Wed, 31 Dec 2014 06:09:35 +0100
Cloned upstream origin tor-browser at tor-browser-31.3.0esr-4.5-1-build1
revision ID fc1c9ff7c1b2defdbc039f12214767608f46423f for hacking purpose.
michael@0 | 1 | /* |
michael@0 | 2 | ****************************************************************************** |
michael@0 | 3 | * |
michael@0 | 4 | * Copyright (C) 1999-2013, International Business Machines |
michael@0 | 5 | * Corporation and others. All Rights Reserved. |
michael@0 | 6 | * |
michael@0 | 7 | ****************************************************************************** |
michael@0 | 8 | * file name: ubidi.h |
michael@0 | 9 | * encoding: US-ASCII |
michael@0 | 10 | * tab size: 8 (not used) |
michael@0 | 11 | * indentation:4 |
michael@0 | 12 | * |
michael@0 | 13 | * created on: 1999jul27 |
michael@0 | 14 | * created by: Markus W. Scherer, updated by Matitiahu Allouche |
michael@0 | 15 | */ |
michael@0 | 16 | |
michael@0 | 17 | #ifndef UBIDI_H |
michael@0 | 18 | #define UBIDI_H |
michael@0 | 19 | |
michael@0 | 20 | #include "unicode/utypes.h" |
michael@0 | 21 | #include "unicode/uchar.h" |
michael@0 | 22 | #include "unicode/localpointer.h" |
michael@0 | 23 | |
michael@0 | 24 | /** |
michael@0 | 25 | *\file |
michael@0 | 26 | * \brief C API: Bidi algorithm |
michael@0 | 27 | * |
michael@0 | 28 | * <h2>Bidi algorithm for ICU</h2> |
michael@0 | 29 | * |
michael@0 | 30 | * This is an implementation of the Unicode Bidirectional Algorithm. |
michael@0 | 31 | * The algorithm is defined in the |
michael@0 | 32 | * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Annex #9</a>.<p> |
michael@0 | 33 | * |
michael@0 | 34 | * Note: Libraries that perform a bidirectional algorithm and |
michael@0 | 35 | * reorder strings accordingly are sometimes called "Storage Layout Engines". |
michael@0 | 36 | * ICU's Bidi and shaping (u_shapeArabic()) APIs can be used at the core of such |
michael@0 | 37 | * "Storage Layout Engines". |
michael@0 | 38 | * |
michael@0 | 39 | * <h3>General remarks about the API:</h3> |
michael@0 | 40 | * |
michael@0 | 41 | * In functions with an error code parameter, |
michael@0 | 42 | * the <code>pErrorCode</code> pointer must be valid |
michael@0 | 43 | * and the value that it points to must not indicate a failure before |
michael@0 | 44 | * the function call. Otherwise, the function returns immediately. |
michael@0 | 45 | * After the function call, the value indicates success or failure.<p> |
michael@0 | 46 | * |
michael@0 | 47 | * The "limit" of a sequence of characters is the position just after their |
michael@0 | 48 | * last character, i.e., one more than that position.<p> |
michael@0 | 49 | * |
michael@0 | 50 | * Some of the API functions provide access to "runs". |
michael@0 | 51 | * Such a "run" is defined as a sequence of characters |
michael@0 | 52 | * that are at the same embedding level |
michael@0 | 53 | * after performing the Bidi algorithm.<p> |
michael@0 | 54 | * |
michael@0 | 55 | * @author Markus W. Scherer |
michael@0 | 56 | * @version 1.0 |
michael@0 | 57 | * |
michael@0 | 58 | * |
michael@0 | 59 | * <h4> Sample code for the ICU Bidi API </h4> |
michael@0 | 60 | * |
michael@0 | 61 | * <h5>Rendering a paragraph with the ICU Bidi API</h5> |
michael@0 | 62 | * |
michael@0 | 63 | * This is (hypothetical) sample code that illustrates |
michael@0 | 64 | * how the ICU Bidi API could be used to render a paragraph of text. |
michael@0 | 65 | * Rendering code depends highly on the graphics system, |
michael@0 | 66 | * therefore this sample code must make a lot of assumptions, |
michael@0 | 67 | * which may or may not match any existing graphics system's properties. |
michael@0 | 68 | * |
michael@0 | 69 | * <p>The basic assumptions are:</p> |
michael@0 | 70 | * <ul> |
michael@0 | 71 | * <li>Rendering is done from left to right on a horizontal line.</li> |
michael@0 | 72 | * <li>A run of single-style, unidirectional text can be rendered at once.</li> |
michael@0 | 73 | * <li>Such a run of text is passed to the graphics system with |
michael@0 | 74 | * characters (code units) in logical order.</li> |
michael@0 | 75 | * <li>The line-breaking algorithm is very complicated |
michael@0 | 76 | * and Locale-dependent - |
michael@0 | 77 | * and therefore its implementation omitted from this sample code.</li> |
michael@0 | 78 | * </ul> |
michael@0 | 79 | * |
michael@0 | 80 | * <pre> |
michael@0 | 81 | * \code |
michael@0 | 82 | *#include "unicode/ubidi.h" |
michael@0 | 83 | * |
michael@0 | 84 | *typedef enum { |
michael@0 | 85 | * styleNormal=0, styleSelected=1, |
michael@0 | 86 | * styleBold=2, styleItalics=4, |
michael@0 | 87 | * styleSuper=8, styleSub=16 |
michael@0 | 88 | *} Style; |
michael@0 | 89 | * |
michael@0 | 90 | *typedef struct { int32_t limit; Style style; } StyleRun; |
michael@0 | 91 | * |
michael@0 | 92 | *int getTextWidth(const UChar *text, int32_t start, int32_t limit, |
michael@0 | 93 | * const StyleRun *styleRuns, int styleRunCount); |
michael@0 | 94 | * |
michael@0 | 95 | * // set *pLimit and *pStyleRunLimit for a line |
michael@0 | 96 | * // from text[start] and from styleRuns[styleRunStart] |
michael@0 | 97 | * // using ubidi_getLogicalRun(para, ...) |
michael@0 | 98 | *void getLineBreak(const UChar *text, int32_t start, int32_t *pLimit, |
michael@0 | 99 | * UBiDi *para, |
michael@0 | 100 | * const StyleRun *styleRuns, int styleRunStart, int *pStyleRunLimit, |
michael@0 | 101 | * int *pLineWidth); |
michael@0 | 102 | * |
michael@0 | 103 | * // render runs on a line sequentially, always from left to right |
michael@0 | 104 | * |
michael@0 | 105 | * // prepare rendering a new line |
michael@0 | 106 | * void startLine(UBiDiDirection textDirection, int lineWidth); |
michael@0 | 107 | * |
michael@0 | 108 | * // render a run of text and advance to the right by the run width |
michael@0 | 109 | * // the text[start..limit-1] is always in logical order |
michael@0 | 110 | * void renderRun(const UChar *text, int32_t start, int32_t limit, |
michael@0 | 111 | * UBiDiDirection textDirection, Style style); |
michael@0 | 112 | * |
michael@0 | 113 | * // We could compute a cross-product |
michael@0 | 114 | * // from the style runs with the directional runs |
michael@0 | 115 | * // and then reorder it. |
michael@0 | 116 | * // Instead, here we iterate over each run type |
michael@0 | 117 | * // and render the intersections - |
michael@0 | 118 | * // with shortcuts in simple (and common) cases. |
michael@0 | 119 | * // renderParagraph() is the main function. |
michael@0 | 120 | * |
michael@0 | 121 | * // render a directional run with |
michael@0 | 122 | * // (possibly) multiple style runs intersecting with it |
michael@0 | 123 | * void renderDirectionalRun(const UChar *text, |
michael@0 | 124 | * int32_t start, int32_t limit, |
michael@0 | 125 | * UBiDiDirection direction, |
michael@0 | 126 | * const StyleRun *styleRuns, int styleRunCount) { |
michael@0 | 127 | * int i; |
michael@0 | 128 | * |
michael@0 | 129 | * // iterate over style runs |
michael@0 | 130 | * if(direction==UBIDI_LTR) { |
michael@0 | 131 | * int styleLimit; |
michael@0 | 132 | * |
michael@0 | 133 | * for(i=0; i<styleRunCount; ++i) { |
michael@0 | 134 | * styleLimit=styleRun[i].limit; |
michael@0 | 135 | * if(start<styleLimit) { |
michael@0 | 136 | * if(styleLimit>limit) { styleLimit=limit; } |
michael@0 | 137 | * renderRun(text, start, styleLimit, |
michael@0 | 138 | * direction, styleRun[i].style); |
michael@0 | 139 | * if(styleLimit==limit) { break; } |
michael@0 | 140 | * start=styleLimit; |
michael@0 | 141 | * } |
michael@0 | 142 | * } |
michael@0 | 143 | * } else { |
michael@0 | 144 | * int styleStart; |
michael@0 | 145 | * |
michael@0 | 146 | * for(i=styleRunCount-1; i>=0; --i) { |
michael@0 | 147 | * if(i>0) { |
michael@0 | 148 | * styleStart=styleRun[i-1].limit; |
michael@0 | 149 | * } else { |
michael@0 | 150 | * styleStart=0; |
michael@0 | 151 | * } |
michael@0 | 152 | * if(limit>=styleStart) { |
michael@0 | 153 | * if(styleStart<start) { styleStart=start; } |
michael@0 | 154 | * renderRun(text, styleStart, limit, |
michael@0 | 155 | * direction, styleRun[i].style); |
michael@0 | 156 | * if(styleStart==start) { break; } |
michael@0 | 157 | * limit=styleStart; |
michael@0 | 158 | * } |
michael@0 | 159 | * } |
michael@0 | 160 | * } |
michael@0 | 161 | * } |
michael@0 | 162 | * |
michael@0 | 163 | * // the line object represents text[start..limit-1] |
michael@0 | 164 | * void renderLine(UBiDi *line, const UChar *text, |
michael@0 | 165 | * int32_t start, int32_t limit, |
michael@0 | 166 | * const StyleRun *styleRuns, int styleRunCount) { |
michael@0 | 167 | * UBiDiDirection direction=ubidi_getDirection(line); |
michael@0 | 168 | * if(direction!=UBIDI_MIXED) { |
michael@0 | 169 | * // unidirectional |
michael@0 | 170 | * if(styleRunCount<=1) { |
michael@0 | 171 | * renderRun(text, start, limit, direction, styleRuns[0].style); |
michael@0 | 172 | * } else { |
michael@0 | 173 | * renderDirectionalRun(text, start, limit, |
michael@0 | 174 | * direction, styleRuns, styleRunCount); |
michael@0 | 175 | * } |
michael@0 | 176 | * } else { |
michael@0 | 177 | * // mixed-directional |
michael@0 | 178 | * int32_t count, i, length; |
michael@0 | 179 | * UBiDiLevel level; |
michael@0 | 180 | * |
michael@0 | 181 | * count=ubidi_countRuns(para, pErrorCode); |
michael@0 | 182 | * if(U_SUCCESS(*pErrorCode)) { |
michael@0 | 183 | * if(styleRunCount<=1) { |
michael@0 | 184 | * Style style=styleRuns[0].style; |
michael@0 | 185 | * |
michael@0 | 186 | * // iterate over directional runs |
michael@0 | 187 | * for(i=0; i<count; ++i) { |
michael@0 | 188 | * direction=ubidi_getVisualRun(para, i, &start, &length); |
michael@0 | 189 | * renderRun(text, start, start+length, direction, style); |
michael@0 | 190 | * } |
michael@0 | 191 | * } else { |
michael@0 | 192 | * int32_t j; |
michael@0 | 193 | * |
michael@0 | 194 | * // iterate over both directional and style runs |
michael@0 | 195 | * for(i=0; i<count; ++i) { |
michael@0 | 196 | * direction=ubidi_getVisualRun(line, i, &start, &length); |
michael@0 | 197 | * renderDirectionalRun(text, start, start+length, |
michael@0 | 198 | * direction, styleRuns, styleRunCount); |
michael@0 | 199 | * } |
michael@0 | 200 | * } |
michael@0 | 201 | * } |
michael@0 | 202 | * } |
michael@0 | 203 | * } |
michael@0 | 204 | * |
michael@0 | 205 | *void renderParagraph(const UChar *text, int32_t length, |
michael@0 | 206 | * UBiDiDirection textDirection, |
michael@0 | 207 | * const StyleRun *styleRuns, int styleRunCount, |
michael@0 | 208 | * int lineWidth, |
michael@0 | 209 | * UErrorCode *pErrorCode) { |
michael@0 | 210 | * UBiDi *para; |
michael@0 | 211 | * |
michael@0 | 212 | * if(pErrorCode==NULL || U_FAILURE(*pErrorCode) || length<=0) { |
michael@0 | 213 | * return; |
michael@0 | 214 | * } |
michael@0 | 215 | * |
michael@0 | 216 | * para=ubidi_openSized(length, 0, pErrorCode); |
michael@0 | 217 | * if(para==NULL) { return; } |
michael@0 | 218 | * |
michael@0 | 219 | * ubidi_setPara(para, text, length, |
michael@0 | 220 | * textDirection ? UBIDI_DEFAULT_RTL : UBIDI_DEFAULT_LTR, |
michael@0 | 221 | * NULL, pErrorCode); |
michael@0 | 222 | * if(U_SUCCESS(*pErrorCode)) { |
michael@0 | 223 | * UBiDiLevel paraLevel=1&ubidi_getParaLevel(para); |
michael@0 | 224 | * StyleRun styleRun={ length, styleNormal }; |
michael@0 | 225 | * int width; |
michael@0 | 226 | * |
michael@0 | 227 | * if(styleRuns==NULL || styleRunCount<=0) { |
michael@0 | 228 | * styleRunCount=1; |
michael@0 | 229 | * styleRuns=&styleRun; |
michael@0 | 230 | * } |
michael@0 | 231 | * |
michael@0 | 232 | * // assume styleRuns[styleRunCount-1].limit>=length |
michael@0 | 233 | * |
michael@0 | 234 | * width=getTextWidth(text, 0, length, styleRuns, styleRunCount); |
michael@0 | 235 | * if(width<=lineWidth) { |
michael@0 | 236 | * // everything fits onto one line |
michael@0 | 237 | * |
michael@0 | 238 | * // prepare rendering a new line from either left or right |
michael@0 | 239 | * startLine(paraLevel, width); |
michael@0 | 240 | * |
michael@0 | 241 | * renderLine(para, text, 0, length, |
michael@0 | 242 | * styleRuns, styleRunCount); |
michael@0 | 243 | * } else { |
michael@0 | 244 | * UBiDi *line; |
michael@0 | 245 | * |
michael@0 | 246 | * // we need to render several lines |
michael@0 | 247 | * line=ubidi_openSized(length, 0, pErrorCode); |
michael@0 | 248 | * if(line!=NULL) { |
michael@0 | 249 | * int32_t start=0, limit; |
michael@0 | 250 | * int styleRunStart=0, styleRunLimit; |
michael@0 | 251 | * |
michael@0 | 252 | * for(;;) { |
michael@0 | 253 | * limit=length; |
michael@0 | 254 | * styleRunLimit=styleRunCount; |
michael@0 | 255 | * getLineBreak(text, start, &limit, para, |
michael@0 | 256 | * styleRuns, styleRunStart, &styleRunLimit, |
michael@0 | 257 | * &width); |
michael@0 | 258 | * ubidi_setLine(para, start, limit, line, pErrorCode); |
michael@0 | 259 | * if(U_SUCCESS(*pErrorCode)) { |
michael@0 | 260 | * // prepare rendering a new line |
michael@0 | 261 | * // from either left or right |
michael@0 | 262 | * startLine(paraLevel, width); |
michael@0 | 263 | * |
michael@0 | 264 | * renderLine(line, text, start, limit, |
michael@0 | 265 | * styleRuns+styleRunStart, |
michael@0 | 266 | * styleRunLimit-styleRunStart); |
michael@0 | 267 | * } |
michael@0 | 268 | * if(limit==length) { break; } |
michael@0 | 269 | * start=limit; |
michael@0 | 270 | * styleRunStart=styleRunLimit-1; |
michael@0 | 271 | * if(start>=styleRuns[styleRunStart].limit) { |
michael@0 | 272 | * ++styleRunStart; |
michael@0 | 273 | * } |
michael@0 | 274 | * } |
michael@0 | 275 | * |
michael@0 | 276 | * ubidi_close(line); |
michael@0 | 277 | * } |
michael@0 | 278 | * } |
michael@0 | 279 | * } |
michael@0 | 280 | * |
michael@0 | 281 | * ubidi_close(para); |
michael@0 | 282 | *} |
michael@0 | 283 | *\endcode |
michael@0 | 284 | * </pre> |
michael@0 | 285 | */ |
michael@0 | 286 | |
michael@0 | 287 | /*DOCXX_TAG*/ |
michael@0 | 288 | /*@{*/ |
michael@0 | 289 | |
michael@0 | 290 | /** |
michael@0 | 291 | * UBiDiLevel is the type of the level values in this |
michael@0 | 292 | * Bidi implementation. |
michael@0 | 293 | * It holds an embedding level and indicates the visual direction |
michael@0 | 294 | * by its bit 0 (even/odd value).<p> |
michael@0 | 295 | * |
michael@0 | 296 | * It can also hold non-level values for the |
michael@0 | 297 | * <code>paraLevel</code> and <code>embeddingLevels</code> |
michael@0 | 298 | * arguments of <code>ubidi_setPara()</code>; there: |
michael@0 | 299 | * <ul> |
michael@0 | 300 | * <li>bit 7 of an <code>embeddingLevels[]</code> |
michael@0 | 301 | * value indicates whether the using application is |
michael@0 | 302 | * specifying the level of a character to <i>override</i> whatever the |
michael@0 | 303 | * Bidi implementation would resolve it to.</li> |
michael@0 | 304 | * <li><code>paraLevel</code> can be set to the |
michael@0 | 305 | * pseudo-level values <code>UBIDI_DEFAULT_LTR</code> |
michael@0 | 306 | * and <code>UBIDI_DEFAULT_RTL</code>.</li> |
michael@0 | 307 | * </ul> |
michael@0 | 308 | * |
michael@0 | 309 | * @see ubidi_setPara |
michael@0 | 310 | * |
michael@0 | 311 | * <p>The related constants are not real, valid level values. |
michael@0 | 312 | * <code>UBIDI_DEFAULT_XXX</code> can be used to specify |
michael@0 | 313 | * a default for the paragraph level for |
michael@0 | 314 | * when the <code>ubidi_setPara()</code> function |
michael@0 | 315 | * shall determine it but there is no |
michael@0 | 316 | * strongly typed character in the input.<p> |
michael@0 | 317 | * |
michael@0 | 318 | * Note that the value for <code>UBIDI_DEFAULT_LTR</code> is even |
michael@0 | 319 | * and the one for <code>UBIDI_DEFAULT_RTL</code> is odd, |
michael@0 | 320 | * just like with normal LTR and RTL level values - |
michael@0 | 321 | * these special values are designed that way. Also, the implementation |
michael@0 | 322 | * assumes that UBIDI_MAX_EXPLICIT_LEVEL is odd. |
michael@0 | 323 | * |
michael@0 | 324 | * @see UBIDI_DEFAULT_LTR |
michael@0 | 325 | * @see UBIDI_DEFAULT_RTL |
michael@0 | 326 | * @see UBIDI_LEVEL_OVERRIDE |
michael@0 | 327 | * @see UBIDI_MAX_EXPLICIT_LEVEL |
michael@0 | 328 | * @stable ICU 2.0 |
michael@0 | 329 | */ |
michael@0 | 330 | typedef uint8_t UBiDiLevel; |
michael@0 | 331 | |
michael@0 | 332 | /** Paragraph level setting.<p> |
michael@0 | 333 | * |
michael@0 | 334 | * Constant indicating that the base direction depends on the first strong |
michael@0 | 335 | * directional character in the text according to the Unicode Bidirectional |
michael@0 | 336 | * Algorithm. If no strong directional character is present, |
michael@0 | 337 | * then set the paragraph level to 0 (left-to-right).<p> |
michael@0 | 338 | * |
michael@0 | 339 | * If this value is used in conjunction with reordering modes |
michael@0 | 340 | * <code>UBIDI_REORDER_INVERSE_LIKE_DIRECT</code> or |
michael@0 | 341 | * <code>UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the text to reorder |
michael@0 | 342 | * is assumed to be visual LTR, and the text after reordering is required |
michael@0 | 343 | * to be the corresponding logical string with appropriate contextual |
michael@0 | 344 | * direction. The direction of the result string will be RTL if either |
michael@0 | 345 | * the righmost or leftmost strong character of the source text is RTL |
michael@0 | 346 | * or Arabic Letter, the direction will be LTR otherwise.<p> |
michael@0 | 347 | * |
michael@0 | 348 | * If reordering option <code>UBIDI_OPTION_INSERT_MARKS</code> is set, an RLM may |
michael@0 | 349 | * be added at the beginning of the result string to ensure round trip |
michael@0 | 350 | * (that the result string, when reordered back to visual, will produce |
michael@0 | 351 | * the original source text). |
michael@0 | 352 | * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT |
michael@0 | 353 | * @see UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL |
michael@0 | 354 | * @stable ICU 2.0 |
michael@0 | 355 | */ |
michael@0 | 356 | #define UBIDI_DEFAULT_LTR 0xfe |
michael@0 | 357 | |
michael@0 | 358 | /** Paragraph level setting.<p> |
michael@0 | 359 | * |
michael@0 | 360 | * Constant indicating that the base direction depends on the first strong |
michael@0 | 361 | * directional character in the text according to the Unicode Bidirectional |
michael@0 | 362 | * Algorithm. If no strong directional character is present, |
michael@0 | 363 | * then set the paragraph level to 1 (right-to-left).<p> |
michael@0 | 364 | * |
michael@0 | 365 | * If this value is used in conjunction with reordering modes |
michael@0 | 366 | * <code>UBIDI_REORDER_INVERSE_LIKE_DIRECT</code> or |
michael@0 | 367 | * <code>UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the text to reorder |
michael@0 | 368 | * is assumed to be visual LTR, and the text after reordering is required |
michael@0 | 369 | * to be the corresponding logical string with appropriate contextual |
michael@0 | 370 | * direction. The direction of the result string will be RTL if either |
michael@0 | 371 | * the righmost or leftmost strong character of the source text is RTL |
michael@0 | 372 | * or Arabic Letter, or if the text contains no strong character; |
michael@0 | 373 | * the direction will be LTR otherwise.<p> |
michael@0 | 374 | * |
michael@0 | 375 | * If reordering option <code>UBIDI_OPTION_INSERT_MARKS</code> is set, an RLM may |
michael@0 | 376 | * be added at the beginning of the result string to ensure round trip |
michael@0 | 377 | * (that the result string, when reordered back to visual, will produce |
michael@0 | 378 | * the original source text). |
michael@0 | 379 | * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT |
michael@0 | 380 | * @see UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL |
michael@0 | 381 | * @stable ICU 2.0 |
michael@0 | 382 | */ |
michael@0 | 383 | #define UBIDI_DEFAULT_RTL 0xff |
michael@0 | 384 | |
michael@0 | 385 | /** |
michael@0 | 386 | * Maximum explicit embedding level. |
michael@0 | 387 | * (The maximum resolved level can be up to <code>UBIDI_MAX_EXPLICIT_LEVEL+1</code>). |
michael@0 | 388 | * @stable ICU 2.0 |
michael@0 | 389 | */ |
michael@0 | 390 | #define UBIDI_MAX_EXPLICIT_LEVEL 125 |
michael@0 | 391 | |
michael@0 | 392 | /** Bit flag for level input. |
michael@0 | 393 | * Overrides directional properties. |
michael@0 | 394 | * @stable ICU 2.0 |
michael@0 | 395 | */ |
michael@0 | 396 | #define UBIDI_LEVEL_OVERRIDE 0x80 |
michael@0 | 397 | |
michael@0 | 398 | /** |
michael@0 | 399 | * Special value which can be returned by the mapping functions when a logical |
michael@0 | 400 | * index has no corresponding visual index or vice-versa. This may happen |
michael@0 | 401 | * for the logical-to-visual mapping of a Bidi control when option |
michael@0 | 402 | * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is specified. This can also happen |
michael@0 | 403 | * for the visual-to-logical mapping of a Bidi mark (LRM or RLM) inserted |
michael@0 | 404 | * by option <code>#UBIDI_OPTION_INSERT_MARKS</code>. |
michael@0 | 405 | * @see ubidi_getVisualIndex |
michael@0 | 406 | * @see ubidi_getVisualMap |
michael@0 | 407 | * @see ubidi_getLogicalIndex |
michael@0 | 408 | * @see ubidi_getLogicalMap |
michael@0 | 409 | * @stable ICU 3.6 |
michael@0 | 410 | */ |
michael@0 | 411 | #define UBIDI_MAP_NOWHERE (-1) |
michael@0 | 412 | |
michael@0 | 413 | /** |
michael@0 | 414 | * <code>UBiDiDirection</code> values indicate the text direction. |
michael@0 | 415 | * @stable ICU 2.0 |
michael@0 | 416 | */ |
michael@0 | 417 | enum UBiDiDirection { |
michael@0 | 418 | /** Left-to-right text. This is a 0 value. |
michael@0 | 419 | * <ul> |
michael@0 | 420 | * <li>As return value for <code>ubidi_getDirection()</code>, it means |
michael@0 | 421 | * that the source string contains no right-to-left characters, or |
michael@0 | 422 | * that the source string is empty and the paragraph level is even. |
michael@0 | 423 | * <li> As return value for <code>ubidi_getBaseDirection()</code>, it |
michael@0 | 424 | * means that the first strong character of the source string has |
michael@0 | 425 | * a left-to-right direction. |
michael@0 | 426 | * </ul> |
michael@0 | 427 | * @stable ICU 2.0 |
michael@0 | 428 | */ |
michael@0 | 429 | UBIDI_LTR, |
michael@0 | 430 | /** Right-to-left text. This is a 1 value. |
michael@0 | 431 | * <ul> |
michael@0 | 432 | * <li>As return value for <code>ubidi_getDirection()</code>, it means |
michael@0 | 433 | * that the source string contains no left-to-right characters, or |
michael@0 | 434 | * that the source string is empty and the paragraph level is odd. |
michael@0 | 435 | * <li> As return value for <code>ubidi_getBaseDirection()</code>, it |
michael@0 | 436 | * means that the first strong character of the source string has |
michael@0 | 437 | * a right-to-left direction. |
michael@0 | 438 | * </ul> |
michael@0 | 439 | * @stable ICU 2.0 |
michael@0 | 440 | */ |
michael@0 | 441 | UBIDI_RTL, |
michael@0 | 442 | /** Mixed-directional text. |
michael@0 | 443 | * <p>As return value for <code>ubidi_getDirection()</code>, it means |
michael@0 | 444 | * that the source string contains both left-to-right and |
michael@0 | 445 | * right-to-left characters. |
michael@0 | 446 | * @stable ICU 2.0 |
michael@0 | 447 | */ |
michael@0 | 448 | UBIDI_MIXED, |
michael@0 | 449 | /** No strongly directional text. |
michael@0 | 450 | * <p>As return value for <code>ubidi_getBaseDirection()</code>, it means |
michael@0 | 451 | * that the source string is missing or empty, or contains neither left-to-right |
michael@0 | 452 | * nor right-to-left characters. |
michael@0 | 453 | * @stable ICU 4.6 |
michael@0 | 454 | */ |
michael@0 | 455 | UBIDI_NEUTRAL |
michael@0 | 456 | }; |
michael@0 | 457 | |
michael@0 | 458 | /** @stable ICU 2.0 */ |
michael@0 | 459 | typedef enum UBiDiDirection UBiDiDirection; |
michael@0 | 460 | |
michael@0 | 461 | /** |
michael@0 | 462 | * Forward declaration of the <code>UBiDi</code> structure for the declaration of |
michael@0 | 463 | * the API functions. Its fields are implementation-specific.<p> |
michael@0 | 464 | * This structure holds information about a paragraph (or multiple paragraphs) |
michael@0 | 465 | * of text with Bidi-algorithm-related details, or about one line of |
michael@0 | 466 | * such a paragraph.<p> |
michael@0 | 467 | * Reordering can be done on a line, or on one or more paragraphs which are |
michael@0 | 468 | * then interpreted each as one single line. |
michael@0 | 469 | * @stable ICU 2.0 |
michael@0 | 470 | */ |
michael@0 | 471 | struct UBiDi; |
michael@0 | 472 | |
michael@0 | 473 | /** @stable ICU 2.0 */ |
michael@0 | 474 | typedef struct UBiDi UBiDi; |
michael@0 | 475 | |
michael@0 | 476 | /** |
michael@0 | 477 | * Allocate a <code>UBiDi</code> structure. |
michael@0 | 478 | * Such an object is initially empty. It is assigned |
michael@0 | 479 | * the Bidi properties of a piece of text containing one or more paragraphs |
michael@0 | 480 | * by <code>ubidi_setPara()</code> |
michael@0 | 481 | * or the Bidi properties of a line within a paragraph by |
michael@0 | 482 | * <code>ubidi_setLine()</code>.<p> |
michael@0 | 483 | * This object can be reused for as long as it is not deallocated |
michael@0 | 484 | * by calling <code>ubidi_close()</code>.<p> |
michael@0 | 485 | * <code>ubidi_setPara()</code> and <code>ubidi_setLine()</code> will allocate |
michael@0 | 486 | * additional memory for internal structures as necessary. |
michael@0 | 487 | * |
michael@0 | 488 | * @return An empty <code>UBiDi</code> object. |
michael@0 | 489 | * @stable ICU 2.0 |
michael@0 | 490 | */ |
michael@0 | 491 | U_STABLE UBiDi * U_EXPORT2 |
michael@0 | 492 | ubidi_open(void); |
michael@0 | 493 | |
michael@0 | 494 | /** |
michael@0 | 495 | * Allocate a <code>UBiDi</code> structure with preallocated memory |
michael@0 | 496 | * for internal structures. |
michael@0 | 497 | * This function provides a <code>UBiDi</code> object like <code>ubidi_open()</code> |
michael@0 | 498 | * with no arguments, but it also preallocates memory for internal structures |
michael@0 | 499 | * according to the sizings supplied by the caller.<p> |
michael@0 | 500 | * Subsequent functions will not allocate any more memory, and are thus |
michael@0 | 501 | * guaranteed not to fail because of lack of memory.<p> |
michael@0 | 502 | * The preallocation can be limited to some of the internal memory |
michael@0 | 503 | * by setting some values to 0 here. That means that if, e.g., |
michael@0 | 504 | * <code>maxRunCount</code> cannot be reasonably predetermined and should not |
michael@0 | 505 | * be set to <code>maxLength</code> (the only failproof value) to avoid |
michael@0 | 506 | * wasting memory, then <code>maxRunCount</code> could be set to 0 here |
michael@0 | 507 | * and the internal structures that are associated with it will be allocated |
michael@0 | 508 | * on demand, just like with <code>ubidi_open()</code>. |
michael@0 | 509 | * |
michael@0 | 510 | * @param maxLength is the maximum text or line length that internal memory |
michael@0 | 511 | * will be preallocated for. An attempt to associate this object with a |
michael@0 | 512 | * longer text will fail, unless this value is 0, which leaves the allocation |
michael@0 | 513 | * up to the implementation. |
michael@0 | 514 | * |
michael@0 | 515 | * @param maxRunCount is the maximum anticipated number of same-level runs |
michael@0 | 516 | * that internal memory will be preallocated for. An attempt to access |
michael@0 | 517 | * visual runs on an object that was not preallocated for as many runs |
michael@0 | 518 | * as the text was actually resolved to will fail, |
michael@0 | 519 | * unless this value is 0, which leaves the allocation up to the implementation.<br><br> |
michael@0 | 520 | * The number of runs depends on the actual text and maybe anywhere between |
michael@0 | 521 | * 1 and <code>maxLength</code>. It is typically small. |
michael@0 | 522 | * |
michael@0 | 523 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 524 | * |
michael@0 | 525 | * @return An empty <code>UBiDi</code> object with preallocated memory. |
michael@0 | 526 | * @stable ICU 2.0 |
michael@0 | 527 | */ |
michael@0 | 528 | U_STABLE UBiDi * U_EXPORT2 |
michael@0 | 529 | ubidi_openSized(int32_t maxLength, int32_t maxRunCount, UErrorCode *pErrorCode); |
michael@0 | 530 | |
michael@0 | 531 | /** |
michael@0 | 532 | * <code>ubidi_close()</code> must be called to free the memory |
michael@0 | 533 | * associated with a UBiDi object.<p> |
michael@0 | 534 | * |
michael@0 | 535 | * <strong>Important: </strong> |
michael@0 | 536 | * A parent <code>UBiDi</code> object must not be destroyed or reused if |
michael@0 | 537 | * it still has children. |
michael@0 | 538 | * If a <code>UBiDi</code> object has become the <i>child</i> |
michael@0 | 539 | * of another one (its <i>parent</i>) by calling |
michael@0 | 540 | * <code>ubidi_setLine()</code>, then the child object must |
michael@0 | 541 | * be destroyed (closed) or reused (by calling |
michael@0 | 542 | * <code>ubidi_setPara()</code> or <code>ubidi_setLine()</code>) |
michael@0 | 543 | * before the parent object. |
michael@0 | 544 | * |
michael@0 | 545 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 546 | * |
michael@0 | 547 | * @see ubidi_setPara |
michael@0 | 548 | * @see ubidi_setLine |
michael@0 | 549 | * @stable ICU 2.0 |
michael@0 | 550 | */ |
michael@0 | 551 | U_STABLE void U_EXPORT2 |
michael@0 | 552 | ubidi_close(UBiDi *pBiDi); |
michael@0 | 553 | |
michael@0 | 554 | #if U_SHOW_CPLUSPLUS_API |
michael@0 | 555 | |
michael@0 | 556 | U_NAMESPACE_BEGIN |
michael@0 | 557 | |
michael@0 | 558 | /** |
michael@0 | 559 | * \class LocalUBiDiPointer |
michael@0 | 560 | * "Smart pointer" class, closes a UBiDi via ubidi_close(). |
michael@0 | 561 | * For most methods see the LocalPointerBase base class. |
michael@0 | 562 | * |
michael@0 | 563 | * @see LocalPointerBase |
michael@0 | 564 | * @see LocalPointer |
michael@0 | 565 | * @stable ICU 4.4 |
michael@0 | 566 | */ |
michael@0 | 567 | U_DEFINE_LOCAL_OPEN_POINTER(LocalUBiDiPointer, UBiDi, ubidi_close); |
michael@0 | 568 | |
michael@0 | 569 | U_NAMESPACE_END |
michael@0 | 570 | |
michael@0 | 571 | #endif |
michael@0 | 572 | |
michael@0 | 573 | /** |
michael@0 | 574 | * Modify the operation of the Bidi algorithm such that it |
michael@0 | 575 | * approximates an "inverse Bidi" algorithm. This function |
michael@0 | 576 | * must be called before <code>ubidi_setPara()</code>. |
michael@0 | 577 | * |
michael@0 | 578 | * <p>The normal operation of the Bidi algorithm as described |
michael@0 | 579 | * in the Unicode Technical Report is to take text stored in logical |
michael@0 | 580 | * (keyboard, typing) order and to determine the reordering of it for visual |
michael@0 | 581 | * rendering. |
michael@0 | 582 | * Some legacy systems store text in visual order, and for operations |
michael@0 | 583 | * with standard, Unicode-based algorithms, the text needs to be transformed |
michael@0 | 584 | * to logical order. This is effectively the inverse algorithm of the |
michael@0 | 585 | * described Bidi algorithm. Note that there is no standard algorithm for |
michael@0 | 586 | * this "inverse Bidi" and that the current implementation provides only an |
michael@0 | 587 | * approximation of "inverse Bidi".</p> |
michael@0 | 588 | * |
michael@0 | 589 | * <p>With <code>isInverse</code> set to <code>TRUE</code>, |
michael@0 | 590 | * this function changes the behavior of some of the subsequent functions |
michael@0 | 591 | * in a way that they can be used for the inverse Bidi algorithm. |
michael@0 | 592 | * Specifically, runs of text with numeric characters will be treated in a |
michael@0 | 593 | * special way and may need to be surrounded with LRM characters when they are |
michael@0 | 594 | * written in reordered sequence.</p> |
michael@0 | 595 | * |
michael@0 | 596 | * <p>Output runs should be retrieved using <code>ubidi_getVisualRun()</code>. |
michael@0 | 597 | * Since the actual input for "inverse Bidi" is visually ordered text and |
michael@0 | 598 | * <code>ubidi_getVisualRun()</code> gets the reordered runs, these are actually |
michael@0 | 599 | * the runs of the logically ordered output.</p> |
michael@0 | 600 | * |
michael@0 | 601 | * <p>Calling this function with argument <code>isInverse</code> set to |
michael@0 | 602 | * <code>TRUE</code> is equivalent to calling |
michael@0 | 603 | * <code>ubidi_setReorderingMode</code> with argument |
michael@0 | 604 | * <code>reorderingMode</code> |
michael@0 | 605 | * set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br> |
michael@0 | 606 | * Calling this function with argument <code>isInverse</code> set to |
michael@0 | 607 | * <code>FALSE</code> is equivalent to calling |
michael@0 | 608 | * <code>ubidi_setReorderingMode</code> with argument |
michael@0 | 609 | * <code>reorderingMode</code> |
michael@0 | 610 | * set to <code>#UBIDI_REORDER_DEFAULT</code>. |
michael@0 | 611 | * |
michael@0 | 612 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 613 | * |
michael@0 | 614 | * @param isInverse specifies "forward" or "inverse" Bidi operation. |
michael@0 | 615 | * |
michael@0 | 616 | * @see ubidi_setPara |
michael@0 | 617 | * @see ubidi_writeReordered |
michael@0 | 618 | * @see ubidi_setReorderingMode |
michael@0 | 619 | * @stable ICU 2.0 |
michael@0 | 620 | */ |
michael@0 | 621 | U_STABLE void U_EXPORT2 |
michael@0 | 622 | ubidi_setInverse(UBiDi *pBiDi, UBool isInverse); |
michael@0 | 623 | |
michael@0 | 624 | /** |
michael@0 | 625 | * Is this Bidi object set to perform the inverse Bidi algorithm? |
michael@0 | 626 | * <p>Note: calling this function after setting the reordering mode with |
michael@0 | 627 | * <code>ubidi_setReorderingMode</code> will return <code>TRUE</code> if the |
michael@0 | 628 | * reordering mode was set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>, |
michael@0 | 629 | * <code>FALSE</code> for all other values.</p> |
michael@0 | 630 | * |
michael@0 | 631 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 632 | * @return TRUE if the Bidi object is set to perform the inverse Bidi algorithm |
michael@0 | 633 | * by handling numbers as L. |
michael@0 | 634 | * |
michael@0 | 635 | * @see ubidi_setInverse |
michael@0 | 636 | * @see ubidi_setReorderingMode |
michael@0 | 637 | * @stable ICU 2.0 |
michael@0 | 638 | */ |
michael@0 | 639 | |
michael@0 | 640 | U_STABLE UBool U_EXPORT2 |
michael@0 | 641 | ubidi_isInverse(UBiDi *pBiDi); |
michael@0 | 642 | |
michael@0 | 643 | /** |
michael@0 | 644 | * Specify whether block separators must be allocated level zero, |
michael@0 | 645 | * so that successive paragraphs will progress from left to right. |
michael@0 | 646 | * This function must be called before <code>ubidi_setPara()</code>. |
michael@0 | 647 | * Paragraph separators (B) may appear in the text. Setting them to level zero |
michael@0 | 648 | * means that all paragraph separators (including one possibly appearing |
michael@0 | 649 | * in the last text position) are kept in the reordered text after the text |
michael@0 | 650 | * that they follow in the source text. |
michael@0 | 651 | * When this feature is not enabled, a paragraph separator at the last |
michael@0 | 652 | * position of the text before reordering will go to the first position |
michael@0 | 653 | * of the reordered text when the paragraph level is odd. |
michael@0 | 654 | * |
michael@0 | 655 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 656 | * |
michael@0 | 657 | * @param orderParagraphsLTR specifies whether paragraph separators (B) must |
michael@0 | 658 | * receive level 0, so that successive paragraphs progress from left to right. |
michael@0 | 659 | * |
michael@0 | 660 | * @see ubidi_setPara |
michael@0 | 661 | * @stable ICU 3.4 |
michael@0 | 662 | */ |
michael@0 | 663 | U_STABLE void U_EXPORT2 |
michael@0 | 664 | ubidi_orderParagraphsLTR(UBiDi *pBiDi, UBool orderParagraphsLTR); |
michael@0 | 665 | |
michael@0 | 666 | /** |
michael@0 | 667 | * Is this Bidi object set to allocate level 0 to block separators so that |
michael@0 | 668 | * successive paragraphs progress from left to right? |
michael@0 | 669 | * |
michael@0 | 670 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 671 | * @return TRUE if the Bidi object is set to allocate level 0 to block |
michael@0 | 672 | * separators. |
michael@0 | 673 | * |
michael@0 | 674 | * @see ubidi_orderParagraphsLTR |
michael@0 | 675 | * @stable ICU 3.4 |
michael@0 | 676 | */ |
michael@0 | 677 | U_STABLE UBool U_EXPORT2 |
michael@0 | 678 | ubidi_isOrderParagraphsLTR(UBiDi *pBiDi); |
michael@0 | 679 | |
michael@0 | 680 | /** |
michael@0 | 681 | * <code>UBiDiReorderingMode</code> values indicate which variant of the Bidi |
michael@0 | 682 | * algorithm to use. |
michael@0 | 683 | * |
michael@0 | 684 | * @see ubidi_setReorderingMode |
michael@0 | 685 | * @stable ICU 3.6 |
michael@0 | 686 | */ |
michael@0 | 687 | typedef enum UBiDiReorderingMode { |
michael@0 | 688 | /** Regular Logical to Visual Bidi algorithm according to Unicode. |
michael@0 | 689 | * This is a 0 value. |
michael@0 | 690 | * @stable ICU 3.6 */ |
michael@0 | 691 | UBIDI_REORDER_DEFAULT = 0, |
michael@0 | 692 | /** Logical to Visual algorithm which handles numbers in a way which |
michael@0 | 693 | * mimicks the behavior of Windows XP. |
michael@0 | 694 | * @stable ICU 3.6 */ |
michael@0 | 695 | UBIDI_REORDER_NUMBERS_SPECIAL, |
michael@0 | 696 | /** Logical to Visual algorithm grouping numbers with adjacent R characters |
michael@0 | 697 | * (reversible algorithm). |
michael@0 | 698 | * @stable ICU 3.6 */ |
michael@0 | 699 | UBIDI_REORDER_GROUP_NUMBERS_WITH_R, |
michael@0 | 700 | /** Reorder runs only to transform a Logical LTR string to the Logical RTL |
michael@0 | 701 | * string with the same display, or vice-versa.<br> |
michael@0 | 702 | * If this mode is set together with option |
michael@0 | 703 | * <code>#UBIDI_OPTION_INSERT_MARKS</code>, some Bidi controls in the source |
michael@0 | 704 | * text may be removed and other controls may be added to produce the |
michael@0 | 705 | * minimum combination which has the required display. |
michael@0 | 706 | * @stable ICU 3.6 */ |
michael@0 | 707 | UBIDI_REORDER_RUNS_ONLY, |
michael@0 | 708 | /** Visual to Logical algorithm which handles numbers like L |
michael@0 | 709 | * (same algorithm as selected by <code>ubidi_setInverse(TRUE)</code>. |
michael@0 | 710 | * @see ubidi_setInverse |
michael@0 | 711 | * @stable ICU 3.6 */ |
michael@0 | 712 | UBIDI_REORDER_INVERSE_NUMBERS_AS_L, |
michael@0 | 713 | /** Visual to Logical algorithm equivalent to the regular Logical to Visual |
michael@0 | 714 | * algorithm. |
michael@0 | 715 | * @stable ICU 3.6 */ |
michael@0 | 716 | UBIDI_REORDER_INVERSE_LIKE_DIRECT, |
michael@0 | 717 | /** Inverse Bidi (Visual to Logical) algorithm for the |
michael@0 | 718 | * <code>UBIDI_REORDER_NUMBERS_SPECIAL</code> Bidi algorithm. |
michael@0 | 719 | * @stable ICU 3.6 */ |
michael@0 | 720 | UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL, |
michael@0 | 721 | /** Number of values for reordering mode. |
michael@0 | 722 | * @stable ICU 3.6 */ |
michael@0 | 723 | UBIDI_REORDER_COUNT |
michael@0 | 724 | } UBiDiReorderingMode; |
michael@0 | 725 | |
michael@0 | 726 | /** |
michael@0 | 727 | * Modify the operation of the Bidi algorithm such that it implements some |
michael@0 | 728 | * variant to the basic Bidi algorithm or approximates an "inverse Bidi" |
michael@0 | 729 | * algorithm, depending on different values of the "reordering mode". |
michael@0 | 730 | * This function must be called before <code>ubidi_setPara()</code>, and stays |
michael@0 | 731 | * in effect until called again with a different argument. |
michael@0 | 732 | * |
michael@0 | 733 | * <p>The normal operation of the Bidi algorithm as described |
michael@0 | 734 | * in the Unicode Standard Annex #9 is to take text stored in logical |
michael@0 | 735 | * (keyboard, typing) order and to determine how to reorder it for visual |
michael@0 | 736 | * rendering.</p> |
michael@0 | 737 | * |
michael@0 | 738 | * <p>With the reordering mode set to a value other than |
michael@0 | 739 | * <code>#UBIDI_REORDER_DEFAULT</code>, this function changes the behavior of |
michael@0 | 740 | * some of the subsequent functions in a way such that they implement an |
michael@0 | 741 | * inverse Bidi algorithm or some other algorithm variants.</p> |
michael@0 | 742 | * |
michael@0 | 743 | * <p>Some legacy systems store text in visual order, and for operations |
michael@0 | 744 | * with standard, Unicode-based algorithms, the text needs to be transformed |
michael@0 | 745 | * into logical order. This is effectively the inverse algorithm of the |
michael@0 | 746 | * described Bidi algorithm. Note that there is no standard algorithm for |
michael@0 | 747 | * this "inverse Bidi", so a number of variants are implemented here.</p> |
michael@0 | 748 | * |
michael@0 | 749 | * <p>In other cases, it may be desirable to emulate some variant of the |
michael@0 | 750 | * Logical to Visual algorithm (e.g. one used in MS Windows), or perform a |
michael@0 | 751 | * Logical to Logical transformation.</p> |
michael@0 | 752 | * |
michael@0 | 753 | * <ul> |
michael@0 | 754 | * <li>When the reordering mode is set to <code>#UBIDI_REORDER_DEFAULT</code>, |
michael@0 | 755 | * the standard Bidi Logical to Visual algorithm is applied.</li> |
michael@0 | 756 | * |
michael@0 | 757 | * <li>When the reordering mode is set to |
michael@0 | 758 | * <code>#UBIDI_REORDER_NUMBERS_SPECIAL</code>, |
michael@0 | 759 | * the algorithm used to perform Bidi transformations when calling |
michael@0 | 760 | * <code>ubidi_setPara</code> should approximate the algorithm used in |
michael@0 | 761 | * Microsoft Windows XP rather than strictly conform to the Unicode Bidi |
michael@0 | 762 | * algorithm. |
michael@0 | 763 | * <br> |
michael@0 | 764 | * The differences between the basic algorithm and the algorithm addressed |
michael@0 | 765 | * by this option are as follows: |
michael@0 | 766 | * <ul> |
michael@0 | 767 | * <li>Within text at an even embedding level, the sequence "123AB" |
michael@0 | 768 | * (where AB represent R or AL letters) is transformed to "123BA" by the |
michael@0 | 769 | * Unicode algorithm and to "BA123" by the Windows algorithm.</li> |
michael@0 | 770 | * <li>Arabic-Indic numbers (AN) are handled by the Windows algorithm just |
michael@0 | 771 | * like regular numbers (EN).</li> |
michael@0 | 772 | * </ul></li> |
michael@0 | 773 | * |
michael@0 | 774 | * <li>When the reordering mode is set to |
michael@0 | 775 | * <code>#UBIDI_REORDER_GROUP_NUMBERS_WITH_R</code>, |
michael@0 | 776 | * numbers located between LTR text and RTL text are associated with the RTL |
michael@0 | 777 | * text. For instance, an LTR paragraph with content "abc 123 DEF" (where |
michael@0 | 778 | * upper case letters represent RTL characters) will be transformed to |
michael@0 | 779 | * "abc FED 123" (and not "abc 123 FED"), "DEF 123 abc" will be transformed |
michael@0 | 780 | * to "123 FED abc" and "123 FED abc" will be transformed to "DEF 123 abc". |
michael@0 | 781 | * This makes the algorithm reversible and makes it useful when round trip |
michael@0 | 782 | * (from visual to logical and back to visual) must be achieved without |
michael@0 | 783 | * adding LRM characters. However, this is a variation from the standard |
michael@0 | 784 | * Unicode Bidi algorithm.<br> |
michael@0 | 785 | * The source text should not contain Bidi control characters other than LRM |
michael@0 | 786 | * or RLM.</li> |
michael@0 | 787 | * |
michael@0 | 788 | * <li>When the reordering mode is set to |
michael@0 | 789 | * <code>#UBIDI_REORDER_RUNS_ONLY</code>, |
michael@0 | 790 | * a "Logical to Logical" transformation must be performed: |
michael@0 | 791 | * <ul> |
michael@0 | 792 | * <li>If the default text level of the source text (argument <code>paraLevel</code> |
michael@0 | 793 | * in <code>ubidi_setPara</code>) is even, the source text will be handled as |
michael@0 | 794 | * LTR logical text and will be transformed to the RTL logical text which has |
michael@0 | 795 | * the same LTR visual display.</li> |
michael@0 | 796 | * <li>If the default level of the source text is odd, the source text |
michael@0 | 797 | * will be handled as RTL logical text and will be transformed to the |
michael@0 | 798 | * LTR logical text which has the same LTR visual display.</li> |
michael@0 | 799 | * </ul> |
michael@0 | 800 | * This mode may be needed when logical text which is basically Arabic or |
michael@0 | 801 | * Hebrew, with possible included numbers or phrases in English, has to be |
michael@0 | 802 | * displayed as if it had an even embedding level (this can happen if the |
michael@0 | 803 | * displaying application treats all text as if it was basically LTR). |
michael@0 | 804 | * <br> |
michael@0 | 805 | * This mode may also be needed in the reverse case, when logical text which is |
michael@0 | 806 | * basically English, with possible included phrases in Arabic or Hebrew, has to |
michael@0 | 807 | * be displayed as if it had an odd embedding level. |
michael@0 | 808 | * <br> |
michael@0 | 809 | * Both cases could be handled by adding LRE or RLE at the head of the text, |
michael@0 | 810 | * if the display subsystem supports these formatting controls. If it does not, |
michael@0 | 811 | * the problem may be handled by transforming the source text in this mode |
michael@0 | 812 | * before displaying it, so that it will be displayed properly.<br> |
michael@0 | 813 | * The source text should not contain Bidi control characters other than LRM |
michael@0 | 814 | * or RLM.</li> |
michael@0 | 815 | * |
michael@0 | 816 | * <li>When the reordering mode is set to |
michael@0 | 817 | * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>, an "inverse Bidi" algorithm |
michael@0 | 818 | * is applied. |
michael@0 | 819 | * Runs of text with numeric characters will be treated like LTR letters and |
michael@0 | 820 | * may need to be surrounded with LRM characters when they are written in |
michael@0 | 821 | * reordered sequence (the option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> can |
michael@0 | 822 | * be used with function <code>ubidi_writeReordered</code> to this end. This |
michael@0 | 823 | * mode is equivalent to calling <code>ubidi_setInverse()</code> with |
michael@0 | 824 | * argument <code>isInverse</code> set to <code>TRUE</code>.</li> |
michael@0 | 825 | * |
michael@0 | 826 | * <li>When the reordering mode is set to |
michael@0 | 827 | * <code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code>, the "direct" Logical to Visual |
michael@0 | 828 | * Bidi algorithm is used as an approximation of an "inverse Bidi" algorithm. |
michael@0 | 829 | * This mode is similar to mode <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> |
michael@0 | 830 | * but is closer to the regular Bidi algorithm. |
michael@0 | 831 | * <br> |
michael@0 | 832 | * For example, an LTR paragraph with the content "FED 123 456 CBA" (where |
michael@0 | 833 | * upper case represents RTL characters) will be transformed to |
michael@0 | 834 | * "ABC 456 123 DEF", as opposed to "DEF 123 456 ABC" |
michael@0 | 835 | * with mode <code>UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br> |
michael@0 | 836 | * When used in conjunction with option |
michael@0 | 837 | * <code>#UBIDI_OPTION_INSERT_MARKS</code>, this mode generally |
michael@0 | 838 | * adds Bidi marks to the output significantly more sparingly than mode |
michael@0 | 839 | * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> with option |
michael@0 | 840 | * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to |
michael@0 | 841 | * <code>ubidi_writeReordered</code>.</li> |
michael@0 | 842 | * |
michael@0 | 843 | * <li>When the reordering mode is set to |
michael@0 | 844 | * <code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the Logical to Visual |
michael@0 | 845 | * Bidi algorithm used in Windows XP is used as an approximation of an "inverse Bidi" algorithm. |
michael@0 | 846 | * <br> |
michael@0 | 847 | * For example, an LTR paragraph with the content "abc FED123" (where |
michael@0 | 848 | * upper case represents RTL characters) will be transformed to "abc 123DEF."</li> |
michael@0 | 849 | * </ul> |
michael@0 | 850 | * |
michael@0 | 851 | * <p>In all the reordering modes specifying an "inverse Bidi" algorithm |
michael@0 | 852 | * (i.e. those with a name starting with <code>UBIDI_REORDER_INVERSE</code>), |
michael@0 | 853 | * output runs should be retrieved using |
michael@0 | 854 | * <code>ubidi_getVisualRun()</code>, and the output text with |
michael@0 | 855 | * <code>ubidi_writeReordered()</code>. The caller should keep in mind that in |
michael@0 | 856 | * "inverse Bidi" modes the input is actually visually ordered text and |
michael@0 | 857 | * reordered output returned by <code>ubidi_getVisualRun()</code> or |
michael@0 | 858 | * <code>ubidi_writeReordered()</code> are actually runs or character string |
michael@0 | 859 | * of logically ordered output.<br> |
michael@0 | 860 | * For all the "inverse Bidi" modes, the source text should not contain |
michael@0 | 861 | * Bidi control characters other than LRM or RLM.</p> |
michael@0 | 862 | * |
michael@0 | 863 | * <p>Note that option <code>#UBIDI_OUTPUT_REVERSE</code> of |
michael@0 | 864 | * <code>ubidi_writeReordered</code> has no useful meaning and should not be |
michael@0 | 865 | * used in conjunction with any value of the reordering mode specifying |
michael@0 | 866 | * "inverse Bidi" or with value <code>UBIDI_REORDER_RUNS_ONLY</code>. |
michael@0 | 867 | * |
michael@0 | 868 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 869 | * @param reorderingMode specifies the required variant of the Bidi algorithm. |
michael@0 | 870 | * |
michael@0 | 871 | * @see UBiDiReorderingMode |
michael@0 | 872 | * @see ubidi_setInverse |
michael@0 | 873 | * @see ubidi_setPara |
michael@0 | 874 | * @see ubidi_writeReordered |
michael@0 | 875 | * @stable ICU 3.6 |
michael@0 | 876 | */ |
michael@0 | 877 | U_STABLE void U_EXPORT2 |
michael@0 | 878 | ubidi_setReorderingMode(UBiDi *pBiDi, UBiDiReorderingMode reorderingMode); |
michael@0 | 879 | |
michael@0 | 880 | /** |
michael@0 | 881 | * What is the requested reordering mode for a given Bidi object? |
michael@0 | 882 | * |
michael@0 | 883 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 884 | * @return the current reordering mode of the Bidi object |
michael@0 | 885 | * @see ubidi_setReorderingMode |
michael@0 | 886 | * @stable ICU 3.6 |
michael@0 | 887 | */ |
michael@0 | 888 | U_STABLE UBiDiReorderingMode U_EXPORT2 |
michael@0 | 889 | ubidi_getReorderingMode(UBiDi *pBiDi); |
michael@0 | 890 | |
michael@0 | 891 | /** |
michael@0 | 892 | * <code>UBiDiReorderingOption</code> values indicate which options are |
michael@0 | 893 | * specified to affect the Bidi algorithm. |
michael@0 | 894 | * |
michael@0 | 895 | * @see ubidi_setReorderingOptions |
michael@0 | 896 | * @stable ICU 3.6 |
michael@0 | 897 | */ |
michael@0 | 898 | typedef enum UBiDiReorderingOption { |
michael@0 | 899 | /** |
michael@0 | 900 | * option value for <code>ubidi_setReorderingOptions</code>: |
michael@0 | 901 | * disable all the options which can be set with this function |
michael@0 | 902 | * @see ubidi_setReorderingOptions |
michael@0 | 903 | * @stable ICU 3.6 |
michael@0 | 904 | */ |
michael@0 | 905 | UBIDI_OPTION_DEFAULT = 0, |
michael@0 | 906 | |
michael@0 | 907 | /** |
michael@0 | 908 | * option bit for <code>ubidi_setReorderingOptions</code>: |
michael@0 | 909 | * insert Bidi marks (LRM or RLM) when needed to ensure correct result of |
michael@0 | 910 | * a reordering to a Logical order |
michael@0 | 911 | * |
michael@0 | 912 | * <p>This option must be set or reset before calling |
michael@0 | 913 | * <code>ubidi_setPara</code>.</p> |
michael@0 | 914 | * |
michael@0 | 915 | * <p>This option is significant only with reordering modes which generate |
michael@0 | 916 | * a result with Logical order, specifically:</p> |
michael@0 | 917 | * <ul> |
michael@0 | 918 | * <li><code>#UBIDI_REORDER_RUNS_ONLY</code></li> |
michael@0 | 919 | * <li><code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code></li> |
michael@0 | 920 | * <li><code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code></li> |
michael@0 | 921 | * <li><code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code></li> |
michael@0 | 922 | * </ul> |
michael@0 | 923 | * |
michael@0 | 924 | * <p>If this option is set in conjunction with reordering mode |
michael@0 | 925 | * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> or with calling |
michael@0 | 926 | * <code>ubidi_setInverse(TRUE)</code>, it implies |
michael@0 | 927 | * option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> |
michael@0 | 928 | * in calls to function <code>ubidi_writeReordered()</code>.</p> |
michael@0 | 929 | * |
michael@0 | 930 | * <p>For other reordering modes, a minimum number of LRM or RLM characters |
michael@0 | 931 | * will be added to the source text after reordering it so as to ensure |
michael@0 | 932 | * round trip, i.e. when applying the inverse reordering mode on the |
michael@0 | 933 | * resulting logical text with removal of Bidi marks |
michael@0 | 934 | * (option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> set before calling |
michael@0 | 935 | * <code>ubidi_setPara()</code> or option <code>#UBIDI_REMOVE_BIDI_CONTROLS</code> |
michael@0 | 936 | * in <code>ubidi_writeReordered</code>), the result will be identical to the |
michael@0 | 937 | * source text in the first transformation. |
michael@0 | 938 | * |
michael@0 | 939 | * <p>This option will be ignored if specified together with option |
michael@0 | 940 | * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>. It inhibits option |
michael@0 | 941 | * <code>UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to function |
michael@0 | 942 | * <code>ubidi_writeReordered()</code> and it implies option |
michael@0 | 943 | * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to function |
michael@0 | 944 | * <code>ubidi_writeReordered()</code> if the reordering mode is |
michael@0 | 945 | * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.</p> |
michael@0 | 946 | * |
michael@0 | 947 | * @see ubidi_setReorderingMode |
michael@0 | 948 | * @see ubidi_setReorderingOptions |
michael@0 | 949 | * @stable ICU 3.6 |
michael@0 | 950 | */ |
michael@0 | 951 | UBIDI_OPTION_INSERT_MARKS = 1, |
michael@0 | 952 | |
michael@0 | 953 | /** |
michael@0 | 954 | * option bit for <code>ubidi_setReorderingOptions</code>: |
michael@0 | 955 | * remove Bidi control characters |
michael@0 | 956 | * |
michael@0 | 957 | * <p>This option must be set or reset before calling |
michael@0 | 958 | * <code>ubidi_setPara</code>.</p> |
michael@0 | 959 | * |
michael@0 | 960 | * <p>This option nullifies option <code>#UBIDI_OPTION_INSERT_MARKS</code>. |
michael@0 | 961 | * It inhibits option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls |
michael@0 | 962 | * to function <code>ubidi_writeReordered()</code> and it implies option |
michael@0 | 963 | * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to that function.</p> |
michael@0 | 964 | * |
michael@0 | 965 | * @see ubidi_setReorderingMode |
michael@0 | 966 | * @see ubidi_setReorderingOptions |
michael@0 | 967 | * @stable ICU 3.6 |
michael@0 | 968 | */ |
michael@0 | 969 | UBIDI_OPTION_REMOVE_CONTROLS = 2, |
michael@0 | 970 | |
michael@0 | 971 | /** |
michael@0 | 972 | * option bit for <code>ubidi_setReorderingOptions</code>: |
michael@0 | 973 | * process the output as part of a stream to be continued |
michael@0 | 974 | * |
michael@0 | 975 | * <p>This option must be set or reset before calling |
michael@0 | 976 | * <code>ubidi_setPara</code>.</p> |
michael@0 | 977 | * |
michael@0 | 978 | * <p>This option specifies that the caller is interested in processing large |
michael@0 | 979 | * text object in parts. |
michael@0 | 980 | * The results of the successive calls are expected to be concatenated by the |
michael@0 | 981 | * caller. Only the call for the last part will have this option bit off.</p> |
michael@0 | 982 | * |
michael@0 | 983 | * <p>When this option bit is on, <code>ubidi_setPara()</code> may process |
michael@0 | 984 | * less than the full source text in order to truncate the text at a meaningful |
michael@0 | 985 | * boundary. The caller should call <code>ubidi_getProcessedLength()</code> |
michael@0 | 986 | * immediately after calling <code>ubidi_setPara()</code> in order to |
michael@0 | 987 | * determine how much of the source text has been processed. |
michael@0 | 988 | * Source text beyond that length should be resubmitted in following calls to |
michael@0 | 989 | * <code>ubidi_setPara</code>. The processed length may be less than |
michael@0 | 990 | * the length of the source text if a character preceding the last character of |
michael@0 | 991 | * the source text constitutes a reasonable boundary (like a block separator) |
michael@0 | 992 | * for text to be continued.<br> |
michael@0 | 993 | * If the last character of the source text constitutes a reasonable |
michael@0 | 994 | * boundary, the whole text will be processed at once.<br> |
michael@0 | 995 | * If nowhere in the source text there exists |
michael@0 | 996 | * such a reasonable boundary, the processed length will be zero.<br> |
michael@0 | 997 | * The caller should check for such an occurrence and do one of the following: |
michael@0 | 998 | * <ul><li>submit a larger amount of text with a better chance to include |
michael@0 | 999 | * a reasonable boundary.</li> |
michael@0 | 1000 | * <li>resubmit the same text after turning off option |
michael@0 | 1001 | * <code>UBIDI_OPTION_STREAMING</code>.</li></ul> |
michael@0 | 1002 | * In all cases, this option should be turned off before processing the last |
michael@0 | 1003 | * part of the text.</p> |
michael@0 | 1004 | * |
michael@0 | 1005 | * <p>When the <code>UBIDI_OPTION_STREAMING</code> option is used, |
michael@0 | 1006 | * it is recommended to call <code>ubidi_orderParagraphsLTR()</code> with |
michael@0 | 1007 | * argument <code>orderParagraphsLTR</code> set to <code>TRUE</code> before |
michael@0 | 1008 | * calling <code>ubidi_setPara</code> so that later paragraphs may be |
michael@0 | 1009 | * concatenated to previous paragraphs on the right.</p> |
michael@0 | 1010 | * |
michael@0 | 1011 | * @see ubidi_setReorderingMode |
michael@0 | 1012 | * @see ubidi_setReorderingOptions |
michael@0 | 1013 | * @see ubidi_getProcessedLength |
michael@0 | 1014 | * @see ubidi_orderParagraphsLTR |
michael@0 | 1015 | * @stable ICU 3.6 |
michael@0 | 1016 | */ |
michael@0 | 1017 | UBIDI_OPTION_STREAMING = 4 |
michael@0 | 1018 | } UBiDiReorderingOption; |
michael@0 | 1019 | |
michael@0 | 1020 | /** |
michael@0 | 1021 | * Specify which of the reordering options |
michael@0 | 1022 | * should be applied during Bidi transformations. |
michael@0 | 1023 | * |
michael@0 | 1024 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 1025 | * @param reorderingOptions is a combination of zero or more of the following |
michael@0 | 1026 | * options: |
michael@0 | 1027 | * <code>#UBIDI_OPTION_DEFAULT</code>, <code>#UBIDI_OPTION_INSERT_MARKS</code>, |
michael@0 | 1028 | * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>, <code>#UBIDI_OPTION_STREAMING</code>. |
michael@0 | 1029 | * |
michael@0 | 1030 | * @see ubidi_getReorderingOptions |
michael@0 | 1031 | * @stable ICU 3.6 |
michael@0 | 1032 | */ |
michael@0 | 1033 | U_STABLE void U_EXPORT2 |
michael@0 | 1034 | ubidi_setReorderingOptions(UBiDi *pBiDi, uint32_t reorderingOptions); |
michael@0 | 1035 | |
michael@0 | 1036 | /** |
michael@0 | 1037 | * What are the reordering options applied to a given Bidi object? |
michael@0 | 1038 | * |
michael@0 | 1039 | * @param pBiDi is a <code>UBiDi</code> object. |
michael@0 | 1040 | * @return the current reordering options of the Bidi object |
michael@0 | 1041 | * @see ubidi_setReorderingOptions |
michael@0 | 1042 | * @stable ICU 3.6 |
michael@0 | 1043 | */ |
michael@0 | 1044 | U_STABLE uint32_t U_EXPORT2 |
michael@0 | 1045 | ubidi_getReorderingOptions(UBiDi *pBiDi); |
michael@0 | 1046 | |
michael@0 | 1047 | /** |
michael@0 | 1048 | * Set the context before a call to ubidi_setPara().<p> |
michael@0 | 1049 | * |
michael@0 | 1050 | * ubidi_setPara() computes the left-right directionality for a given piece |
michael@0 | 1051 | * of text which is supplied as one of its arguments. Sometimes this piece |
michael@0 | 1052 | * of text (the "main text") should be considered in context, because text |
michael@0 | 1053 | * appearing before ("prologue") and/or after ("epilogue") the main text |
michael@0 | 1054 | * may affect the result of this computation.<p> |
michael@0 | 1055 | * |
michael@0 | 1056 | * This function specifies the prologue and/or the epilogue for the next |
michael@0 | 1057 | * call to ubidi_setPara(). The characters specified as prologue and |
michael@0 | 1058 | * epilogue should not be modified by the calling program until the call |
michael@0 | 1059 | * to ubidi_setPara() has returned. If successive calls to ubidi_setPara() |
michael@0 | 1060 | * all need specification of a context, ubidi_setContext() must be called |
michael@0 | 1061 | * before each call to ubidi_setPara(). In other words, a context is not |
michael@0 | 1062 | * "remembered" after the following successful call to ubidi_setPara().<p> |
michael@0 | 1063 | * |
michael@0 | 1064 | * If a call to ubidi_setPara() specifies UBIDI_DEFAULT_LTR or |
michael@0 | 1065 | * UBIDI_DEFAULT_RTL as paraLevel and is preceded by a call to |
michael@0 | 1066 | * ubidi_setContext() which specifies a prologue, the paragraph level will |
michael@0 | 1067 | * be computed taking in consideration the text in the prologue.<p> |
michael@0 | 1068 | * |
michael@0 | 1069 | * When ubidi_setPara() is called without a previous call to |
michael@0 | 1070 | * ubidi_setContext, the main text is handled as if preceded and followed |
michael@0 | 1071 | * by strong directional characters at the current paragraph level. |
michael@0 | 1072 | * Calling ubidi_setContext() with specification of a prologue will change |
michael@0 | 1073 | * this behavior by handling the main text as if preceded by the last |
michael@0 | 1074 | * strong character appearing in the prologue, if any. |
michael@0 | 1075 | * Calling ubidi_setContext() with specification of an epilogue will change |
michael@0 | 1076 | * the behavior of ubidi_setPara() by handling the main text as if followed |
michael@0 | 1077 | * by the first strong character or digit appearing in the epilogue, if any.<p> |
michael@0 | 1078 | * |
michael@0 | 1079 | * Note 1: if <code>ubidi_setContext</code> is called repeatedly without |
michael@0 | 1080 | * calling <code>ubidi_setPara</code>, the earlier calls have no effect, |
michael@0 | 1081 | * only the last call will be remembered for the next call to |
michael@0 | 1082 | * <code>ubidi_setPara</code>.<p> |
michael@0 | 1083 | * |
michael@0 | 1084 | * Note 2: calling <code>ubidi_setContext(pBiDi, NULL, 0, NULL, 0, &errorCode)</code> |
michael@0 | 1085 | * cancels any previous setting of non-empty prologue or epilogue. |
michael@0 | 1086 | * The next call to <code>ubidi_setPara()</code> will process no |
michael@0 | 1087 | * prologue or epilogue.<p> |
michael@0 | 1088 | * |
michael@0 | 1089 | * Note 3: users must be aware that even after setting the context |
michael@0 | 1090 | * before a call to ubidi_setPara() to perform e.g. a logical to visual |
michael@0 | 1091 | * transformation, the resulting string may not be identical to what it |
michael@0 | 1092 | * would have been if all the text, including prologue and epilogue, had |
michael@0 | 1093 | * been processed together.<br> |
michael@0 | 1094 | * Example (upper case letters represent RTL characters):<br> |
michael@0 | 1095 | * prologue = "<code>abc DE</code>"<br> |
michael@0 | 1096 | * epilogue = none<br> |
michael@0 | 1097 | * main text = "<code>FGH xyz</code>"<br> |
michael@0 | 1098 | * paraLevel = UBIDI_LTR<br> |
michael@0 | 1099 | * display without prologue = "<code>HGF xyz</code>" |
michael@0 | 1100 | * ("HGF" is adjacent to "xyz")<br> |
michael@0 | 1101 | * display with prologue = "<code>abc HGFED xyz</code>" |
michael@0 | 1102 | * ("HGF" is not adjacent to "xyz")<br> |
michael@0 | 1103 | * |
michael@0 | 1104 | * @param pBiDi is a paragraph <code>UBiDi</code> object. |
michael@0 | 1105 | * |
michael@0 | 1106 | * @param prologue is a pointer to the text which precedes the text that |
michael@0 | 1107 | * will be specified in a coming call to ubidi_setPara(). |
michael@0 | 1108 | * If there is no prologue to consider, then <code>proLength</code> |
michael@0 | 1109 | * must be zero and this pointer can be NULL. |
michael@0 | 1110 | * |
michael@0 | 1111 | * @param proLength is the length of the prologue; if <code>proLength==-1</code> |
michael@0 | 1112 | * then the prologue must be zero-terminated. |
michael@0 | 1113 | * Otherwise proLength must be >= 0. If <code>proLength==0</code>, it means |
michael@0 | 1114 | * that there is no prologue to consider. |
michael@0 | 1115 | * |
michael@0 | 1116 | * @param epilogue is a pointer to the text which follows the text that |
michael@0 | 1117 | * will be specified in a coming call to ubidi_setPara(). |
michael@0 | 1118 | * If there is no epilogue to consider, then <code>epiLength</code> |
michael@0 | 1119 | * must be zero and this pointer can be NULL. |
michael@0 | 1120 | * |
michael@0 | 1121 | * @param epiLength is the length of the epilogue; if <code>epiLength==-1</code> |
michael@0 | 1122 | * then the epilogue must be zero-terminated. |
michael@0 | 1123 | * Otherwise epiLength must be >= 0. If <code>epiLength==0</code>, it means |
michael@0 | 1124 | * that there is no epilogue to consider. |
michael@0 | 1125 | * |
michael@0 | 1126 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1127 | * |
michael@0 | 1128 | * @see ubidi_setPara |
michael@0 | 1129 | * @stable ICU 4.8 |
michael@0 | 1130 | */ |
michael@0 | 1131 | U_STABLE void U_EXPORT2 |
michael@0 | 1132 | ubidi_setContext(UBiDi *pBiDi, |
michael@0 | 1133 | const UChar *prologue, int32_t proLength, |
michael@0 | 1134 | const UChar *epilogue, int32_t epiLength, |
michael@0 | 1135 | UErrorCode *pErrorCode); |
michael@0 | 1136 | |
michael@0 | 1137 | /** |
michael@0 | 1138 | * Perform the Unicode Bidi algorithm. It is defined in the |
michael@0 | 1139 | * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Anned #9</a>, |
michael@0 | 1140 | * version 13, |
michael@0 | 1141 | * also described in The Unicode Standard, Version 4.0 .<p> |
michael@0 | 1142 | * |
michael@0 | 1143 | * This function takes a piece of plain text containing one or more paragraphs, |
michael@0 | 1144 | * with or without externally specified embedding levels from <i>styled</i> |
michael@0 | 1145 | * text and computes the left-right-directionality of each character.<p> |
michael@0 | 1146 | * |
michael@0 | 1147 | * If the entire text is all of the same directionality, then |
michael@0 | 1148 | * the function may not perform all the steps described by the algorithm, |
michael@0 | 1149 | * i.e., some levels may not be the same as if all steps were performed. |
michael@0 | 1150 | * This is not relevant for unidirectional text.<br> |
michael@0 | 1151 | * For example, in pure LTR text with numbers the numbers would get |
michael@0 | 1152 | * a resolved level of 2 higher than the surrounding text according to |
michael@0 | 1153 | * the algorithm. This implementation may set all resolved levels to |
michael@0 | 1154 | * the same value in such a case.<p> |
michael@0 | 1155 | * |
michael@0 | 1156 | * The text can be composed of multiple paragraphs. Occurrence of a block |
michael@0 | 1157 | * separator in the text terminates a paragraph, and whatever comes next starts |
michael@0 | 1158 | * a new paragraph. The exception to this rule is when a Carriage Return (CR) |
michael@0 | 1159 | * is followed by a Line Feed (LF). Both CR and LF are block separators, but |
michael@0 | 1160 | * in that case, the pair of characters is considered as terminating the |
michael@0 | 1161 | * preceding paragraph, and a new paragraph will be started by a character |
michael@0 | 1162 | * coming after the LF. |
michael@0 | 1163 | * |
michael@0 | 1164 | * @param pBiDi A <code>UBiDi</code> object allocated with <code>ubidi_open()</code> |
michael@0 | 1165 | * which will be set to contain the reordering information, |
michael@0 | 1166 | * especially the resolved levels for all the characters in <code>text</code>. |
michael@0 | 1167 | * |
michael@0 | 1168 | * @param text is a pointer to the text that the Bidi algorithm will be performed on. |
michael@0 | 1169 | * This pointer is stored in the UBiDi object and can be retrieved |
michael@0 | 1170 | * with <code>ubidi_getText()</code>.<br> |
michael@0 | 1171 | * <strong>Note:</strong> the text must be (at least) <code>length</code> long. |
michael@0 | 1172 | * |
michael@0 | 1173 | * @param length is the length of the text; if <code>length==-1</code> then |
michael@0 | 1174 | * the text must be zero-terminated. |
michael@0 | 1175 | * |
michael@0 | 1176 | * @param paraLevel specifies the default level for the text; |
michael@0 | 1177 | * it is typically 0 (LTR) or 1 (RTL). |
michael@0 | 1178 | * If the function shall determine the paragraph level from the text, |
michael@0 | 1179 | * then <code>paraLevel</code> can be set to |
michael@0 | 1180 | * either <code>#UBIDI_DEFAULT_LTR</code> |
michael@0 | 1181 | * or <code>#UBIDI_DEFAULT_RTL</code>; if the text contains multiple |
michael@0 | 1182 | * paragraphs, the paragraph level shall be determined separately for |
michael@0 | 1183 | * each paragraph; if a paragraph does not include any strongly typed |
michael@0 | 1184 | * character, then the desired default is used (0 for LTR or 1 for RTL). |
michael@0 | 1185 | * Any other value between 0 and <code>#UBIDI_MAX_EXPLICIT_LEVEL</code> |
michael@0 | 1186 | * is also valid, with odd levels indicating RTL. |
michael@0 | 1187 | * |
michael@0 | 1188 | * @param embeddingLevels (in) may be used to preset the embedding and override levels, |
michael@0 | 1189 | * ignoring characters like LRE and PDF in the text. |
michael@0 | 1190 | * A level overrides the directional property of its corresponding |
michael@0 | 1191 | * (same index) character if the level has the |
michael@0 | 1192 | * <code>#UBIDI_LEVEL_OVERRIDE</code> bit set.<br><br> |
michael@0 | 1193 | * Except for that bit, it must be |
michael@0 | 1194 | * <code>paraLevel<=embeddingLevels[]<=UBIDI_MAX_EXPLICIT_LEVEL</code>, |
michael@0 | 1195 | * with one exception: a level of zero may be specified for a paragraph |
michael@0 | 1196 | * separator even if <code>paraLevel>0</code> when multiple paragraphs |
michael@0 | 1197 | * are submitted in the same call to <code>ubidi_setPara()</code>.<br><br> |
michael@0 | 1198 | * <strong>Caution: </strong>A copy of this pointer, not of the levels, |
michael@0 | 1199 | * will be stored in the <code>UBiDi</code> object; |
michael@0 | 1200 | * the <code>embeddingLevels</code> array must not be |
michael@0 | 1201 | * deallocated before the <code>UBiDi</code> structure is destroyed or reused, |
michael@0 | 1202 | * and the <code>embeddingLevels</code> |
michael@0 | 1203 | * should not be modified to avoid unexpected results on subsequent Bidi operations. |
michael@0 | 1204 | * However, the <code>ubidi_setPara()</code> and |
michael@0 | 1205 | * <code>ubidi_setLine()</code> functions may modify some or all of the levels.<br><br> |
michael@0 | 1206 | * After the <code>UBiDi</code> object is reused or destroyed, the caller |
michael@0 | 1207 | * must take care of the deallocation of the <code>embeddingLevels</code> array.<br><br> |
michael@0 | 1208 | * <strong>Note:</strong> the <code>embeddingLevels</code> array must be |
michael@0 | 1209 | * at least <code>length</code> long. |
michael@0 | 1210 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1211 | * value is not necessary. |
michael@0 | 1212 | * |
michael@0 | 1213 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1214 | * @stable ICU 2.0 |
michael@0 | 1215 | */ |
michael@0 | 1216 | U_STABLE void U_EXPORT2 |
michael@0 | 1217 | ubidi_setPara(UBiDi *pBiDi, const UChar *text, int32_t length, |
michael@0 | 1218 | UBiDiLevel paraLevel, UBiDiLevel *embeddingLevels, |
michael@0 | 1219 | UErrorCode *pErrorCode); |
michael@0 | 1220 | |
michael@0 | 1221 | /** |
michael@0 | 1222 | * <code>ubidi_setLine()</code> sets a <code>UBiDi</code> to |
michael@0 | 1223 | * contain the reordering information, especially the resolved levels, |
michael@0 | 1224 | * for all the characters in a line of text. This line of text is |
michael@0 | 1225 | * specified by referring to a <code>UBiDi</code> object representing |
michael@0 | 1226 | * this information for a piece of text containing one or more paragraphs, |
michael@0 | 1227 | * and by specifying a range of indexes in this text.<p> |
michael@0 | 1228 | * In the new line object, the indexes will range from 0 to <code>limit-start-1</code>.<p> |
michael@0 | 1229 | * |
michael@0 | 1230 | * This is used after calling <code>ubidi_setPara()</code> |
michael@0 | 1231 | * for a piece of text, and after line-breaking on that text. |
michael@0 | 1232 | * It is not necessary if each paragraph is treated as a single line.<p> |
michael@0 | 1233 | * |
michael@0 | 1234 | * After line-breaking, rules (L1) and (L2) for the treatment of |
michael@0 | 1235 | * trailing WS and for reordering are performed on |
michael@0 | 1236 | * a <code>UBiDi</code> object that represents a line.<p> |
michael@0 | 1237 | * |
michael@0 | 1238 | * <strong>Important: </strong><code>pLineBiDi</code> shares data with |
michael@0 | 1239 | * <code>pParaBiDi</code>. |
michael@0 | 1240 | * You must destroy or reuse <code>pLineBiDi</code> before <code>pParaBiDi</code>. |
michael@0 | 1241 | * In other words, you must destroy or reuse the <code>UBiDi</code> object for a line |
michael@0 | 1242 | * before the object for its parent paragraph.<p> |
michael@0 | 1243 | * |
michael@0 | 1244 | * The text pointer that was stored in <code>pParaBiDi</code> is also copied, |
michael@0 | 1245 | * and <code>start</code> is added to it so that it points to the beginning of the |
michael@0 | 1246 | * line for this object. |
michael@0 | 1247 | * |
michael@0 | 1248 | * @param pParaBiDi is the parent paragraph object. It must have been set |
michael@0 | 1249 | * by a successful call to ubidi_setPara. |
michael@0 | 1250 | * |
michael@0 | 1251 | * @param start is the line's first index into the text. |
michael@0 | 1252 | * |
michael@0 | 1253 | * @param limit is just behind the line's last index into the text |
michael@0 | 1254 | * (its last index +1).<br> |
michael@0 | 1255 | * It must be <code>0<=start<limit<=</code>containing paragraph limit. |
michael@0 | 1256 | * If the specified line crosses a paragraph boundary, the function |
michael@0 | 1257 | * will terminate with error code U_ILLEGAL_ARGUMENT_ERROR. |
michael@0 | 1258 | * |
michael@0 | 1259 | * @param pLineBiDi is the object that will now represent a line of the text. |
michael@0 | 1260 | * |
michael@0 | 1261 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1262 | * |
michael@0 | 1263 | * @see ubidi_setPara |
michael@0 | 1264 | * @see ubidi_getProcessedLength |
michael@0 | 1265 | * @stable ICU 2.0 |
michael@0 | 1266 | */ |
michael@0 | 1267 | U_STABLE void U_EXPORT2 |
michael@0 | 1268 | ubidi_setLine(const UBiDi *pParaBiDi, |
michael@0 | 1269 | int32_t start, int32_t limit, |
michael@0 | 1270 | UBiDi *pLineBiDi, |
michael@0 | 1271 | UErrorCode *pErrorCode); |
michael@0 | 1272 | |
michael@0 | 1273 | /** |
michael@0 | 1274 | * Get the directionality of the text. |
michael@0 | 1275 | * |
michael@0 | 1276 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1277 | * |
michael@0 | 1278 | * @return a value of <code>UBIDI_LTR</code>, <code>UBIDI_RTL</code> |
michael@0 | 1279 | * or <code>UBIDI_MIXED</code> |
michael@0 | 1280 | * that indicates if the entire text |
michael@0 | 1281 | * represented by this object is unidirectional, |
michael@0 | 1282 | * and which direction, or if it is mixed-directional. |
michael@0 | 1283 | * Note - The value <code>UBIDI_NEUTRAL</code> is never returned from this method. |
michael@0 | 1284 | * |
michael@0 | 1285 | * @see UBiDiDirection |
michael@0 | 1286 | * @stable ICU 2.0 |
michael@0 | 1287 | */ |
michael@0 | 1288 | U_STABLE UBiDiDirection U_EXPORT2 |
michael@0 | 1289 | ubidi_getDirection(const UBiDi *pBiDi); |
michael@0 | 1290 | |
michael@0 | 1291 | /** |
michael@0 | 1292 | * Gets the base direction of the text provided according |
michael@0 | 1293 | * to the Unicode Bidirectional Algorithm. The base direction |
michael@0 | 1294 | * is derived from the first character in the string with bidirectional |
michael@0 | 1295 | * character type L, R, or AL. If the first such character has type L, |
michael@0 | 1296 | * <code>UBIDI_LTR</code> is returned. If the first such character has |
michael@0 | 1297 | * type R or AL, <code>UBIDI_RTL</code> is returned. If the string does |
michael@0 | 1298 | * not contain any character of these types, then |
michael@0 | 1299 | * <code>UBIDI_NEUTRAL</code> is returned. |
michael@0 | 1300 | * |
michael@0 | 1301 | * This is a lightweight function for use when only the base direction |
michael@0 | 1302 | * is needed and no further bidi processing of the text is needed. |
michael@0 | 1303 | * |
michael@0 | 1304 | * @param text is a pointer to the text whose base |
michael@0 | 1305 | * direction is needed. |
michael@0 | 1306 | * Note: the text must be (at least) @c length long. |
michael@0 | 1307 | * |
michael@0 | 1308 | * @param length is the length of the text; |
michael@0 | 1309 | * if <code>length==-1</code> then the text |
michael@0 | 1310 | * must be zero-terminated. |
michael@0 | 1311 | * |
michael@0 | 1312 | * @return <code>UBIDI_LTR</code>, <code>UBIDI_RTL</code>, |
michael@0 | 1313 | * <code>UBIDI_NEUTRAL</code> |
michael@0 | 1314 | * |
michael@0 | 1315 | * @see UBiDiDirection |
michael@0 | 1316 | * @stable ICU 4.6 |
michael@0 | 1317 | */ |
michael@0 | 1318 | U_STABLE UBiDiDirection U_EXPORT2 |
michael@0 | 1319 | ubidi_getBaseDirection(const UChar *text, int32_t length ); |
michael@0 | 1320 | |
michael@0 | 1321 | /** |
michael@0 | 1322 | * Get the pointer to the text. |
michael@0 | 1323 | * |
michael@0 | 1324 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1325 | * |
michael@0 | 1326 | * @return The pointer to the text that the UBiDi object was created for. |
michael@0 | 1327 | * |
michael@0 | 1328 | * @see ubidi_setPara |
michael@0 | 1329 | * @see ubidi_setLine |
michael@0 | 1330 | * @stable ICU 2.0 |
michael@0 | 1331 | */ |
michael@0 | 1332 | U_STABLE const UChar * U_EXPORT2 |
michael@0 | 1333 | ubidi_getText(const UBiDi *pBiDi); |
michael@0 | 1334 | |
michael@0 | 1335 | /** |
michael@0 | 1336 | * Get the length of the text. |
michael@0 | 1337 | * |
michael@0 | 1338 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1339 | * |
michael@0 | 1340 | * @return The length of the text that the UBiDi object was created for. |
michael@0 | 1341 | * @stable ICU 2.0 |
michael@0 | 1342 | */ |
michael@0 | 1343 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 1344 | ubidi_getLength(const UBiDi *pBiDi); |
michael@0 | 1345 | |
michael@0 | 1346 | /** |
michael@0 | 1347 | * Get the paragraph level of the text. |
michael@0 | 1348 | * |
michael@0 | 1349 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1350 | * |
michael@0 | 1351 | * @return The paragraph level. If there are multiple paragraphs, their |
michael@0 | 1352 | * level may vary if the required paraLevel is UBIDI_DEFAULT_LTR or |
michael@0 | 1353 | * UBIDI_DEFAULT_RTL. In that case, the level of the first paragraph |
michael@0 | 1354 | * is returned. |
michael@0 | 1355 | * |
michael@0 | 1356 | * @see UBiDiLevel |
michael@0 | 1357 | * @see ubidi_getParagraph |
michael@0 | 1358 | * @see ubidi_getParagraphByIndex |
michael@0 | 1359 | * @stable ICU 2.0 |
michael@0 | 1360 | */ |
michael@0 | 1361 | U_STABLE UBiDiLevel U_EXPORT2 |
michael@0 | 1362 | ubidi_getParaLevel(const UBiDi *pBiDi); |
michael@0 | 1363 | |
michael@0 | 1364 | /** |
michael@0 | 1365 | * Get the number of paragraphs. |
michael@0 | 1366 | * |
michael@0 | 1367 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1368 | * |
michael@0 | 1369 | * @return The number of paragraphs. |
michael@0 | 1370 | * @stable ICU 3.4 |
michael@0 | 1371 | */ |
michael@0 | 1372 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 1373 | ubidi_countParagraphs(UBiDi *pBiDi); |
michael@0 | 1374 | |
michael@0 | 1375 | /** |
michael@0 | 1376 | * Get a paragraph, given a position within the text. |
michael@0 | 1377 | * This function returns information about a paragraph.<br> |
michael@0 | 1378 | * Note: if the paragraph index is known, it is more efficient to |
michael@0 | 1379 | * retrieve the paragraph information using ubidi_getParagraphByIndex().<p> |
michael@0 | 1380 | * |
michael@0 | 1381 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1382 | * |
michael@0 | 1383 | * @param charIndex is the index of a character within the text, in the |
michael@0 | 1384 | * range <code>[0..ubidi_getProcessedLength(pBiDi)-1]</code>. |
michael@0 | 1385 | * |
michael@0 | 1386 | * @param pParaStart will receive the index of the first character of the |
michael@0 | 1387 | * paragraph in the text. |
michael@0 | 1388 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1389 | * value is not necessary. |
michael@0 | 1390 | * |
michael@0 | 1391 | * @param pParaLimit will receive the limit of the paragraph. |
michael@0 | 1392 | * The l-value that you point to here may be the |
michael@0 | 1393 | * same expression (variable) as the one for |
michael@0 | 1394 | * <code>charIndex</code>. |
michael@0 | 1395 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1396 | * value is not necessary. |
michael@0 | 1397 | * |
michael@0 | 1398 | * @param pParaLevel will receive the level of the paragraph. |
michael@0 | 1399 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1400 | * value is not necessary. |
michael@0 | 1401 | * |
michael@0 | 1402 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1403 | * |
michael@0 | 1404 | * @return The index of the paragraph containing the specified position. |
michael@0 | 1405 | * |
michael@0 | 1406 | * @see ubidi_getProcessedLength |
michael@0 | 1407 | * @stable ICU 3.4 |
michael@0 | 1408 | */ |
michael@0 | 1409 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 1410 | ubidi_getParagraph(const UBiDi *pBiDi, int32_t charIndex, int32_t *pParaStart, |
michael@0 | 1411 | int32_t *pParaLimit, UBiDiLevel *pParaLevel, |
michael@0 | 1412 | UErrorCode *pErrorCode); |
michael@0 | 1413 | |
michael@0 | 1414 | /** |
michael@0 | 1415 | * Get a paragraph, given the index of this paragraph. |
michael@0 | 1416 | * |
michael@0 | 1417 | * This function returns information about a paragraph.<p> |
michael@0 | 1418 | * |
michael@0 | 1419 | * @param pBiDi is the paragraph <code>UBiDi</code> object. |
michael@0 | 1420 | * |
michael@0 | 1421 | * @param paraIndex is the number of the paragraph, in the |
michael@0 | 1422 | * range <code>[0..ubidi_countParagraphs(pBiDi)-1]</code>. |
michael@0 | 1423 | * |
michael@0 | 1424 | * @param pParaStart will receive the index of the first character of the |
michael@0 | 1425 | * paragraph in the text. |
michael@0 | 1426 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1427 | * value is not necessary. |
michael@0 | 1428 | * |
michael@0 | 1429 | * @param pParaLimit will receive the limit of the paragraph. |
michael@0 | 1430 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1431 | * value is not necessary. |
michael@0 | 1432 | * |
michael@0 | 1433 | * @param pParaLevel will receive the level of the paragraph. |
michael@0 | 1434 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1435 | * value is not necessary. |
michael@0 | 1436 | * |
michael@0 | 1437 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1438 | * |
michael@0 | 1439 | * @stable ICU 3.4 |
michael@0 | 1440 | */ |
michael@0 | 1441 | U_STABLE void U_EXPORT2 |
michael@0 | 1442 | ubidi_getParagraphByIndex(const UBiDi *pBiDi, int32_t paraIndex, |
michael@0 | 1443 | int32_t *pParaStart, int32_t *pParaLimit, |
michael@0 | 1444 | UBiDiLevel *pParaLevel, UErrorCode *pErrorCode); |
michael@0 | 1445 | |
michael@0 | 1446 | /** |
michael@0 | 1447 | * Get the level for one character. |
michael@0 | 1448 | * |
michael@0 | 1449 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1450 | * |
michael@0 | 1451 | * @param charIndex the index of a character. It must be in the range |
michael@0 | 1452 | * [0..ubidi_getProcessedLength(pBiDi)]. |
michael@0 | 1453 | * |
michael@0 | 1454 | * @return The level for the character at charIndex (0 if charIndex is not |
michael@0 | 1455 | * in the valid range). |
michael@0 | 1456 | * |
michael@0 | 1457 | * @see UBiDiLevel |
michael@0 | 1458 | * @see ubidi_getProcessedLength |
michael@0 | 1459 | * @stable ICU 2.0 |
michael@0 | 1460 | */ |
michael@0 | 1461 | U_STABLE UBiDiLevel U_EXPORT2 |
michael@0 | 1462 | ubidi_getLevelAt(const UBiDi *pBiDi, int32_t charIndex); |
michael@0 | 1463 | |
michael@0 | 1464 | /** |
michael@0 | 1465 | * Get an array of levels for each character.<p> |
michael@0 | 1466 | * |
michael@0 | 1467 | * Note that this function may allocate memory under some |
michael@0 | 1468 | * circumstances, unlike <code>ubidi_getLevelAt()</code>. |
michael@0 | 1469 | * |
michael@0 | 1470 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object, whose |
michael@0 | 1471 | * text length must be strictly positive. |
michael@0 | 1472 | * |
michael@0 | 1473 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1474 | * |
michael@0 | 1475 | * @return The levels array for the text, |
michael@0 | 1476 | * or <code>NULL</code> if an error occurs. |
michael@0 | 1477 | * |
michael@0 | 1478 | * @see UBiDiLevel |
michael@0 | 1479 | * @see ubidi_getProcessedLength |
michael@0 | 1480 | * @stable ICU 2.0 |
michael@0 | 1481 | */ |
michael@0 | 1482 | U_STABLE const UBiDiLevel * U_EXPORT2 |
michael@0 | 1483 | ubidi_getLevels(UBiDi *pBiDi, UErrorCode *pErrorCode); |
michael@0 | 1484 | |
michael@0 | 1485 | /** |
michael@0 | 1486 | * Get a logical run. |
michael@0 | 1487 | * This function returns information about a run and is used |
michael@0 | 1488 | * to retrieve runs in logical order.<p> |
michael@0 | 1489 | * This is especially useful for line-breaking on a paragraph. |
michael@0 | 1490 | * |
michael@0 | 1491 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1492 | * |
michael@0 | 1493 | * @param logicalPosition is a logical position within the source text. |
michael@0 | 1494 | * |
michael@0 | 1495 | * @param pLogicalLimit will receive the limit of the corresponding run. |
michael@0 | 1496 | * The l-value that you point to here may be the |
michael@0 | 1497 | * same expression (variable) as the one for |
michael@0 | 1498 | * <code>logicalPosition</code>. |
michael@0 | 1499 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1500 | * value is not necessary. |
michael@0 | 1501 | * |
michael@0 | 1502 | * @param pLevel will receive the level of the corresponding run. |
michael@0 | 1503 | * This pointer can be <code>NULL</code> if this |
michael@0 | 1504 | * value is not necessary. |
michael@0 | 1505 | * |
michael@0 | 1506 | * @see ubidi_getProcessedLength |
michael@0 | 1507 | * @stable ICU 2.0 |
michael@0 | 1508 | */ |
michael@0 | 1509 | U_STABLE void U_EXPORT2 |
michael@0 | 1510 | ubidi_getLogicalRun(const UBiDi *pBiDi, int32_t logicalPosition, |
michael@0 | 1511 | int32_t *pLogicalLimit, UBiDiLevel *pLevel); |
michael@0 | 1512 | |
michael@0 | 1513 | /** |
michael@0 | 1514 | * Get the number of runs. |
michael@0 | 1515 | * This function may invoke the actual reordering on the |
michael@0 | 1516 | * <code>UBiDi</code> object, after <code>ubidi_setPara()</code> |
michael@0 | 1517 | * may have resolved only the levels of the text. Therefore, |
michael@0 | 1518 | * <code>ubidi_countRuns()</code> may have to allocate memory, |
michael@0 | 1519 | * and may fail doing so. |
michael@0 | 1520 | * |
michael@0 | 1521 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1522 | * |
michael@0 | 1523 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1524 | * |
michael@0 | 1525 | * @return The number of runs. |
michael@0 | 1526 | * @stable ICU 2.0 |
michael@0 | 1527 | */ |
michael@0 | 1528 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 1529 | ubidi_countRuns(UBiDi *pBiDi, UErrorCode *pErrorCode); |
michael@0 | 1530 | |
michael@0 | 1531 | /** |
michael@0 | 1532 | * Get one run's logical start, length, and directionality, |
michael@0 | 1533 | * which can be 0 for LTR or 1 for RTL. |
michael@0 | 1534 | * In an RTL run, the character at the logical start is |
michael@0 | 1535 | * visually on the right of the displayed run. |
michael@0 | 1536 | * The length is the number of characters in the run.<p> |
michael@0 | 1537 | * <code>ubidi_countRuns()</code> should be called |
michael@0 | 1538 | * before the runs are retrieved. |
michael@0 | 1539 | * |
michael@0 | 1540 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1541 | * |
michael@0 | 1542 | * @param runIndex is the number of the run in visual order, in the |
michael@0 | 1543 | * range <code>[0..ubidi_countRuns(pBiDi)-1]</code>. |
michael@0 | 1544 | * |
michael@0 | 1545 | * @param pLogicalStart is the first logical character index in the text. |
michael@0 | 1546 | * The pointer may be <code>NULL</code> if this index is not needed. |
michael@0 | 1547 | * |
michael@0 | 1548 | * @param pLength is the number of characters (at least one) in the run. |
michael@0 | 1549 | * The pointer may be <code>NULL</code> if this is not needed. |
michael@0 | 1550 | * |
michael@0 | 1551 | * @return the directionality of the run, |
michael@0 | 1552 | * <code>UBIDI_LTR==0</code> or <code>UBIDI_RTL==1</code>, |
michael@0 | 1553 | * never <code>UBIDI_MIXED</code>, |
michael@0 | 1554 | * never <code>UBIDI_NEUTRAL</code>. |
michael@0 | 1555 | * |
michael@0 | 1556 | * @see ubidi_countRuns |
michael@0 | 1557 | * |
michael@0 | 1558 | * Example: |
michael@0 | 1559 | * <pre> |
michael@0 | 1560 | * \code |
michael@0 | 1561 | * int32_t i, count=ubidi_countRuns(pBiDi), |
michael@0 | 1562 | * logicalStart, visualIndex=0, length; |
michael@0 | 1563 | * for(i=0; i<count; ++i) { |
michael@0 | 1564 | * if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, i, &logicalStart, &length)) { |
michael@0 | 1565 | * do { // LTR |
michael@0 | 1566 | * show_char(text[logicalStart++], visualIndex++); |
michael@0 | 1567 | * } while(--length>0); |
michael@0 | 1568 | * } else { |
michael@0 | 1569 | * logicalStart+=length; // logicalLimit |
michael@0 | 1570 | * do { // RTL |
michael@0 | 1571 | * show_char(text[--logicalStart], visualIndex++); |
michael@0 | 1572 | * } while(--length>0); |
michael@0 | 1573 | * } |
michael@0 | 1574 | * } |
michael@0 | 1575 | *\endcode |
michael@0 | 1576 | * </pre> |
michael@0 | 1577 | * |
michael@0 | 1578 | * Note that in right-to-left runs, code like this places |
michael@0 | 1579 | * second surrogates before first ones (which is generally a bad idea) |
michael@0 | 1580 | * and combining characters before base characters. |
michael@0 | 1581 | * <p> |
michael@0 | 1582 | * Use of <code>ubidi_writeReordered()</code>, optionally with the |
michael@0 | 1583 | * <code>#UBIDI_KEEP_BASE_COMBINING</code> option, can be considered in order |
michael@0 | 1584 | * to avoid these issues. |
michael@0 | 1585 | * @stable ICU 2.0 |
michael@0 | 1586 | */ |
michael@0 | 1587 | U_STABLE UBiDiDirection U_EXPORT2 |
michael@0 | 1588 | ubidi_getVisualRun(UBiDi *pBiDi, int32_t runIndex, |
michael@0 | 1589 | int32_t *pLogicalStart, int32_t *pLength); |
michael@0 | 1590 | |
michael@0 | 1591 | /** |
michael@0 | 1592 | * Get the visual position from a logical text position. |
michael@0 | 1593 | * If such a mapping is used many times on the same |
michael@0 | 1594 | * <code>UBiDi</code> object, then calling |
michael@0 | 1595 | * <code>ubidi_getLogicalMap()</code> is more efficient.<p> |
michael@0 | 1596 | * |
michael@0 | 1597 | * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no |
michael@0 | 1598 | * visual position because the corresponding text character is a Bidi control |
michael@0 | 1599 | * removed from output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>. |
michael@0 | 1600 | * <p> |
michael@0 | 1601 | * When the visual output is altered by using options of |
michael@0 | 1602 | * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>, |
michael@0 | 1603 | * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>, |
michael@0 | 1604 | * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the visual position returned may not |
michael@0 | 1605 | * be correct. It is advised to use, when possible, reordering options |
michael@0 | 1606 | * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>. |
michael@0 | 1607 | * <p> |
michael@0 | 1608 | * Note that in right-to-left runs, this mapping places |
michael@0 | 1609 | * second surrogates before first ones (which is generally a bad idea) |
michael@0 | 1610 | * and combining characters before base characters. |
michael@0 | 1611 | * Use of <code>ubidi_writeReordered()</code>, optionally with the |
michael@0 | 1612 | * <code>#UBIDI_KEEP_BASE_COMBINING</code> option can be considered instead |
michael@0 | 1613 | * of using the mapping, in order to avoid these issues. |
michael@0 | 1614 | * |
michael@0 | 1615 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1616 | * |
michael@0 | 1617 | * @param logicalIndex is the index of a character in the text. |
michael@0 | 1618 | * |
michael@0 | 1619 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1620 | * |
michael@0 | 1621 | * @return The visual position of this character. |
michael@0 | 1622 | * |
michael@0 | 1623 | * @see ubidi_getLogicalMap |
michael@0 | 1624 | * @see ubidi_getLogicalIndex |
michael@0 | 1625 | * @see ubidi_getProcessedLength |
michael@0 | 1626 | * @stable ICU 2.0 |
michael@0 | 1627 | */ |
michael@0 | 1628 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 1629 | ubidi_getVisualIndex(UBiDi *pBiDi, int32_t logicalIndex, UErrorCode *pErrorCode); |
michael@0 | 1630 | |
michael@0 | 1631 | /** |
michael@0 | 1632 | * Get the logical text position from a visual position. |
michael@0 | 1633 | * If such a mapping is used many times on the same |
michael@0 | 1634 | * <code>UBiDi</code> object, then calling |
michael@0 | 1635 | * <code>ubidi_getVisualMap()</code> is more efficient.<p> |
michael@0 | 1636 | * |
michael@0 | 1637 | * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no |
michael@0 | 1638 | * logical position because the corresponding text character is a Bidi mark |
michael@0 | 1639 | * inserted in the output by option <code>#UBIDI_OPTION_INSERT_MARKS</code>. |
michael@0 | 1640 | * <p> |
michael@0 | 1641 | * This is the inverse function to <code>ubidi_getVisualIndex()</code>. |
michael@0 | 1642 | * <p> |
michael@0 | 1643 | * When the visual output is altered by using options of |
michael@0 | 1644 | * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>, |
michael@0 | 1645 | * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>, |
michael@0 | 1646 | * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the logical position returned may not |
michael@0 | 1647 | * be correct. It is advised to use, when possible, reordering options |
michael@0 | 1648 | * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>. |
michael@0 | 1649 | * |
michael@0 | 1650 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1651 | * |
michael@0 | 1652 | * @param visualIndex is the visual position of a character. |
michael@0 | 1653 | * |
michael@0 | 1654 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1655 | * |
michael@0 | 1656 | * @return The index of this character in the text. |
michael@0 | 1657 | * |
michael@0 | 1658 | * @see ubidi_getVisualMap |
michael@0 | 1659 | * @see ubidi_getVisualIndex |
michael@0 | 1660 | * @see ubidi_getResultLength |
michael@0 | 1661 | * @stable ICU 2.0 |
michael@0 | 1662 | */ |
michael@0 | 1663 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 1664 | ubidi_getLogicalIndex(UBiDi *pBiDi, int32_t visualIndex, UErrorCode *pErrorCode); |
michael@0 | 1665 | |
michael@0 | 1666 | /** |
michael@0 | 1667 | * Get a logical-to-visual index map (array) for the characters in the UBiDi |
michael@0 | 1668 | * (paragraph or line) object. |
michael@0 | 1669 | * <p> |
michael@0 | 1670 | * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the |
michael@0 | 1671 | * corresponding text characters are Bidi controls removed from the visual |
michael@0 | 1672 | * output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>. |
michael@0 | 1673 | * <p> |
michael@0 | 1674 | * When the visual output is altered by using options of |
michael@0 | 1675 | * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>, |
michael@0 | 1676 | * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>, |
michael@0 | 1677 | * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the visual positions returned may not |
michael@0 | 1678 | * be correct. It is advised to use, when possible, reordering options |
michael@0 | 1679 | * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>. |
michael@0 | 1680 | * <p> |
michael@0 | 1681 | * Note that in right-to-left runs, this mapping places |
michael@0 | 1682 | * second surrogates before first ones (which is generally a bad idea) |
michael@0 | 1683 | * and combining characters before base characters. |
michael@0 | 1684 | * Use of <code>ubidi_writeReordered()</code>, optionally with the |
michael@0 | 1685 | * <code>#UBIDI_KEEP_BASE_COMBINING</code> option can be considered instead |
michael@0 | 1686 | * of using the mapping, in order to avoid these issues. |
michael@0 | 1687 | * |
michael@0 | 1688 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1689 | * |
michael@0 | 1690 | * @param indexMap is a pointer to an array of <code>ubidi_getProcessedLength()</code> |
michael@0 | 1691 | * indexes which will reflect the reordering of the characters. |
michael@0 | 1692 | * If option <code>#UBIDI_OPTION_INSERT_MARKS</code> is set, the number |
michael@0 | 1693 | * of elements allocated in <code>indexMap</code> must be no less than |
michael@0 | 1694 | * <code>ubidi_getResultLength()</code>. |
michael@0 | 1695 | * The array does not need to be initialized.<br><br> |
michael@0 | 1696 | * The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>. |
michael@0 | 1697 | * |
michael@0 | 1698 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1699 | * |
michael@0 | 1700 | * @see ubidi_getVisualMap |
michael@0 | 1701 | * @see ubidi_getVisualIndex |
michael@0 | 1702 | * @see ubidi_getProcessedLength |
michael@0 | 1703 | * @see ubidi_getResultLength |
michael@0 | 1704 | * @stable ICU 2.0 |
michael@0 | 1705 | */ |
michael@0 | 1706 | U_STABLE void U_EXPORT2 |
michael@0 | 1707 | ubidi_getLogicalMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode); |
michael@0 | 1708 | |
michael@0 | 1709 | /** |
michael@0 | 1710 | * Get a visual-to-logical index map (array) for the characters in the UBiDi |
michael@0 | 1711 | * (paragraph or line) object. |
michael@0 | 1712 | * <p> |
michael@0 | 1713 | * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the |
michael@0 | 1714 | * corresponding text characters are Bidi marks inserted in the visual output |
michael@0 | 1715 | * by the option <code>#UBIDI_OPTION_INSERT_MARKS</code>. |
michael@0 | 1716 | * <p> |
michael@0 | 1717 | * When the visual output is altered by using options of |
michael@0 | 1718 | * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>, |
michael@0 | 1719 | * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>, |
michael@0 | 1720 | * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the logical positions returned may not |
michael@0 | 1721 | * be correct. It is advised to use, when possible, reordering options |
michael@0 | 1722 | * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>. |
michael@0 | 1723 | * |
michael@0 | 1724 | * @param pBiDi is the paragraph or line <code>UBiDi</code> object. |
michael@0 | 1725 | * |
michael@0 | 1726 | * @param indexMap is a pointer to an array of <code>ubidi_getResultLength()</code> |
michael@0 | 1727 | * indexes which will reflect the reordering of the characters. |
michael@0 | 1728 | * If option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is set, the number |
michael@0 | 1729 | * of elements allocated in <code>indexMap</code> must be no less than |
michael@0 | 1730 | * <code>ubidi_getProcessedLength()</code>. |
michael@0 | 1731 | * The array does not need to be initialized.<br><br> |
michael@0 | 1732 | * The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>. |
michael@0 | 1733 | * |
michael@0 | 1734 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 1735 | * |
michael@0 | 1736 | * @see ubidi_getLogicalMap |
michael@0 | 1737 | * @see ubidi_getLogicalIndex |
michael@0 | 1738 | * @see ubidi_getProcessedLength |
michael@0 | 1739 | * @see ubidi_getResultLength |
michael@0 | 1740 | * @stable ICU 2.0 |
michael@0 | 1741 | */ |
michael@0 | 1742 | U_STABLE void U_EXPORT2 |
michael@0 | 1743 | ubidi_getVisualMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode); |
michael@0 | 1744 | |
michael@0 | 1745 | /** |
michael@0 | 1746 | * This is a convenience function that does not use a UBiDi object. |
michael@0 | 1747 | * It is intended to be used for when an application has determined the levels |
michael@0 | 1748 | * of objects (character sequences) and just needs to have them reordered (L2). |
michael@0 | 1749 | * This is equivalent to using <code>ubidi_getLogicalMap()</code> on a |
michael@0 | 1750 | * <code>UBiDi</code> object. |
michael@0 | 1751 | * |
michael@0 | 1752 | * @param levels is an array with <code>length</code> levels that have been determined by |
michael@0 | 1753 | * the application. |
michael@0 | 1754 | * |
michael@0 | 1755 | * @param length is the number of levels in the array, or, semantically, |
michael@0 | 1756 | * the number of objects to be reordered. |
michael@0 | 1757 | * It must be <code>length>0</code>. |
michael@0 | 1758 | * |
michael@0 | 1759 | * @param indexMap is a pointer to an array of <code>length</code> |
michael@0 | 1760 | * indexes which will reflect the reordering of the characters. |
michael@0 | 1761 | * The array does not need to be initialized.<p> |
michael@0 | 1762 | * The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>. |
michael@0 | 1763 | * @stable ICU 2.0 |
michael@0 | 1764 | */ |
michael@0 | 1765 | U_STABLE void U_EXPORT2 |
michael@0 | 1766 | ubidi_reorderLogical(const UBiDiLevel *levels, int32_t length, int32_t *indexMap); |
michael@0 | 1767 | |
michael@0 | 1768 | /** |
michael@0 | 1769 | * This is a convenience function that does not use a UBiDi object. |
michael@0 | 1770 | * It is intended to be used for when an application has determined the levels |
michael@0 | 1771 | * of objects (character sequences) and just needs to have them reordered (L2). |
michael@0 | 1772 | * This is equivalent to using <code>ubidi_getVisualMap()</code> on a |
michael@0 | 1773 | * <code>UBiDi</code> object. |
michael@0 | 1774 | * |
michael@0 | 1775 | * @param levels is an array with <code>length</code> levels that have been determined by |
michael@0 | 1776 | * the application. |
michael@0 | 1777 | * |
michael@0 | 1778 | * @param length is the number of levels in the array, or, semantically, |
michael@0 | 1779 | * the number of objects to be reordered. |
michael@0 | 1780 | * It must be <code>length>0</code>. |
michael@0 | 1781 | * |
michael@0 | 1782 | * @param indexMap is a pointer to an array of <code>length</code> |
michael@0 | 1783 | * indexes which will reflect the reordering of the characters. |
michael@0 | 1784 | * The array does not need to be initialized.<p> |
michael@0 | 1785 | * The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>. |
michael@0 | 1786 | * @stable ICU 2.0 |
michael@0 | 1787 | */ |
michael@0 | 1788 | U_STABLE void U_EXPORT2 |
michael@0 | 1789 | ubidi_reorderVisual(const UBiDiLevel *levels, int32_t length, int32_t *indexMap); |
michael@0 | 1790 | |
michael@0 | 1791 | /** |
michael@0 | 1792 | * Invert an index map. |
michael@0 | 1793 | * The index mapping of the first map is inverted and written to |
michael@0 | 1794 | * the second one. |
michael@0 | 1795 | * |
michael@0 | 1796 | * @param srcMap is an array with <code>length</code> elements |
michael@0 | 1797 | * which defines the original mapping from a source array containing |
michael@0 | 1798 | * <code>length</code> elements to a destination array. |
michael@0 | 1799 | * Some elements of the source array may have no mapping in the |
michael@0 | 1800 | * destination array. In that case, their value will be |
michael@0 | 1801 | * the special value <code>UBIDI_MAP_NOWHERE</code>. |
michael@0 | 1802 | * All elements must be >=0 or equal to <code>UBIDI_MAP_NOWHERE</code>. |
michael@0 | 1803 | * Some elements may have a value >= <code>length</code>, if the |
michael@0 | 1804 | * destination array has more elements than the source array. |
michael@0 | 1805 | * There must be no duplicate indexes (two or more elements with the |
michael@0 | 1806 | * same value except <code>UBIDI_MAP_NOWHERE</code>). |
michael@0 | 1807 | * |
michael@0 | 1808 | * @param destMap is an array with a number of elements equal to 1 + the highest |
michael@0 | 1809 | * value in <code>srcMap</code>. |
michael@0 | 1810 | * <code>destMap</code> will be filled with the inverse mapping. |
michael@0 | 1811 | * If element with index i in <code>srcMap</code> has a value k different |
michael@0 | 1812 | * from <code>UBIDI_MAP_NOWHERE</code>, this means that element i of |
michael@0 | 1813 | * the source array maps to element k in the destination array. |
michael@0 | 1814 | * The inverse map will have value i in its k-th element. |
michael@0 | 1815 | * For all elements of the destination array which do not map to |
michael@0 | 1816 | * an element in the source array, the corresponding element in the |
michael@0 | 1817 | * inverse map will have a value equal to <code>UBIDI_MAP_NOWHERE</code>. |
michael@0 | 1818 | * |
michael@0 | 1819 | * @param length is the length of each array. |
michael@0 | 1820 | * @see UBIDI_MAP_NOWHERE |
michael@0 | 1821 | * @stable ICU 2.0 |
michael@0 | 1822 | */ |
michael@0 | 1823 | U_STABLE void U_EXPORT2 |
michael@0 | 1824 | ubidi_invertMap(const int32_t *srcMap, int32_t *destMap, int32_t length); |
michael@0 | 1825 | |
michael@0 | 1826 | /** option flags for ubidi_writeReordered() */ |
michael@0 | 1827 | |
michael@0 | 1828 | /** |
michael@0 | 1829 | * option bit for ubidi_writeReordered(): |
michael@0 | 1830 | * keep combining characters after their base characters in RTL runs |
michael@0 | 1831 | * |
michael@0 | 1832 | * @see ubidi_writeReordered |
michael@0 | 1833 | * @stable ICU 2.0 |
michael@0 | 1834 | */ |
michael@0 | 1835 | #define UBIDI_KEEP_BASE_COMBINING 1 |
michael@0 | 1836 | |
michael@0 | 1837 | /** |
michael@0 | 1838 | * option bit for ubidi_writeReordered(): |
michael@0 | 1839 | * replace characters with the "mirrored" property in RTL runs |
michael@0 | 1840 | * by their mirror-image mappings |
michael@0 | 1841 | * |
michael@0 | 1842 | * @see ubidi_writeReordered |
michael@0 | 1843 | * @stable ICU 2.0 |
michael@0 | 1844 | */ |
michael@0 | 1845 | #define UBIDI_DO_MIRRORING 2 |
michael@0 | 1846 | |
michael@0 | 1847 | /** |
michael@0 | 1848 | * option bit for ubidi_writeReordered(): |
michael@0 | 1849 | * surround the run with LRMs if necessary; |
michael@0 | 1850 | * this is part of the approximate "inverse Bidi" algorithm |
michael@0 | 1851 | * |
michael@0 | 1852 | * <p>This option does not imply corresponding adjustment of the index |
michael@0 | 1853 | * mappings.</p> |
michael@0 | 1854 | * |
michael@0 | 1855 | * @see ubidi_setInverse |
michael@0 | 1856 | * @see ubidi_writeReordered |
michael@0 | 1857 | * @stable ICU 2.0 |
michael@0 | 1858 | */ |
michael@0 | 1859 | #define UBIDI_INSERT_LRM_FOR_NUMERIC 4 |
michael@0 | 1860 | |
michael@0 | 1861 | /** |
michael@0 | 1862 | * option bit for ubidi_writeReordered(): |
michael@0 | 1863 | * remove Bidi control characters |
michael@0 | 1864 | * (this does not affect #UBIDI_INSERT_LRM_FOR_NUMERIC) |
michael@0 | 1865 | * |
michael@0 | 1866 | * <p>This option does not imply corresponding adjustment of the index |
michael@0 | 1867 | * mappings.</p> |
michael@0 | 1868 | * |
michael@0 | 1869 | * @see ubidi_writeReordered |
michael@0 | 1870 | * @stable ICU 2.0 |
michael@0 | 1871 | */ |
michael@0 | 1872 | #define UBIDI_REMOVE_BIDI_CONTROLS 8 |
michael@0 | 1873 | |
michael@0 | 1874 | /** |
michael@0 | 1875 | * option bit for ubidi_writeReordered(): |
michael@0 | 1876 | * write the output in reverse order |
michael@0 | 1877 | * |
michael@0 | 1878 | * <p>This has the same effect as calling <code>ubidi_writeReordered()</code> |
michael@0 | 1879 | * first without this option, and then calling |
michael@0 | 1880 | * <code>ubidi_writeReverse()</code> without mirroring. |
michael@0 | 1881 | * Doing this in the same step is faster and avoids a temporary buffer. |
michael@0 | 1882 | * An example for using this option is output to a character terminal that |
michael@0 | 1883 | * is designed for RTL scripts and stores text in reverse order.</p> |
michael@0 | 1884 | * |
michael@0 | 1885 | * @see ubidi_writeReordered |
michael@0 | 1886 | * @stable ICU 2.0 |
michael@0 | 1887 | */ |
michael@0 | 1888 | #define UBIDI_OUTPUT_REVERSE 16 |
michael@0 | 1889 | |
michael@0 | 1890 | /** |
michael@0 | 1891 | * Get the length of the source text processed by the last call to |
michael@0 | 1892 | * <code>ubidi_setPara()</code>. This length may be different from the length |
michael@0 | 1893 | * of the source text if option <code>#UBIDI_OPTION_STREAMING</code> |
michael@0 | 1894 | * has been set. |
michael@0 | 1895 | * <br> |
michael@0 | 1896 | * Note that whenever the length of the text affects the execution or the |
michael@0 | 1897 | * result of a function, it is the processed length which must be considered, |
michael@0 | 1898 | * except for <code>ubidi_setPara</code> (which receives unprocessed source |
michael@0 | 1899 | * text) and <code>ubidi_getLength</code> (which returns the original length |
michael@0 | 1900 | * of the source text).<br> |
michael@0 | 1901 | * In particular, the processed length is the one to consider in the following |
michael@0 | 1902 | * cases: |
michael@0 | 1903 | * <ul> |
michael@0 | 1904 | * <li>maximum value of the <code>limit</code> argument of |
michael@0 | 1905 | * <code>ubidi_setLine</code></li> |
michael@0 | 1906 | * <li>maximum value of the <code>charIndex</code> argument of |
michael@0 | 1907 | * <code>ubidi_getParagraph</code></li> |
michael@0 | 1908 | * <li>maximum value of the <code>charIndex</code> argument of |
michael@0 | 1909 | * <code>ubidi_getLevelAt</code></li> |
michael@0 | 1910 | * <li>number of elements in the array returned by <code>ubidi_getLevels</code></li> |
michael@0 | 1911 | * <li>maximum value of the <code>logicalStart</code> argument of |
michael@0 | 1912 | * <code>ubidi_getLogicalRun</code></li> |
michael@0 | 1913 | * <li>maximum value of the <code>logicalIndex</code> argument of |
michael@0 | 1914 | * <code>ubidi_getVisualIndex</code></li> |
michael@0 | 1915 | * <li>number of elements filled in the <code>*indexMap</code> argument of |
michael@0 | 1916 | * <code>ubidi_getLogicalMap</code></li> |
michael@0 | 1917 | * <li>length of text processed by <code>ubidi_writeReordered</code></li> |
michael@0 | 1918 | * </ul> |
michael@0 | 1919 | * |
michael@0 | 1920 | * @param pBiDi is the paragraph <code>UBiDi</code> object. |
michael@0 | 1921 | * |
michael@0 | 1922 | * @return The length of the part of the source text processed by |
michael@0 | 1923 | * the last call to <code>ubidi_setPara</code>. |
michael@0 | 1924 | * @see ubidi_setPara |
michael@0 | 1925 | * @see UBIDI_OPTION_STREAMING |
michael@0 | 1926 | * @stable ICU 3.6 |
michael@0 | 1927 | */ |
michael@0 | 1928 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 1929 | ubidi_getProcessedLength(const UBiDi *pBiDi); |
michael@0 | 1930 | |
michael@0 | 1931 | /** |
michael@0 | 1932 | * Get the length of the reordered text resulting from the last call to |
michael@0 | 1933 | * <code>ubidi_setPara()</code>. This length may be different from the length |
michael@0 | 1934 | * of the source text if option <code>#UBIDI_OPTION_INSERT_MARKS</code> |
michael@0 | 1935 | * or option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> has been set. |
michael@0 | 1936 | * <br> |
michael@0 | 1937 | * This resulting length is the one to consider in the following cases: |
michael@0 | 1938 | * <ul> |
michael@0 | 1939 | * <li>maximum value of the <code>visualIndex</code> argument of |
michael@0 | 1940 | * <code>ubidi_getLogicalIndex</code></li> |
michael@0 | 1941 | * <li>number of elements of the <code>*indexMap</code> argument of |
michael@0 | 1942 | * <code>ubidi_getVisualMap</code></li> |
michael@0 | 1943 | * </ul> |
michael@0 | 1944 | * Note that this length stays identical to the source text length if |
michael@0 | 1945 | * Bidi marks are inserted or removed using option bits of |
michael@0 | 1946 | * <code>ubidi_writeReordered</code>, or if option |
michael@0 | 1947 | * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> has been set. |
michael@0 | 1948 | * |
michael@0 | 1949 | * @param pBiDi is the paragraph <code>UBiDi</code> object. |
michael@0 | 1950 | * |
michael@0 | 1951 | * @return The length of the reordered text resulting from |
michael@0 | 1952 | * the last call to <code>ubidi_setPara</code>. |
michael@0 | 1953 | * @see ubidi_setPara |
michael@0 | 1954 | * @see UBIDI_OPTION_INSERT_MARKS |
michael@0 | 1955 | * @see UBIDI_OPTION_REMOVE_CONTROLS |
michael@0 | 1956 | * @stable ICU 3.6 |
michael@0 | 1957 | */ |
michael@0 | 1958 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 1959 | ubidi_getResultLength(const UBiDi *pBiDi); |
michael@0 | 1960 | |
michael@0 | 1961 | U_CDECL_BEGIN |
michael@0 | 1962 | /** |
michael@0 | 1963 | * value returned by <code>UBiDiClassCallback</code> callbacks when |
michael@0 | 1964 | * there is no need to override the standard Bidi class for a given code point. |
michael@0 | 1965 | * @see UBiDiClassCallback |
michael@0 | 1966 | * @stable ICU 3.6 |
michael@0 | 1967 | */ |
michael@0 | 1968 | #define U_BIDI_CLASS_DEFAULT U_CHAR_DIRECTION_COUNT |
michael@0 | 1969 | |
michael@0 | 1970 | /** |
michael@0 | 1971 | * Callback type declaration for overriding default Bidi class values with |
michael@0 | 1972 | * custom ones. |
michael@0 | 1973 | * <p>Usually, the function pointer will be propagated to a <code>UBiDi</code> |
michael@0 | 1974 | * object by calling the <code>ubidi_setClassCallback()</code> function; |
michael@0 | 1975 | * then the callback will be invoked by the UBA implementation any time the |
michael@0 | 1976 | * class of a character is to be determined.</p> |
michael@0 | 1977 | * |
michael@0 | 1978 | * @param context is a pointer to the callback private data. |
michael@0 | 1979 | * |
michael@0 | 1980 | * @param c is the code point to get a Bidi class for. |
michael@0 | 1981 | * |
michael@0 | 1982 | * @return The directional property / Bidi class for the given code point |
michael@0 | 1983 | * <code>c</code> if the default class has been overridden, or |
michael@0 | 1984 | * <code>#U_BIDI_CLASS_DEFAULT</code> if the standard Bidi class value |
michael@0 | 1985 | * for <code>c</code> is to be used. |
michael@0 | 1986 | * @see ubidi_setClassCallback |
michael@0 | 1987 | * @see ubidi_getClassCallback |
michael@0 | 1988 | * @stable ICU 3.6 |
michael@0 | 1989 | */ |
michael@0 | 1990 | typedef UCharDirection U_CALLCONV |
michael@0 | 1991 | UBiDiClassCallback(const void *context, UChar32 c); |
michael@0 | 1992 | |
michael@0 | 1993 | U_CDECL_END |
michael@0 | 1994 | |
michael@0 | 1995 | /** |
michael@0 | 1996 | * Retrieve the Bidi class for a given code point. |
michael@0 | 1997 | * <p>If a <code>#UBiDiClassCallback</code> callback is defined and returns a |
michael@0 | 1998 | * value other than <code>#U_BIDI_CLASS_DEFAULT</code>, that value is used; |
michael@0 | 1999 | * otherwise the default class determination mechanism is invoked.</p> |
michael@0 | 2000 | * |
michael@0 | 2001 | * @param pBiDi is the paragraph <code>UBiDi</code> object. |
michael@0 | 2002 | * |
michael@0 | 2003 | * @param c is the code point whose Bidi class must be retrieved. |
michael@0 | 2004 | * |
michael@0 | 2005 | * @return The Bidi class for character <code>c</code> based |
michael@0 | 2006 | * on the given <code>pBiDi</code> instance. |
michael@0 | 2007 | * @see UBiDiClassCallback |
michael@0 | 2008 | * @stable ICU 3.6 |
michael@0 | 2009 | */ |
michael@0 | 2010 | U_STABLE UCharDirection U_EXPORT2 |
michael@0 | 2011 | ubidi_getCustomizedClass(UBiDi *pBiDi, UChar32 c); |
michael@0 | 2012 | |
michael@0 | 2013 | /** |
michael@0 | 2014 | * Set the callback function and callback data used by the UBA |
michael@0 | 2015 | * implementation for Bidi class determination. |
michael@0 | 2016 | * <p>This may be useful for assigning Bidi classes to PUA characters, or |
michael@0 | 2017 | * for special application needs. For instance, an application may want to |
michael@0 | 2018 | * handle all spaces like L or R characters (according to the base direction) |
michael@0 | 2019 | * when creating the visual ordering of logical lines which are part of a report |
michael@0 | 2020 | * organized in columns: there should not be interaction between adjacent |
michael@0 | 2021 | * cells.<p> |
michael@0 | 2022 | * |
michael@0 | 2023 | * @param pBiDi is the paragraph <code>UBiDi</code> object. |
michael@0 | 2024 | * |
michael@0 | 2025 | * @param newFn is the new callback function pointer. |
michael@0 | 2026 | * |
michael@0 | 2027 | * @param newContext is the new callback context pointer. This can be NULL. |
michael@0 | 2028 | * |
michael@0 | 2029 | * @param oldFn fillin: Returns the old callback function pointer. This can be |
michael@0 | 2030 | * NULL. |
michael@0 | 2031 | * |
michael@0 | 2032 | * @param oldContext fillin: Returns the old callback's context. This can be |
michael@0 | 2033 | * NULL. |
michael@0 | 2034 | * |
michael@0 | 2035 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 2036 | * |
michael@0 | 2037 | * @see ubidi_getClassCallback |
michael@0 | 2038 | * @stable ICU 3.6 |
michael@0 | 2039 | */ |
michael@0 | 2040 | U_STABLE void U_EXPORT2 |
michael@0 | 2041 | ubidi_setClassCallback(UBiDi *pBiDi, UBiDiClassCallback *newFn, |
michael@0 | 2042 | const void *newContext, UBiDiClassCallback **oldFn, |
michael@0 | 2043 | const void **oldContext, UErrorCode *pErrorCode); |
michael@0 | 2044 | |
michael@0 | 2045 | /** |
michael@0 | 2046 | * Get the current callback function used for Bidi class determination. |
michael@0 | 2047 | * |
michael@0 | 2048 | * @param pBiDi is the paragraph <code>UBiDi</code> object. |
michael@0 | 2049 | * |
michael@0 | 2050 | * @param fn fillin: Returns the callback function pointer. |
michael@0 | 2051 | * |
michael@0 | 2052 | * @param context fillin: Returns the callback's private context. |
michael@0 | 2053 | * |
michael@0 | 2054 | * @see ubidi_setClassCallback |
michael@0 | 2055 | * @stable ICU 3.6 |
michael@0 | 2056 | */ |
michael@0 | 2057 | U_STABLE void U_EXPORT2 |
michael@0 | 2058 | ubidi_getClassCallback(UBiDi *pBiDi, UBiDiClassCallback **fn, const void **context); |
michael@0 | 2059 | |
michael@0 | 2060 | /** |
michael@0 | 2061 | * Take a <code>UBiDi</code> object containing the reordering |
michael@0 | 2062 | * information for a piece of text (one or more paragraphs) set by |
michael@0 | 2063 | * <code>ubidi_setPara()</code> or for a line of text set by |
michael@0 | 2064 | * <code>ubidi_setLine()</code> and write a reordered string to the |
michael@0 | 2065 | * destination buffer. |
michael@0 | 2066 | * |
michael@0 | 2067 | * This function preserves the integrity of characters with multiple |
michael@0 | 2068 | * code units and (optionally) combining characters. |
michael@0 | 2069 | * Characters in RTL runs can be replaced by mirror-image characters |
michael@0 | 2070 | * in the destination buffer. Note that "real" mirroring has |
michael@0 | 2071 | * to be done in a rendering engine by glyph selection |
michael@0 | 2072 | * and that for many "mirrored" characters there are no |
michael@0 | 2073 | * Unicode characters as mirror-image equivalents. |
michael@0 | 2074 | * There are also options to insert or remove Bidi control |
michael@0 | 2075 | * characters; see the description of the <code>destSize</code> |
michael@0 | 2076 | * and <code>options</code> parameters and of the option bit flags. |
michael@0 | 2077 | * |
michael@0 | 2078 | * @param pBiDi A pointer to a <code>UBiDi</code> object that |
michael@0 | 2079 | * is set by <code>ubidi_setPara()</code> or |
michael@0 | 2080 | * <code>ubidi_setLine()</code> and contains the reordering |
michael@0 | 2081 | * information for the text that it was defined for, |
michael@0 | 2082 | * as well as a pointer to that text.<br><br> |
michael@0 | 2083 | * The text was aliased (only the pointer was stored |
michael@0 | 2084 | * without copying the contents) and must not have been modified |
michael@0 | 2085 | * since the <code>ubidi_setPara()</code> call. |
michael@0 | 2086 | * |
michael@0 | 2087 | * @param dest A pointer to where the reordered text is to be copied. |
michael@0 | 2088 | * The source text and <code>dest[destSize]</code> |
michael@0 | 2089 | * must not overlap. |
michael@0 | 2090 | * |
michael@0 | 2091 | * @param destSize The size of the <code>dest</code> buffer, |
michael@0 | 2092 | * in number of UChars. |
michael@0 | 2093 | * If the <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code> |
michael@0 | 2094 | * option is set, then the destination length could be |
michael@0 | 2095 | * as large as |
michael@0 | 2096 | * <code>ubidi_getLength(pBiDi)+2*ubidi_countRuns(pBiDi)</code>. |
michael@0 | 2097 | * If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option |
michael@0 | 2098 | * is set, then the destination length may be less than |
michael@0 | 2099 | * <code>ubidi_getLength(pBiDi)</code>. |
michael@0 | 2100 | * If none of these options is set, then the destination length |
michael@0 | 2101 | * will be exactly <code>ubidi_getProcessedLength(pBiDi)</code>. |
michael@0 | 2102 | * |
michael@0 | 2103 | * @param options A bit set of options for the reordering that control |
michael@0 | 2104 | * how the reordered text is written. |
michael@0 | 2105 | * The options include mirroring the characters on a code |
michael@0 | 2106 | * point basis and inserting LRM characters, which is used |
michael@0 | 2107 | * especially for transforming visually stored text |
michael@0 | 2108 | * to logically stored text (although this is still an |
michael@0 | 2109 | * imperfect implementation of an "inverse Bidi" algorithm |
michael@0 | 2110 | * because it uses the "forward Bidi" algorithm at its core). |
michael@0 | 2111 | * The available options are: |
michael@0 | 2112 | * <code>#UBIDI_DO_MIRRORING</code>, |
michael@0 | 2113 | * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>, |
michael@0 | 2114 | * <code>#UBIDI_KEEP_BASE_COMBINING</code>, |
michael@0 | 2115 | * <code>#UBIDI_OUTPUT_REVERSE</code>, |
michael@0 | 2116 | * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code> |
michael@0 | 2117 | * |
michael@0 | 2118 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 2119 | * |
michael@0 | 2120 | * @return The length of the output string. |
michael@0 | 2121 | * |
michael@0 | 2122 | * @see ubidi_getProcessedLength |
michael@0 | 2123 | * @stable ICU 2.0 |
michael@0 | 2124 | */ |
michael@0 | 2125 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 2126 | ubidi_writeReordered(UBiDi *pBiDi, |
michael@0 | 2127 | UChar *dest, int32_t destSize, |
michael@0 | 2128 | uint16_t options, |
michael@0 | 2129 | UErrorCode *pErrorCode); |
michael@0 | 2130 | |
michael@0 | 2131 | /** |
michael@0 | 2132 | * Reverse a Right-To-Left run of Unicode text. |
michael@0 | 2133 | * |
michael@0 | 2134 | * This function preserves the integrity of characters with multiple |
michael@0 | 2135 | * code units and (optionally) combining characters. |
michael@0 | 2136 | * Characters can be replaced by mirror-image characters |
michael@0 | 2137 | * in the destination buffer. Note that "real" mirroring has |
michael@0 | 2138 | * to be done in a rendering engine by glyph selection |
michael@0 | 2139 | * and that for many "mirrored" characters there are no |
michael@0 | 2140 | * Unicode characters as mirror-image equivalents. |
michael@0 | 2141 | * There are also options to insert or remove Bidi control |
michael@0 | 2142 | * characters. |
michael@0 | 2143 | * |
michael@0 | 2144 | * This function is the implementation for reversing RTL runs as part |
michael@0 | 2145 | * of <code>ubidi_writeReordered()</code>. For detailed descriptions |
michael@0 | 2146 | * of the parameters, see there. |
michael@0 | 2147 | * Since no Bidi controls are inserted here, the output string length |
michael@0 | 2148 | * will never exceed <code>srcLength</code>. |
michael@0 | 2149 | * |
michael@0 | 2150 | * @see ubidi_writeReordered |
michael@0 | 2151 | * |
michael@0 | 2152 | * @param src A pointer to the RTL run text. |
michael@0 | 2153 | * |
michael@0 | 2154 | * @param srcLength The length of the RTL run. |
michael@0 | 2155 | * |
michael@0 | 2156 | * @param dest A pointer to where the reordered text is to be copied. |
michael@0 | 2157 | * <code>src[srcLength]</code> and <code>dest[destSize]</code> |
michael@0 | 2158 | * must not overlap. |
michael@0 | 2159 | * |
michael@0 | 2160 | * @param destSize The size of the <code>dest</code> buffer, |
michael@0 | 2161 | * in number of UChars. |
michael@0 | 2162 | * If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option |
michael@0 | 2163 | * is set, then the destination length may be less than |
michael@0 | 2164 | * <code>srcLength</code>. |
michael@0 | 2165 | * If this option is not set, then the destination length |
michael@0 | 2166 | * will be exactly <code>srcLength</code>. |
michael@0 | 2167 | * |
michael@0 | 2168 | * @param options A bit set of options for the reordering that control |
michael@0 | 2169 | * how the reordered text is written. |
michael@0 | 2170 | * See the <code>options</code> parameter in <code>ubidi_writeReordered()</code>. |
michael@0 | 2171 | * |
michael@0 | 2172 | * @param pErrorCode must be a valid pointer to an error code value. |
michael@0 | 2173 | * |
michael@0 | 2174 | * @return The length of the output string. |
michael@0 | 2175 | * @stable ICU 2.0 |
michael@0 | 2176 | */ |
michael@0 | 2177 | U_STABLE int32_t U_EXPORT2 |
michael@0 | 2178 | ubidi_writeReverse(const UChar *src, int32_t srcLength, |
michael@0 | 2179 | UChar *dest, int32_t destSize, |
michael@0 | 2180 | uint16_t options, |
michael@0 | 2181 | UErrorCode *pErrorCode); |
michael@0 | 2182 | |
michael@0 | 2183 | /*#define BIDI_SAMPLE_CODE*/ |
michael@0 | 2184 | /*@}*/ |
michael@0 | 2185 | |
michael@0 | 2186 | #endif |