiDi</code> object.
499 U_STABLE UBiDi
* U_EXPORT2
503 * Allocate a <code>UBiDi</code> structure with preallocated memory
504 * for internal structures.
505 * This function provides a <code>UBiDi</code> object like <code>ubidi_open()</code>
506 * with no arguments, but it also preallocates memory for internal structures
507 * according to the sizings supplied by the caller.<p>
508 * Subsequent functions will not allocate any more memory, and are thus
509 * guaranteed not to fail because of lack of memory.<p>
510 * The preallocation can be limited to some of the internal memory
511 * by setting some values to 0 here. That means that if, e.g.,
512 * <code>maxRunCount</code> cannot be reasonably predetermined and should not
513 * be set to <code>maxLength</code> (the only failproof value) to avoid
514 * wasting memory, then <code>maxRunCount</code> could be set to 0 here
515 * and the internal structures that are associated with it will be allocated
516 * on demand, just like with <code>ubidi_open()</code>.
518 * @param maxLength is the maximum text or line length that internal memory
519 * will be preallocated for. An attempt to associate this object with a
520 * longer text will fail, unless this value is 0, which leaves the allocation
521 * up to the implementation.
523 * @param maxRunCount is the maximum anticipated number of same-level runs
524 * that internal memory will be preallocated for. An attempt to access
525 * visual runs on an object that was not preallocated for as many runs
526 * as the text was actually resolved to will fail,
527 * unless this value is 0, which leaves the allocation up to the implementation.<br><br>
528 * The number of runs depends on the actual text and maybe anywhere between
529 * 1 and <code>maxLength</code>. It is typically small.
531 * @param pErrorCode must be a valid pointer to an error code value.
533 * @return An empty <code>UBiDi</code> object with preallocated memory.
536 U_STABLE UBiDi
* U_EXPORT2
537 ubidi_openSized(int32_t maxLength
, int32_t maxRunCount
, UErrorCode
*pErrorCode
);
540 * <code>ubidi_close()</code> must be called to free the memory
541 * associated with a UBiDi object.<p>
543 * <strong>Important: </strong>
544 * A parent <code>UBiDi</code> object must not be destroyed or reused if
545 * it still has children.
546 * If a <code>UBiDi</code> object has become the <i>child</i>
547 * of another one (its <i>parent</i>) by calling
548 * <code>ubidi_setLine()</code>, then the child object must
549 * be destroyed (closed) or reused (by calling
550 * <code>ubidi_setPara()</code> or <code>ubidi_setLine()</code>)
551 * before the parent object.
553 * @param pBiDi is a <code>UBiDi</code> object.
559 U_STABLE
void U_EXPORT2
560 ubidi_close(UBiDi
*pBiDi
);
562 #if U_SHOW_CPLUSPLUS_API
567 * \class LocalUBiDiPointer
568 * "Smart pointer" class, closes a UBiDi via ubidi_close().
569 * For most methods see the LocalPointerBase base class.
571 * @see LocalPointerBase
575 U_DEFINE_LOCAL_OPEN_POINTER(LocalUBiDiPointer
, UBiDi
, ubidi_close
);
579 #endif // U_SHOW_CPLUSPLUS_API
582 * Modify the operation of the Bidi algorithm such that it
583 * approximates an "inverse Bidi" algorithm. This function
584 * must be called before <code>ubidi_setPara()</code>.
586 * <p>The normal operation of the Bidi algorithm as described
587 * in the Unicode Technical Report is to take text stored in logical
588 * (keyboard, typing) order and to determine the reordering of it for visual
590 * Some legacy systems store text in visual order, and for operations
591 * with standard, Unicode-based algorithms, the text needs to be transformed
592 * to logical order. This is effectively the inverse algorithm of the
593 * described Bidi algorithm. Note that there is no standard algorithm for
594 * this "inverse Bidi" and that the current implementation provides only an
595 * approximation of "inverse Bidi".</p>
597 * <p>With <code>isInverse</code> set to <code>TRUE</code>,
598 * this function changes the behavior of some of the subsequent functions
599 * in a way that they can be used for the inverse Bidi algorithm.
600 * Specifically, runs of text with numeric characters will be treated in a
601 * special way and may need to be surrounded with LRM characters when they are
602 * written in reordered sequence.</p>
604 * <p>Output runs should be retrieved using <code>ubidi_getVisualRun()</code>.
605 * Since the actual input for "inverse Bidi" is visually ordered text and
606 * <code>ubidi_getVisualRun()</code> gets the reordered runs, these are actually
607 * the runs of the logically ordered output.</p>
609 * <p>Calling this function with argument <code>isInverse</code> set to
610 * <code>TRUE</code> is equivalent to calling
611 * <code>ubidi_setReorderingMode</code> with argument
612 * <code>reorderingMode</code>
613 * set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br>
614 * Calling this function with argument <code>isInverse</code> set to
615 * <code>FALSE</code> is equivalent to calling
616 * <code>ubidi_setReorderingMode</code> with argument
617 * <code>reorderingMode</code>
618 * set to <code>#UBIDI_REORDER_DEFAULT</code>.
620 * @param pBiDi is a <code>UBiDi</code> object.
622 * @param isInverse specifies "forward" or "inverse" Bidi operation.
625 * @see ubidi_writeReordered
626 * @see ubidi_setReorderingMode
629 U_STABLE
void U_EXPORT2
630 ubidi_setInverse(UBiDi
*pBiDi
, UBool isInverse
);
633 * Is this Bidi object set to perform the inverse Bidi algorithm?
634 * <p>Note: calling this function after setting the reordering mode with
635 * <code>ubidi_setReorderingMode</code> will return <code>TRUE</code> if the
636 * reordering mode was set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>,
637 * <code>FALSE</code> for all other values.</p>
639 * @param pBiDi is a <code>UBiDi</code> object.
640 * @return TRUE if the Bidi object is set to perform the inverse Bidi algorithm
641 * by handling numbers as L.
643 * @see ubidi_setInverse
644 * @see ubidi_setReorderingMode
648 U_STABLE UBool U_EXPORT2
649 ubidi_isInverse(UBiDi
*pBiDi
);
652 * Specify whether block separators must be allocated level zero,
653 * so that successive paragraphs will progress from left to right.
654 * This function must be called before <code>ubidi_setPara()</code>.
655 * Paragraph separators (B) may appear in the text. Setting them to level zero
656 * means that all paragraph separators (including one possibly appearing
657 * in the last text position) are kept in the reordered text after the text
658 * that they follow in the source text.
659 * When this feature is not enabled, a paragraph separator at the last
660 * position of the text before reordering will go to the first position
661 * of the reordered text when the paragraph level is odd.
663 * @param pBiDi is a <code>UBiDi</code> object.
665 * @param orderParagraphsLTR specifies whether paragraph separators (B) must
666 * receive level 0, so that successive paragraphs progress from left to right.
671 U_STABLE
void U_EXPORT2
672 ubidi_orderParagraphsLTR(UBiDi
*pBiDi
, UBool orderParagraphsLTR
);
675 * Is this Bidi object set to allocate level 0 to block separators so that
676 * successive paragraphs progress from left to right?
678 * @param pBiDi is a <code>UBiDi</code> object.
679 * @return TRUE if the Bidi object is set to allocate level 0 to block
682 * @see ubidi_orderParagraphsLTR
685 U_STABLE UBool U_EXPORT2
686 ubidi_isOrderParagraphsLTR(UBiDi
*pBiDi
);
689 * <code>UBiDiReorderingMode</code> values indicate which variant of the Bidi
692 * @see ubidi_setReorderingMode
695 typedef enum UBiDiReorderingMode
{
696 /** Regular Logical to Visual Bidi algorithm according to Unicode.
699 UBIDI_REORDER_DEFAULT
= 0,
700 /** Logical to Visual algorithm which handles numbers in a way which
701 * mimics the behavior of Windows XP.
703 UBIDI_REORDER_NUMBERS_SPECIAL
,
704 /** Logical to Visual algorithm grouping numbers with adjacent R characters
705 * (reversible algorithm).
707 UBIDI_REORDER_GROUP_NUMBERS_WITH_R
,
708 /** Reorder runs only to transform a Logical LTR string to the Logical RTL
709 * string with the same display, or vice-versa.<br>
710 * If this mode is set together with option
711 * <code>#UBIDI_OPTION_INSERT_MARKS</code>, some Bidi controls in the source
712 * text may be removed and other controls may be added to produce the
713 * minimum combination which has the required display.
715 UBIDI_REORDER_RUNS_ONLY
,
716 /** Visual to Logical algorithm which handles numbers like L
717 * (same algorithm as selected by <code>ubidi_setInverse(TRUE)</code>.
718 * @see ubidi_setInverse
720 UBIDI_REORDER_INVERSE_NUMBERS_AS_L
,
721 /** Visual to Logical algorithm equivalent to the regular Logical to Visual
724 UBIDI_REORDER_INVERSE_LIKE_DIRECT
,
725 /** Inverse Bidi (Visual to Logical) algorithm for the
726 * <code>UBIDI_REORDER_NUMBERS_SPECIAL</code> Bidi algorithm.
728 UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL
,
729 #ifndef U_HIDE_DEPRECATED_API
731 * Number of values for reordering mode.
732 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
735 #endif // U_HIDE_DEPRECATED_API
736 } UBiDiReorderingMode
;
739 * Modify the operation of the Bidi algorithm such that it implements some
740 * variant to the basic Bidi algorithm or approximates an "inverse Bidi"
741 * algorithm, depending on different values of the "reordering mode".
742 * This function must be called before <code>ubidi_setPara()</code>, and stays
743 * in effect until called again with a different argument.
745 * <p>The normal operation of the Bidi algorithm as described
746 * in the Unicode Standard Annex #9 is to take text stored in logical
747 * (keyboard, typing) order and to determine how to reorder it for visual
750 * <p>With the reordering mode set to a value other than
751 * <code>#UBIDI_REORDER_DEFAULT</code>, this function changes the behavior of
752 * some of the subsequent functions in a way such that they implement an
753 * inverse Bidi algorithm or some other algorithm variants.</p>
755 * <p>Some legacy systems store text in visual order, and for operations
756 * with standard, Unicode-based algorithms, the text needs to be transformed
757 * into logical order. This is effectively the inverse algorithm of the
758 * described Bidi algorithm. Note that there is no standard algorithm for
759 * this "inverse Bidi", so a number of variants are implemented here.</p>
761 * <p>In other cases, it may be desirable to emulate some variant of the
762 * Logical to Visual algorithm (e.g. one used in MS Windows), or perform a
763 * Logical to Logical transformation.</p>
766 * <li>When the reordering mode is set to <code>#UBIDI_REORDER_DEFAULT</code>,
767 * the standard Bidi Logical to Visual algorithm is applied.</li>
769 * <li>When the reordering mode is set to
770 * <code>#UBIDI_REORDER_NUMBERS_SPECIAL</code>,
771 * the algorithm used to perform Bidi transformations when calling
772 * <code>ubidi_setPara</code> should approximate the algorithm used in
773 * Microsoft Windows XP rather than strictly conform to the Unicode Bidi
776 * The differences between the basic algorithm and the algorithm addressed
777 * by this option are as follows:
779 * <li>Within text at an even embedding level, the sequence "123AB"
780 * (where AB represent R or AL letters) is transformed to "123BA" by the
781 * Unicode algorithm and to "BA123" by the Windows algorithm.</li>
782 * <li>Arabic-Indic numbers (AN) are handled by the Windows algorithm just
783 * like regular numbers (EN).</li>
786 * <li>When the reordering mode is set to
787 * <code>#UBIDI_REORDER_GROUP_NUMBERS_WITH_R</code>,
788 * numbers located between LTR text and RTL text are associated with the RTL
789 * text. For instance, an LTR paragraph with content "abc 123 DEF" (where
790 * upper case letters represent RTL characters) will be transformed to
791 * "abc FED 123" (and not "abc 123 FED"), "DEF 123 abc" will be transformed
792 * to "123 FED abc" and "123 FED abc" will be transformed to "DEF 123 abc".
793 * This makes the algorithm reversible and makes it useful when round trip
794 * (from visual to logical and back to visual) must be achieved without
795 * adding LRM characters. However, this is a variation from the standard
796 * Unicode Bidi algorithm.<br>
797 * The source text should not contain Bidi control characters other than LRM
800 * <li>When the reordering mode is set to
801 * <code>#UBIDI_REORDER_RUNS_ONLY</code>,
802 * a "Logical to Logical" transformation must be performed:
804 * <li>If the default text level of the source text (argument <code>paraLevel</code>
805 * in <code>ubidi_setPara</code>) is even, the source text will be handled as
806 * LTR logical text and will be transformed to the RTL logical text which has
807 * the same LTR visual display.</li>
808 * <li>If the default level of the source text is odd, the source text
809 * will be handled as RTL logical text and will be transformed to the
810 * LTR logical text which has the same LTR visual display.</li>
812 * This mode may be needed when logical text which is basically Arabic or
813 * Hebrew, with possible included numbers or phrases in English, has to be
814 * displayed as if it had an even embedding level (this can happen if the
815 * displaying application treats all text as if it was basically LTR).
817 * This mode may also be needed in the reverse case, when logical text which is
818 * basically English, with possible included phrases in Arabic or Hebrew, has to
819 * be displayed as if it had an odd embedding level.
821 * Both cases could be handled by adding LRE or RLE at the head of the text,
822 * if the display subsystem supports these formatting controls. If it does not,
823 * the problem may be handled by transforming the source text in this mode
824 * before displaying it, so that it will be displayed properly.<br>
825 * The source text should not contain Bidi control characters other than LRM
828 * <li>When the reordering mode is set to
829 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>, an "inverse Bidi" algorithm
831 * Runs of text with numeric characters will be treated like LTR letters and
832 * may need to be surrounded with LRM characters when they are written in
833 * reordered sequence (the option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> can
834 * be used with function <code>ubidi_writeReordered</code> to this end. This
835 * mode is equivalent to calling <code>ubidi_setInverse()</code> with
836 * argument <code>isInverse</code> set to <code>TRUE</code>.</li>
838 * <li>When the reordering mode is set to
839 * <code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code>, the "direct" Logical to Visual
840 * Bidi algorithm is used as an approximation of an "inverse Bidi" algorithm.
841 * This mode is similar to mode <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>
842 * but is closer to the regular Bidi algorithm.
844 * For example, an LTR paragraph with the content "FED 123 456 CBA" (where
845 * upper case represents RTL characters) will be transformed to
846 * "ABC 456 123 DEF", as opposed to "DEF 123 456 ABC"
847 * with mode <code>UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br>
848 * When used in conjunction with option
849 * <code>#UBIDI_OPTION_INSERT_MARKS</code>, this mode generally
850 * adds Bidi marks to the output significantly more sparingly than mode
851 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> with option
852 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to
853 * <code>ubidi_writeReordered</code>.</li>
855 * <li>When the reordering mode is set to
856 * <code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the Logical to Visual
857 * Bidi algorithm used in Windows XP is used as an approximation of an "inverse Bidi" algorithm.
859 * For example, an LTR paragraph with the content "abc FED123" (where
860 * upper case represents RTL characters) will be transformed to "abc 123DEF."</li>
863 * <p>In all the reordering modes specifying an "inverse Bidi" algorithm
864 * (i.e. those with a name starting with <code>UBIDI_REORDER_INVERSE</code>),
865 * output runs should be retrieved using
866 * <code>ubidi_getVisualRun()</code>, and the output text with
867 * <code>ubidi_writeReordered()</code>. The caller should keep in mind that in
868 * "inverse Bidi" modes the input is actually visually ordered text and
869 * reordered output returned by <code>ubidi_getVisualRun()</code> or
870 * <code>ubidi_writeReordered()</code> are actually runs or character string
871 * of logically ordered output.<br>
872 * For all the "inverse Bidi" modes, the source text should not contain
873 * Bidi control characters other than LRM or RLM.</p>
875 * <p>Note that option <code>#UBIDI_OUTPUT_REVERSE</code> of
876 * <code>ubidi_writeReordered</code> has no useful meaning and should not be
877 * used in conjunction with any value of the reordering mode specifying
878 * "inverse Bidi" or with value <code>UBIDI_REORDER_RUNS_ONLY</code>.
880 * @param pBiDi is a <code>UBiDi</code> object.
881 * @param reorderingMode specifies the required variant of the Bidi algorithm.
883 * @see UBiDiReorderingMode
884 * @see ubidi_setInverse
886 * @see ubidi_writeReordered
889 U_STABLE
void U_EXPORT2
890 ubidi_setReorderingMode(UBiDi
*pBiDi
, UBiDiReorderingMode reorderingMode
);
893 * What is the requested reordering mode for a given Bidi object?
895 * @param pBiDi is a <code>UBiDi</code> object.
896 * @return the current reordering mode of the Bidi object
897 * @see ubidi_setReorderingMode
900 U_STABLE UBiDiReorderingMode U_EXPORT2
901 ubidi_getReorderingMode(UBiDi
*pBiDi
);
904 * <code>UBiDiReorderingOption</code> values indicate which options are
905 * specified to affect the Bidi algorithm.
907 * @see ubidi_setReorderingOptions
910 typedef enum UBiDiReorderingOption
{
912 * option value for <code>ubidi_setReorderingOptions</code>:
913 * disable all the options which can be set with this function
914 * @see ubidi_setReorderingOptions
917 UBIDI_OPTION_DEFAULT
= 0,
920 * option bit for <code>ubidi_setReorderingOptions</code>:
921 * insert Bidi marks (LRM or RLM) when needed to ensure correct result of
922 * a reordering to a Logical order
924 * <p>This option must be set or reset before calling
925 * <code>ubidi_setPara</code>.</p>
927 * <p>This option is significant only with reordering modes which generate
928 * a result with Logical order, specifically:</p>
930 * <li><code>#UBIDI_REORDER_RUNS_ONLY</code></li>
931 * <li><code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code></li>
932 * <li><code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code></li>
933 * <li><code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code></li>
936 * <p>If this option is set in conjunction with reordering mode
937 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> or with calling
938 * <code>ubidi_setInverse(TRUE)</code>, it implies
939 * option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>
940 * in calls to function <code>ubidi_writeReordered()</code>.</p>
942 * <p>For other reordering modes, a minimum number of LRM or RLM characters
943 * will be added to the source text after reordering it so as to ensure
944 * round trip, i.e. when applying the inverse reordering mode on the
945 * resulting logical text with removal of Bidi marks
946 * (option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> set before calling
947 * <code>ubidi_setPara()</code> or option <code>#UBIDI_REMOVE_BIDI_CONTROLS</code>
948 * in <code>ubidi_writeReordered</code>), the result will be identical to the
949 * source text in the first transformation.
951 * <p>This option will be ignored if specified together with option
952 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>. It inhibits option
953 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to function
954 * <code>ubidi_writeReordered()</code> and it implies option
955 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to function
956 * <code>ubidi_writeReordered()</code> if the reordering mode is
957 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.</p>
959 * @see ubidi_setReorderingMode
960 * @see ubidi_setReorderingOptions
963 UBIDI_OPTION_INSERT_MARKS
= 1,
966 * option bit for <code>ubidi_setReorderingOptions</code>:
967 * remove Bidi control characters
969 * <p>This option must be set or reset before calling
970 * <code>ubidi_setPara</code>.</p>
972 * <p>This option nullifies option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
973 * It inhibits option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls
974 * to function <code>ubidi_writeReordered()</code> and it implies option
975 * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to that function.</p>
977 * @see ubidi_setReorderingMode
978 * @see ubidi_setReorderingOptions
981 UBIDI_OPTION_REMOVE_CONTROLS
= 2,
984 * option bit for <code>ubidi_setReorderingOptions</code>:
985 * process the output as part of a stream to be continued
987 * <p>This option must be set or reset before calling
988 * <code>ubidi_setPara</code>.</p>
990 * <p>This option specifies that the caller is interested in processing large
991 * text object in parts.
992 * The results of the successive calls are expected to be concatenated by the
993 * caller. Only the call for the last part will have this option bit off.</p>
995 * <p>When this option bit is on, <code>ubidi_setPara()</code> may process
996 * less than the full source text in order to truncate the text at a meaningful
997 * boundary. The caller should call <code>ubidi_getProcessedLength()</code>
998 * immediately after calling <code>ubidi_setPara()</code> in order to
999 * determine how much of the source text has been processed.
1000 * Source text beyond that length should be resubmitted in following calls to
1001 * <code>ubidi_setPara</code>. The processed length may be less than
1002 * the length of the source text if a character preceding the last character of
1003 * the source text constitutes a reasonable boundary (like a block separator)
1004 * for text to be continued.<br>
1005 * If the last character of the source text constitutes a reasonable
1006 * boundary, the whole text will be processed at once.<br>
1007 * If nowhere in the source text there exists
1008 * such a reasonable boundary, the processed length will be zero.<br>
1009 * The caller should check for such an occurrence and do one of the following:
1010 * <ul><li>submit a larger amount of text with a better chance to include
1011 * a reasonable boundary.</li>
1012 * <li>resubmit the same text after turning off option
1013 * <code>UBIDI_OPTION_STREAMING</code>.</li></ul>
1014 * In all cases, this option should be turned off before processing the last
1015 * part of the text.</p>
1017 * <p>When the <code>UBIDI_OPTION_STREAMING</code> option is used,
1018 * it is recommended to call <code>ubidi_orderParagraphsLTR()</code> with
1019 * argument <code>orderParagraphsLTR</code> set to <code>TRUE</code> before
1020 * calling <code>ubidi_setPara</code> so that later paragraphs may be
1021 * concatenated to previous paragraphs on the right.</p>
1023 * @see ubidi_setReorderingMode
1024 * @see ubidi_setReorderingOptions
1025 * @see ubidi_getProcessedLength
1026 * @see ubidi_orderParagraphsLTR
1029 UBIDI_OPTION_STREAMING
= 4
1030 } UBiDiReorderingOption
;
1033 * Specify which of the reordering options
1034 * should be applied during Bidi transformations.
1036 * @param pBiDi is a <code>UBiDi</code> object.
1037 * @param reorderingOptions is a combination of zero or more of the following
1039 * <code>#UBIDI_OPTION_DEFAULT</code>, <code>#UBIDI_OPTION_INSERT_MARKS</code>,
1040 * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>, <code>#UBIDI_OPTION_STREAMING</code>.
1042 * @see ubidi_getReorderingOptions
1045 U_STABLE
void U_EXPORT2
1046 ubidi_setReorderingOptions(UBiDi
*pBiDi
, uint32_t reorderingOptions
);
1049 * What are the reordering options applied to a given Bidi object?
1051 * @param pBiDi is a <code>UBiDi</code> object.
1052 * @return the current reordering options of the Bidi object
1053 * @see ubidi_setReorderingOptions
1056 U_STABLE
uint32_t U_EXPORT2
1057 ubidi_getReorderingOptions(UBiDi
*pBiDi
);
1060 * Set the context before a call to ubidi_setPara().<p>
1062 * ubidi_setPara() computes the left-right directionality for a given piece
1063 * of text which is supplied as one of its arguments. Sometimes this piece
1064 * of text (the "main text") should be considered in context, because text
1065 * appearing before ("prologue") and/or after ("epilogue") the main text
1066 * may affect the result of this computation.<p>
1068 * This function specifies the prologue and/or the epilogue for the next
1069 * call to ubidi_setPara(). The characters specified as prologue and
1070 * epilogue should not be modified by the calling program until the call
1071 * to ubidi_setPara() has returned. If successive calls to ubidi_setPara()
1072 * all need specification of a context, ubidi_setContext() must be called
1073 * before each call to ubidi_setPara(). In other words, a context is not
1074 * "remembered" after the following successful call to ubidi_setPara().<p>
1076 * If a call to ubidi_setPara() specifies UBIDI_DEFAULT_LTR or
1077 * UBIDI_DEFAULT_RTL as paraLevel and is preceded by a call to
1078 * ubidi_setContext() which specifies a prologue, the paragraph level will
1079 * be computed taking in consideration the text in the prologue.<p>
1081 * When ubidi_setPara() is called without a previous call to
1082 * ubidi_setContext, the main text is handled as if preceded and followed
1083 * by strong directional characters at the current paragraph level.
1084 * Calling ubidi_setContext() with specification of a prologue will change
1085 * this behavior by handling the main text as if preceded by the last
1086 * strong character appearing in the prologue, if any.
1087 * Calling ubidi_setContext() with specification of an epilogue will change
1088 * the behavior of ubidi_setPara() by handling the main text as if followed
1089 * by the first strong character or digit appearing in the epilogue, if any.<p>
1091 * Note 1: if <code>ubidi_setContext</code> is called repeatedly without
1092 * calling <code>ubidi_setPara</code>, the earlier calls have no effect,
1093 * only the last call will be remembered for the next call to
1094 * <code>ubidi_setPara</code>.<p>
1096 * Note 2: calling <code>ubidi_setContext(pBiDi, NULL, 0, NULL, 0, &errorCode)</code>
1097 * cancels any previous setting of non-empty prologue or epilogue.
1098 * The next call to <code>ubidi_setPara()</code> will process no
1099 * prologue or epilogue.<p>
1101 * Note 3: users must be aware that even after setting the context
1102 * before a call to ubidi_setPara() to perform e.g. a logical to visual
1103 * transformation, the resulting string may not be identical to what it
1104 * would have been if all the text, including prologue and epilogue, had
1105 * been processed together.<br>
1106 * Example (upper case letters represent RTL characters):<br>
1107 * prologue = "<code>abc DE</code>"<br>
1108 * epilogue = none<br>
1109 * main text = "<code>FGH xyz</code>"<br>
1110 * paraLevel = UBIDI_LTR<br>
1111 * display without prologue = "<code>HGF xyz</code>"
1112 * ("HGF" is adjacent to "xyz")<br>
1113 * display with prologue = "<code>abc HGFED xyz</code>"
1114 * ("HGF" is not adjacent to "xyz")<br>
1116 * @param pBiDi is a paragraph <code>UBiDi</code> object.
1118 * @param prologue is a pointer to the text which precedes the text that
1119 * will be specified in a coming call to ubidi_setPara().
1120 * If there is no prologue to consider, then <code>proLength</code>
1121 * must be zero and this pointer can be NULL.
1123 * @param proLength is the length of the prologue; if <code>proLength==-1</code>
1124 * then the prologue must be zero-terminated.
1125 * Otherwise proLength must be >= 0. If <code>proLength==0</code>, it means
1126 * that there is no prologue to consider.
1128 * @param epilogue is a pointer to the text which follows the text that
1129 * will be specified in a coming call to ubidi_setPara().
1130 * If there is no epilogue to consider, then <code>epiLength</code>
1131 * must be zero and this pointer can be NULL.
1133 * @param epiLength is the length of the epilogue; if <code>epiLength==-1</code>
1134 * then the epilogue must be zero-terminated.
1135 * Otherwise epiLength must be >= 0. If <code>epiLength==0</code>, it means
1136 * that there is no epilogue to consider.
1138 * @param pErrorCode must be a valid pointer to an error code value.
1140 * @see ubidi_setPara
1143 U_STABLE
void U_EXPORT2
1144 ubidi_setContext(UBiDi
*pBiDi
,
1145 const UChar
*prologue
, int32_t proLength
,
1146 const UChar
*epilogue
, int32_t epiLength
,
1147 UErrorCode
*pErrorCode
);
1150 * Perform the Unicode Bidi algorithm. It is defined in the
1151 * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Annex #9</a>,
1152 * Unicode 8.0.0 / revision 33,
1153 * also described in The Unicode Standard, Version 8.0 .<p>
1155 * This function takes a piece of plain text containing one or more paragraphs,
1156 * with or without externally specified embedding levels from <i>styled</i>
1157 * text and computes the left-right-directionality of each character.<p>
1159 * If the entire text is all of the same directionality, then
1160 * the function may not perform all the steps described by the algorithm,
1161 * i.e., some levels may not be the same as if all steps were performed.
1162 * This is not relevant for unidirectional text.<br>
1163 * For example, in pure LTR text with numbers the numbers would get
1164 * a resolved level of 2 higher than the surrounding text according to
1165 * the algorithm. This implementation may set all resolved levels to
1166 * the same value in such a case.<p>
1168 * The text can be composed of multiple paragraphs. Occurrence of a block
1169 * separator in the text terminates a paragraph, and whatever comes next starts
1170 * a new paragraph. The exception to this rule is when a Carriage Return (CR)
1171 * is followed by a Line Feed (LF). Both CR and LF are block separators, but
1172 * in that case, the pair of characters is considered as terminating the
1173 * preceding paragraph, and a new paragraph will be started by a character
1174 * coming after the LF.
1176 * @param pBiDi A <code>UBiDi</code> object allocated with <code>ubidi_open()</code>
1177 * which will be set to contain the reordering information,
1178 * especially the resolved levels for all the characters in <code>text</code>.
1180 * @param text is a pointer to the text that the Bidi algorithm will be performed on.
1181 * This pointer is stored in the UBiDi object and can be retrieved
1182 * with <code>ubidi_getText()</code>.<br>
1183 * <strong>Note:</strong> the text must be (at least) <code>length</code> long.
1185 * @param length is the length of the text; if <code>length==-1</code> then
1186 * the text must be zero-terminated.
1188 * @param paraLevel specifies the default level for the text;
1189 * it is typically 0 (LTR) or 1 (RTL).
1190 * If the function shall determine the paragraph level from the text,
1191 * then <code>paraLevel</code> can be set to
1192 * either <code>#UBIDI_DEFAULT_LTR</code>
1193 * or <code>#UBIDI_DEFAULT_RTL</code>; if the text contains multiple
1194 * paragraphs, the paragraph level shall be determined separately for
1195 * each paragraph; if a paragraph does not include any strongly typed
1196 * character, then the desired default is used (0 for LTR or 1 for RTL).
1197 * Any other value between 0 and <code>#UBIDI_MAX_EXPLICIT_LEVEL</code>
1198 * is also valid, with odd levels indicating RTL.
1200 * @param embeddingLevels (in) may be used to preset the embedding and override levels,
1201 * ignoring characters like LRE and PDF in the text.
1202 * A level overrides the directional property of its corresponding
1203 * (same index) character if the level has the
1204 * <code>#UBIDI_LEVEL_OVERRIDE</code> bit set.<br><br>
1205 * Aside from that bit, it must be
1206 * <code>paraLevel<=embeddingLevels[]<=UBIDI_MAX_EXPLICIT_LEVEL</code>,
1207 * except that level 0 is always allowed.
1208 * Level 0 for a paragraph separator prevents reordering of paragraphs;
1209 * this only works reliably if <code>#UBIDI_LEVEL_OVERRIDE</code>
1210 * is also set for paragraph separators.
1211 * Level 0 for other characters is treated as a wildcard
1212 * and is lifted up to the resolved level of the surrounding paragraph.<br><br>
1213 * <strong>Caution: </strong>A copy of this pointer, not of the levels,
1214 * will be stored in the <code>UBiDi</code> object;
1215 * the <code>embeddingLevels</code> array must not be
1216 * deallocated before the <code>UBiDi</code> structure is destroyed or reused,
1217 * and the <code>embeddingLevels</code>
1218 * should not be modified to avoid unexpected results on subsequent Bidi operations.
1219 * However, the <code>ubidi_setPara()</code> and
1220 * <code>ubidi_setLine()</code> functions may modify some or all of the levels.<br><br>
1221 * After the <code>UBiDi</code> object is reused or destroyed, the caller
1222 * must take care of the deallocation of the <code>embeddingLevels</code> array.<br><br>
1223 * <strong>Note:</strong> the <code>embeddingLevels</code> array must be
1224 * at least <code>length</code> long.
1225 * This pointer can be <code>NULL</code> if this
1226 * value is not necessary.
1228 * @param pErrorCode must be a valid pointer to an error code value.
1231 U_STABLE
void U_EXPORT2
1232 ubidi_setPara(UBiDi
*pBiDi
, const UChar
*text
, int32_t length
,
1233 UBiDiLevel paraLevel
, UBiDiLevel
*embeddingLevels
,
1234 UErrorCode
*pErrorCode
);
1236 #ifndef U_HIDE_INTERNAL_API
1238 * Perform the Unicode Bidi algorithm. It is defined in the
1239 * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Annex #9</a>,
1240 * Unicode 8.0.0 / revision 33,
1241 * also described in The Unicode Standard, Version 8.0 .<p>
1243 * This function takes a piece of plain text containing one or more paragraphs,
1244 * with or without externally specified direction overrides (in the form of
1245 * sequences of one or more bidi control characters for
1246 * embeddings/overrides/isolates to be effectively inserted at specified points
1247 * in the text), and computes the left-right-directionality of each character.
1248 * Note that ubidi_setContext may be used to set the context before or after the
1249 * text passed to ubidi_setPara, so ubidi_setParaWithControls is only needed if
1250 * externally specified direction overrides need to be effectively inserted at
1251 * other locations in the text.<p>
1253 * Note: Currently the external specified direction overrides are only supported
1254 * for the Logical to Visual values of UBiDiReorderingMode: UBIDI_REORDER_DEFAULT,
1255 * UBIDI_REORDER_NUMBERS_SPECIAL, UBIDI_REORDER_GROUP_NUMBERS_WITH_R. With other
1256 * UBiDiReorderingMode settings, this function behaves as if offsetCount is 0.<p>
1258 * If the entire text is all of the same directionality, then the function may
1259 * not perform all the steps described by the algorithm, i.e., some levels may
1260 * not be the same as if all steps were performed. This is not relevant for
1261 * unidirectional text.<br>
1262 * For example, in pure LTR text with numbers the numbers would get a resolved
1263 * level of 2 higher than the surrounding text according to the algorithm. This
1264 * implementation may set all resolved levels to the same value in such a case.<p>
1266 * The text can be composed of multiple paragraphs. Occurrence of a block
1267 * separator in the text terminates a paragraph, and whatever comes next starts
1268 * a new paragraph. The exception to this rule is when a Carriage Return (CR)
1269 * is followed by a Line Feed (LF). Both CR and LF are block separators, but
1270 * in that case, the pair of characters is considered as terminating the
1271 * preceding paragraph, and a new paragraph will be started by a character
1272 * coming after the LF.<p>
1274 * @param pBiDi A <code>UBiDi</code> object allocated with <code>ubidi_open()</code>
1275 * which will be set to contain the reordering information,
1276 * especially the resolved levels for all the characters in <code>text</code>.
1278 * @param text is a pointer to the text that the Bidi algorithm will be performed on.
1279 * This pointer is stored in the UBiDi object and can be retrieved
1280 * with <code>ubidi_getText()</code>.<br>
1281 * <strong>Note:</strong> the text must be (at least) <code>length</code> long.
1283 * @param length is the length of the text; if <code>length==-1</code> then
1284 * the text must be zero-terminated.
1286 * @param paraLevel specifies the default level for the text;
1287 * it is typically 0 (LTR) or 1 (RTL).
1288 * If the function shall determine the paragraph level from the text,
1289 * then <code>paraLevel</code> can be set to
1290 * either <code>#UBIDI_DEFAULT_LTR</code>
1291 * or <code>#UBIDI_DEFAULT_RTL</code>; if the text contains multiple
1292 * paragraphs, the paragraph level shall be determined separately for
1293 * each paragraph; if a paragraph does not include any strongly typed
1294 * character, then the desired default is used (0 for LTR or 1 for RTL).
1295 * Any other value between 0 and <code>#UBIDI_MAX_EXPLICIT_LEVEL</code>
1296 * is also valid, with odd levels indicating RTL.
1298 * @param offsets Array of text offsets at which sequences of one or more
1299 * bidi controls are to be effectively inserted. The offset values must
1300 * be >= 0 and < <code>length</code> (use <code>ubidi_setContext</code>
1301 * to provide the effect of inserting controls after the last character
1302 * of the text). This must be non-NULL if <code>offsetCount</code> > 0.
1304 * @param offsetCount The number of entries in the offsets array, and in the
1305 * controlStringIndices array if the latter is present (non NULL). If
1306 * <code>offsetCount</code> is 0, then no controls will be inserted and
1307 * the parameters <code>offsets</code>, <code>controlStringIndices</code>
1308 * and <code>controlStrings</code> will be ignored.
1310 * @param controlStringIndices If not NULL, this array must have the same
1311 * number of entries as the offsets array; each entry in this array
1312 * maps from the corresponding offset to the index in controlStrings
1313 * of the control sequence that is to be effectively inserted at that
1314 * offset. This indirection is useful when certain control sequences
1315 * are to be effectively inserted in many different places in the text.
1316 * If this array is NULL, then the entries in controlStrings correspond
1317 * directly to the entries in the offsets array.
1319 * @param controlStrings Array of const pointers to zero-terminated
1320 * const UChar strings each consisting of zero or more characters that
1321 * are bidi controls for embeddings, overrides, or isolates (see list
1322 * below). Other characters that might be supported in the future
1323 * (depending on need) include bidi marks an characters with
1324 * bidi class B (block separator) or class S (segment separator).
1325 * The characters in these strings only affect the bidi levels assigned
1326 * to the characters in he text array, they are not used for any other
1328 * If controlStringIndices is NULL, then controlStrings must have the
1329 * same number of entries as the offsets array, and each entry provides
1330 * the UChar string that is effectively inserted at the corresponding
1331 * offset. If controlStringIndices is not NULL, then controlStrings must
1332 * have at least enough entries to accommodate to all of the index values
1333 * in the controlStringIndices array. This must be non-NULL if
1334 * offsetCount > 0.<br>
1335 * Current limitations:<br>
1336 * Each zero-terminated const UChar string is limited a maximum length
1337 * of 4, not including the zero terminator.<br>
1338 * Each zero-terminated const UChar string may contain at most one
1339 * instance of FSI, LRI, or RLI.<br>
1341 * @param pErrorCode must be a valid pointer to an error code value.
1345 * Supported bidi controls for embeddings / overrides / isolates as of Unicode 8.0:
1346 * LRE U+202A LEFT-TO-RIGHT EMBEDDING
1347 * RLE U+202B RIGHT-TO-LEFT EMBEDDING
1348 * PDF U+202C POP DIRECTIONAL FORMATTING
1349 * LRO U+202D LEFT-TO-RIGHT OVERRIDE
1350 * RLO U+202E RIGHT-TO-LEFT OVERRIDE
1352 * LRI U+2066 LEFT‑TO‑RIGHT ISOLATE
1353 * RLI U+2067 RIGHT‑TO‑LEFT ISOLATE
1354 * FSI U+2068 FIRST STRONG ISOLATE
1355 * PDI U+2069 POP DIRECTIONAL ISOLATE
1357 * Bidi marks as of Unicode 8.0:
1358 * ALM U+061C ARABIC LETTER MARK (bidi class AL)
1359 * LRM U+200E LEFT-TO-RIGHT MARK (bidi class L)
1360 * RLM U+200F RIGHT-TO-LEFT MARK (bidi class R)
1361 * Characters with bidi class B (block separator) as of Unicode 8.0:
1362 * B U+000A LINE FEED (LF)
1363 * B U+000D CARRIAGE RETURN (CR)
1364 * B U+001C INFORMATION SEPARATOR FOUR
1365 * B U+001D INFORMATION SEPARATOR THREE
1366 * B U+001E INFORMATION SEPARATOR TWO
1367 * B U+0085 NEXT LINE (NEL)
1368 * B U+2029 PARAGRAPH SEPARATOR
1369 * Characters with bidi class S (segment separator) as of Unicode 8.0:
1370 * S U+0009 CHARACTER TABULATION
1371 * S U+000B LINE TABULATION
1372 * S U+001F INFORMATION SEPARATOR ONE
1375 * @see ubidi_setContext
1376 * @internal technology preview as of ICU 57
1378 U_INTERNAL
void U_EXPORT2
1379 ubidi_setParaWithControls(UBiDi
*pBiDi
,
1380 const UChar
*text
, int32_t length
,
1381 UBiDiLevel paraLevel
,
1382 const int32_t *offsets
, int32_t offsetCount
,
1383 const int32_t *controlStringIndices
,
1384 const UChar
* const * controlStrings
,
1385 UErrorCode
*pErrorCode
);
1387 #endif /* U_HIDE_INTERNAL_API */
1390 * <code>ubidi_setLine()</code> sets a <code>UBiDi</code> to
1391 * contain the reordering information, especially the resolved levels,
1392 * for all the characters in a line of text. This line of text is
1393 * specified by referring to a <code>UBiDi</code> object representing
1394 * this information for a piece of text containing one or more paragraphs,
1395 * and by specifying a range of indexes in this text.<p>
1396 * In the new line object, the indexes will range from 0 to <code>limit-start-1</code>.<p>
1398 * This is used after calling <code>ubidi_setPara()</code>
1399 * for a piece of text, and after line-breaking on that text.
1400 * It is not necessary if each paragraph is treated as a single line.<p>
1402 * After line-breaking, rules (L1) and (L2) for the treatment of
1403 * trailing WS and for reordering are performed on
1404 * a <code>UBiDi</code> object that represents a line.<p>
1406 * <strong>Important: </strong><code>pLineBiDi</code> shares data with
1407 * <code>pParaBiDi</code>.
1408 * You must destroy or reuse <code>pLineBiDi</code> before <code>pParaBiDi</code>.
1409 * In other words, you must destroy or reuse the <code>UBiDi</code> object for a line
1410 * before the object for its parent paragraph.<p>
1412 * The text pointer that was stored in <code>pParaBiDi</code> is also copied,
1413 * and <code>start</code> is added to it so that it points to the beginning of the
1414 * line for this object.
1416 * @param pParaBiDi is the parent paragraph object. It must have been set
1417 * by a successful call to ubidi_setPara.
1419 * @param start is the line's first index into the text.
1421 * @param limit is just behind the line's last index into the text
1422 * (its last index +1).<br>
1423 * It must be <code>0<=start<limit<=</code>containing paragraph limit.
1424 * If the specified line crosses a paragraph boundary, the function
1425 * will terminate with error code U_ILLEGAL_ARGUMENT_ERROR.
1427 * @param pLineBiDi is the object that will now represent a line of the text.
1429 * @param pErrorCode must be a valid pointer to an error code value.
1431 * @see ubidi_setPara
1432 * @see ubidi_getProcessedLength
1435 U_STABLE
void U_EXPORT2
1436 ubidi_setLine(const UBiDi
*pParaBiDi
,
1437 int32_t start
, int32_t limit
,
1439 UErrorCode
*pErrorCode
);
1442 * Get the directionality of the text.
1444 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1446 * @return a value of <code>UBIDI_LTR</code>, <code>UBIDI_RTL</code>
1447 * or <code>UBIDI_MIXED</code>
1448 * that indicates if the entire text
1449 * represented by this object is unidirectional,
1450 * and which direction, or if it is mixed-directional.
1451 * Note - The value <code>UBIDI_NEUTRAL</code> is never returned from this method.
1453 * @see UBiDiDirection
1456 U_STABLE UBiDiDirection U_EXPORT2
1457 ubidi_getDirection(const UBiDi
*pBiDi
);
1460 * Gets the base direction of the text provided according
1461 * to the Unicode Bidirectional Algorithm. The base direction
1462 * is derived from the first character in the string with bidirectional
1463 * character type L, R, or AL. If the first such character has type L,
1464 * <code>UBIDI_LTR</code> is returned. If the first such character has
1465 * type R or AL, <code>UBIDI_RTL</code> is returned. If the string does
1466 * not contain any character of these types, then
1467 * <code>UBIDI_NEUTRAL</code> is returned.
1469 * This is a lightweight function for use when only the base direction
1470 * is needed and no further bidi processing of the text is needed.
1472 * @param text is a pointer to the text whose base
1473 * direction is needed.
1474 * Note: the text must be (at least) @c length long.
1476 * @param length is the length of the text;
1477 * if <code>length==-1</code> then the text
1478 * must be zero-terminated.
1480 * @return <code>UBIDI_LTR</code>, <code>UBIDI_RTL</code>,
1481 * <code>UBIDI_NEUTRAL</code>
1483 * @see UBiDiDirection
1486 U_STABLE UBiDiDirection U_EXPORT2
1487 ubidi_getBaseDirection(const UChar
*text
, int32_t length
);
1490 * Get the pointer to the text.
1492 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1494 * @return The pointer to the text that the UBiDi object was created for.
1496 * @see ubidi_setPara
1497 * @see ubidi_setLine
1500 U_STABLE
const UChar
* U_EXPORT2
1501 ubidi_getText(const UBiDi
*pBiDi
);
1504 * Get the length of the text.
1506 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1508 * @return The length of the text that the UBiDi object was created for.
1511 U_STABLE
int32_t U_EXPORT2
1512 ubidi_getLength(const UBiDi
*pBiDi
);
1515 * Get the paragraph level of the text.
1517 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1519 * @return The paragraph level. If there are multiple paragraphs, their
1520 * level may vary if the required paraLevel is UBIDI_DEFAULT_LTR or
1521 * UBIDI_DEFAULT_RTL. In that case, the level of the first paragraph
1525 * @see ubidi_getParagraph
1526 * @see ubidi_getParagraphByIndex
1529 U_STABLE UBiDiLevel U_EXPORT2
1530 ubidi_getParaLevel(const UBiDi
*pBiDi
);
1533 * Get the number of paragraphs.
1535 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1537 * @return The number of paragraphs.
1540 U_STABLE
int32_t U_EXPORT2
1541 ubidi_countParagraphs(UBiDi
*pBiDi
);
1544 * Get a paragraph, given a position within the text.
1545 * This function returns information about a paragraph.<br>
1546 * Note: if the paragraph index is known, it is more efficient to
1547 * retrieve the paragraph information using ubidi_getParagraphByIndex().<p>
1549 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1551 * @param charIndex is the index of a character within the text, in the
1552 * range <code>[0..ubidi_getProcessedLength(pBiDi)-1]</code>.
1554 * @param pParaStart will receive the index of the first character of the
1555 * paragraph in the text.
1556 * This pointer can be <code>NULL</code> if this
1557 * value is not necessary.
1559 * @param pParaLimit will receive the limit of the paragraph.
1560 * The l-value that you point to here may be the
1561 * same expression (variable) as the one for
1562 * <code>charIndex</code>.
1563 * This pointer can be <code>NULL</code> if this
1564 * value is not necessary.
1566 * @param pParaLevel will receive the level of the paragraph.
1567 * This pointer can be <code>NULL</code> if this
1568 * value is not necessary.
1570 * @param pErrorCode must be a valid pointer to an error code value.
1572 * @return The index of the paragraph containing the specified position.
1574 * @see ubidi_getProcessedLength
1577 U_STABLE
int32_t U_EXPORT2
1578 ubidi_getParagraph(const UBiDi
*pBiDi
, int32_t charIndex
, int32_t *pParaStart
,
1579 int32_t *pParaLimit
, UBiDiLevel
*pParaLevel
,
1580 UErrorCode
*pErrorCode
);
1583 * Get a paragraph, given the index of this paragraph.
1585 * This function returns information about a paragraph.<p>
1587 * @param pBiDi is the paragraph <code>UBiDi</code> object.
1589 * @param paraIndex is the number of the paragraph, in the
1590 * range <code>[0..ubidi_countParagraphs(pBiDi)-1]</code>.
1592 * @param pParaStart will receive the index of the first character of the
1593 * paragraph in the text.
1594 * This pointer can be <code>NULL</code> if this
1595 * value is not necessary.
1597 * @param pParaLimit will receive the limit of the paragraph.
1598 * This pointer can be <code>NULL</code> if this
1599 * value is not necessary.
1601 * @param pParaLevel will receive the level of the paragraph.
1602 * This pointer can be <code>NULL</code> if this
1603 * value is not necessary.
1605 * @param pErrorCode must be a valid pointer to an error code value.
1609 U_STABLE
void U_EXPORT2
1610 ubidi_getParagraphByIndex(const UBiDi
*pBiDi
, int32_t paraIndex
,
1611 int32_t *pParaStart
, int32_t *pParaLimit
,
1612 UBiDiLevel
*pParaLevel
, UErrorCode
*pErrorCode
);
1615 * Get the level for one character.
1617 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1619 * @param charIndex the index of a character. It must be in the range
1620 * [0..ubidi_getProcessedLength(pBiDi)].
1622 * @return The level for the character at charIndex (0 if charIndex is not
1623 * in the valid range).
1626 * @see ubidi_getProcessedLength
1629 U_STABLE UBiDiLevel U_EXPORT2
1630 ubidi_getLevelAt(const UBiDi
*pBiDi
, int32_t charIndex
);
1633 * Get an array of levels for each character.<p>
1635 * Note that this function may allocate memory under some
1636 * circumstances, unlike <code>ubidi_getLevelAt()</code>.
1638 * @param pBiDi is the paragraph or line <code>UBiDi</code> object, whose
1639 * text length must be strictly positive.
1641 * @param pErrorCode must be a valid pointer to an error code value.
1643 * @return The levels array for the text,
1644 * or <code>NULL</code> if an error occurs.
1647 * @see ubidi_getProcessedLength
1650 U_STABLE
const UBiDiLevel
* U_EXPORT2
1651 ubidi_getLevels(UBiDi
*pBiDi
, UErrorCode
*pErrorCode
);
1654 * Get a logical run.
1655 * This function returns information about a run and is used
1656 * to retrieve runs in logical order.<p>
1657 * This is especially useful for line-breaking on a paragraph.
1659 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1661 * @param logicalPosition is a logical position within the source text.
1663 * @param pLogicalLimit will receive the limit of the corresponding run.
1664 * The l-value that you point to here may be the
1665 * same expression (variable) as the one for
1666 * <code>logicalPosition</code>.
1667 * This pointer can be <code>NULL</code> if this
1668 * value is not necessary.
1670 * @param pLevel will receive the level of the corresponding run.
1671 * This pointer can be <code>NULL</code> if this
1672 * value is not necessary.
1674 * @see ubidi_getProcessedLength
1677 U_STABLE
void U_EXPORT2
1678 ubidi_getLogicalRun(const UBiDi
*pBiDi
, int32_t logicalPosition
,
1679 int32_t *pLogicalLimit
, UBiDiLevel
*pLevel
);
1682 * Get the number of runs.
1683 * This function may invoke the actual reordering on the
1684 * <code>UBiDi</code> object, after <code>ubidi_setPara()</code>
1685 * may have resolved only the levels of the text. Therefore,
1686 * <code>ubidi_countRuns()</code> may have to allocate memory,
1687 * and may fail doing so.
1689 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1691 * @param pErrorCode must be a valid pointer to an error code value.
1693 * @return The number of runs.
1696 U_STABLE
int32_t U_EXPORT2
1697 ubidi_countRuns(UBiDi
*pBiDi
, UErrorCode
*pErrorCode
);
1700 * Get one run's logical start, length, and directionality,
1701 * which can be 0 for LTR or 1 for RTL.
1702 * In an RTL run, the character at the logical start is
1703 * visually on the right of the displayed run.
1704 * The length is the number of characters in the run.<p>
1705 * <code>ubidi_countRuns()</code> should be called
1706 * before the runs are retrieved.
1708 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1710 * @param runIndex is the number of the run in visual order, in the
1711 * range <code>[0..ubidi_countRuns(pBiDi)-1]</code>.
1713 * @param pLogicalStart is the first logical character index in the text.
1714 * The pointer may be <code>NULL</code> if this index is not needed.
1716 * @param pLength is the number of characters (at least one) in the run.
1717 * The pointer may be <code>NULL</code> if this is not needed.
1719 * @return the directionality of the run,
1720 * <code>UBIDI_LTR==0</code> or <code>UBIDI_RTL==1</code>,
1721 * never <code>UBIDI_MIXED</code>,
1722 * never <code>UBIDI_NEUTRAL</code>.
1724 * @see ubidi_countRuns
1729 * int32_t i, count=ubidi_countRuns(pBiDi),
1730 * logicalStart, visualIndex=0, length;
1731 * for(i=0; i<count; ++i) {
1732 * if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, i, &logicalStart, &length)) {
1734 * show_char(text[logicalStart++], visualIndex++);
1735 * } while(--length>0);
1737 * logicalStart+=length; // logicalLimit
1739 * show_char(text[--logicalStart], visualIndex++);
1740 * } while(--length>0);
1746 * Note that in right-to-left runs, code like this places
1747 * second surrogates before first ones (which is generally a bad idea)
1748 * and combining characters before base characters.
1750 * Use of <code>ubidi_writeReordered()</code>, optionally with the
1751 * <code>#UBIDI_KEEP_BASE_COMBINING</code> option, can be considered in order
1752 * to avoid these issues.
1755 U_STABLE UBiDiDirection U_EXPORT2
1756 ubidi_getVisualRun(UBiDi
*pBiDi
, int32_t runIndex
,
1757 int32_t *pLogicalStart
, int32_t *pLength
);
1760 * Get the visual position from a logical text position.
1761 * If such a mapping is used many times on the same
1762 * <code>UBiDi</code> object, then calling
1763 * <code>ubidi_getLogicalMap()</code> is more efficient.<p>
1765 * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no
1766 * visual position because the corresponding text character is a Bidi control
1767 * removed from output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>.
1769 * When the visual output is altered by using options of
1770 * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1771 * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
1772 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the visual position returned may not
1773 * be correct. It is advised to use, when possible, reordering options
1774 * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
1776 * Note that in right-to-left runs, this mapping places
1777 * second surrogates before first ones (which is generally a bad idea)
1778 * and combining characters before base characters.
1779 * Use of <code>ubidi_writeReordered()</code>, optionally with the
1780 * <code>#UBIDI_KEEP_BASE_COMBINING</code> option can be considered instead
1781 * of using the mapping, in order to avoid these issues.
1783 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1785 * @param logicalIndex is the index of a character in the text.
1787 * @param pErrorCode must be a valid pointer to an error code value.
1789 * @return The visual position of this character.
1791 * @see ubidi_getLogicalMap
1792 * @see ubidi_getLogicalIndex
1793 * @see ubidi_getProcessedLength
1796 U_STABLE
int32_t U_EXPORT2
1797 ubidi_getVisualIndex(UBiDi
*pBiDi
, int32_t logicalIndex
, UErrorCode
*pErrorCode
);
1800 * Get the logical text position from a visual position.
1801 * If such a mapping is used many times on the same
1802 * <code>UBiDi</code> object, then calling
1803 * <code>ubidi_getVisualMap()</code> is more efficient.<p>
1805 * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no
1806 * logical position because the corresponding text character is a Bidi mark
1807 * inserted in the output by option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
1809 * This is the inverse function to <code>ubidi_getVisualIndex()</code>.
1811 * When the visual output is altered by using options of
1812 * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1813 * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
1814 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the logical position returned may not
1815 * be correct. It is advised to use, when possible, reordering options
1816 * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
1818 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1820 * @param visualIndex is the visual position of a character.
1822 * @param pErrorCode must be a valid pointer to an error code value.
1824 * @return The index of this character in the text.
1826 * @see ubidi_getVisualMap
1827 * @see ubidi_getVisualIndex
1828 * @see ubidi_getResultLength
1831 U_STABLE
int32_t U_EXPORT2
1832 ubidi_getLogicalIndex(UBiDi
*pBiDi
, int32_t visualIndex
, UErrorCode
*pErrorCode
);
1835 * Get a logical-to-visual index map (array) for the characters in the UBiDi
1836 * (paragraph or line) object.
1838 * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the
1839 * corresponding text characters are Bidi controls removed from the visual
1840 * output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>.
1842 * When the visual output is altered by using options of
1843 * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1844 * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
1845 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the visual positions returned may not
1846 * be correct. It is advised to use, when possible, reordering options
1847 * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
1849 * Note that in right-to-left runs, this mapping places
1850 * second surrogates before first ones (which is generally a bad idea)
1851 * and combining characters before base characters.
1852 * Use of <code>ubidi_writeReordered()</code>, optionally with the
1853 * <code>#UBIDI_KEEP_BASE_COMBINING</code> option can be considered instead
1854 * of using the mapping, in order to avoid these issues.
1856 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1858 * @param indexMap is a pointer to an array of <code>ubidi_getProcessedLength()</code>
1859 * indexes which will reflect the reordering of the characters.
1860 * If option <code>#UBIDI_OPTION_INSERT_MARKS</code> is set, the number
1861 * of elements allocated in <code>indexMap</code> must be no less than
1862 * <code>ubidi_getResultLength()</code>.
1863 * The array does not need to be initialized.<br><br>
1864 * The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>.
1866 * @param pErrorCode must be a valid pointer to an error code value.
1868 * @see ubidi_getVisualMap
1869 * @see ubidi_getVisualIndex
1870 * @see ubidi_getProcessedLength
1871 * @see ubidi_getResultLength
1874 U_STABLE
void U_EXPORT2
1875 ubidi_getLogicalMap(UBiDi
*pBiDi
, int32_t *indexMap
, UErrorCode
*pErrorCode
);
1878 * Get a visual-to-logical index map (array) for the characters in the UBiDi
1879 * (paragraph or line) object.
1881 * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the
1882 * corresponding text characters are Bidi marks inserted in the visual output
1883 * by the option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
1885 * When the visual output is altered by using options of
1886 * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
1887 * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
1888 * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the logical positions returned may not
1889 * be correct. It is advised to use, when possible, reordering options
1890 * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
1892 * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
1894 * @param indexMap is a pointer to an array of <code>ubidi_getResultLength()</code>
1895 * indexes which will reflect the reordering of the characters.
1896 * If option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is set, the number
1897 * of elements allocated in <code>indexMap</code> must be no less than
1898 * <code>ubidi_getProcessedLength()</code>.
1899 * The array does not need to be initialized.<br><br>
1900 * The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>.
1902 * @param pErrorCode must be a valid pointer to an error code value.
1904 * @see ubidi_getLogicalMap
1905 * @see ubidi_getLogicalIndex
1906 * @see ubidi_getProcessedLength
1907 * @see ubidi_getResultLength
1910 U_STABLE
void U_EXPORT2
1911 ubidi_getVisualMap(UBiDi
*pBiDi
, int32_t *indexMap
, UErrorCode
*pErrorCode
);
1914 * This is a convenience function that does not use a UBiDi object.
1915 * It is intended to be used for when an application has determined the levels
1916 * of objects (character sequences) and just needs to have them reordered (L2).
1917 * This is equivalent to using <code>ubidi_getLogicalMap()</code> on a
1918 * <code>UBiDi</code> object.
1920 * @param levels is an array with <code>length</code> levels that have been determined by
1923 * @param length is the number of levels in the array, or, semantically,
1924 * the number of objects to be reordered.
1925 * It must be <code>length>0</code>.
1927 * @param indexMap is a pointer to an array of <code>length</code>
1928 * indexes which will reflect the reordering of the characters.
1929 * The array does not need to be initialized.<p>
1930 * The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>.
1933 U_STABLE
void U_EXPORT2
1934 ubidi_reorderLogical(const UBiDiLevel
*levels
, int32_t length
, int32_t *indexMap
);
1937 * This is a convenience function that does not use a UBiDi object.
1938 * It is intended to be used for when an application has determined the levels
1939 * of objects (character sequences) and just needs to have them reordered (L2).
1940 * This is equivalent to using <code>ubidi_getVisualMap()</code> on a
1941 * <code>UBiDi</code> object.
1943 * @param levels is an array with <code>length</code> levels that have been determined by
1946 * @param length is the number of levels in the array, or, semantically,
1947 * the number of objects to be reordered.
1948 * It must be <code>length>0</code>.
1950 * @param indexMap is a pointer to an array of <code>length</code>
1951 * indexes which will reflect the reordering of the characters.
1952 * The array does not need to be initialized.<p>
1953 * The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>.
1956 U_STABLE
void U_EXPORT2
1957 ubidi_reorderVisual(const UBiDiLevel
*levels
, int32_t length
, int32_t *indexMap
);
1960 * Invert an index map.
1961 * The index mapping of the first map is inverted and written to
1964 * @param srcMap is an array with <code>length</code> elements
1965 * which defines the original mapping from a source array containing
1966 * <code>length</code> elements to a destination array.
1967 * Some elements of the source array may have no mapping in the
1968 * destination array. In that case, their value will be
1969 * the special value <code>UBIDI_MAP_NOWHERE</code>.
1970 * All elements must be >=0 or equal to <code>UBIDI_MAP_NOWHERE</code>.
1971 * Some elements may have a value >= <code>length</code>, if the
1972 * destination array has more elements than the source array.
1973 * There must be no duplicate indexes (two or more elements with the
1974 * same value except <code>UBIDI_MAP_NOWHERE</code>).
1976 * @param destMap is an array with a number of elements equal to 1 + the highest
1977 * value in <code>srcMap</code>.
1978 * <code>destMap</code> will be filled with the inverse mapping.
1979 * If element with index i in <code>srcMap</code> has a value k different
1980 * from <code>UBIDI_MAP_NOWHERE</code>, this means that element i of
1981 * the source array maps to element k in the destination array.
1982 * The inverse map will have value i in its k-th element.
1983 * For all elements of the destination array which do not map to
1984 * an element in the source array, the corresponding element in the
1985 * inverse map will have a value equal to <code>UBIDI_MAP_NOWHERE</code>.
1987 * @param length is the length of each array.
1988 * @see UBIDI_MAP_NOWHERE
1991 U_STABLE
void U_EXPORT2
1992 ubidi_invertMap(const int32_t *srcMap
, int32_t *destMap
, int32_t length
);
1994 /** option flags for ubidi_writeReordered() */
1997 * option bit for ubidi_writeReordered():
1998 * keep combining characters after their base characters in RTL runs
2000 * @see ubidi_writeReordered
2003 #define UBIDI_KEEP_BASE_COMBINING 1
2006 * option bit for ubidi_writeReordered():
2007 * replace characters with the "mirrored" property in RTL runs
2008 * by their mirror-image mappings
2010 * @see ubidi_writeReordered
2013 #define UBIDI_DO_MIRRORING 2
2016 * option bit for ubidi_writeReordered():
2017 * surround the run with LRMs if necessary;
2018 * this is part of the approximate "inverse Bidi" algorithm
2020 * <p>This option does not imply corresponding adjustment of the index
2023 * @see ubidi_setInverse
2024 * @see ubidi_writeReordered
2027 #define UBIDI_INSERT_LRM_FOR_NUMERIC 4
2030 * option bit for ubidi_writeReordered():
2031 * remove Bidi control characters
2032 * (this does not affect #UBIDI_INSERT_LRM_FOR_NUMERIC)
2034 * <p>This option does not imply corresponding adjustment of the index
2037 * @see ubidi_writeReordered
2040 #define UBIDI_REMOVE_BIDI_CONTROLS 8
2043 * option bit for ubidi_writeReordered():
2044 * write the output in reverse order
2046 * <p>This has the same effect as calling <code>ubidi_writeReordered()</code>
2047 * first without this option, and then calling
2048 * <code>ubidi_writeReverse()</code> without mirroring.
2049 * Doing this in the same step is faster and avoids a temporary buffer.
2050 * An example for using this option is output to a character terminal that
2051 * is designed for RTL scripts and stores text in reverse order.</p>
2053 * @see ubidi_writeReordered
2056 #define UBIDI_OUTPUT_REVERSE 16
2059 * Get the length of the source text processed by the last call to
2060 * <code>ubidi_setPara()</code>. This length may be different from the length
2061 * of the source text if option <code>#UBIDI_OPTION_STREAMING</code>
2064 * Note that whenever the length of the text affects the execution or the
2065 * result of a function, it is the processed length which must be considered,
2066 * except for <code>ubidi_setPara</code> (which receives unprocessed source
2067 * text) and <code>ubidi_getLength</code> (which returns the original length
2068 * of the source text).<br>
2069 * In particular, the processed length is the one to consider in the following
2072 * <li>maximum value of the <code>limit</code> argument of
2073 * <code>ubidi_setLine</code></li>
2074 * <li>maximum value of the <code>charIndex</code> argument of
2075 * <code>ubidi_getParagraph</code></li>
2076 * <li>maximum value of the <code>charIndex</code> argument of
2077 * <code>ubidi_getLevelAt</code></li>
2078 * <li>number of elements in the array returned by <code>ubidi_getLevels</code></li>
2079 * <li>maximum value of the <code>logicalStart</code> argument of
2080 * <code>ubidi_getLogicalRun</code></li>
2081 * <li>maximum value of the <code>logicalIndex</code> argument of
2082 * <code>ubidi_getVisualIndex</code></li>
2083 * <li>number of elements filled in the <code>*indexMap</code> argument of
2084 * <code>ubidi_getLogicalMap</code></li>
2085 * <li>length of text processed by <code>ubidi_writeReordered</code></li>
2088 * @param pBiDi is the paragraph <code>UBiDi</code> object.
2090 * @return The length of the part of the source text processed by
2091 * the last call to <code>ubidi_setPara</code>.
2092 * @see ubidi_setPara
2093 * @see UBIDI_OPTION_STREAMING
2096 U_STABLE
int32_t U_EXPORT2
2097 ubidi_getProcessedLength(const UBiDi
*pBiDi
);
2100 * Get the length of the reordered text resulting from the last call to
2101 * <code>ubidi_setPara()</code>. This length may be different from the length
2102 * of the source text if option <code>#UBIDI_OPTION_INSERT_MARKS</code>
2103 * or option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> has been set.
2105 * This resulting length is the one to consider in the following cases:
2107 * <li>maximum value of the <code>visualIndex</code> argument of
2108 * <code>ubidi_getLogicalIndex</code></li>
2109 * <li>number of elements of the <code>*indexMap</code> argument of
2110 * <code>ubidi_getVisualMap</code></li>
2112 * Note that this length stays identical to the source text length if
2113 * Bidi marks are inserted or removed using option bits of
2114 * <code>ubidi_writeReordered</code>, or if option
2115 * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> has been set.
2117 * @param pBiDi is the paragraph <code>UBiDi</code> object.
2119 * @return The length of the reordered text resulting from
2120 * the last call to <code>ubidi_setPara</code>.
2121 * @see ubidi_setPara
2122 * @see UBIDI_OPTION_INSERT_MARKS
2123 * @see UBIDI_OPTION_REMOVE_CONTROLS
2126 U_STABLE
int32_t U_EXPORT2
2127 ubidi_getResultLength(const UBiDi
*pBiDi
);
2131 #ifndef U_HIDE_DEPRECATED_API
2133 * Value returned by <code>UBiDiClassCallback</code> callbacks when
2134 * there is no need to override the standard Bidi class for a given code point.
2136 * This constant is deprecated; use u_getIntPropertyMaxValue(UCHAR_BIDI_CLASS)+1 instead.
2138 * @see UBiDiClassCallback
2139 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
2141 #define U_BIDI_CLASS_DEFAULT U_CHAR_DIRECTION_COUNT
2142 #endif // U_HIDE_DEPRECATED_API
2145 * Callback type declaration for overriding default Bidi class values with
2147 * <p>Usually, the function pointer will be propagated to a <code>UBiDi</code>
2148 * object by calling the <code>ubidi_setClassCallback()</code> function;
2149 * then the callback will be invoked by the UBA implementation any time the
2150 * class of a character is to be determined.</p>
2152 * @param context is a pointer to the callback private data.
2154 * @param c is the code point to get a Bidi class for.
2156 * @return The directional property / Bidi class for the given code point
2157 * <code>c</code> if the default class has been overridden, or
2158 * <code>u_getIntPropertyMaxValue(UCHAR_BIDI_CLASS)+1</code>
2159 * if the standard Bidi class value for <code>c</code> is to be used.
2160 * @see ubidi_setClassCallback
2161 * @see ubidi_getClassCallback
2164 typedef UCharDirection U_CALLCONV
2165 UBiDiClassCallback(const void *context
, UChar32 c
);
2170 * Retrieve the Bidi class for a given code point.
2171 * <p>If a <code>#UBiDiClassCallback</code> callback is defined and returns a
2172 * value other than <code>u_getIntPropertyMaxValue(UCHAR_BIDI_CLASS)+1</code>,
2173 * that value is used; otherwise the default class determination mechanism is invoked.</p>
2175 * @param pBiDi is the paragraph <code>UBiDi</code> object.
2177 * @param c is the code point whose Bidi class must be retrieved.
2179 * @return The Bidi class for character <code>c</code> based
2180 * on the given <code>pBiDi</code> instance.
2181 * @see UBiDiClassCallback
2184 U_STABLE UCharDirection U_EXPORT2
2185 ubidi_getCustomizedClass(UBiDi
*pBiDi
, UChar32 c
);
2188 * Set the callback function and callback data used by the UBA
2189 * implementation for Bidi class determination.
2190 * <p>This may be useful for assigning Bidi classes to PUA characters, or
2191 * for special application needs. For instance, an application may want to
2192 * handle all spaces like L or R characters (according to the base direction)
2193 * when creating the visual ordering of logical lines which are part of a report
2194 * organized in columns: there should not be interaction between adjacent
2197 * @param pBiDi is the paragraph <code>UBiDi</code> object.
2199 * @param newFn is the new callback function pointer.
2201 * @param newContext is the new callback context pointer. This can be NULL.
2203 * @param oldFn fillin: Returns the old callback function pointer. This can be
2206 * @param oldContext fillin: Returns the old callback's context. This can be
2209 * @param pErrorCode must be a valid pointer to an error code value.
2211 * @see ubidi_getClassCallback
2214 U_STABLE
void U_EXPORT2
2215 ubidi_setClassCallback(UBiDi
*pBiDi
, UBiDiClassCallback
*newFn
,
2216 const void *newContext
, UBiDiClassCallback
**oldFn
,
2217 const void **oldContext
, UErrorCode
*pErrorCode
);
2220 * Get the current callback function used for Bidi class determination.
2222 * @param pBiDi is the paragraph <code>UBiDi</code> object.
2224 * @param fn fillin: Returns the callback function pointer.
2226 * @param context fillin: Returns the callback's private context.
2228 * @see ubidi_setClassCallback
2231 U_STABLE
void U_EXPORT2
2232 ubidi_getClassCallback(UBiDi
*pBiDi
, UBiDiClassCallback
**fn
, const void **context
);
2235 * Take a <code>UBiDi</code> object containing the reordering
2236 * information for a piece of text (one or more paragraphs) set by
2237 * <code>ubidi_setPara()</code> or for a line of text set by
2238 * <code>ubidi_setLine()</code> and write a reordered string to the
2239 * destination buffer.
2241 * This function preserves the integrity of characters with multiple
2242 * code units and (optionally) combining characters.
2243 * Characters in RTL runs can be replaced by mirror-image characters
2244 * in the destination buffer. Note that "real" mirroring has
2245 * to be done in a rendering engine by glyph selection
2246 * and that for many "mirrored" characters there are no
2247 * Unicode characters as mirror-image equivalents.
2248 * There are also options to insert or remove Bidi control
2249 * characters; see the description of the <code>destSize</code>
2250 * and <code>options</code> parameters and of the option bit flags.
2252 * @param pBiDi A pointer to a <code>UBiDi</code> object that
2253 * is set by <code>ubidi_setPara()</code> or
2254 * <code>ubidi_setLine()</code> and contains the reordering
2255 * information for the text that it was defined for,
2256 * as well as a pointer to that text.<br><br>
2257 * The text was aliased (only the pointer was stored
2258 * without copying the contents) and must not have been modified
2259 * since the <code>ubidi_setPara()</code> call.
2261 * @param dest A pointer to where the reordered text is to be copied.
2262 * The source text and <code>dest[destSize]</code>
2265 * @param destSize The size of the <code>dest</code> buffer,
2266 * in number of UChars.
2267 * If the <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>
2268 * option is set, then the destination length could be
2270 * <code>ubidi_getLength(pBiDi)+2*ubidi_countRuns(pBiDi)</code>.
2271 * If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option
2272 * is set, then the destination length may be less than
2273 * <code>ubidi_getLength(pBiDi)</code>.
2274 * If none of these options is set, then the destination length
2275 * will be exactly <code>ubidi_getProcessedLength(pBiDi)</code>.
2277 * @param options A bit set of options for the reordering that control
2278 * how the reordered text is written.
2279 * The options include mirroring the characters on a code
2280 * point basis and inserting LRM characters, which is used
2281 * especially for transforming visually stored text
2282 * to logically stored text (although this is still an
2283 * imperfect implementation of an "inverse Bidi" algorithm
2284 * because it uses the "forward Bidi" algorithm at its core).
2285 * The available options are:
2286 * <code>#UBIDI_DO_MIRRORING</code>,
2287 * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
2288 * <code>#UBIDI_KEEP_BASE_COMBINING</code>,
2289 * <code>#UBIDI_OUTPUT_REVERSE</code>,
2290 * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code>
2292 * @param pErrorCode must be a valid pointer to an error code value.
2294 * @return The length of the output string.
2296 * @see ubidi_getProcessedLength
2299 U_STABLE
int32_t U_EXPORT2
2300 ubidi_writeReordered(UBiDi
*pBiDi
,
2301 UChar
*dest
, int32_t destSize
,
2303 UErrorCode
*pErrorCode
);
2306 * Reverse a Right-To-Left run of Unicode text.
2308 * This function preserves the integrity of characters with multiple
2309 * code units and (optionally) combining characters.
2310 * Characters can be replaced by mirror-image characters
2311 * in the destination buffer. Note that "real" mirroring has
2312 * to be done in a rendering engine by glyph selection
2313 * and that for many "mirrored" characters there are no
2314 * Unicode characters as mirror-image equivalents.
2315 * There are also options to insert or remove Bidi control
2318 * This function is the implementation for reversing RTL runs as part
2319 * of <code>ubidi_writeReordered()</code>. For detailed descriptions
2320 * of the parameters, see there.
2321 * Since no Bidi controls are inserted here, the output string length
2322 * will never exceed <code>srcLength</code>.
2324 * @see ubidi_writeReordered
2326 * @param src A pointer to the RTL run text.
2328 * @param srcLength The length of the RTL run.
2330 * @param dest A pointer to where the reordered text is to be copied.
2331 * <code>src[srcLength]</code> and <code>dest[destSize]</code>
2334 * @param destSize The size of the <code>dest</code> buffer,
2335 * in number of UChars.
2336 * If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option
2337 * is set, then the destination length may be less than
2338 * <code>srcLength</code>.
2339 * If this option is not set, then the destination length
2340 * will be exactly <code>srcLength</code>.
2342 * @param options A bit set of options for the reordering that control
2343 * how the reordered text is written.
2344 * See the <code>options</code> parameter in <code>ubidi_writeReordered()</code>.
2346 * @param pErrorCode must be a valid pointer to an error code value.
2348 * @return The length of the output string.
2351 U_STABLE
int32_t U_EXPORT2
2352 ubidi_writeReverse(const UChar
*src
, int32_t srcLength
,
2353 UChar
*dest
, int32_t destSize
,
2355 UErrorCode
*pErrorCode
);
2357 /*#define BIDI_SAMPLE_CODE*/