1/* GENERATED SOURCE. DO NOT MODIFY. */ 2// © 2016 and later: Unicode, Inc. and others. 3// License & terms of use: http://www.unicode.org/copyright.html#License 4/* 5******************************************************************************* 6* Copyright (C) 2001-2012, International Business Machines 7* Corporation and others. All Rights Reserved. 8******************************************************************************* 9*/ 10 11package android.icu.text; 12 13import android.icu.impl.UBiDiProps; 14import android.icu.lang.UCharacterDirection; 15 16/** 17 * Shape Arabic text on a character basis. 18 * 19 * <p>ArabicShaping performs basic operations for "shaping" Arabic text. It is most 20 * useful for use with legacy data formats and legacy display technology 21 * (simple terminals). All operations are performed on Unicode characters.</p> 22 * 23 * <p>Text-based shaping means that some character code points in the text are 24 * replaced by others depending on the context. It transforms one kind of text 25 * into another. In comparison, modern displays for Arabic text select 26 * appropriate, context-dependent font glyphs for each text element, which means 27 * that they transform text into a glyph vector.</p> 28 * 29 * <p>Text transformations are necessary when modern display technology is not 30 * available or when text needs to be transformed to or from legacy formats that 31 * use "shaped" characters. Since the Arabic script is cursive, connecting 32 * adjacent letters to each other, computers select images for each letter based 33 * on the surrounding letters. This usually results in four images per Arabic 34 * letter: initial, middle, final, and isolated forms. In Unicode, on the other 35 * hand, letters are normally stored abstract, and a display system is expected 36 * to select the necessary glyphs. (This makes searching and other text 37 * processing easier because the same letter has only one code.) It is possible 38 * to mimic this with text transformations because there are characters in 39 * Unicode that are rendered as letters with a specific shape 40 * (or cursive connectivity). They were included for interoperability with 41 * legacy systems and codepages, and for unsophisticated display systems.</p> 42 * 43 * <p>A second kind of text transformations is supported for Arabic digits: 44 * For compatibility with legacy codepages that only include European digits, 45 * it is possible to replace one set of digits by another, changing the 46 * character code points. These operations can be performed for either 47 * Arabic-Indic Digits (U+0660...U+0669) or Eastern (Extended) Arabic-Indic 48 * digits (U+06f0...U+06f9).</p> 49 * 50 * <p>Some replacements may result in more or fewer characters (code points). 51 * By default, this means that the destination buffer may receive text with a 52 * length different from the source length. Some legacy systems rely on the 53 * length of the text to be constant. They expect extra spaces to be added 54 * or consumed either next to the affected character or at the end of the 55 * text.</p> 56 * @hide Only a subset of ICU is exposed in Android 57 */ 58public final class ArabicShaping { 59 private final int options; 60 private boolean isLogical; // convenience 61 private boolean spacesRelativeToTextBeginEnd; 62 private char tailChar; 63 64 /** 65 * Convert a range of text in the source array, putting the result 66 * into a range of text in the destination array, and return the number 67 * of characters written. 68 * 69 * @param source An array containing the input text 70 * @param sourceStart The start of the range of text to convert 71 * @param sourceLength The length of the range of text to convert 72 * @param dest The destination array that will receive the result. 73 * It may be <code>NULL</code> only if <code>destSize</code> is 0. 74 * @param destStart The start of the range of the destination buffer to use. 75 * @param destSize The size (capacity) of the destination buffer. 76 * If <code>destSize</code> is 0, then no output is produced, 77 * but the necessary buffer size is returned ("preflighting"). This 78 * does not validate the text against the options, for example, 79 * if letters are being unshaped, and spaces are being consumed 80 * following lamalef, this will not detect a lamalef without a 81 * corresponding space. An error will be thrown when the actual 82 * conversion is attempted. 83 * @return The number of chars written to the destination buffer. 84 * If an error occurs, then no output was written, or it may be 85 * incomplete. 86 * @throws ArabicShapingException if the text cannot be converted according to the options. 87 */ 88 public int shape(char[] source, int sourceStart, int sourceLength, 89 char[] dest, int destStart, int destSize) throws ArabicShapingException { 90 if (source == null) { 91 throw new IllegalArgumentException("source can not be null"); 92 } 93 if (sourceStart < 0 || sourceLength < 0 || sourceStart + sourceLength > source.length) { 94 throw new IllegalArgumentException("bad source start (" + sourceStart + 95 ") or length (" + sourceLength + 96 ") for buffer of length " + source.length); 97 } 98 if (dest == null && destSize != 0) { 99 throw new IllegalArgumentException("null dest requires destSize == 0"); 100 } 101 if ((destSize != 0) && 102 (destStart < 0 || destSize < 0 || destStart + destSize > dest.length)) { 103 throw new IllegalArgumentException("bad dest start (" + destStart + 104 ") or size (" + destSize + 105 ") for buffer of length " + dest.length); 106 } 107 /* Validate input options */ 108 if ( ((options&TASHKEEL_MASK) != 0) && 109 !(((options & TASHKEEL_MASK)==TASHKEEL_BEGIN) || 110 ((options & TASHKEEL_MASK)==TASHKEEL_END) || 111 ((options & TASHKEEL_MASK)==TASHKEEL_RESIZE) || 112 ((options & TASHKEEL_MASK)==TASHKEEL_REPLACE_BY_TATWEEL))) { 113 throw new IllegalArgumentException("Wrong Tashkeel argument"); 114 } 115 116 ///CLOVER:OFF 117 //According to Steven Loomis, the code is unreachable when you OR all the constants within the if statements 118 if(((options&LAMALEF_MASK) != 0) && 119 !(((options & LAMALEF_MASK)==LAMALEF_BEGIN) || 120 ((options & LAMALEF_MASK)==LAMALEF_END) || 121 ((options & LAMALEF_MASK)==LAMALEF_RESIZE) || 122 ((options & LAMALEF_MASK)==LAMALEF_AUTO) || 123 ((options & LAMALEF_MASK)==LAMALEF_NEAR))) { 124 throw new IllegalArgumentException("Wrong Lam Alef argument"); 125 } 126 ///CLOVER:ON 127 128 /* Validate Tashkeel (Tashkeel replacement options should be enabled in shaping mode only)*/ 129 if(((options&TASHKEEL_MASK) != 0) && (options&LETTERS_MASK) == LETTERS_UNSHAPE) { 130 throw new IllegalArgumentException("Tashkeel replacement should not be enabled in deshaping mode "); 131 } 132 return internalShape(source, sourceStart, sourceLength, dest, destStart, destSize); 133 } 134 135 /** 136 * Convert a range of text in place. This may only be used if the Length option 137 * does not grow or shrink the text. 138 * 139 * @param source An array containing the input text 140 * @param start The start of the range of text to convert 141 * @param length The length of the range of text to convert 142 * @throws ArabicShapingException if the text cannot be converted according to the options. 143 */ 144 public void shape(char[] source, int start, int length) throws ArabicShapingException { 145 if ((options & LAMALEF_MASK) == LAMALEF_RESIZE) { 146 throw new ArabicShapingException("Cannot shape in place with length option resize."); 147 } 148 shape(source, start, length, source, start, length); 149 } 150 151 /** 152 * Convert a string, returning the new string. 153 * 154 * @param text the string to convert 155 * @return the converted string 156 * @throws ArabicShapingException if the string cannot be converted according to the options. 157 */ 158 public String shape(String text) throws ArabicShapingException { 159 char[] src = text.toCharArray(); 160 char[] dest = src; 161 if (((options & LAMALEF_MASK) == LAMALEF_RESIZE) && 162 ((options & LETTERS_MASK) == LETTERS_UNSHAPE)) { 163 164 dest = new char[src.length * 2]; // max 165 } 166 int len = shape(src, 0, src.length, dest, 0, dest.length); 167 168 return new String(dest, 0, len); 169 } 170 171 /** 172 * Construct ArabicShaping using the options flags. 173 * The flags are as follows:<br> 174 * 'LENGTH' flags control whether the text can change size, and if not, 175 * how to maintain the size of the text when LamAlef ligatures are 176 * formed or broken.<br> 177 * 'TEXT_DIRECTION' flags control whether the text is read and written 178 * in visual order or in logical order.<br> 179 * 'LETTERS_SHAPE' flags control whether conversion is to or from 180 * presentation forms.<br> 181 * 'DIGITS' flags control whether digits are shaped, and whether from 182 * European to Arabic-Indic or vice-versa.<br> 183 * 'DIGIT_TYPE' flags control whether standard or extended Arabic-Indic 184 * digits are used when performing digit conversion. 185 */ 186 public ArabicShaping(int options) { 187 this.options = options; 188 if ((options & DIGITS_MASK) > 0x80) { 189 throw new IllegalArgumentException("bad DIGITS options"); 190 } 191 192 isLogical = ( (options & TEXT_DIRECTION_MASK) == TEXT_DIRECTION_LOGICAL ); 193 /* Validate options */ 194 spacesRelativeToTextBeginEnd = ( (options & SPACES_RELATIVE_TO_TEXT_MASK) == SPACES_RELATIVE_TO_TEXT_BEGIN_END ); 195 if ( (options&SHAPE_TAIL_TYPE_MASK) == SHAPE_TAIL_NEW_UNICODE){ 196 tailChar = NEW_TAIL_CHAR; 197 } else { 198 tailChar = OLD_TAIL_CHAR; 199 } 200 } 201 202 /* Seen Tail options */ 203 /** 204 * Memory option: the result must have the same length as the source. 205 * Shaping mode: The SEEN family character will expand into two characters using space near 206 * the SEEN family character(i.e. the space after the character). 207 * if there are no spaces found, ArabicShapingException will be thrown 208 * 209 * De-shaping mode: Any Seen character followed by Tail character will be 210 * replaced by one cell Seen and a space will replace the Tail. 211 * Affects: Seen options 212 */ 213 public static final int SEEN_TWOCELL_NEAR = 0x200000; 214 215 /** Bit mask for Seen memory options. 216 */ 217 public static final int SEEN_MASK = 0x700000; 218 219 /* YehHamza options */ 220 /** 221 * Memory option: the result must have the same length as the source. 222 * Shaping mode: The YEHHAMZA character will expand into two characters using space near it 223 * (i.e. the space after the character) 224 * if there are no spaces found, ArabicShapingException will be thrown 225 * 226 * De-shaping mode: Any Yeh (final or isolated) character followed by Hamza character will be 227 * replaced by one cell YehHamza and space will replace the Hamza. 228 * Affects: YehHamza options 229 */ 230 public static final int YEHHAMZA_TWOCELL_NEAR = 0x1000000; 231 232 233 /** Bit mask for YehHamza memory options. 234 */ 235 public static final int YEHHAMZA_MASK = 0x3800000; 236 237 /* New Tashkeel options */ 238 /** 239 * Memory option: the result must have the same length as the source. 240 * Shaping mode: Tashkeel characters will be replaced by spaces. 241 * Spaces will be placed at beginning of the buffer 242 * 243 * De-shaping mode: N/A 244 * Affects: Tashkeel options 245 */ 246 public static final int TASHKEEL_BEGIN = 0x40000; 247 248 /** 249 * Memory option: the result must have the same length as the source. 250 * Shaping mode: Tashkeel characters will be replaced by spaces. 251 * Spaces will be placed at end of the buffer 252 * 253 * De-shaping mode: N/A 254 * Affects: Tashkeel options 255 */ 256 public static final int TASHKEEL_END = 0x60000; 257 258 /** 259 * Memory option: allow the result to have a different length than the source. 260 * Shaping mode: Tashkeel characters will be removed, buffer length will shrink. 261 * De-shaping mode: N/A 262 * 263 * Affects: Tashkeel options 264 */ 265 public static final int TASHKEEL_RESIZE = 0x80000; 266 267 /** 268 * Memory option: the result must have the same length as the source. 269 * Shaping mode: Tashkeel characters will be replaced by Tatweel if it is connected to adjacent 270 * characters (i.e. shaped on Tatweel) or replaced by space if it is not connected. 271 * 272 * De-shaping mode: N/A 273 * Affects: YehHamza options 274 */ 275 public static final int TASHKEEL_REPLACE_BY_TATWEEL = 0xC0000; 276 277 /** Bit mask for Tashkeel replacement with Space or Tatweel memory options. 278 */ 279 public static final int TASHKEEL_MASK = 0xE0000; 280 281 /* Space location Control options */ 282 /** 283 * This option effects the meaning of BEGIN and END options. if this option is not used the default 284 * for BEGIN and END will be as following: 285 * The Default (for both Visual LTR, Visual RTL and Logical Text) 286 * 1. BEGIN always refers to the start address of physical memory. 287 * 2. END always refers to the end address of physical memory. 288 * 289 * If this option is used it will swap the meaning of BEGIN and END only for Visual LTR text. 290 * 291 * The affect on BEGIN and END Memory Options will be as following: 292 * A. BEGIN For Visual LTR text: This will be the beginning (right side) of the visual text 293 * (corresponding to the physical memory address end, same as END in default behavior) 294 * B. BEGIN For Logical text: Same as BEGIN in default behavior. 295 * C. END For Visual LTR text: This will be the end (left side) of the visual text. (corresponding to 296 * the physical memory address beginning, same as BEGIN in default behavior) 297 * D. END For Logical text: Same as END in default behavior. 298 * Affects: All LamAlef BEGIN, END and AUTO options. 299 */ 300 public static final int SPACES_RELATIVE_TO_TEXT_BEGIN_END = 0x4000000; 301 302 /** Bit mask for swapping BEGIN and END for Visual LTR text 303 */ 304 public static final int SPACES_RELATIVE_TO_TEXT_MASK = 0x4000000; 305 306 /** 307 * If this option is used, shaping will use the new Unicode code point for TAIL (i.e. 0xFE73). 308 * If this option is not specified (Default), old unofficial Unicode TAIL code point is used (i.e. 0x200B) 309 * De-shaping will not use this option as it will always search for both the new Unicode code point for the 310 * TAIL (i.e. 0xFE73) or the old unofficial Unicode TAIL code point (i.e. 0x200B) and de-shape the 311 * Seen-Family letter accordingly. 312 * 313 * Shaping Mode: Only shaping. 314 * De-shaping Mode: N/A. 315 * Affects: All Seen options 316 */ 317 public static final int SHAPE_TAIL_NEW_UNICODE = 0x8000000; 318 319 /** Bit mask for new Unicode Tail option 320 */ 321 public static final int SHAPE_TAIL_TYPE_MASK = 0x8000000; 322 323 /** 324 * Memory option: allow the result to have a different length than the source. 325 */ 326 public static final int LENGTH_GROW_SHRINK = 0; 327 328 /** 329 * Memory option: allow the result to have a different length than the source. 330 * Affects: LamAlef options 331 * This option is an alias to LENGTH_GROW_SHRINK 332 */ 333 public static final int LAMALEF_RESIZE = 0; 334 335 /** 336 * Memory option: the result must have the same length as the source. 337 * If more room is necessary, then try to consume spaces next to modified characters. 338 */ 339 public static final int LENGTH_FIXED_SPACES_NEAR = 1; 340 341 /** 342 * Memory option: the result must have the same length as the source. 343 * If more room is necessary, then try to consume spaces next to modified characters. 344 * Affects: LamAlef options 345 * This option is an alias to LENGTH_FIXED_SPACES_NEAR 346 */ 347 public static final int LAMALEF_NEAR = 1 ; 348 349 /** 350 * Memory option: the result must have the same length as the source. 351 * If more room is necessary, then try to consume spaces at the end of the text. 352 */ 353 public static final int LENGTH_FIXED_SPACES_AT_END = 2; 354 355 356 /** 357 * Memory option: the result must have the same length as the source. 358 * If more room is necessary, then try to consume spaces at the end of the text. 359 * Affects: LamAlef options 360 * This option is an alias to LENGTH_FIXED_SPACES_AT_END 361 */ 362 public static final int LAMALEF_END = 2; 363 364 /** 365 * Memory option: the result must have the same length as the source. 366 * If more room is necessary, then try to consume spaces at the beginning of the text. 367 */ 368 public static final int LENGTH_FIXED_SPACES_AT_BEGINNING = 3; 369 370 /** 371 * Memory option: the result must have the same length as the source. 372 * If more room is necessary, then try to consume spaces at the beginning of the text. 373 * Affects: LamAlef options 374 * This option is an alias to LENGTH_FIXED_SPACES_AT_BEGINNING 375 */ 376 public static final int LAMALEF_BEGIN = 3; 377 378 /** 379 * Memory option: the result must have the same length as the source. 380 * Shaping Mode: For each LAMALEF character found, expand LAMALEF using space at end. 381 * If there is no space at end, use spaces at beginning of the buffer. If there 382 * is no space at beginning of the buffer, use spaces at the near (i.e. the space 383 * after the LAMALEF character). 384 * 385 * Deshaping Mode: Perform the same function as the flag equals LAMALEF_END. 386 * Affects: LamAlef options 387 */ 388 public static final int LAMALEF_AUTO = 0x10000; 389 390 /** 391 * Bit mask for memory options. 392 */ 393 public static final int LENGTH_MASK = 0x10003; 394 395 /** Bit mask for LamAlef memory options. 396 */ 397 398 public static final int LAMALEF_MASK = 0x10003; 399 400 /** 401 * Direction indicator: the source is in logical (keyboard) order. 402 */ 403 public static final int TEXT_DIRECTION_LOGICAL = 0; 404 405 /** 406 * Direction indicator:the source is in visual RTL order, 407 * the rightmost displayed character stored first. 408 * This option is an alias to U_SHAPE_TEXT_DIRECTION_LOGICAL 409 */ 410 public static final int TEXT_DIRECTION_VISUAL_RTL = 0; 411 412 /** 413 * Direction indicator: the source is in visual (display) order, that is, 414 * the leftmost displayed character is stored first. 415 */ 416 public static final int TEXT_DIRECTION_VISUAL_LTR = 4; 417 418 /** 419 * Bit mask for direction indicators. 420 */ 421 public static final int TEXT_DIRECTION_MASK = 4; 422 423 424 /** 425 * Letter shaping option: do not perform letter shaping. 426 */ 427 public static final int LETTERS_NOOP = 0; 428 429 /** 430 * Letter shaping option: replace normative letter characters in the U+0600 (Arabic) block, 431 * by shaped ones in the U+FE70 (Presentation Forms B) block. Performs Lam-Alef ligature 432 * substitution. 433 */ 434 public static final int LETTERS_SHAPE = 8; 435 436 /** 437 * Letter shaping option: replace shaped letter characters in the U+FE70 (Presentation Forms B) block 438 * by normative ones in the U+0600 (Arabic) block. Converts Lam-Alef ligatures to pairs of Lam and 439 * Alef characters, consuming spaces if required. 440 */ 441 public static final int LETTERS_UNSHAPE = 0x10; 442 443 /** 444 * Letter shaping option: replace normative letter characters in the U+0600 (Arabic) block, 445 * except for the TASHKEEL characters at U+064B...U+0652, by shaped ones in the U+Fe70 446 * (Presentation Forms B) block. The TASHKEEL characters will always be converted to 447 * the isolated forms rather than to their correct shape. 448 */ 449 public static final int LETTERS_SHAPE_TASHKEEL_ISOLATED = 0x18; 450 451 /** 452 * Bit mask for letter shaping options. 453 */ 454 public static final int LETTERS_MASK = 0x18; 455 456 457 /** 458 * Digit shaping option: do not perform digit shaping. 459 */ 460 public static final int DIGITS_NOOP = 0; 461 462 /** 463 * Digit shaping option: Replace European digits (U+0030...U+0039) by Arabic-Indic digits. 464 */ 465 public static final int DIGITS_EN2AN = 0x20; 466 467 /** 468 * Digit shaping option: Replace Arabic-Indic digits by European digits (U+0030...U+0039). 469 */ 470 public static final int DIGITS_AN2EN = 0x40; 471 472 /** 473 * Digit shaping option: 474 * Replace European digits (U+0030...U+0039) by Arabic-Indic digits 475 * if the most recent strongly directional character 476 * is an Arabic letter (its Bidi direction value is RIGHT_TO_LEFT_ARABIC). 477 * The initial state at the start of the text is assumed to be not an Arabic, 478 * letter, so European digits at the start of the text will not change. 479 * Compare to DIGITS_ALEN2AN_INIT_AL. 480 */ 481 public static final int DIGITS_EN2AN_INIT_LR = 0x60; 482 483 /** 484 * Digit shaping option: 485 * Replace European digits (U+0030...U+0039) by Arabic-Indic digits 486 * if the most recent strongly directional character 487 * is an Arabic letter (its Bidi direction value is RIGHT_TO_LEFT_ARABIC). 488 * The initial state at the start of the text is assumed to be an Arabic, 489 * letter, so European digits at the start of the text will change. 490 * Compare to DIGITS_ALEN2AN_INT_LR. 491 */ 492 public static final int DIGITS_EN2AN_INIT_AL = 0x80; 493 494 /** Not a valid option value. */ 495 //private static final int DIGITS_RESERVED = 0xa0; 496 497 /** 498 * Bit mask for digit shaping options. 499 */ 500 public static final int DIGITS_MASK = 0xe0; 501 502 /** 503 * Digit type option: Use Arabic-Indic digits (U+0660...U+0669). 504 */ 505 public static final int DIGIT_TYPE_AN = 0; 506 507 /** 508 * Digit type option: Use Eastern (Extended) Arabic-Indic digits (U+06f0...U+06f9). 509 */ 510 public static final int DIGIT_TYPE_AN_EXTENDED = 0x100; 511 512 /** 513 * Bit mask for digit type options. 514 */ 515 public static final int DIGIT_TYPE_MASK = 0x0100; // 0x3f00? 516 517 /** 518 * some constants 519 */ 520 private static final char HAMZAFE_CHAR = '\ufe80'; 521 private static final char HAMZA06_CHAR = '\u0621'; 522 private static final char YEH_HAMZA_CHAR = '\u0626'; 523 private static final char YEH_HAMZAFE_CHAR = '\uFE89'; 524 private static final char LAMALEF_SPACE_SUB = '\uffff'; 525 private static final char TASHKEEL_SPACE_SUB = '\ufffe'; 526 private static final char LAM_CHAR = '\u0644'; 527 private static final char SPACE_CHAR = '\u0020'; 528 private static final char SHADDA_CHAR = '\uFE7C'; 529 private static final char SHADDA06_CHAR = '\u0651'; 530 private static final char TATWEEL_CHAR = '\u0640'; 531 private static final char SHADDA_TATWEEL_CHAR = '\uFE7D'; 532 private static final char NEW_TAIL_CHAR = '\uFE73'; 533 private static final char OLD_TAIL_CHAR = '\u200B'; 534 private static final int SHAPE_MODE = 0; 535 private static final int DESHAPE_MODE = 1; 536 537 /** 538 */ 539 @Override 540 public boolean equals(Object rhs) { 541 return rhs != null && 542 rhs.getClass() == ArabicShaping.class && 543 options == ((ArabicShaping)rhs).options; 544 } 545 546 /** 547 */ 548 ///CLOVER:OFF 549 @Override 550 public int hashCode() { 551 return options; 552 } 553 554 /** 555 */ 556 @Override 557 public String toString() { 558 StringBuilder buf = new StringBuilder(super.toString()); 559 buf.append('['); 560 561 switch (options & LAMALEF_MASK) { 562 case LAMALEF_RESIZE: buf.append("LamAlef resize"); break; 563 case LAMALEF_NEAR: buf.append("LamAlef spaces at near"); break; 564 case LAMALEF_BEGIN: buf.append("LamAlef spaces at begin"); break; 565 case LAMALEF_END: buf.append("LamAlef spaces at end"); break; 566 case LAMALEF_AUTO: buf.append("lamAlef auto"); break; 567 } 568 switch (options & TEXT_DIRECTION_MASK) { 569 case TEXT_DIRECTION_LOGICAL: buf.append(", logical"); break; 570 case TEXT_DIRECTION_VISUAL_LTR: buf.append(", visual"); break; 571 } 572 switch (options & LETTERS_MASK) { 573 case LETTERS_NOOP: buf.append(", no letter shaping"); break; 574 case LETTERS_SHAPE: buf.append(", shape letters"); break; 575 case LETTERS_SHAPE_TASHKEEL_ISOLATED: buf.append(", shape letters tashkeel isolated"); break; 576 case LETTERS_UNSHAPE: buf.append(", unshape letters"); break; 577 } 578 switch (options & SEEN_MASK) { 579 case SEEN_TWOCELL_NEAR: buf.append(", Seen at near"); break; 580 } 581 switch (options & YEHHAMZA_MASK) { 582 case YEHHAMZA_TWOCELL_NEAR: buf.append(", Yeh Hamza at near"); break; 583 } 584 switch (options & TASHKEEL_MASK) { 585 case TASHKEEL_BEGIN: buf.append(", Tashkeel at begin"); break; 586 case TASHKEEL_END: buf.append(", Tashkeel at end"); break; 587 case TASHKEEL_REPLACE_BY_TATWEEL: buf.append(", Tashkeel replace with tatweel"); break; 588 case TASHKEEL_RESIZE: buf.append(", Tashkeel resize"); break; 589 } 590 591 switch (options & DIGITS_MASK) { 592 case DIGITS_NOOP: buf.append(", no digit shaping"); break; 593 case DIGITS_EN2AN: buf.append(", shape digits to AN"); break; 594 case DIGITS_AN2EN: buf.append(", shape digits to EN"); break; 595 case DIGITS_EN2AN_INIT_LR: buf.append(", shape digits to AN contextually: default EN"); break; 596 case DIGITS_EN2AN_INIT_AL: buf.append(", shape digits to AN contextually: default AL"); break; 597 } 598 switch (options & DIGIT_TYPE_MASK) { 599 case DIGIT_TYPE_AN: buf.append(", standard Arabic-Indic digits"); break; 600 case DIGIT_TYPE_AN_EXTENDED: buf.append(", extended Arabic-Indic digits"); break; 601 } 602 buf.append("]"); 603 604 return buf.toString(); 605 } 606 ///CLOVER:ON 607 608 // 609 // ported api 610 // 611 612 private static final int IRRELEVANT = 4; 613 private static final int LAMTYPE = 16; 614 private static final int ALEFTYPE = 32; 615 616 private static final int LINKR = 1; 617 private static final int LINKL = 2; 618 private static final int LINK_MASK = 3; 619 620 private static final int irrelevantPos[] = { 621 0x0, 0x2, 0x4, 0x6, 0x8, 0xA, 0xC, 0xE 622 }; 623 624/* 625 private static final char convertLamAlef[] = { 626 '\u0622', // FEF5 627 '\u0622', // FEF6 628 '\u0623', // FEF7 629 '\u0623', // FEF8 630 '\u0625', // FEF9 631 '\u0625', // FEFA 632 '\u0627', // FEFB 633 '\u0627' // FEFC 634 }; 635*/ 636 637 private static final int tailFamilyIsolatedFinal[] = { 638 /* FEB1 */ 1, 639 /* FEB2 */ 1, 640 /* FEB3 */ 0, 641 /* FEB4 */ 0, 642 /* FEB5 */ 1, 643 /* FEB6 */ 1, 644 /* FEB7 */ 0, 645 /* FEB8 */ 0, 646 /* FEB9 */ 1, 647 /* FEBA */ 1, 648 /* FEBB */ 0, 649 /* FEBC */ 0, 650 /* FEBD */ 1, 651 /* FEBE */ 1 652 }; 653 654 private static final int tashkeelMedial[] = { 655 /* FE70 */ 0, 656 /* FE71 */ 1, 657 /* FE72 */ 0, 658 /* FE73 */ 0, 659 /* FE74 */ 0, 660 /* FE75 */ 0, 661 /* FE76 */ 0, 662 /* FE77 */ 1, 663 /* FE78 */ 0, 664 /* FE79 */ 1, 665 /* FE7A */ 0, 666 /* FE7B */ 1, 667 /* FE7C */ 0, 668 /* FE7D */ 1, 669 /* FE7E */ 0, 670 /* FE7F */ 1 671 }; 672 673 private static final char yehHamzaToYeh[] = 674 { 675 /* isolated*/ 0xFEEF, 676 /* final */ 0xFEF0 677 }; 678 679 private static final char convertNormalizedLamAlef[] = { 680 '\u0622', // 065C 681 '\u0623', // 065D 682 '\u0625', // 065E 683 '\u0627', // 065F 684 }; 685 686 private static final int[] araLink = { 687 1 + 32 + 256 * 0x11, /*0x0622*/ 688 1 + 32 + 256 * 0x13, /*0x0623*/ 689 1 + 256 * 0x15, /*0x0624*/ 690 1 + 32 + 256 * 0x17, /*0x0625*/ 691 1 + 2 + 256 * 0x19, /*0x0626*/ 692 1 + 32 + 256 * 0x1D, /*0x0627*/ 693 1 + 2 + 256 * 0x1F, /*0x0628*/ 694 1 + 256 * 0x23, /*0x0629*/ 695 1 + 2 + 256 * 0x25, /*0x062A*/ 696 1 + 2 + 256 * 0x29, /*0x062B*/ 697 1 + 2 + 256 * 0x2D, /*0x062C*/ 698 1 + 2 + 256 * 0x31, /*0x062D*/ 699 1 + 2 + 256 * 0x35, /*0x062E*/ 700 1 + 256 * 0x39, /*0x062F*/ 701 1 + 256 * 0x3B, /*0x0630*/ 702 1 + 256 * 0x3D, /*0x0631*/ 703 1 + 256 * 0x3F, /*0x0632*/ 704 1 + 2 + 256 * 0x41, /*0x0633*/ 705 1 + 2 + 256 * 0x45, /*0x0634*/ 706 1 + 2 + 256 * 0x49, /*0x0635*/ 707 1 + 2 + 256 * 0x4D, /*0x0636*/ 708 1 + 2 + 256 * 0x51, /*0x0637*/ 709 1 + 2 + 256 * 0x55, /*0x0638*/ 710 1 + 2 + 256 * 0x59, /*0x0639*/ 711 1 + 2 + 256 * 0x5D, /*0x063A*/ 712 0, 0, 0, 0, 0, /*0x063B-0x063F*/ 713 1 + 2, /*0x0640*/ 714 1 + 2 + 256 * 0x61, /*0x0641*/ 715 1 + 2 + 256 * 0x65, /*0x0642*/ 716 1 + 2 + 256 * 0x69, /*0x0643*/ 717 1 + 2 + 16 + 256 * 0x6D, /*0x0644*/ 718 1 + 2 + 256 * 0x71, /*0x0645*/ 719 1 + 2 + 256 * 0x75, /*0x0646*/ 720 1 + 2 + 256 * 0x79, /*0x0647*/ 721 1 + 256 * 0x7D, /*0x0648*/ 722 1 + 256 * 0x7F, /*0x0649*/ 723 1 + 2 + 256 * 0x81, /*0x064A*/ 724 4, 4, 4, 4, /*0x064B-0x064E*/ 725 4, 4, 4, 4, /*0x064F-0x0652*/ 726 4, 4, 4, 0, 0, /*0x0653-0x0657*/ 727 0, 0, 0, 0, /*0x0658-0x065B*/ 728 1 + 256 * 0x85, /*0x065C*/ 729 1 + 256 * 0x87, /*0x065D*/ 730 1 + 256 * 0x89, /*0x065E*/ 731 1 + 256 * 0x8B, /*0x065F*/ 732 0, 0, 0, 0, 0, /*0x0660-0x0664*/ 733 0, 0, 0, 0, 0, /*0x0665-0x0669*/ 734 0, 0, 0, 0, 0, 0, /*0x066A-0x066F*/ 735 4, /*0x0670*/ 736 0, /*0x0671*/ 737 1 + 32, /*0x0672*/ 738 1 + 32, /*0x0673*/ 739 0, /*0x0674*/ 740 1 + 32, /*0x0675*/ 741 1, 1, /*0x0676-0x0677*/ 742 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x0678-0x067D*/ 743 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x067E-0x0683*/ 744 1+2, 1+2, 1+2, 1+2, /*0x0684-0x0687*/ 745 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /*0x0688-0x0691*/ 746 1, 1, 1, 1, 1, 1, 1, 1, /*0x0692-0x0699*/ 747 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x069A-0x06A3*/ 748 1+2, 1+2, 1+2, 1+2, /*0x069A-0x06A3*/ 749 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06A4-0x06AD*/ 750 1+2, 1+2, 1+2, 1+2, /*0x06A4-0x06AD*/ 751 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06AE-0x06B7*/ 752 1+2, 1+2, 1+2, 1+2, /*0x06AE-0x06B7*/ 753 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06B8-0x06BF*/ 754 1+2, 1+2, /*0x06B8-0x06BF*/ 755 1, /*0x06C0*/ 756 1+2, /*0x06C1*/ 757 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /*0x06C2-0x06CB*/ 758 1+2, /*0x06CC*/ 759 1, /*0x06CD*/ 760 1+2, 1+2, 1+2, 1+2, /*0x06CE-0x06D1*/ 761 1, 1 /*0x06D2-0x06D3*/ 762 }; 763 764 private static final int[] presLink = { 765 1 + 2, /*0xFE70*/ 766 1 + 2, /*0xFE71*/ 767 1 + 2, 0, 1+ 2, 0, 1+ 2, /*0xFE72-0xFE76*/ 768 1 + 2, /*0xFE77*/ 769 1+ 2, 1 + 2, 1+2, 1 + 2, /*0xFE78-0xFE81*/ 770 1+ 2, 1 + 2, 1+2, 1 + 2, /*0xFE82-0xFE85*/ 771 0, 0 + 32, 1 + 32, 0 + 32, /*0xFE86-0xFE89*/ 772 1 + 32, 0, 1, 0 + 32, /*0xFE8A-0xFE8D*/ 773 1 + 32, 0, 2, 1 + 2, /*0xFE8E-0xFE91*/ 774 1, 0 + 32, 1 + 32, 0, /*0xFE92-0xFE95*/ 775 2, 1 + 2, 1, 0, /*0xFE96-0xFE99*/ 776 1, 0, 2, 1 + 2, /*0xFE9A-0xFE9D*/ 777 1, 0, 2, 1 + 2, /*0xFE9E-0xFEA1*/ 778 1, 0, 2, 1 + 2, /*0xFEA2-0xFEA5*/ 779 1, 0, 2, 1 + 2, /*0xFEA6-0xFEA9*/ 780 1, 0, 2, 1 + 2, /*0xFEAA-0xFEAD*/ 781 1, 0, 1, 0, /*0xFEAE-0xFEB1*/ 782 1, 0, 1, 0, /*0xFEB2-0xFEB5*/ 783 1, 0, 2, 1+2, /*0xFEB6-0xFEB9*/ 784 1, 0, 2, 1+2, /*0xFEBA-0xFEBD*/ 785 1, 0, 2, 1+2, /*0xFEBE-0xFEC1*/ 786 1, 0, 2, 1+2, /*0xFEC2-0xFEC5*/ 787 1, 0, 2, 1+2, /*0xFEC6-0xFEC9*/ 788 1, 0, 2, 1+2, /*0xFECA-0xFECD*/ 789 1, 0, 2, 1+2, /*0xFECE-0xFED1*/ 790 1, 0, 2, 1+2, /*0xFED2-0xFED5*/ 791 1, 0, 2, 1+2, /*0xFED6-0xFED9*/ 792 1, 0, 2, 1+2, /*0xFEDA-0xFEDD*/ 793 1, 0, 2, 1+2, /*0xFEDE-0xFEE1*/ 794 1, 0 + 16, 2 + 16, 1 + 2 +16, /*0xFEE2-0xFEE5*/ 795 1 + 16, 0, 2, 1+2, /*0xFEE6-0xFEE9*/ 796 1, 0, 2, 1+2, /*0xFEEA-0xFEED*/ 797 1, 0, 2, 1+2, /*0xFEEE-0xFEF1*/ 798 1, 0, 1, 0, /*0xFEF2-0xFEF5*/ 799 1, 0, 2, 1+2, /*0xFEF6-0xFEF9*/ 800 1, 0, 1, 0, /*0xFEFA-0xFEFD*/ 801 1, 0, 1, 0, 802 1 803 }; 804 805 private static int[] convertFEto06 = { 806 /***********0******1******2******3******4******5******6******7******8******9******A******B******C******D******E******F***/ 807 /*FE7*/ 0x64B, 0x64B, 0x64C, 0x64C, 0x64D, 0x64D, 0x64E, 0x64E, 0x64F, 0x64F, 0x650, 0x650, 0x651, 0x651, 0x652, 0x652, 808 /*FE8*/ 0x621, 0x622, 0x622, 0x623, 0x623, 0x624, 0x624, 0x625, 0x625, 0x626, 0x626, 0x626, 0x626, 0x627, 0x627, 0x628, 809 /*FE9*/ 0x628, 0x628, 0x628, 0x629, 0x629, 0x62A, 0x62A, 0x62A, 0x62A, 0x62B, 0x62B, 0x62B, 0x62B, 0x62C, 0x62C, 0x62C, 810 /*FEA*/ 0x62C, 0x62D, 0x62D, 0x62D, 0x62D, 0x62E, 0x62E, 0x62E, 0x62E, 0x62F, 0x62F, 0x630, 0x630, 0x631, 0x631, 0x632, 811 /*FEB*/ 0x632, 0x633, 0x633, 0x633, 0x633, 0x634, 0x634, 0x634, 0x634, 0x635, 0x635, 0x635, 0x635, 0x636, 0x636, 0x636, 812 /*FEC*/ 0x636, 0x637, 0x637, 0x637, 0x637, 0x638, 0x638, 0x638, 0x638, 0x639, 0x639, 0x639, 0x639, 0x63A, 0x63A, 0x63A, 813 /*FED*/ 0x63A, 0x641, 0x641, 0x641, 0x641, 0x642, 0x642, 0x642, 0x642, 0x643, 0x643, 0x643, 0x643, 0x644, 0x644, 0x644, 814 /*FEE*/ 0x644, 0x645, 0x645, 0x645, 0x645, 0x646, 0x646, 0x646, 0x646, 0x647, 0x647, 0x647, 0x647, 0x648, 0x648, 0x649, 815 /*FEF*/ 0x649, 0x64A, 0x64A, 0x64A, 0x64A, 0x65C, 0x65C, 0x65D, 0x65D, 0x65E, 0x65E, 0x65F, 0x65F 816 }; 817 818 private static final int shapeTable[][][] = { 819 { {0,0,0,0}, {0,0,0,0}, {0,1,0,3}, {0,1,0,1} }, 820 { {0,0,2,2}, {0,0,1,2}, {0,1,1,2}, {0,1,1,3} }, 821 { {0,0,0,0}, {0,0,0,0}, {0,1,0,3}, {0,1,0,3} }, 822 { {0,0,1,2}, {0,0,1,2}, {0,1,1,2}, {0,1,1,3} } 823 }; 824 825 /* 826 * This function shapes European digits to Arabic-Indic digits 827 * in-place, writing over the input characters. Data is in visual 828 * order. 829 */ 830 private void shapeToArabicDigitsWithContext(char[] dest, 831 int start, 832 int length, 833 char digitBase, 834 boolean lastStrongWasAL) { 835 UBiDiProps bdp=UBiDiProps.INSTANCE; 836 digitBase -= '0'; // move common adjustment out of loop 837 838 for(int i = start + length; --i >= start;) { 839 char ch = dest[i]; 840 switch (bdp.getClass(ch)) { 841 case UCharacterDirection.LEFT_TO_RIGHT: 842 case UCharacterDirection.RIGHT_TO_LEFT: 843 lastStrongWasAL = false; 844 break; 845 case UCharacterDirection.RIGHT_TO_LEFT_ARABIC: 846 lastStrongWasAL = true; 847 break; 848 case UCharacterDirection.EUROPEAN_NUMBER: 849 if (lastStrongWasAL && ch <= '\u0039') { 850 dest[i] = (char)(ch + digitBase); 851 } 852 break; 853 default: 854 break; 855 } 856 } 857 } 858 859 /* 860 * Name : invertBuffer 861 * Function: This function inverts the buffer, it's used 862 * in case the user specifies the buffer to be 863 * TEXT_DIRECTION_LOGICAL 864 */ 865 private static void invertBuffer(char[] buffer, 866 int start, 867 int length) { 868 869 for(int i = start, j = start + length - 1; i < j; i++, --j) { 870 char temp = buffer[i]; 871 buffer[i] = buffer[j]; 872 buffer[j] = temp; 873 } 874 } 875 876 /* 877 * Name : changeLamAlef 878 * Function: Converts the Alef characters into an equivalent 879 * LamAlef location in the 0x06xx Range, this is an 880 * intermediate stage in the operation of the program 881 * later it'll be converted into the 0xFExx LamAlefs 882 * in the shaping function. 883 */ 884 private static char changeLamAlef(char ch) { 885 switch(ch) { 886 case '\u0622': return '\u065C'; 887 case '\u0623': return '\u065D'; 888 case '\u0625': return '\u065E'; 889 case '\u0627': return '\u065F'; 890 default: return '\u0000'; // not a lamalef 891 } 892 } 893 894 /* 895 * Name : specialChar 896 * Function: Special Arabic characters need special handling in the shapeUnicode 897 * function, this function returns 1 or 2 for these special characters 898 */ 899 private static int specialChar(char ch) { 900 if ((ch > '\u0621' && ch < '\u0626') || 901 (ch == '\u0627') || 902 (ch > '\u062E' && ch < '\u0633') || 903 (ch > '\u0647' && ch < '\u064A') || 904 (ch == '\u0629')) { 905 return 1; 906 } else if (ch >= '\u064B' && ch<= '\u0652') { 907 return 2; 908 } else if (ch >= 0x0653 && ch <= 0x0655 || 909 ch == 0x0670 || 910 ch >= 0xFE70 && ch <= 0xFE7F) { 911 return 3; 912 } else { 913 return 0; 914 } 915 } 916 917 /* 918 * Name : getLink 919 * Function: Resolves the link between the characters as 920 * Arabic characters have four forms : 921 * Isolated, Initial, Middle and Final Form 922 */ 923 private static int getLink(char ch) { 924 if (ch >= '\u0622' && ch <= '\u06D3') { 925 return araLink[ch - '\u0622']; 926 } else if (ch == '\u200D') { 927 return 3; 928 } else if (ch >= '\u206D' && ch <= '\u206F') { 929 return 4; 930 } else if (ch >= '\uFE70' && ch <= '\uFEFC') { 931 return presLink[ch - '\uFE70']; 932 } else { 933 return 0; 934 } 935 } 936 937 /* 938 * Name : countSpaces 939 * Function: Counts the number of spaces 940 * at each end of the logical buffer 941 */ 942 private static int countSpacesLeft(char[] dest, 943 int start, 944 int count) { 945 for (int i = start, e = start + count; i < e; ++i) { 946 if (dest[i] != SPACE_CHAR) { 947 return i - start; 948 } 949 } 950 return count; 951 } 952 953 private static int countSpacesRight(char[] dest, 954 int start, 955 int count) { 956 957 for (int i = start + count; --i >= start;) { 958 if (dest[i] != SPACE_CHAR) { 959 return start + count - 1 - i; 960 } 961 } 962 return count; 963 } 964 965 /* 966 * Name : isTashkeelChar 967 * Function: Returns true for Tashkeel characters else return false 968 */ 969 private static boolean isTashkeelChar(char ch) { 970 return ( ch >='\u064B' && ch <= '\u0652' ); 971 } 972 973 /* 974 *Name : isSeenTailFamilyChar 975 *Function : returns 1 if the character is a seen family isolated character 976 * in the FE range otherwise returns 0 977 */ 978 979 private static int isSeenTailFamilyChar(char ch) { 980 if (ch >= 0xfeb1 && ch < 0xfebf){ 981 return tailFamilyIsolatedFinal [ch - 0xFEB1]; 982 } else { 983 return 0; 984 } 985 } 986 987 /* Name : isSeenFamilyChar 988 * Function : returns 1 if the character is a seen family character in the Unicode 989 * 06 range otherwise returns 0 990 */ 991 992 private static int isSeenFamilyChar(char ch){ 993 if (ch >= 0x633 && ch <= 0x636){ 994 return 1; 995 }else { 996 return 0; 997 } 998 } 999 1000 /* 1001 *Name : isTailChar 1002 *Function : returns true if the character matches one of the tail characters 1003 * (0xfe73 or 0x200b) otherwise returns false 1004 */ 1005 1006 private static boolean isTailChar(char ch) { 1007 if(ch == OLD_TAIL_CHAR || ch == NEW_TAIL_CHAR){ 1008 return true; 1009 }else{ 1010 return false; 1011 } 1012 } 1013 1014 /* 1015 *Name : isAlefMaksouraChar 1016 *Function : returns true if the character is a Alef Maksoura Final or isolated 1017 * otherwise returns false 1018 */ 1019 private static boolean isAlefMaksouraChar(char ch) { 1020 return ( (ch == 0xFEEF) || ( ch == 0xFEF0) || (ch == 0x0649)); 1021 } 1022 1023 /* 1024 * Name : isYehHamzaChar 1025 * Function : returns true if the character is a yehHamza isolated or yehhamza 1026 * final is found otherwise returns false 1027 */ 1028 private static boolean isYehHamzaChar(char ch) { 1029 if((ch==0xFE89)||(ch==0xFE8A)){ 1030 return true; 1031 }else{ 1032 return false; 1033 } 1034 } 1035 1036 /* 1037 *Name : isTashkeelCharFE 1038 *Function : Returns true for Tashkeel characters in FE range else return false 1039 */ 1040 1041 private static boolean isTashkeelCharFE(char ch) { 1042 return ( ch!=0xFE75 &&(ch>=0xFE70 && ch<= 0xFE7F) ); 1043 } 1044 1045 /* 1046 * Name: isTashkeelOnTatweelChar 1047 * Function: Checks if the Tashkeel Character is on Tatweel or not,if the 1048 * Tashkeel on tatweel (FE range), it returns 1 else if the 1049 * Tashkeel with shadda on tatweel (FC range)return 2 otherwise 1050 * returns 0 1051 */ 1052 private static int isTashkeelOnTatweelChar(char ch){ 1053 if (ch >= 0xfe70 && ch <= 0xfe7f && ch != NEW_TAIL_CHAR && ch != 0xFE75 && ch != SHADDA_TATWEEL_CHAR) 1054 { 1055 return tashkeelMedial [ch - 0xFE70]; 1056 } else if( (ch >= 0xfcf2 && ch <= 0xfcf4) || (ch == SHADDA_TATWEEL_CHAR)) { 1057 return 2; 1058 } else { 1059 return 0; 1060 } 1061 } 1062 1063 /* 1064 * Name: isIsolatedTashkeelChar 1065 * Function: Checks if the Tashkeel Character is in the isolated form 1066 * (i.e. Unicode FE range) returns 1 else if the Tashkeel 1067 * with shadda is in the isolated form (i.e. Unicode FC range) 1068 * returns 1 otherwise returns 0 1069 */ 1070 private static int isIsolatedTashkeelChar(char ch){ 1071 if (ch >= 0xfe70 && ch <= 0xfe7f && ch != NEW_TAIL_CHAR && ch != 0xFE75){ 1072 return (1 - tashkeelMedial [ch - 0xFE70]); 1073 } else if(ch >= 0xfc5e && ch <= 0xfc63){ 1074 return 1; 1075 } else{ 1076 return 0; 1077 } 1078 } 1079 1080 /* 1081 * Name : isAlefChar 1082 * Function: Returns 1 for Alef characters else return 0 1083 */ 1084 private static boolean isAlefChar(char ch) { 1085 return ch == '\u0622' || ch == '\u0623' || ch == '\u0625' || ch == '\u0627'; 1086 } 1087 1088 /* 1089 * Name : isLamAlefChar 1090 * Function: Returns true for LamAlef characters else return false 1091 */ 1092 private static boolean isLamAlefChar(char ch) { 1093 return ch >= '\uFEF5' && ch <= '\uFEFC'; 1094 } 1095 1096 private static boolean isNormalizedLamAlefChar(char ch) { 1097 return ch >= '\u065C' && ch <= '\u065F'; 1098 } 1099 1100 /* 1101 * Name : calculateSize 1102 * Function: This function calculates the destSize to be used in preflighting 1103 * when the destSize is equal to 0 1104 */ 1105 private int calculateSize(char[] source, 1106 int sourceStart, 1107 int sourceLength) { 1108 1109 int destSize = sourceLength; 1110 1111 switch (options & LETTERS_MASK) { 1112 case LETTERS_SHAPE: 1113 case LETTERS_SHAPE_TASHKEEL_ISOLATED: 1114 if (isLogical) { 1115 for (int i = sourceStart, e = sourceStart + sourceLength - 1; i < e; ++i) { 1116 if ((source[i] == LAM_CHAR && isAlefChar(source[i+1])) || isTashkeelCharFE(source[i])){ 1117 --destSize; 1118 } 1119 } 1120 } else { // visual 1121 for(int i = sourceStart + 1, e = sourceStart + sourceLength; i < e; ++i) { 1122 if ((source[i] == LAM_CHAR && isAlefChar(source[i-1])) || isTashkeelCharFE(source[i])) { 1123 --destSize; 1124 } 1125 } 1126 } 1127 break; 1128 1129 case LETTERS_UNSHAPE: 1130 for(int i = sourceStart, e = sourceStart + sourceLength; i < e; ++i) { 1131 if (isLamAlefChar(source[i])) { 1132 destSize++; 1133 } 1134 } 1135 break; 1136 1137 default: 1138 break; 1139 } 1140 1141 return destSize; 1142 } 1143 1144 1145 /* 1146 * Name : countSpaceSub 1147 * Function: Counts number of times the subChar appears in the array 1148 */ 1149 private static int countSpaceSub(char [] dest,int length, char subChar){ 1150 int i = 0; 1151 int count = 0; 1152 while (i < length) { 1153 if (dest[i] == subChar) { 1154 count++; 1155 } 1156 i++; 1157 } 1158 return count; 1159 } 1160 1161 /* 1162 * Name : shiftArray 1163 * Function: Shifts characters to replace space sub characters 1164 */ 1165 private static void shiftArray(char [] dest,int start, int e, char subChar){ 1166 int w = e; 1167 int r = e; 1168 while (--r >= start) { 1169 char ch = dest[r]; 1170 if (ch != subChar) { 1171 --w; 1172 if (w != r) { 1173 dest[w] = ch; 1174 } 1175 } 1176 } 1177 } 1178 1179 /* 1180 * Name : flipArray 1181 * Function: inverts array, so that start becomes end and vice versa 1182 */ 1183 private static int flipArray(char [] dest, int start, int e, int w){ 1184 int r; 1185 if (w > start) { 1186 // shift, assume small buffer size so don't use arraycopy 1187 r = w; 1188 w = start; 1189 while (r < e) { 1190 dest[w++] = dest[r++]; 1191 } 1192 } else { 1193 w = e; 1194 } 1195 return w; 1196 } 1197 1198 /* 1199 * Name : handleTashkeelWithTatweel 1200 * Function : Replaces Tashkeel as following: 1201 * Case 1 :if the Tashkeel on tatweel, replace it with Tatweel. 1202 * Case 2 :if the Tashkeel aggregated with Shadda on Tatweel, replace 1203 * it with Shadda on Tatweel. 1204 * Case 3: if the Tashkeel is isolated replace it with Space. 1205 * 1206 */ 1207 private static int handleTashkeelWithTatweel(char[] dest, int sourceLength) { 1208 int i; 1209 for(i = 0; i < sourceLength; i++){ 1210 if((isTashkeelOnTatweelChar(dest[i]) == 1)){ 1211 dest[i] = TATWEEL_CHAR; 1212 }else if((isTashkeelOnTatweelChar(dest[i]) == 2)){ 1213 dest[i] = SHADDA_TATWEEL_CHAR; 1214 }else if((isIsolatedTashkeelChar(dest[i])==1) && dest[i] != SHADDA_CHAR){ 1215 dest[i] = SPACE_CHAR; 1216 } 1217 } 1218 return sourceLength; 1219 } 1220 1221 /* 1222 *Name : handleGeneratedSpaces 1223 *Function : The shapeUnicode function converts Lam + Alef into LamAlef + space, 1224 * and Tashkeel to space. 1225 * handleGeneratedSpaces function puts these generated spaces 1226 * according to the options the user specifies. LamAlef and Tashkeel 1227 * spaces can be replaced at begin, at end, at near or decrease the 1228 * buffer size. 1229 * 1230 * There is also Auto option for LamAlef and tashkeel, which will put 1231 * the spaces at end of the buffer (or end of text if the user used 1232 * the option SPACES_RELATIVE_TO_TEXT_BEGIN_END). 1233 * 1234 * If the text type was visual_LTR and the option 1235 * SPACES_RELATIVE_TO_TEXT_BEGIN_END was selected the END 1236 * option will place the space at the beginning of the buffer and 1237 * BEGIN will place the space at the end of the buffer. 1238 */ 1239 private int handleGeneratedSpaces(char[] dest, 1240 int start, 1241 int length) { 1242 1243 int lenOptionsLamAlef = options & LAMALEF_MASK; 1244 int lenOptionsTashkeel = options & TASHKEEL_MASK; 1245 boolean lamAlefOn = false; 1246 boolean tashkeelOn = false; 1247 1248 if (!isLogical & !spacesRelativeToTextBeginEnd) { 1249 switch (lenOptionsLamAlef) { 1250 case LAMALEF_BEGIN: lenOptionsLamAlef = LAMALEF_END; break; 1251 case LAMALEF_END: lenOptionsLamAlef = LAMALEF_BEGIN; break; 1252 default: break; 1253 } 1254 switch (lenOptionsTashkeel){ 1255 case TASHKEEL_BEGIN: lenOptionsTashkeel = TASHKEEL_END; break; 1256 case TASHKEEL_END: lenOptionsTashkeel = TASHKEEL_BEGIN; break; 1257 default: break; 1258 } 1259 } 1260 1261 1262 if (lenOptionsLamAlef == LAMALEF_NEAR) { 1263 for (int i = start, e = i + length; i < e; ++i) { 1264 if (dest[i] == LAMALEF_SPACE_SUB) { 1265 dest[i] = SPACE_CHAR; 1266 } 1267 } 1268 1269 } else { 1270 1271 final int e = start + length; 1272 int wL = countSpaceSub(dest, length, LAMALEF_SPACE_SUB); 1273 int wT = countSpaceSub(dest, length, TASHKEEL_SPACE_SUB); 1274 1275 if (lenOptionsLamAlef == LAMALEF_END){ 1276 lamAlefOn = true; 1277 } 1278 if (lenOptionsTashkeel == TASHKEEL_END){ 1279 tashkeelOn = true; 1280 } 1281 1282 1283 if (lamAlefOn && (lenOptionsLamAlef == LAMALEF_END)) { 1284 shiftArray(dest, start, e, LAMALEF_SPACE_SUB); 1285 while (wL > start) { 1286 dest[--wL] = SPACE_CHAR; 1287 } 1288 } 1289 1290 if (tashkeelOn && (lenOptionsTashkeel == TASHKEEL_END)){ 1291 shiftArray(dest, start, e, TASHKEEL_SPACE_SUB); 1292 while (wT > start) { 1293 dest[--wT] = SPACE_CHAR; 1294 } 1295 } 1296 1297 lamAlefOn = false; 1298 tashkeelOn = false; 1299 1300 if (lenOptionsLamAlef == LAMALEF_RESIZE){ 1301 lamAlefOn = true; 1302 } 1303 if (lenOptionsTashkeel == TASHKEEL_RESIZE){ 1304 tashkeelOn = true; 1305 } 1306 1307 if (lamAlefOn && (lenOptionsLamAlef == LAMALEF_RESIZE)){ 1308 shiftArray(dest, start, e, LAMALEF_SPACE_SUB); 1309 wL = flipArray(dest,start,e, wL); 1310 length = wL - start; 1311 } 1312 if (tashkeelOn && (lenOptionsTashkeel == TASHKEEL_RESIZE)) { 1313 shiftArray(dest, start, e, TASHKEEL_SPACE_SUB); 1314 wT = flipArray(dest,start,e, wT); 1315 length = wT - start; 1316 } 1317 1318 lamAlefOn = false; 1319 tashkeelOn = false; 1320 1321 if ((lenOptionsLamAlef == LAMALEF_BEGIN) || 1322 (lenOptionsLamAlef == LAMALEF_AUTO)){ 1323 lamAlefOn = true; 1324 } 1325 if (lenOptionsTashkeel == TASHKEEL_BEGIN){ 1326 tashkeelOn = true; 1327 } 1328 1329 if (lamAlefOn && ((lenOptionsLamAlef == LAMALEF_BEGIN)|| 1330 (lenOptionsLamAlef == LAMALEF_AUTO))) { // spaces at beginning 1331 shiftArray(dest, start, e, LAMALEF_SPACE_SUB); 1332 wL = flipArray(dest,start,e, wL); 1333 while (wL < e) { 1334 dest[wL++] = SPACE_CHAR; 1335 } 1336 } 1337 if(tashkeelOn && (lenOptionsTashkeel == TASHKEEL_BEGIN)){ 1338 shiftArray(dest, start, e, TASHKEEL_SPACE_SUB); 1339 wT = flipArray(dest,start,e, wT); 1340 while (wT < e) { 1341 dest[wT++] = SPACE_CHAR; 1342 } 1343 } 1344 } 1345 1346 return length; 1347 } 1348 1349 1350 /* 1351 *Name :expandCompositCharAtBegin 1352 *Function :Expands the LamAlef character to Lam and Alef consuming the required 1353 * space from beginning of the buffer. If the text type was visual_LTR 1354 * and the option SPACES_RELATIVE_TO_TEXT_BEGIN_END was selected 1355 * the spaces will be located at end of buffer. 1356 * If there are no spaces to expand the LamAlef, an exception is thrown. 1357*/ 1358 private boolean expandCompositCharAtBegin(char[] dest,int start, int length, 1359 int lacount) { 1360 boolean spaceNotFound = false; 1361 1362 if (lacount > countSpacesRight(dest, start, length)) { 1363 spaceNotFound = true; 1364 return spaceNotFound; 1365 } 1366 for (int r = start + length - lacount, w = start + length; --r >= start;) { 1367 char ch = dest[r]; 1368 if (isNormalizedLamAlefChar(ch)) { 1369 dest[--w] = LAM_CHAR; 1370 dest[--w] = convertNormalizedLamAlef[ch - '\u065C']; 1371 } else { 1372 dest[--w] = ch; 1373 } 1374 } 1375 return spaceNotFound; 1376 1377 } 1378 1379 /* 1380 *Name : expandCompositCharAtEnd 1381 *Function : Expands the LamAlef character to Lam and Alef consuming the 1382 * required space from end of the buffer. If the text type was 1383 * Visual LTR and the option SPACES_RELATIVE_TO_TEXT_BEGIN_END 1384 * was used, the spaces will be consumed from begin of buffer. If 1385 * there are no spaces to expand the LamAlef, an exception is thrown. 1386 */ 1387 1388 private boolean expandCompositCharAtEnd(char[] dest,int start, int length, 1389 int lacount){ 1390 boolean spaceNotFound = false; 1391 1392 if (lacount > countSpacesLeft(dest, start, length)) { 1393 spaceNotFound = true; 1394 return spaceNotFound; 1395 } 1396 for (int r = start + lacount, w = start, e = start + length; r < e; ++r) { 1397 char ch = dest[r]; 1398 if (isNormalizedLamAlefChar(ch)) { 1399 dest[w++] = convertNormalizedLamAlef[ch - '\u065C']; 1400 dest[w++] = LAM_CHAR; 1401 } else { 1402 dest[w++] = ch; 1403 } 1404 } 1405 return spaceNotFound; 1406 } 1407 1408 /* 1409 *Name : expandCompositCharAtNear 1410 *Function : Expands the LamAlef character into Lam + Alef, YehHamza character 1411 * into Yeh + Hamza, SeenFamily character into SeenFamily character 1412 * + Tail, while consuming the space next to the character. 1413 */ 1414 1415 private boolean expandCompositCharAtNear(char[] dest,int start, int length, 1416 int yehHamzaOption, int seenTailOption, int lamAlefOption){ 1417 1418 boolean spaceNotFound = false; 1419 1420 1421 1422 if (isNormalizedLamAlefChar(dest[start])) { 1423 spaceNotFound = true; 1424 return spaceNotFound; 1425 } 1426 for (int i = start + length; --i >=start;) { 1427 char ch = dest[i]; 1428 if (lamAlefOption == 1 && isNormalizedLamAlefChar(ch)) { 1429 if (i>start &&dest[i-1] == SPACE_CHAR) { 1430 dest[i] = LAM_CHAR; 1431 dest[--i] = convertNormalizedLamAlef[ch - '\u065C']; 1432 } else { 1433 spaceNotFound = true; 1434 return spaceNotFound; 1435 } 1436 }else if(seenTailOption == 1 && isSeenTailFamilyChar(ch) == 1){ 1437 if(i>start &&dest[i-1] == SPACE_CHAR){ 1438 dest[i-1] = tailChar; 1439 } else{ 1440 spaceNotFound = true; 1441 return spaceNotFound; 1442 } 1443 }else if(yehHamzaOption == 1 && isYehHamzaChar(ch)){ 1444 1445 if(i>start &&dest[i-1] == SPACE_CHAR){ 1446 dest[i] = yehHamzaToYeh[ch - YEH_HAMZAFE_CHAR]; 1447 dest[i-1] = HAMZAFE_CHAR; 1448 }else{ 1449 spaceNotFound = true; 1450 return spaceNotFound; 1451 } 1452 1453 1454 } 1455 } 1456 return false; 1457 1458 } 1459 1460 /* 1461 * Name : expandCompositChar 1462 * Function: LamAlef needs special handling as the LamAlef is 1463 * one character while expanding it will give two 1464 * characters Lam + Alef, so we need to expand the LamAlef 1465 * in near or far spaces according to the options the user 1466 * specifies or increase the buffer size. 1467 * Dest has enough room for the expansion if we are growing. 1468 * lamalef are normalized to the 'special characters' 1469 */ 1470 private int expandCompositChar(char[] dest, 1471 int start, 1472 int length, 1473 int lacount, 1474 int shapingMode) throws ArabicShapingException { 1475 1476 int lenOptionsLamAlef = options & LAMALEF_MASK; 1477 int lenOptionsSeen = options & SEEN_MASK; 1478 int lenOptionsYehHamza = options & YEHHAMZA_MASK; 1479 boolean spaceNotFound = false; 1480 1481 if (!isLogical && !spacesRelativeToTextBeginEnd) { 1482 switch (lenOptionsLamAlef) { 1483 case LAMALEF_BEGIN: lenOptionsLamAlef = LAMALEF_END; break; 1484 case LAMALEF_END: lenOptionsLamAlef = LAMALEF_BEGIN; break; 1485 default: break; 1486 } 1487 } 1488 1489 if(shapingMode == 1){ 1490 if(lenOptionsLamAlef == LAMALEF_AUTO){ 1491 if(isLogical){ 1492 spaceNotFound = expandCompositCharAtEnd(dest, start, length, lacount); 1493 if(spaceNotFound){ 1494 spaceNotFound = expandCompositCharAtBegin(dest, start, length, lacount); 1495 } 1496 if(spaceNotFound){ 1497 spaceNotFound = expandCompositCharAtNear(dest, start, length,0,0,1); 1498 } 1499 if(spaceNotFound){ 1500 throw new ArabicShapingException("No spacefor lamalef"); 1501 } 1502 }else{ 1503 spaceNotFound = expandCompositCharAtBegin(dest, start, length, lacount); 1504 if(spaceNotFound){ 1505 spaceNotFound = expandCompositCharAtEnd(dest, start, length, lacount); 1506 } 1507 if(spaceNotFound){ 1508 spaceNotFound = expandCompositCharAtNear(dest, start, length,0,0,1); 1509 } 1510 if(spaceNotFound){ 1511 throw new ArabicShapingException("No spacefor lamalef"); 1512 } 1513 } 1514 }else if(lenOptionsLamAlef == LAMALEF_END){ 1515 spaceNotFound = expandCompositCharAtEnd(dest, start, length, lacount); 1516 if(spaceNotFound){ 1517 throw new ArabicShapingException("No spacefor lamalef"); 1518 } 1519 }else if(lenOptionsLamAlef == LAMALEF_BEGIN){ 1520 spaceNotFound = expandCompositCharAtBegin(dest, start, length, lacount); 1521 if(spaceNotFound){ 1522 throw new ArabicShapingException("No spacefor lamalef"); 1523 } 1524 }else if(lenOptionsLamAlef == LAMALEF_NEAR){ 1525 spaceNotFound = expandCompositCharAtNear(dest, start, length,0,0,1); 1526 if(spaceNotFound){ 1527 throw new ArabicShapingException("No spacefor lamalef"); 1528 } 1529 }else if(lenOptionsLamAlef == LAMALEF_RESIZE){ 1530 for (int r = start + length, w = r + lacount; --r >= start;) { 1531 char ch = dest[r]; 1532 if (isNormalizedLamAlefChar(ch)) { 1533 dest[--w] = '\u0644'; 1534 dest[--w] = convertNormalizedLamAlef[ch - '\u065C']; 1535 } else { 1536 dest[--w] = ch; 1537 } 1538 } 1539 length += lacount; 1540 } 1541 }else{ 1542 if(lenOptionsSeen == SEEN_TWOCELL_NEAR){ 1543 spaceNotFound = expandCompositCharAtNear(dest, start, length,0,1,0); 1544 if(spaceNotFound){ 1545 throw new ArabicShapingException("No space for Seen tail expansion"); 1546 } 1547 } 1548 if(lenOptionsYehHamza == YEHHAMZA_TWOCELL_NEAR){ 1549 spaceNotFound = expandCompositCharAtNear(dest, start, length,1,0,0); 1550 if(spaceNotFound){ 1551 throw new ArabicShapingException("No space for YehHamza expansion"); 1552 } 1553 } 1554 } 1555 return length; 1556 } 1557 1558 1559 /* Convert the input buffer from FExx Range into 06xx Range 1560 * to put all characters into the 06xx range 1561 * even the lamalef is converted to the special region in 1562 * the 06xx range. Return the number of lamalef chars found. 1563 */ 1564 private int normalize(char[] dest, int start, int length) { 1565 int lacount = 0; 1566 for (int i = start, e = i + length; i < e; ++i) { 1567 char ch = dest[i]; 1568 if (ch >= '\uFE70' && ch <= '\uFEFC') { 1569 if (isLamAlefChar(ch)) { 1570 ++lacount; 1571 } 1572 dest[i] = (char)convertFEto06[ch - '\uFE70']; 1573 } 1574 } 1575 return lacount; 1576 } 1577 1578 /* 1579 * Name : deshapeNormalize 1580 * Function: Convert the input buffer from FExx Range into 06xx Range 1581 * even the lamalef is converted to the special region in the 06xx range. 1582 * According to the options the user enters, all seen family characters 1583 * followed by a tail character are merged to seen tail family character and 1584 * any yeh followed by a hamza character are merged to yehhamza character. 1585 * Method returns the number of lamalef chars found. 1586 */ 1587 private int deshapeNormalize(char[] dest, int start, int length) { 1588 int lacount = 0; 1589 int yehHamzaComposeEnabled = 0; 1590 int seenComposeEnabled = 0; 1591 1592 yehHamzaComposeEnabled = ((options&YEHHAMZA_MASK) == YEHHAMZA_TWOCELL_NEAR) ? 1 : 0; 1593 seenComposeEnabled = ((options&SEEN_MASK) == SEEN_TWOCELL_NEAR)? 1 : 0; 1594 1595 for (int i = start, e = i + length; i < e; ++i) { 1596 char ch = dest[i]; 1597 1598 if( (yehHamzaComposeEnabled == 1) && ((ch == HAMZA06_CHAR) || (ch == HAMZAFE_CHAR)) 1599 && (i < (length - 1)) && isAlefMaksouraChar(dest[i+1] )) { 1600 dest[i] = SPACE_CHAR; 1601 dest[i+1] = YEH_HAMZA_CHAR; 1602 } else if ( (seenComposeEnabled == 1) && (isTailChar(ch)) && (i< (length - 1)) 1603 && (isSeenTailFamilyChar(dest[i+1])==1) ) { 1604 dest[i] = SPACE_CHAR; 1605 } 1606 else if (ch >= '\uFE70' && ch <= '\uFEFC') { 1607 if (isLamAlefChar(ch)) { 1608 ++lacount; 1609 } 1610 dest[i] = (char)convertFEto06[ch - '\uFE70']; 1611 } 1612 } 1613 return lacount; 1614 } 1615 1616 /* 1617 * Name : shapeUnicode 1618 * Function: Converts an Arabic Unicode buffer in 06xx Range into a shaped 1619 * arabic Unicode buffer in FExx Range 1620 */ 1621 private int shapeUnicode(char[] dest, 1622 int start, 1623 int length, 1624 int destSize, 1625 int tashkeelFlag)throws ArabicShapingException { 1626 1627 int lamalef_count = normalize(dest, start, length); 1628 1629 // resolve the link between the characters. 1630 // Arabic characters have four forms: Isolated, Initial, Medial and Final. 1631 // Tashkeel characters have two, isolated or medial, and sometimes only isolated. 1632 // tashkeelFlag == 0: shape normally, 1: shape isolated, 2: don't shape 1633 1634 boolean lamalef_found = false, seenfam_found = false; 1635 boolean yehhamza_found = false, tashkeel_found = false; 1636 int i = start + length - 1; 1637 int currLink = getLink(dest[i]); 1638 int nextLink = 0; 1639 int prevLink = 0; 1640 int lastLink = 0; 1641 //int prevPos = i; 1642 int lastPos = i; 1643 int nx = -2; 1644 int nw = 0; 1645 1646 while (i >= 0) { 1647 // If high byte of currLink != 0 then there might be more than one shape 1648 if ((currLink & '\uFF00') != 0 || isTashkeelChar(dest[i])) { 1649 nw = i - 1; 1650 nx = -2; 1651 while (nx < 0) { // we need to know about next char 1652 if (nw == -1) { 1653 nextLink = 0; 1654 nx = Integer.MAX_VALUE; 1655 } else { 1656 nextLink = getLink(dest[nw]); 1657 if ((nextLink & IRRELEVANT) == 0) { 1658 nx = nw; 1659 } else { 1660 --nw; 1661 } 1662 } 1663 } 1664 1665 if (((currLink & ALEFTYPE) > 0) && ((lastLink & LAMTYPE) > 0)) { 1666 lamalef_found = true; 1667 char wLamalef = changeLamAlef(dest[i]); // get from 0x065C-0x065f 1668 if (wLamalef != '\u0000') { 1669 // replace alef by marker, it will be removed later 1670 dest[i] = '\uffff'; 1671 dest[lastPos] = wLamalef; 1672 i = lastPos; 1673 } 1674 1675 lastLink = prevLink; 1676 currLink = getLink(wLamalef); // requires '\u0000', unfortunately 1677 } 1678 if ((i > 0) && (dest[i-1] == SPACE_CHAR)) 1679 { 1680 if ( isSeenFamilyChar(dest[i]) == 1){ 1681 seenfam_found = true; 1682 } else if (dest[i] == YEH_HAMZA_CHAR) { 1683 yehhamza_found = true; 1684 } 1685 } 1686 else if(i==0){ 1687 if ( isSeenFamilyChar(dest[i]) == 1){ 1688 seenfam_found = true; 1689 } else if (dest[i] == YEH_HAMZA_CHAR) { 1690 yehhamza_found = true; 1691 } 1692 } 1693 1694 1695 // get the proper shape according to link ability of neighbors 1696 // and of character; depends on the order of the shapes 1697 // (isolated, initial, middle, final) in the compatibility area 1698 1699 int flag = specialChar(dest[i]); 1700 1701 int shape = shapeTable[nextLink & LINK_MASK] 1702 [lastLink & LINK_MASK] 1703 [currLink & LINK_MASK]; 1704 1705 if (flag == 1) { 1706 shape &= 0x1; 1707 } else if (flag == 2) { 1708 if (tashkeelFlag == 0 && 1709 ((lastLink & LINKL) != 0) && 1710 ((nextLink & LINKR) != 0) && 1711 dest[i] != '\u064C' && 1712 dest[i] != '\u064D' && 1713 !((nextLink & ALEFTYPE) == ALEFTYPE && 1714 (lastLink & LAMTYPE) == LAMTYPE)) { 1715 1716 shape = 1; 1717 1718 } else if(tashkeelFlag == 2 && dest[i] == SHADDA06_CHAR){ 1719 shape = 1; 1720 1721 } else { 1722 shape = 0; 1723 } 1724 } 1725 if (flag == 2) { 1726 if (tashkeelFlag == 2 && dest[i] != SHADDA06_CHAR) { 1727 dest[i] = TASHKEEL_SPACE_SUB; 1728 tashkeel_found = true; 1729 } 1730 else{ 1731 dest[i] = (char)('\uFE70' + irrelevantPos[dest[i] - '\u064B'] + shape); 1732 } 1733 // else leave tashkeel alone 1734 } else { 1735 dest[i] = (char)('\uFE70' + (currLink >> 8) + shape); 1736 } 1737 } 1738 1739 // move one notch forward 1740 if ((currLink & IRRELEVANT) == 0) { 1741 prevLink = lastLink; 1742 lastLink = currLink; 1743 //prevPos = lastPos; 1744 lastPos = i; 1745 } 1746 1747 --i; 1748 if (i == nx) { 1749 currLink = nextLink; 1750 nx = -2; 1751 } else if (i != -1) { 1752 currLink = getLink(dest[i]); 1753 } 1754 } 1755 1756 // If we found a lam/alef pair in the buffer 1757 // call handleGeneratedSpaces to remove the spaces that were added 1758 1759 destSize = length; 1760 if (lamalef_found || tashkeel_found) { 1761 destSize = handleGeneratedSpaces(dest, start, length); 1762 } 1763 if (seenfam_found || yehhamza_found){ 1764 destSize = expandCompositChar(dest, start, destSize, lamalef_count, SHAPE_MODE); 1765 } 1766 return destSize; 1767 } 1768 1769 /* 1770 * Name : deShapeUnicode 1771 * Function: Converts an Arabic Unicode buffer in FExx Range into unshaped 1772 * arabic Unicode buffer in 06xx Range 1773 */ 1774 private int deShapeUnicode(char[] dest, 1775 int start, 1776 int length, 1777 int destSize) throws ArabicShapingException { 1778 1779 int lamalef_count = deshapeNormalize(dest, start, length); 1780 1781 // If there was a lamalef in the buffer call expandLamAlef 1782 if (lamalef_count != 0) { 1783 // need to adjust dest to fit expanded buffer... !!! 1784 destSize = expandCompositChar(dest, start, length, lamalef_count,DESHAPE_MODE); 1785 } else { 1786 destSize = length; 1787 } 1788 1789 return destSize; 1790 } 1791 1792 private int internalShape(char[] source, 1793 int sourceStart, 1794 int sourceLength, 1795 char[] dest, 1796 int destStart, 1797 int destSize) throws ArabicShapingException { 1798 1799 if (sourceLength == 0) { 1800 return 0; 1801 } 1802 1803 if (destSize == 0) { 1804 if (((options & LETTERS_MASK) != LETTERS_NOOP) && 1805 ((options & LAMALEF_MASK) == LAMALEF_RESIZE)) { 1806 1807 return calculateSize(source, sourceStart, sourceLength); 1808 } else { 1809 return sourceLength; // by definition 1810 } 1811 } 1812 1813 // always use temp buffer 1814 char[] temp = new char[sourceLength * 2]; // all lamalefs requiring expansion 1815 System.arraycopy(source, sourceStart, temp, 0, sourceLength); 1816 1817 if (isLogical) { 1818 invertBuffer(temp, 0, sourceLength); 1819 } 1820 1821 int outputSize = sourceLength; 1822 1823 switch (options & LETTERS_MASK) { 1824 case LETTERS_SHAPE_TASHKEEL_ISOLATED: 1825 outputSize = shapeUnicode(temp, 0, sourceLength, destSize, 1); 1826 break; 1827 1828 case LETTERS_SHAPE: 1829 if( ((options&TASHKEEL_MASK) != 0) && 1830 ((options&TASHKEEL_MASK) !=TASHKEEL_REPLACE_BY_TATWEEL)) { 1831 /* Call the shaping function with tashkeel flag == 2 for removal of tashkeel */ 1832 outputSize = shapeUnicode(temp, 0, sourceLength, destSize, 2); 1833 }else { 1834 //default Call the shaping function with tashkeel flag == 1 */ 1835 outputSize = shapeUnicode(temp, 0, sourceLength, destSize, 0); 1836 1837 /*After shaping text check if user wants to remove tashkeel and replace it with tatweel*/ 1838 if( (options&TASHKEEL_MASK) == TASHKEEL_REPLACE_BY_TATWEEL){ 1839 outputSize = handleTashkeelWithTatweel(temp,sourceLength); 1840 } 1841 } 1842 break; 1843 1844 case LETTERS_UNSHAPE: 1845 outputSize = deShapeUnicode(temp, 0, sourceLength, destSize); 1846 break; 1847 1848 default: 1849 break; 1850 } 1851 1852 if (outputSize > destSize) { 1853 throw new ArabicShapingException("not enough room for result data"); 1854 } 1855 1856 if ((options & DIGITS_MASK) != DIGITS_NOOP) { 1857 char digitBase = '\u0030'; // European digits 1858 switch (options & DIGIT_TYPE_MASK) { 1859 case DIGIT_TYPE_AN: 1860 digitBase = '\u0660'; // Arabic-Indic digits 1861 break; 1862 1863 case DIGIT_TYPE_AN_EXTENDED: 1864 digitBase = '\u06f0'; // Eastern Arabic-Indic digits (Persian and Urdu) 1865 break; 1866 1867 default: 1868 break; 1869 } 1870 1871 switch (options & DIGITS_MASK) { 1872 case DIGITS_EN2AN: 1873 { 1874 int digitDelta = digitBase - '\u0030'; 1875 for (int i = 0; i < outputSize; ++i) { 1876 char ch = temp[i]; 1877 if (ch <= '\u0039' && ch >= '\u0030') { 1878 temp[i] += digitDelta; 1879 } 1880 } 1881 } 1882 break; 1883 1884 case DIGITS_AN2EN: 1885 { 1886 char digitTop = (char)(digitBase + 9); 1887 int digitDelta = '\u0030' - digitBase; 1888 for (int i = 0; i < outputSize; ++i) { 1889 char ch = temp[i]; 1890 if (ch <= digitTop && ch >= digitBase) { 1891 temp[i] += digitDelta; 1892 } 1893 } 1894 } 1895 break; 1896 1897 case DIGITS_EN2AN_INIT_LR: 1898 shapeToArabicDigitsWithContext(temp, 0, outputSize, digitBase, false); 1899 break; 1900 1901 case DIGITS_EN2AN_INIT_AL: 1902 shapeToArabicDigitsWithContext(temp, 0, outputSize, digitBase, true); 1903 break; 1904 1905 default: 1906 break; 1907 } 1908 } 1909 1910 if (isLogical) { 1911 invertBuffer(temp, 0, outputSize); 1912 } 1913 1914 System.arraycopy(temp, 0, dest, destStart, outputSize); 1915 1916 return outputSize; 1917 } 1918} 1919