1 /* 2 ******************************************************************************* 3 * Copyright (C) 2001-2009, International Business Machines 4 * Corporation and others. All Rights Reserved. 5 ******************************************************************************* 6 */ 7 8 /* 9 * Ported with minor modifications from ICU4J 4.2's 10 * com.ibm.icu.text.ArabicShaping class. 11 */ 12 13 package android.icu.text; 14 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 * @stable ICU 2.0 57 * 58 * @hide 59 */ 60 public class ArabicShaping { 61 private final int options; 62 private boolean isLogical; // convenience 63 private boolean spacesRelativeToTextBeginEnd; 64 private char tailChar; 65 66 public static final ArabicShaping SHAPER = new ArabicShaping( 67 ArabicShaping.TEXT_DIRECTION_LOGICAL | 68 ArabicShaping.LENGTH_FIXED_SPACES_NEAR | 69 ArabicShaping.LETTERS_SHAPE | 70 ArabicShaping.DIGITS_NOOP); 71 72 /** 73 * Convert a range of text in the source array, putting the result 74 * into a range of text in the destination array, and return the number 75 * of characters written. 76 * 77 * @param source An array containing the input text 78 * @param sourceStart The start of the range of text to convert 79 * @param sourceLength The length of the range of text to convert 80 * @param dest The destination array that will receive the result. 81 * It may be <code>NULL</code> only if <code>destSize</code> is 0. 82 * @param destStart The start of the range of the destination buffer to use. 83 * @param destSize The size (capacity) of the destination buffer. 84 * If <code>destSize</code> is 0, then no output is produced, 85 * but the necessary buffer size is returned ("preflighting"). This 86 * does not validate the text against the options, for example, 87 * if letters are being unshaped, and spaces are being consumed 88 * following lamalef, this will not detect a lamalef without a 89 * corresponding space. An error will be thrown when the actual 90 * conversion is attempted. 91 * @return The number of chars written to the destination buffer. 92 * If an error occurs, then no output was written, or it may be 93 * incomplete. 94 * @throws ArabicShapingException if the text cannot be converted according to the options. 95 * @stable ICU 2.0 96 */ shape(char[] source, int sourceStart, int sourceLength, char[] dest, int destStart, int destSize)97 public int shape(char[] source, int sourceStart, int sourceLength, 98 char[] dest, int destStart, int destSize) throws ArabicShapingException { 99 if (source == null) { 100 throw new IllegalArgumentException("source can not be null"); 101 } 102 if (sourceStart < 0 || sourceLength < 0 || sourceStart + sourceLength > source.length) { 103 throw new IllegalArgumentException("bad source start (" + sourceStart + 104 ") or length (" + sourceLength + 105 ") for buffer of length " + source.length); 106 } 107 if (dest == null && destSize != 0) { 108 throw new IllegalArgumentException("null dest requires destSize == 0"); 109 } 110 if ((destSize != 0) && 111 (destStart < 0 || destSize < 0 || destStart + destSize > dest.length)) { 112 throw new IllegalArgumentException("bad dest start (" + destStart + 113 ") or size (" + destSize + 114 ") for buffer of length " + dest.length); 115 } 116 /* Validate input options */ 117 if ( ((options&TASHKEEL_MASK) > 0) && 118 !(((options & TASHKEEL_MASK)==TASHKEEL_BEGIN) || 119 ((options & TASHKEEL_MASK)==TASHKEEL_END ) || 120 ((options & TASHKEEL_MASK)==TASHKEEL_RESIZE )|| 121 ((options & TASHKEEL_MASK)==TASHKEEL_REPLACE_BY_TATWEEL)) ){ 122 throw new IllegalArgumentException("Wrong Tashkeel argument"); 123 } 124 125 ///CLOVER:OFF 126 //According to Steven Loomis, the code is unreachable when you OR all the constants within the if statements 127 if(((options&LAMALEF_MASK) > 0)&& 128 !(((options & LAMALEF_MASK)==LAMALEF_BEGIN) || 129 ((options & LAMALEF_MASK)==LAMALEF_END ) || 130 ((options & LAMALEF_MASK)==LAMALEF_RESIZE )|| 131 ((options & LAMALEF_MASK)==LAMALEF_AUTO) || 132 ((options & LAMALEF_MASK)==LAMALEF_NEAR))){ 133 throw new IllegalArgumentException("Wrong Lam Alef argument"); 134 } 135 ///CLOVER:ON 136 137 /* Validate Tashkeel (Tashkeel replacement options should be enabled in shaping mode only)*/ 138 if(((options&TASHKEEL_MASK) > 0) && (options&LETTERS_MASK) == LETTERS_UNSHAPE) { 139 throw new IllegalArgumentException("Tashkeel replacement should not be enabled in deshaping mode "); 140 } 141 return internalShape(source, sourceStart, sourceLength, dest, destStart, destSize); 142 } 143 144 /** 145 * Convert a range of text in place. This may only be used if the Length option 146 * does not grow or shrink the text. 147 * 148 * @param source An array containing the input text 149 * @param start The start of the range of text to convert 150 * @param length The length of the range of text to convert 151 * @throws ArabicShapingException if the text cannot be converted according to the options. 152 * @stable ICU 2.0 153 */ shape(char[] source, int start, int length)154 public void shape(char[] source, int start, int length) throws ArabicShapingException { 155 if ((options & LAMALEF_MASK) == LAMALEF_RESIZE) { 156 throw new ArabicShapingException("Cannot shape in place with length option resize."); 157 } 158 shape(source, start, length, source, start, length); 159 } 160 161 /** 162 * Convert a string, returning the new string. 163 * 164 * @param text the string to convert 165 * @return the converted string 166 * @throws ArabicShapingException if the string cannot be converted according to the options. 167 * @stable ICU 2.0 168 */ shape(String text)169 public String shape(String text) throws ArabicShapingException { 170 char[] src = text.toCharArray(); 171 char[] dest = src; 172 if (((options & LAMALEF_MASK) == LAMALEF_RESIZE) && 173 ((options & LETTERS_MASK) == LETTERS_UNSHAPE)) { 174 175 dest = new char[src.length * 2]; // max 176 } 177 int len = shape(src, 0, src.length, dest, 0, dest.length); 178 179 return new String(dest, 0, len); 180 } 181 182 /** 183 * Construct ArabicShaping using the options flags. 184 * The flags are as follows:<br> 185 * 'LENGTH' flags control whether the text can change size, and if not, 186 * how to maintain the size of the text when LamAlef ligatures are 187 * formed or broken.<br> 188 * 'TEXT_DIRECTION' flags control whether the text is read and written 189 * in visual order or in logical order.<br> 190 * 'LETTERS_SHAPE' flags control whether conversion is to or from 191 * presentation forms.<br> 192 * 'DIGITS' flags control whether digits are shaped, and whether from 193 * European to Arabic-Indic or vice-versa.<br> 194 * 'DIGIT_TYPE' flags control whether standard or extended Arabic-Indic 195 * digits are used when performing digit conversion. 196 * @stable ICU 2.0 197 */ ArabicShaping(int options)198 public ArabicShaping(int options) { 199 this.options = options; 200 if ((options & DIGITS_MASK) > 0x80) { 201 throw new IllegalArgumentException("bad DIGITS options"); 202 } 203 204 isLogical = ( (options & TEXT_DIRECTION_MASK) == TEXT_DIRECTION_LOGICAL ); 205 /* Validate options */ 206 spacesRelativeToTextBeginEnd = ( (options & SPACES_RELATIVE_TO_TEXT_MASK) == SPACES_RELATIVE_TO_TEXT_BEGIN_END ); 207 if ( (options&SHAPE_TAIL_TYPE_MASK) == SHAPE_TAIL_NEW_UNICODE){ 208 tailChar = NEW_TAIL_CHAR; 209 } else { 210 tailChar = OLD_TAIL_CHAR; 211 } 212 } 213 214 /* Seen Tail options */ 215 /** 216 * Memory option: the result must have the same length as the source. 217 * Shaping mode: The SEEN family character will expand into two characters using space near 218 * the SEEN family character(i.e. the space after the character). 219 * if there are no spaces found, ArabicShapingException will be thrown 220 * 221 * De-shaping mode: Any Seen character followed by Tail character will be 222 * replaced by one cell Seen and a space will replace the Tail. 223 * Affects: Seen options 224 */ 225 public static final int SEEN_TWOCELL_NEAR = 0x200000; 226 227 /** Bit mask for Seen memory options. */ 228 public static final int SEEN_MASK = 0x700000; 229 230 /* YehHamza options */ 231 /** 232 * Memory option: the result must have the same length as the source. 233 * Shaping mode: The YEHHAMZA character will expand into two characters using space near it 234 * (i.e. the space after the character) 235 * if there are no spaces found, ArabicShapingException will be thrown 236 * 237 * De-shaping mode: Any Yeh (final or isolated) character followed by Hamza character will be 238 * replaced by one cell YehHamza and space will replace the Hamza. 239 * Affects: YehHamza options 240 */ 241 public static final int YEHHAMZA_TWOCELL_NEAR = 0x1000000; 242 243 244 /** Bit mask for YehHamza memory options. */ 245 public static final int YEHHAMZA_MASK = 0x3800000; 246 247 /* New Tashkeel options */ 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 beginning of the buffer 252 * 253 * De-shaping mode: N/A 254 * Affects: Tashkeel options 255 */ 256 public static final int TASHKEEL_BEGIN = 0x40000; 257 258 /** 259 * Memory option: the result must have the same length as the source. 260 * Shaping mode: Tashkeel characters will be replaced by spaces. 261 * Spaces will be placed at end of the buffer 262 * 263 * De-shaping mode: N/A 264 * Affects: Tashkeel options 265 */ 266 public static final int TASHKEEL_END = 0x60000; 267 268 /** 269 * Memory option: allow the result to have a different length than the source. 270 * Shaping mode: Tashkeel characters will be removed, buffer length will shrink. 271 * De-shaping mode: N/A 272 * 273 * Affects: Tashkeel options 274 */ 275 public static final int TASHKEEL_RESIZE = 0x80000; 276 277 /** 278 * Memory option: the result must have the same length as the source. 279 * Shaping mode: Tashkeel characters will be replaced by Tatweel if it is connected to adjacent 280 * characters (i.e. shaped on Tatweel) or replaced by space if it is not connected. 281 * 282 * De-shaping mode: N/A 283 * Affects: YehHamza options 284 */ 285 public static final int TASHKEEL_REPLACE_BY_TATWEEL = 0xC0000; 286 287 /** Bit mask for Tashkeel replacement with Space or Tatweel memory options. */ 288 public static final int TASHKEEL_MASK = 0xE0000; 289 290 /* Space location Control options */ 291 /** 292 * This option effects the meaning of BEGIN and END options. if this option is not used the default 293 * for BEGIN and END will be as following: 294 * The Default (for both Visual LTR, Visual RTL and Logical Text) 295 * 1. BEGIN always refers to the start address of physical memory. 296 * 2. END always refers to the end address of physical memory. 297 * 298 * If this option is used it will swap the meaning of BEGIN and END only for Visual LTR text. 299 * 300 * The affect on BEGIN and END Memory Options will be as following: 301 * A. BEGIN For Visual LTR text: This will be the beginning (right side) of the visual text 302 * (corresponding to the physical memory address end, same as END in default behavior) 303 * B. BEGIN For Logical text: Same as BEGIN in default behavior. 304 * C. END For Visual LTR text: This will be the end (left side) of the visual text. (corresponding to 305 * the physical memory address beginning, same as BEGIN in default behavior) 306 * D. END For Logical text: Same as END in default behavior. 307 * Affects: All LamAlef BEGIN, END and AUTO options. 308 */ 309 public static final int SPACES_RELATIVE_TO_TEXT_BEGIN_END = 0x4000000; 310 311 /** Bit mask for swapping BEGIN and END for Visual LTR text */ 312 public static final int SPACES_RELATIVE_TO_TEXT_MASK = 0x4000000; 313 314 /** 315 * If this option is used, shaping will use the new Unicode code point for TAIL (i.e. 0xFE73). 316 * If this option is not specified (Default), old unofficial Unicode TAIL code point is used (i.e. 0x200B) 317 * De-shaping will not use this option as it will always search for both the new Unicode code point for the 318 * TAIL (i.e. 0xFE73) or the old unofficial Unicode TAIL code point (i.e. 0x200B) and de-shape the 319 * Seen-Family letter accordingly. 320 * 321 * Shaping Mode: Only shaping. 322 * De-shaping Mode: N/A. 323 * Affects: All Seen options 324 */ 325 public static final int SHAPE_TAIL_NEW_UNICODE = 0x8000000; 326 327 /** Bit mask for new Unicode Tail option */ 328 public static final int SHAPE_TAIL_TYPE_MASK = 0x8000000; 329 330 /** 331 * Memory option: allow the result to have a different length than the source. 332 * @stable ICU 2.0 333 */ 334 public static final int LENGTH_GROW_SHRINK = 0; 335 336 /** 337 * Memory option: allow the result to have a different length than the source. 338 * Affects: LamAlef options 339 * This option is an alias to LENGTH_GROW_SHRINK 340 */ 341 public static final int LAMALEF_RESIZE = 0; 342 343 /** 344 * Memory option: the result must have the same length as the source. 345 * If more room is necessary, then try to consume spaces next to modified characters. 346 * @stable ICU 2.0 347 */ 348 public static final int LENGTH_FIXED_SPACES_NEAR = 1; 349 350 /** 351 * Memory option: the result must have the same length as the source. 352 * If more room is necessary, then try to consume spaces next to modified characters. 353 * Affects: LamAlef options 354 * This option is an alias to LENGTH_FIXED_SPACES_NEAR 355 */ 356 public static final int LAMALEF_NEAR = 1 ; 357 358 /** 359 * Memory option: the result must have the same length as the source. 360 * If more room is necessary, then try to consume spaces at the end of the text. 361 * @stable ICU 2.0 362 */ 363 public static final int LENGTH_FIXED_SPACES_AT_END = 2; 364 365 366 /** 367 * Memory option: the result must have the same length as the source. 368 * If more room is necessary, then try to consume spaces at the end of the text. 369 * Affects: LamAlef options 370 * This option is an alias to LENGTH_FIXED_SPACES_AT_END 371 */ 372 public static final int LAMALEF_END = 2; 373 374 /** 375 * Memory option: the result must have the same length as the source. 376 * If more room is necessary, then try to consume spaces at the beginning of the text. 377 * @stable ICU 2.0 378 */ 379 public static final int LENGTH_FIXED_SPACES_AT_BEGINNING = 3; 380 381 /** 382 * Memory option: the result must have the same length as the source. 383 * If more room is necessary, then try to consume spaces at the beginning of the text. 384 * Affects: LamAlef options 385 * This option is an alias to LENGTH_FIXED_SPACES_AT_BEGINNING 386 */ 387 public static final int LAMALEF_BEGIN = 3; 388 389 /** 390 * Memory option: the result must have the same length as the source. 391 * Shaping Mode: For each LAMALEF character found, expand LAMALEF using space at end. 392 * If there is no space at end, use spaces at beginning of the buffer. If there 393 * is no space at beginning of the buffer, use spaces at the near (i.e. the space 394 * after the LAMALEF character). 395 * 396 * Deshaping Mode: Perform the same function as the flag equals LAMALEF_END. 397 * Affects: LamAlef options 398 */ 399 public static final int LAMALEF_AUTO = 0x10000; 400 401 /** 402 * Bit mask for memory options. 403 * @stable ICU 2.0 404 */ 405 public static final int LENGTH_MASK = 0x10003; 406 407 /** Bit mask for LamAlef memory options. */ 408 409 public static final int LAMALEF_MASK = 0x10003; 410 411 /** 412 * Direction indicator: the source is in logical (keyboard) order. 413 * @stable ICU 2.0 414 */ 415 public static final int TEXT_DIRECTION_LOGICAL = 0; 416 417 /** 418 * Direction indicator:the source is in visual RTL order, 419 * the rightmost displayed character stored first. 420 * This option is an alias to U_SHAPE_TEXT_DIRECTION_LOGICAL 421 */ 422 public static final int TEXT_DIRECTION_VISUAL_RTL = 0; 423 424 /** 425 * Direction indicator: the source is in visual (display) order, that is, 426 * the leftmost displayed character is stored first. 427 * @stable ICU 2.0 428 */ 429 public static final int TEXT_DIRECTION_VISUAL_LTR = 4; 430 431 /** 432 * Bit mask for direction indicators. 433 * @stable ICU 2.0 434 */ 435 public static final int TEXT_DIRECTION_MASK = 4; 436 437 438 /** 439 * Letter shaping option: do not perform letter shaping. 440 * @stable ICU 2.0 441 */ 442 public static final int LETTERS_NOOP = 0; 443 444 /** 445 * Letter shaping option: replace normative letter characters in the U+0600 (Arabic) block, 446 * by shaped ones in the U+FE70 (Presentation Forms B) block. Performs Lam-Alef ligature 447 * substitution. 448 * @stable ICU 2.0 449 */ 450 public static final int LETTERS_SHAPE = 8; 451 452 /** 453 * Letter shaping option: replace shaped letter characters in the U+FE70 (Presentation Forms B) block 454 * by normative ones in the U+0600 (Arabic) block. Converts Lam-Alef ligatures to pairs of Lam and 455 * Alef characters, consuming spaces if required. 456 * @stable ICU 2.0 457 */ 458 public static final int LETTERS_UNSHAPE = 0x10; 459 460 /** 461 * Letter shaping option: replace normative letter characters in the U+0600 (Arabic) block, 462 * except for the TASHKEEL characters at U+064B...U+0652, by shaped ones in the U+Fe70 463 * (Presentation Forms B) block. The TASHKEEL characters will always be converted to 464 * the isolated forms rather than to their correct shape. 465 * @stable ICU 2.0 466 */ 467 public static final int LETTERS_SHAPE_TASHKEEL_ISOLATED = 0x18; 468 469 /** 470 * Bit mask for letter shaping options. 471 * @stable ICU 2.0 472 */ 473 public static final int LETTERS_MASK = 0x18; 474 475 476 /** 477 * Digit shaping option: do not perform digit shaping. 478 * @stable ICU 2.0 479 */ 480 public static final int DIGITS_NOOP = 0; 481 482 /** 483 * Digit shaping option: Replace European digits (U+0030...U+0039) by Arabic-Indic digits. 484 * @stable ICU 2.0 485 */ 486 public static final int DIGITS_EN2AN = 0x20; 487 488 /** 489 * Digit shaping option: Replace Arabic-Indic digits by European digits (U+0030...U+0039). 490 * @stable ICU 2.0 491 */ 492 public static final int DIGITS_AN2EN = 0x40; 493 494 /** 495 * Digit shaping option: 496 * Replace European digits (U+0030...U+0039) by Arabic-Indic digits 497 * if the most recent strongly directional character 498 * is an Arabic letter (its Bidi direction value is RIGHT_TO_LEFT_ARABIC). 499 * The initial state at the start of the text is assumed to be not an Arabic, 500 * letter, so European digits at the start of the text will not change. 501 * Compare to DIGITS_ALEN2AN_INIT_AL. 502 * @stable ICU 2.0 503 */ 504 public static final int DIGITS_EN2AN_INIT_LR = 0x60; 505 506 /** 507 * Digit shaping option: 508 * Replace European digits (U+0030...U+0039) by Arabic-Indic digits 509 * if the most recent strongly directional character 510 * is an Arabic letter (its Bidi direction value is RIGHT_TO_LEFT_ARABIC). 511 * The initial state at the start of the text is assumed to be an Arabic, 512 * letter, so European digits at the start of the text will change. 513 * Compare to DIGITS_ALEN2AN_INT_LR. 514 * @stable ICU 2.0 515 */ 516 public static final int DIGITS_EN2AN_INIT_AL = 0x80; 517 518 /** Not a valid option value. */ 519 //private static final int DIGITS_RESERVED = 0xa0; 520 521 /** 522 * Bit mask for digit shaping options. 523 * @stable ICU 2.0 524 */ 525 public static final int DIGITS_MASK = 0xe0; 526 527 /** 528 * Digit type option: Use Arabic-Indic digits (U+0660...U+0669). 529 * @stable ICU 2.0 530 */ 531 public static final int DIGIT_TYPE_AN = 0; 532 533 /** 534 * Digit type option: Use Eastern (Extended) Arabic-Indic digits (U+06f0...U+06f9). 535 * @stable ICU 2.0 536 */ 537 public static final int DIGIT_TYPE_AN_EXTENDED = 0x100; 538 539 /** 540 * Bit mask for digit type options. 541 * @stable ICU 2.0 542 */ 543 public static final int DIGIT_TYPE_MASK = 0x0100; // 0x3f00? 544 545 /** 546 * some constants 547 */ 548 private static final char HAMZAFE_CHAR = '\ufe80'; 549 private static final char HAMZA06_CHAR = '\u0621'; 550 private static final char YEH_HAMZA_CHAR = '\u0626'; 551 private static final char YEH_HAMZAFE_CHAR = '\uFE89'; 552 private static final char LAMALEF_SPACE_SUB = '\uffff'; 553 private static final char TASHKEEL_SPACE_SUB = '\ufffe'; 554 private static final char LAM_CHAR = '\u0644'; 555 private static final char SPACE_CHAR = '\u0020'; 556 private static final char SPACE_CHAR_FOR_LAMALEF = '\ufeff'; // XXX: tweak for TextLine use 557 private static final char SHADDA_CHAR = '\uFE7C'; 558 private static final char TATWEEL_CHAR = '\u0640'; 559 private static final char SHADDA_TATWEEL_CHAR = '\uFE7D'; 560 private static final char NEW_TAIL_CHAR = '\uFE73'; 561 private static final char OLD_TAIL_CHAR = '\u200B'; 562 private static final int SHAPE_MODE = 0; 563 private static final int DESHAPE_MODE = 1; 564 565 /** 566 * @stable ICU 2.0 567 */ equals(Object rhs)568 public boolean equals(Object rhs) { 569 return rhs != null && 570 rhs.getClass() == ArabicShaping.class && 571 options == ((ArabicShaping)rhs).options; 572 } 573 574 /** 575 * @stable ICU 2.0 576 */ 577 ///CLOVER:OFF hashCode()578 public int hashCode() { 579 return options; 580 } 581 582 /** 583 * @stable ICU 2.0 584 */ toString()585 public String toString() { 586 StringBuffer buf = new StringBuffer(super.toString()); 587 buf.append('['); 588 589 switch (options & LAMALEF_MASK) { 590 case LAMALEF_RESIZE: buf.append("LamAlef resize"); break; 591 case LAMALEF_NEAR: buf.append("LamAlef spaces at near"); break; 592 case LAMALEF_BEGIN: buf.append("LamAlef spaces at begin"); break; 593 case LAMALEF_END: buf.append("LamAlef spaces at end"); break; 594 case LAMALEF_AUTO: buf.append("lamAlef auto"); break; 595 } 596 switch (options & TEXT_DIRECTION_MASK) { 597 case TEXT_DIRECTION_LOGICAL: buf.append(", logical"); break; 598 case TEXT_DIRECTION_VISUAL_LTR: buf.append(", visual"); break; 599 } 600 switch (options & LETTERS_MASK) { 601 case LETTERS_NOOP: buf.append(", no letter shaping"); break; 602 case LETTERS_SHAPE: buf.append(", shape letters"); break; 603 case LETTERS_SHAPE_TASHKEEL_ISOLATED: buf.append(", shape letters tashkeel isolated"); break; 604 case LETTERS_UNSHAPE: buf.append(", unshape letters"); break; 605 } 606 switch (options & SEEN_MASK) { 607 case SEEN_TWOCELL_NEAR: buf.append(", Seen at near"); break; 608 } 609 switch (options & YEHHAMZA_MASK) { 610 case YEHHAMZA_TWOCELL_NEAR: buf.append(", Yeh Hamza at near"); break; 611 } 612 switch (options & TASHKEEL_MASK) { 613 case TASHKEEL_BEGIN: buf.append(", Tashkeel at begin"); break; 614 case TASHKEEL_END: buf.append(", Tashkeel at end"); break; 615 case TASHKEEL_REPLACE_BY_TATWEEL: buf.append(", Tashkeel replace with tatweel"); break; 616 case TASHKEEL_RESIZE: buf.append(", Tashkeel resize"); break; 617 } 618 619 switch (options & DIGITS_MASK) { 620 case DIGITS_NOOP: buf.append(", no digit shaping"); break; 621 case DIGITS_EN2AN: buf.append(", shape digits to AN"); break; 622 case DIGITS_AN2EN: buf.append(", shape digits to EN"); break; 623 case DIGITS_EN2AN_INIT_LR: buf.append(", shape digits to AN contextually: default EN"); break; 624 case DIGITS_EN2AN_INIT_AL: buf.append(", shape digits to AN contextually: default AL"); break; 625 } 626 switch (options & DIGIT_TYPE_MASK) { 627 case DIGIT_TYPE_AN: buf.append(", standard Arabic-Indic digits"); break; 628 case DIGIT_TYPE_AN_EXTENDED: buf.append(", extended Arabic-Indic digits"); break; 629 } 630 buf.append("]"); 631 632 return buf.toString(); 633 } 634 ///CLOVER:ON 635 636 // 637 // ported api 638 // 639 640 private static final int IRRELEVANT = 4; 641 private static final int LAMTYPE = 16; 642 private static final int ALEFTYPE = 32; 643 644 private static final int LINKR = 1; 645 private static final int LINKL = 2; 646 private static final int LINK_MASK = 3; 647 648 private static final int irrelevantPos[] = { 649 0x0, 0x2, 0x4, 0x6, 0x8, 0xA, 0xC, 0xE 650 }; 651 652 /* 653 private static final char convertLamAlef[] = { 654 '\u0622', // FEF5 655 '\u0622', // FEF6 656 '\u0623', // FEF7 657 '\u0623', // FEF8 658 '\u0625', // FEF9 659 '\u0625', // FEFA 660 '\u0627', // FEFB 661 '\u0627' // FEFC 662 }; 663 */ 664 665 private static final int tailFamilyIsolatedFinal[] = { 666 /* FEB1 */ 1, 667 /* FEB2 */ 1, 668 /* FEB3 */ 0, 669 /* FEB4 */ 0, 670 /* FEB5 */ 1, 671 /* FEB6 */ 1, 672 /* FEB7 */ 0, 673 /* FEB8 */ 0, 674 /* FEB9 */ 1, 675 /* FEBA */ 1, 676 /* FEBB */ 0, 677 /* FEBC */ 0, 678 /* FEBD */ 1, 679 /* FEBE */ 1 680 }; 681 682 private static final int tashkeelMedial[] = { 683 /* FE70 */ 0, 684 /* FE71 */ 1, 685 /* FE72 */ 0, 686 /* FE73 */ 0, 687 /* FE74 */ 0, 688 /* FE75 */ 0, 689 /* FE76 */ 0, 690 /* FE77 */ 1, 691 /* FE78 */ 0, 692 /* FE79 */ 1, 693 /* FE7A */ 0, 694 /* FE7B */ 1, 695 /* FE7C */ 0, 696 /* FE7D */ 1, 697 /* FE7E */ 0, 698 /* FE7F */ 1 699 }; 700 701 private static final char yehHamzaToYeh[] = 702 { 703 /* isolated*/ 0xFEEF, 704 /* final */ 0xFEF0 705 }; 706 707 private static final char convertNormalizedLamAlef[] = { 708 '\u0622', // 065C 709 '\u0623', // 065D 710 '\u0625', // 065E 711 '\u0627', // 065F 712 }; 713 714 private static final int[] araLink = { 715 1 + 32 + 256 * 0x11, /*0x0622*/ 716 1 + 32 + 256 * 0x13, /*0x0623*/ 717 1 + 256 * 0x15, /*0x0624*/ 718 1 + 32 + 256 * 0x17, /*0x0625*/ 719 1 + 2 + 256 * 0x19, /*0x0626*/ 720 1 + 32 + 256 * 0x1D, /*0x0627*/ 721 1 + 2 + 256 * 0x1F, /*0x0628*/ 722 1 + 256 * 0x23, /*0x0629*/ 723 1 + 2 + 256 * 0x25, /*0x062A*/ 724 1 + 2 + 256 * 0x29, /*0x062B*/ 725 1 + 2 + 256 * 0x2D, /*0x062C*/ 726 1 + 2 + 256 * 0x31, /*0x062D*/ 727 1 + 2 + 256 * 0x35, /*0x062E*/ 728 1 + 256 * 0x39, /*0x062F*/ 729 1 + 256 * 0x3B, /*0x0630*/ 730 1 + 256 * 0x3D, /*0x0631*/ 731 1 + 256 * 0x3F, /*0x0632*/ 732 1 + 2 + 256 * 0x41, /*0x0633*/ 733 1 + 2 + 256 * 0x45, /*0x0634*/ 734 1 + 2 + 256 * 0x49, /*0x0635*/ 735 1 + 2 + 256 * 0x4D, /*0x0636*/ 736 1 + 2 + 256 * 0x51, /*0x0637*/ 737 1 + 2 + 256 * 0x55, /*0x0638*/ 738 1 + 2 + 256 * 0x59, /*0x0639*/ 739 1 + 2 + 256 * 0x5D, /*0x063A*/ 740 0, 0, 0, 0, 0, /*0x063B-0x063F*/ 741 1 + 2, /*0x0640*/ 742 1 + 2 + 256 * 0x61, /*0x0641*/ 743 1 + 2 + 256 * 0x65, /*0x0642*/ 744 1 + 2 + 256 * 0x69, /*0x0643*/ 745 1 + 2 + 16 + 256 * 0x6D, /*0x0644*/ 746 1 + 2 + 256 * 0x71, /*0x0645*/ 747 1 + 2 + 256 * 0x75, /*0x0646*/ 748 1 + 2 + 256 * 0x79, /*0x0647*/ 749 1 + 256 * 0x7D, /*0x0648*/ 750 1 + 256 * 0x7F, /*0x0649*/ 751 1 + 2 + 256 * 0x81, /*0x064A*/ 752 4, 4, 4, 4, /*0x064B-0x064E*/ 753 4, 4, 4, 4, /*0x064F-0x0652*/ 754 4, 4, 4, 0, 0, /*0x0653-0x0657*/ 755 0, 0, 0, 0, /*0x0658-0x065B*/ 756 1 + 256 * 0x85, /*0x065C*/ 757 1 + 256 * 0x87, /*0x065D*/ 758 1 + 256 * 0x89, /*0x065E*/ 759 1 + 256 * 0x8B, /*0x065F*/ 760 0, 0, 0, 0, 0, /*0x0660-0x0664*/ 761 0, 0, 0, 0, 0, /*0x0665-0x0669*/ 762 0, 0, 0, 0, 0, 0, /*0x066A-0x066F*/ 763 4, /*0x0670*/ 764 0, /*0x0671*/ 765 1 + 32, /*0x0672*/ 766 1 + 32, /*0x0673*/ 767 0, /*0x0674*/ 768 1 + 32, /*0x0675*/ 769 1, 1, /*0x0676-0x0677*/ 770 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x0678-0x067D*/ 771 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x067E-0x0683*/ 772 1+2, 1+2, 1+2, 1+2, /*0x0684-0x0687*/ 773 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /*0x0688-0x0691*/ 774 1, 1, 1, 1, 1, 1, 1, 1, /*0x0692-0x0699*/ 775 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x069A-0x06A3*/ 776 1+2, 1+2, 1+2, 1+2, /*0x069A-0x06A3*/ 777 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06A4-0x06AD*/ 778 1+2, 1+2, 1+2, 1+2, /*0x06A4-0x06AD*/ 779 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06AE-0x06B7*/ 780 1+2, 1+2, 1+2, 1+2, /*0x06AE-0x06B7*/ 781 1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06B8-0x06BF*/ 782 1+2, 1+2, /*0x06B8-0x06BF*/ 783 1, /*0x06C0*/ 784 1+2, /*0x06C1*/ 785 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /*0x06C2-0x06CB*/ 786 1+2, /*0x06CC*/ 787 1, /*0x06CD*/ 788 1+2, 1+2, 1+2, 1+2, /*0x06CE-0x06D1*/ 789 1, 1 /*0x06D2-0x06D3*/ 790 }; 791 792 private static final int[] presLink = { 793 1 + 2, /*0xFE70*/ 794 1 + 2, /*0xFE71*/ 795 1 + 2, 0, 1+ 2, 0, 1+ 2, /*0xFE72-0xFE76*/ 796 1 + 2, /*0xFE77*/ 797 1+ 2, 1 + 2, 1+2, 1 + 2, /*0xFE78-0xFE81*/ 798 1+ 2, 1 + 2, 1+2, 1 + 2, /*0xFE82-0xFE85*/ 799 0, 0 + 32, 1 + 32, 0 + 32, /*0xFE86-0xFE89*/ 800 1 + 32, 0, 1, 0 + 32, /*0xFE8A-0xFE8D*/ 801 1 + 32, 0, 2, 1 + 2, /*0xFE8E-0xFE91*/ 802 1, 0 + 32, 1 + 32, 0, /*0xFE92-0xFE95*/ 803 2, 1 + 2, 1, 0, /*0xFE96-0xFE99*/ 804 1, 0, 2, 1 + 2, /*0xFE9A-0xFE9D*/ 805 1, 0, 2, 1 + 2, /*0xFE9E-0xFEA1*/ 806 1, 0, 2, 1 + 2, /*0xFEA2-0xFEA5*/ 807 1, 0, 2, 1 + 2, /*0xFEA6-0xFEA9*/ 808 1, 0, 2, 1 + 2, /*0xFEAA-0xFEAD*/ 809 1, 0, 1, 0, /*0xFEAE-0xFEB1*/ 810 1, 0, 1, 0, /*0xFEB2-0xFEB5*/ 811 1, 0, 2, 1+2, /*0xFEB6-0xFEB9*/ 812 1, 0, 2, 1+2, /*0xFEBA-0xFEBD*/ 813 1, 0, 2, 1+2, /*0xFEBE-0xFEC1*/ 814 1, 0, 2, 1+2, /*0xFEC2-0xFEC5*/ 815 1, 0, 2, 1+2, /*0xFEC6-0xFEC9*/ 816 1, 0, 2, 1+2, /*0xFECA-0xFECD*/ 817 1, 0, 2, 1+2, /*0xFECE-0xFED1*/ 818 1, 0, 2, 1+2, /*0xFED2-0xFED5*/ 819 1, 0, 2, 1+2, /*0xFED6-0xFED9*/ 820 1, 0, 2, 1+2, /*0xFEDA-0xFEDD*/ 821 1, 0, 2, 1+2, /*0xFEDE-0xFEE1*/ 822 1, 0 + 16, 2 + 16, 1 + 2 +16, /*0xFEE2-0xFEE5*/ 823 1 + 16, 0, 2, 1+2, /*0xFEE6-0xFEE9*/ 824 1, 0, 2, 1+2, /*0xFEEA-0xFEED*/ 825 1, 0, 2, 1+2, /*0xFEEE-0xFEF1*/ 826 1, 0, 1, 0, /*0xFEF2-0xFEF5*/ 827 1, 0, 2, 1+2, /*0xFEF6-0xFEF9*/ 828 1, 0, 1, 0, /*0xFEFA-0xFEFD*/ 829 1, 0, 1, 0, 830 1 831 }; 832 833 private static int[] convertFEto06 = { 834 /***********0******1******2******3******4******5******6******7******8******9******A******B******C******D******E******F***/ 835 /*FE7*/ 0x64B, 0x64B, 0x64C, 0x64C, 0x64D, 0x64D, 0x64E, 0x64E, 0x64F, 0x64F, 0x650, 0x650, 0x651, 0x651, 0x652, 0x652, 836 /*FE8*/ 0x621, 0x622, 0x622, 0x623, 0x623, 0x624, 0x624, 0x625, 0x625, 0x626, 0x626, 0x626, 0x626, 0x627, 0x627, 0x628, 837 /*FE9*/ 0x628, 0x628, 0x628, 0x629, 0x629, 0x62A, 0x62A, 0x62A, 0x62A, 0x62B, 0x62B, 0x62B, 0x62B, 0x62C, 0x62C, 0x62C, 838 /*FEA*/ 0x62C, 0x62D, 0x62D, 0x62D, 0x62D, 0x62E, 0x62E, 0x62E, 0x62E, 0x62F, 0x62F, 0x630, 0x630, 0x631, 0x631, 0x632, 839 /*FEB*/ 0x632, 0x633, 0x633, 0x633, 0x633, 0x634, 0x634, 0x634, 0x634, 0x635, 0x635, 0x635, 0x635, 0x636, 0x636, 0x636, 840 /*FEC*/ 0x636, 0x637, 0x637, 0x637, 0x637, 0x638, 0x638, 0x638, 0x638, 0x639, 0x639, 0x639, 0x639, 0x63A, 0x63A, 0x63A, 841 /*FED*/ 0x63A, 0x641, 0x641, 0x641, 0x641, 0x642, 0x642, 0x642, 0x642, 0x643, 0x643, 0x643, 0x643, 0x644, 0x644, 0x644, 842 /*FEE*/ 0x644, 0x645, 0x645, 0x645, 0x645, 0x646, 0x646, 0x646, 0x646, 0x647, 0x647, 0x647, 0x647, 0x648, 0x648, 0x649, 843 /*FEF*/ 0x649, 0x64A, 0x64A, 0x64A, 0x64A, 0x65C, 0x65C, 0x65D, 0x65D, 0x65E, 0x65E, 0x65F, 0x65F 844 }; 845 846 private static final int shapeTable[][][] = { 847 { {0,0,0,0}, {0,0,0,0}, {0,1,0,3}, {0,1,0,1} }, 848 { {0,0,2,2}, {0,0,1,2}, {0,1,1,2}, {0,1,1,3} }, 849 { {0,0,0,0}, {0,0,0,0}, {0,1,0,3}, {0,1,0,3} }, 850 { {0,0,1,2}, {0,0,1,2}, {0,1,1,2}, {0,1,1,3} } 851 }; 852 853 /* 854 * This function shapes European digits to Arabic-Indic digits 855 * in-place, writing over the input characters. Data is in visual 856 * order. 857 */ shapeToArabicDigitsWithContext(char[] dest, int start, int length, char digitBase, boolean lastStrongWasAL)858 private void shapeToArabicDigitsWithContext(char[] dest, 859 int start, 860 int length, 861 char digitBase, 862 boolean lastStrongWasAL) { 863 digitBase -= '0'; // move common adjustment out of loop 864 865 for(int i = start + length; --i >= start;) { 866 char ch = dest[i]; 867 switch (Character.getDirectionality(ch)) { 868 case Character.DIRECTIONALITY_LEFT_TO_RIGHT: 869 case Character.DIRECTIONALITY_RIGHT_TO_LEFT: 870 lastStrongWasAL = false; 871 break; 872 case Character.DIRECTIONALITY_RIGHT_TO_LEFT_ARABIC: 873 lastStrongWasAL = true; 874 break; 875 case Character.DIRECTIONALITY_EUROPEAN_NUMBER: 876 if (lastStrongWasAL && ch <= '\u0039') { 877 dest[i] = (char)(ch + digitBase); 878 } 879 break; 880 default: 881 break; 882 } 883 } 884 } 885 886 /* 887 * Name : invertBuffer 888 * Function: This function inverts the buffer, it's used 889 * in case the user specifies the buffer to be 890 * TEXT_DIRECTION_LOGICAL 891 */ invertBuffer(char[] buffer, int start, int length)892 private static void invertBuffer(char[] buffer, 893 int start, 894 int length) { 895 896 for(int i = start, j = start + length - 1; i < j; i++, --j) { 897 char temp = buffer[i]; 898 buffer[i] = buffer[j]; 899 buffer[j] = temp; 900 } 901 } 902 903 /* 904 * Name : changeLamAlef 905 * Function: Converts the Alef characters into an equivalent 906 * LamAlef location in the 0x06xx Range, this is an 907 * intermediate stage in the operation of the program 908 * later it'll be converted into the 0xFExx LamAlefs 909 * in the shaping function. 910 */ changeLamAlef(char ch)911 private static char changeLamAlef(char ch) { 912 switch(ch) { 913 case '\u0622': return '\u065C'; 914 case '\u0623': return '\u065D'; 915 case '\u0625': return '\u065E'; 916 case '\u0627': return '\u065F'; 917 default: return '\u0000'; // not a lamalef 918 } 919 } 920 921 /* 922 * Name : specialChar 923 * Function: Special Arabic characters need special handling in the shapeUnicode 924 * function, this function returns 1 or 2 for these special characters 925 */ specialChar(char ch)926 private static int specialChar(char ch) { 927 if ((ch > '\u0621' && ch < '\u0626') || 928 (ch == '\u0627') || 929 (ch > '\u062E' && ch < '\u0633') || 930 (ch > '\u0647' && ch < '\u064A') || 931 (ch == '\u0629')) { 932 return 1; 933 } else if (ch >= '\u064B' && ch<= '\u0652') { 934 return 2; 935 } else if (ch >= 0x0653 && ch <= 0x0655 || 936 ch == 0x0670 || 937 ch >= 0xFE70 && ch <= 0xFE7F) { 938 return 3; 939 } else { 940 return 0; 941 } 942 } 943 944 /* 945 * Name : getLink 946 * Function: Resolves the link between the characters as 947 * Arabic characters have four forms : 948 * Isolated, Initial, Middle and Final Form 949 */ getLink(char ch)950 private static int getLink(char ch) { 951 if (ch >= '\u0622' && ch <= '\u06D3') { 952 return araLink[ch - '\u0622']; 953 } else if (ch == '\u200D') { 954 return 3; 955 } else if (ch >= '\u206D' && ch <= '\u206F') { 956 return 4; 957 } else if (ch >= '\uFE70' && ch <= '\uFEFC') { 958 return presLink[ch - '\uFE70']; 959 } else { 960 return 0; 961 } 962 } 963 964 /* 965 * Name : countSpaces 966 * Function: Counts the number of spaces 967 * at each end of the logical buffer 968 */ countSpacesLeft(char[] dest, int start, int count)969 private static int countSpacesLeft(char[] dest, 970 int start, 971 int count) { 972 for (int i = start, e = start + count; i < e; ++i) { 973 if (dest[i] != SPACE_CHAR) { 974 return i - start; 975 } 976 } 977 return count; 978 } 979 countSpacesRight(char[] dest, int start, int count)980 private static int countSpacesRight(char[] dest, 981 int start, 982 int count) { 983 984 for (int i = start + count; --i >= start;) { 985 if (dest[i] != SPACE_CHAR) { 986 return start + count - 1 - i; 987 } 988 } 989 return count; 990 } 991 992 /* 993 * Name : isTashkeelChar 994 * Function: Returns true for Tashkeel characters else return false 995 */ isTashkeelChar(char ch)996 private static boolean isTashkeelChar(char ch) { 997 return ( ch >='\u064B' && ch <= '\u0652' ); 998 } 999 1000 /* 1001 *Name : isSeenTailFamilyChar 1002 *Function : returns 1 if the character is a seen family isolated character 1003 * in the FE range otherwise returns 0 1004 */ 1005 isSeenTailFamilyChar(char ch)1006 private static int isSeenTailFamilyChar(char ch) { 1007 if (ch >= 0xfeb1 && ch < 0xfebf){ 1008 return tailFamilyIsolatedFinal [ch - 0xFEB1]; 1009 } else { 1010 return 0; 1011 } 1012 } 1013 1014 /* Name : isSeenFamilyChar 1015 * Function : returns 1 if the character is a seen family character in the Unicode 1016 * 06 range otherwise returns 0 1017 */ 1018 isSeenFamilyChar(char ch)1019 private static int isSeenFamilyChar(char ch){ 1020 if (ch >= 0x633 && ch <= 0x636){ 1021 return 1; 1022 }else { 1023 return 0; 1024 } 1025 } 1026 1027 /* 1028 *Name : isTailChar 1029 *Function : returns true if the character matches one of the tail characters 1030 * (0xfe73 or 0x200b) otherwise returns false 1031 */ 1032 isTailChar(char ch)1033 private static boolean isTailChar(char ch) { 1034 if(ch == OLD_TAIL_CHAR || ch == NEW_TAIL_CHAR){ 1035 return true; 1036 }else{ 1037 return false; 1038 } 1039 } 1040 1041 /* 1042 *Name : isAlefMaksouraChar 1043 *Function : returns true if the character is a Alef Maksoura Final or isolated 1044 * otherwise returns false 1045 */ isAlefMaksouraChar(char ch)1046 private static boolean isAlefMaksouraChar(char ch) { 1047 return ( (ch == 0xFEEF) || ( ch == 0xFEF0) || (ch == 0x0649)); 1048 } 1049 1050 /* 1051 * Name : isYehHamzaChar 1052 * Function : returns true if the character is a yehHamza isolated or yehhamza 1053 * final is found otherwise returns false 1054 */ isYehHamzaChar(char ch)1055 private static boolean isYehHamzaChar(char ch) { 1056 if((ch==0xFE89)||(ch==0xFE8A)){ 1057 return true; 1058 }else{ 1059 return false; 1060 } 1061 } 1062 1063 /* 1064 *Name : isTashkeelCharFE 1065 *Function : Returns true for Tashkeel characters in FE range else return false 1066 */ 1067 isTashkeelCharFE(char ch)1068 private static boolean isTashkeelCharFE(char ch) { 1069 return ( ch!=0xFE75 &&(ch>=0xFE70 && ch<= 0xFE7F) ); 1070 } 1071 1072 /* 1073 * Name: isTashkeelOnTatweelChar 1074 * Function: Checks if the Tashkeel Character is on Tatweel or not,if the 1075 * Tashkeel on tatweel (FE range), it returns 1 else if the 1076 * Tashkeel with shadda on tatweel (FC range)return 2 otherwise 1077 * returns 0 1078 */ isTashkeelOnTatweelChar(char ch)1079 private static int isTashkeelOnTatweelChar(char ch){ 1080 if (ch >= 0xfe70 && ch <= 0xfe7f && ch != NEW_TAIL_CHAR && ch != 0xFE75 && ch != SHADDA_TATWEEL_CHAR) 1081 { 1082 return tashkeelMedial [ch - 0xFE70]; 1083 } else if( (ch >= 0xfcf2 && ch <= 0xfcf4) || (ch == SHADDA_TATWEEL_CHAR)) { 1084 return 2; 1085 } else { 1086 return 0; 1087 } 1088 } 1089 1090 /* 1091 * Name: isIsolatedTashkeelChar 1092 * Function: Checks if the Tashkeel Character is in the isolated form 1093 * (i.e. Unicode FE range) returns 1 else if the Tashkeel 1094 * with shadda is in the isolated form (i.e. Unicode FC range) 1095 * returns 1 otherwise returns 0 1096 */ isIsolatedTashkeelChar(char ch)1097 private static int isIsolatedTashkeelChar(char ch){ 1098 if (ch >= 0xfe70 && ch <= 0xfe7f && ch != NEW_TAIL_CHAR && ch != 0xFE75){ 1099 return (1 - tashkeelMedial [ch - 0xFE70]); 1100 } else if(ch >= 0xfc5e && ch <= 0xfc63){ 1101 return 1; 1102 } else{ 1103 return 0; 1104 } 1105 } 1106 1107 /* 1108 * Name : isAlefChar 1109 * Function: Returns 1 for Alef characters else return 0 1110 */ isAlefChar(char ch)1111 private static boolean isAlefChar(char ch) { 1112 return ch == '\u0622' || ch == '\u0623' || ch == '\u0625' || ch == '\u0627'; 1113 } 1114 1115 /* 1116 * Name : isLamAlefChar 1117 * Function: Returns true for LamAlef characters else return false 1118 */ isLamAlefChar(char ch)1119 private static boolean isLamAlefChar(char ch) { 1120 return ch >= '\uFEF5' && ch <= '\uFEFC'; 1121 } 1122 isNormalizedLamAlefChar(char ch)1123 private static boolean isNormalizedLamAlefChar(char ch) { 1124 return ch >= '\u065C' && ch <= '\u065F'; 1125 } 1126 1127 /* 1128 * Name : calculateSize 1129 * Function: This function calculates the destSize to be used in preflighting 1130 * when the destSize is equal to 0 1131 */ calculateSize(char[] source, int sourceStart, int sourceLength)1132 private int calculateSize(char[] source, 1133 int sourceStart, 1134 int sourceLength) { 1135 1136 int destSize = sourceLength; 1137 1138 switch (options & LETTERS_MASK) { 1139 case LETTERS_SHAPE: 1140 case LETTERS_SHAPE_TASHKEEL_ISOLATED: 1141 if (isLogical) { 1142 for (int i = sourceStart, e = sourceStart + sourceLength - 1; i < e; ++i) { 1143 if ((source[i] == LAM_CHAR && isAlefChar(source[i+1])) || isTashkeelCharFE(source[i])){ 1144 --destSize; 1145 } 1146 } 1147 } else { // visual 1148 for(int i = sourceStart + 1, e = sourceStart + sourceLength; i < e; ++i) { 1149 if ((source[i] == LAM_CHAR && isAlefChar(source[i-1])) || isTashkeelCharFE(source[i])) { 1150 --destSize; 1151 } 1152 } 1153 } 1154 break; 1155 1156 case LETTERS_UNSHAPE: 1157 for(int i = sourceStart, e = sourceStart + sourceLength; i < e; ++i) { 1158 if (isLamAlefChar(source[i])) { 1159 destSize++; 1160 } 1161 } 1162 break; 1163 1164 default: 1165 break; 1166 } 1167 1168 return destSize; 1169 } 1170 1171 1172 /* 1173 * Name : countSpaceSub 1174 * Function: Counts number of times the subChar appears in the array 1175 */ countSpaceSub(char [] dest,int length, char subChar)1176 public static int countSpaceSub(char [] dest,int length, char subChar){ 1177 int i = 0; 1178 int count = 0; 1179 while (i < length) { 1180 if (dest[i] == subChar) { 1181 count++; 1182 } 1183 i++; 1184 } 1185 return count; 1186 } 1187 1188 /* 1189 * Name : shiftArray 1190 * Function: Shifts characters to replace space sub characters 1191 */ shiftArray(char [] dest,int start, int e, char subChar)1192 public static void shiftArray(char [] dest,int start, int e, char subChar){ 1193 int w = e; 1194 int r = e; 1195 while (--r >= start) { 1196 char ch = dest[r]; 1197 if (ch != subChar) { 1198 --w; 1199 if (w != r) { 1200 dest[w] = ch; 1201 } 1202 } 1203 } 1204 } 1205 1206 /* 1207 * Name : flipArray 1208 * Function: inverts array, so that start becomes end and vice versa 1209 */ flipArray(char [] dest, int start, int e, int w)1210 public static int flipArray(char [] dest, int start, int e, int w){ 1211 int r; 1212 if (w > start) { 1213 // shift, assume small buffer size so don't use arraycopy 1214 r = w; 1215 w = start; 1216 while (r < e) { 1217 dest[w++] = dest[r++]; 1218 } 1219 } else { 1220 w = e; 1221 } 1222 return w; 1223 } 1224 1225 /* 1226 * Name : handleTashkeelWithTatweel 1227 * Function : Replaces Tashkeel as following: 1228 * Case 1 :if the Tashkeel on tatweel, replace it with Tatweel. 1229 * Case 2 :if the Tashkeel aggregated with Shadda on Tatweel, replace 1230 * it with Shadda on Tatweel. 1231 * Case 3: if the Tashkeel is isolated replace it with Space. 1232 * 1233 */ handleTashkeelWithTatweel(char[] dest, int sourceLength)1234 private static int handleTashkeelWithTatweel(char[] dest, int sourceLength) { 1235 int i; 1236 for(i = 0; i < sourceLength; i++){ 1237 if((isTashkeelOnTatweelChar(dest[i]) == 1)){ 1238 dest[i] = TATWEEL_CHAR; 1239 }else if((isTashkeelOnTatweelChar(dest[i]) == 2)){ 1240 dest[i] = SHADDA_TATWEEL_CHAR; 1241 }else if((isIsolatedTashkeelChar(dest[i])==1) && dest[i] != SHADDA_CHAR){ 1242 dest[i] = SPACE_CHAR; 1243 } 1244 } 1245 return sourceLength; 1246 } 1247 1248 /* 1249 *Name : handleGeneratedSpaces 1250 *Function : The shapeUnicode function converts Lam + Alef into LamAlef + space, 1251 * and Tashkeel to space. 1252 * handleGeneratedSpaces function puts these generated spaces 1253 * according to the options the user specifies. LamAlef and Tashkeel 1254 * spaces can be replaced at begin, at end, at near or decrease the 1255 * buffer size. 1256 * 1257 * There is also Auto option for LamAlef and tashkeel, which will put 1258 * the spaces at end of the buffer (or end of text if the user used 1259 * the option SPACES_RELATIVE_TO_TEXT_BEGIN_END). 1260 * 1261 * If the text type was visual_LTR and the option 1262 * SPACES_RELATIVE_TO_TEXT_BEGIN_END was selected the END 1263 * option will place the space at the beginning of the buffer and 1264 * BEGIN will place the space at the end of the buffer. 1265 */ handleGeneratedSpaces(char[] dest, int start, int length)1266 private int handleGeneratedSpaces(char[] dest, 1267 int start, 1268 int length) { 1269 1270 int lenOptionsLamAlef = options & LAMALEF_MASK; 1271 int lenOptionsTashkeel = options & TASHKEEL_MASK; 1272 boolean lamAlefOn = false; 1273 boolean tashkeelOn = false; 1274 1275 if (!isLogical & !spacesRelativeToTextBeginEnd) { 1276 switch (lenOptionsLamAlef) { 1277 case LAMALEF_BEGIN: lenOptionsLamAlef = LAMALEF_END; break; 1278 case LAMALEF_END: lenOptionsLamAlef = LAMALEF_BEGIN; break; 1279 default: break; 1280 } 1281 switch (lenOptionsTashkeel){ 1282 case TASHKEEL_BEGIN: lenOptionsTashkeel = TASHKEEL_END; break; 1283 case TASHKEEL_END: lenOptionsTashkeel = TASHKEEL_BEGIN; break; 1284 default: break; 1285 } 1286 } 1287 1288 1289 if (lenOptionsLamAlef == LAMALEF_NEAR) { 1290 for (int i = start, e = i + length; i < e; ++i) { 1291 if (dest[i] == LAMALEF_SPACE_SUB) { 1292 dest[i] = SPACE_CHAR_FOR_LAMALEF; 1293 } 1294 } 1295 1296 } else { 1297 1298 final int e = start + length; 1299 int wL = countSpaceSub(dest, length, LAMALEF_SPACE_SUB); 1300 int wT = countSpaceSub(dest, length, TASHKEEL_SPACE_SUB); 1301 1302 if (lenOptionsLamAlef == LAMALEF_END){ 1303 lamAlefOn = true; 1304 } 1305 if (lenOptionsTashkeel == TASHKEEL_END){ 1306 tashkeelOn = true; 1307 } 1308 1309 1310 if (lamAlefOn && (lenOptionsLamAlef == LAMALEF_END)) { 1311 shiftArray(dest, start, e, LAMALEF_SPACE_SUB); 1312 while (wL > start) { 1313 dest[--wL] = SPACE_CHAR; 1314 } 1315 } 1316 1317 if (tashkeelOn && (lenOptionsTashkeel == TASHKEEL_END)){ 1318 shiftArray(dest, start, e, TASHKEEL_SPACE_SUB); 1319 while (wT > start) { 1320 dest[--wT] = SPACE_CHAR; 1321 } 1322 } 1323 1324 lamAlefOn = false; 1325 tashkeelOn = false; 1326 1327 if (lenOptionsLamAlef == LAMALEF_RESIZE){ 1328 lamAlefOn = true; 1329 } 1330 if (lenOptionsTashkeel == TASHKEEL_RESIZE){ 1331 tashkeelOn = true; 1332 } 1333 1334 if (lamAlefOn && (lenOptionsLamAlef == LAMALEF_RESIZE)){ 1335 shiftArray(dest, start, e, LAMALEF_SPACE_SUB); 1336 wL = flipArray(dest,start,e, wL); 1337 length = wL - start; 1338 } 1339 if (tashkeelOn && (lenOptionsTashkeel == TASHKEEL_RESIZE)) { 1340 shiftArray(dest, start, e, TASHKEEL_SPACE_SUB); 1341 wT = flipArray(dest,start,e, wT); 1342 length = wT - start; 1343 } 1344 1345 lamAlefOn = false; 1346 tashkeelOn = false; 1347 1348 if ((lenOptionsLamAlef == LAMALEF_BEGIN) || 1349 (lenOptionsLamAlef == LAMALEF_AUTO)){ 1350 lamAlefOn = true; 1351 } 1352 if (lenOptionsTashkeel == TASHKEEL_BEGIN){ 1353 tashkeelOn = true; 1354 } 1355 1356 if (lamAlefOn && ((lenOptionsLamAlef == LAMALEF_BEGIN)|| 1357 (lenOptionsLamAlef == LAMALEF_AUTO))) { // spaces at beginning 1358 shiftArray(dest, start, e, LAMALEF_SPACE_SUB); 1359 wL = flipArray(dest,start,e, wL); 1360 while (wL < e) { 1361 dest[wL++] = SPACE_CHAR; 1362 } 1363 } 1364 if(tashkeelOn && (lenOptionsTashkeel == TASHKEEL_BEGIN)){ 1365 shiftArray(dest, start, e, TASHKEEL_SPACE_SUB); 1366 wT = flipArray(dest,start,e, wT); 1367 while (wT < e) { 1368 dest[wT++] = SPACE_CHAR; 1369 } 1370 } 1371 } 1372 1373 return length; 1374 } 1375 1376 1377 /* 1378 *Name :expandCompositCharAtBegin 1379 *Function :Expands the LamAlef character to Lam and Alef consuming the required 1380 * space from beginning of the buffer. If the text type was visual_LTR 1381 * and the option SPACES_RELATIVE_TO_TEXT_BEGIN_END was selected 1382 * the spaces will be located at end of buffer. 1383 * If there are no spaces to expand the LamAlef, an exception is thrown. 1384 */ expandCompositCharAtBegin(char[] dest,int start, int length, int lacount)1385 private boolean expandCompositCharAtBegin(char[] dest,int start, int length, 1386 int lacount) { 1387 boolean spaceNotFound = false; 1388 1389 if (lacount > countSpacesRight(dest, start, length)) { 1390 spaceNotFound = true; 1391 return spaceNotFound; 1392 } 1393 for (int r = start + length - lacount, w = start + length; --r >= start;) { 1394 char ch = dest[r]; 1395 if (isNormalizedLamAlefChar(ch)) { 1396 dest[--w] = LAM_CHAR; 1397 dest[--w] = convertNormalizedLamAlef[ch - '\u065C']; 1398 } else { 1399 dest[--w] = ch; 1400 } 1401 } 1402 return spaceNotFound; 1403 1404 } 1405 1406 /* 1407 *Name : expandCompositCharAtEnd 1408 *Function : Expands the LamAlef character to Lam and Alef consuming the 1409 * required space from end of the buffer. If the text type was 1410 * Visual LTR and the option SPACES_RELATIVE_TO_TEXT_BEGIN_END 1411 * was used, the spaces will be consumed from begin of buffer. If 1412 * there are no spaces to expand the LamAlef, an exception is thrown. 1413 */ 1414 expandCompositCharAtEnd(char[] dest,int start, int length, int lacount)1415 private boolean expandCompositCharAtEnd(char[] dest,int start, int length, 1416 int lacount){ 1417 boolean spaceNotFound = false; 1418 1419 if (lacount > countSpacesLeft(dest, start, length)) { 1420 spaceNotFound = true; 1421 return spaceNotFound; 1422 } 1423 for (int r = start + lacount, w = start, e = start + length; r < e; ++r) { 1424 char ch = dest[r]; 1425 if (isNormalizedLamAlefChar(ch)) { 1426 dest[w++] = convertNormalizedLamAlef[ch - '\u065C']; 1427 dest[w++] = LAM_CHAR; 1428 } else { 1429 dest[w++] = ch; 1430 } 1431 } 1432 return spaceNotFound; 1433 } 1434 1435 /* 1436 *Name : expandCompositCharAtNear 1437 *Function : Expands the LamAlef character into Lam + Alef, YehHamza character 1438 * into Yeh + Hamza, SeenFamily character into SeenFamily character 1439 * + Tail, while consuming the space next to the character. 1440 */ 1441 expandCompositCharAtNear(char[] dest,int start, int length, int yehHamzaOption, int seenTailOption, int lamAlefOption)1442 private boolean expandCompositCharAtNear(char[] dest,int start, int length, 1443 int yehHamzaOption, int seenTailOption, int lamAlefOption){ 1444 1445 boolean spaceNotFound = false; 1446 1447 1448 1449 if (isNormalizedLamAlefChar(dest[start])) { 1450 spaceNotFound = true; 1451 return spaceNotFound; 1452 } 1453 for (int i = start + length; --i >=start;) { 1454 char ch = dest[i]; 1455 if (lamAlefOption == 1 && isNormalizedLamAlefChar(ch)) { 1456 if (i>start &&dest[i-1] == SPACE_CHAR) { 1457 dest[i] = LAM_CHAR; 1458 dest[--i] = convertNormalizedLamAlef[ch - '\u065C']; 1459 } else { 1460 spaceNotFound = true; 1461 return spaceNotFound; 1462 } 1463 }else if(seenTailOption == 1 && isSeenTailFamilyChar(ch) == 1){ 1464 if(i>start &&dest[i-1] == SPACE_CHAR){ 1465 dest[i-1] = tailChar; 1466 } else{ 1467 spaceNotFound = true; 1468 return spaceNotFound; 1469 } 1470 }else if(yehHamzaOption == 1 && isYehHamzaChar(ch)){ 1471 1472 if(i>start &&dest[i-1] == SPACE_CHAR){ 1473 dest[i] = yehHamzaToYeh[ch - YEH_HAMZAFE_CHAR]; 1474 dest[i-1] = HAMZAFE_CHAR; 1475 }else{ 1476 spaceNotFound = true; 1477 return spaceNotFound; 1478 } 1479 1480 1481 } 1482 } 1483 return false; 1484 1485 } 1486 1487 /* 1488 * Name : expandCompositChar 1489 * Function: LamAlef needs special handling as the LamAlef is 1490 * one character while expanding it will give two 1491 * characters Lam + Alef, so we need to expand the LamAlef 1492 * in near or far spaces according to the options the user 1493 * specifies or increase the buffer size. 1494 * Dest has enough room for the expansion if we are growing. 1495 * lamalef are normalized to the 'special characters' 1496 */ expandCompositChar(char[] dest, int start, int length, int lacount, int shapingMode)1497 private int expandCompositChar(char[] dest, 1498 int start, 1499 int length, 1500 int lacount, 1501 int shapingMode) throws ArabicShapingException { 1502 1503 int lenOptionsLamAlef = options & LAMALEF_MASK; 1504 int lenOptionsSeen = options & SEEN_MASK; 1505 int lenOptionsYehHamza = options & YEHHAMZA_MASK; 1506 boolean spaceNotFound = false; 1507 1508 if (!isLogical && !spacesRelativeToTextBeginEnd) { 1509 switch (lenOptionsLamAlef) { 1510 case LAMALEF_BEGIN: lenOptionsLamAlef = LAMALEF_END; break; 1511 case LAMALEF_END: lenOptionsLamAlef = LAMALEF_BEGIN; break; 1512 default: break; 1513 } 1514 } 1515 1516 if(shapingMode == 1){ 1517 if(lenOptionsLamAlef == LAMALEF_AUTO){ 1518 if(isLogical){ 1519 spaceNotFound = expandCompositCharAtEnd(dest, start, length, lacount); 1520 if(spaceNotFound){ 1521 spaceNotFound = expandCompositCharAtBegin(dest, start, length, lacount); 1522 } 1523 if(spaceNotFound){ 1524 spaceNotFound = expandCompositCharAtNear(dest, start, length,0,0,1); 1525 } 1526 if(spaceNotFound){ 1527 throw new ArabicShapingException("No spacefor lamalef"); 1528 } 1529 }else{ 1530 spaceNotFound = expandCompositCharAtBegin(dest, start, length, lacount); 1531 if(spaceNotFound){ 1532 spaceNotFound = expandCompositCharAtEnd(dest, start, length, lacount); 1533 } 1534 if(spaceNotFound){ 1535 spaceNotFound = expandCompositCharAtNear(dest, start, length,0,0,1); 1536 } 1537 if(spaceNotFound){ 1538 throw new ArabicShapingException("No spacefor lamalef"); 1539 } 1540 } 1541 }else if(lenOptionsLamAlef == LAMALEF_END){ 1542 spaceNotFound = expandCompositCharAtEnd(dest, start, length, lacount); 1543 if(spaceNotFound){ 1544 throw new ArabicShapingException("No spacefor lamalef"); 1545 } 1546 }else if(lenOptionsLamAlef == LAMALEF_BEGIN){ 1547 spaceNotFound = expandCompositCharAtBegin(dest, start, length, lacount); 1548 if(spaceNotFound){ 1549 throw new ArabicShapingException("No spacefor lamalef"); 1550 } 1551 }else if(lenOptionsLamAlef == LAMALEF_NEAR){ 1552 spaceNotFound = expandCompositCharAtNear(dest, start, length,0,0,1); 1553 if(spaceNotFound){ 1554 throw new ArabicShapingException("No spacefor lamalef"); 1555 } 1556 }else if(lenOptionsLamAlef == LAMALEF_RESIZE){ 1557 for (int r = start + length, w = r + lacount; --r >= start;) { 1558 char ch = dest[r]; 1559 if (isNormalizedLamAlefChar(ch)) { 1560 dest[--w] = '\u0644'; 1561 dest[--w] = convertNormalizedLamAlef[ch - '\u065C']; 1562 } else { 1563 dest[--w] = ch; 1564 } 1565 } 1566 length += lacount; 1567 } 1568 }else{ 1569 if(lenOptionsSeen == SEEN_TWOCELL_NEAR){ 1570 spaceNotFound = expandCompositCharAtNear(dest, start, length,0,1,0); 1571 if(spaceNotFound){ 1572 throw new ArabicShapingException("No space for Seen tail expansion"); 1573 } 1574 } 1575 if(lenOptionsYehHamza == YEHHAMZA_TWOCELL_NEAR){ 1576 spaceNotFound = expandCompositCharAtNear(dest, start, length,1,0,0); 1577 if(spaceNotFound){ 1578 throw new ArabicShapingException("No space for YehHamza expansion"); 1579 } 1580 } 1581 } 1582 return length; 1583 } 1584 1585 1586 /* Convert the input buffer from FExx Range into 06xx Range 1587 * to put all characters into the 06xx range 1588 * even the lamalef is converted to the special region in 1589 * the 06xx range. Return the number of lamalef chars found. 1590 */ normalize(char[] dest, int start, int length)1591 private int normalize(char[] dest, int start, int length) { 1592 int lacount = 0; 1593 for (int i = start, e = i + length; i < e; ++i) { 1594 char ch = dest[i]; 1595 if (ch >= '\uFE70' && ch <= '\uFEFC') { 1596 if (isLamAlefChar(ch)) { 1597 ++lacount; 1598 } 1599 dest[i] = (char)convertFEto06[ch - '\uFE70']; 1600 } 1601 } 1602 return lacount; 1603 } 1604 1605 /* 1606 * Name : deshapeNormalize 1607 * Function: Convert the input buffer from FExx Range into 06xx Range 1608 * even the lamalef is converted to the special region in the 06xx range. 1609 * According to the options the user enters, all seen family characters 1610 * followed by a tail character are merged to seen tail family character and 1611 * any yeh followed by a hamza character are merged to yehhamza character. 1612 * Method returns the number of lamalef chars found. 1613 */ deshapeNormalize(char[] dest, int start, int length)1614 private int deshapeNormalize(char[] dest, int start, int length) { 1615 int lacount = 0; 1616 int yehHamzaComposeEnabled = 0; 1617 int seenComposeEnabled = 0; 1618 1619 yehHamzaComposeEnabled = ((options&YEHHAMZA_MASK) == YEHHAMZA_TWOCELL_NEAR) ? 1 : 0; 1620 seenComposeEnabled = ((options&SEEN_MASK) == SEEN_TWOCELL_NEAR)? 1 : 0; 1621 1622 for (int i = start, e = i + length; i < e; ++i) { 1623 char ch = dest[i]; 1624 1625 if( (yehHamzaComposeEnabled == 1) && ((ch == HAMZA06_CHAR) || (ch == HAMZAFE_CHAR)) 1626 && (i < (length - 1)) && isAlefMaksouraChar(dest[i+1] )) { 1627 dest[i] = SPACE_CHAR; 1628 dest[i+1] = YEH_HAMZA_CHAR; 1629 } else if ( (seenComposeEnabled == 1) && (isTailChar(ch)) && (i< (length - 1)) 1630 && (isSeenTailFamilyChar(dest[i+1])==1) ) { 1631 dest[i] = SPACE_CHAR; 1632 } 1633 else if (ch >= '\uFE70' && ch <= '\uFEFC') { 1634 if (isLamAlefChar(ch)) { 1635 ++lacount; 1636 } 1637 dest[i] = (char)convertFEto06[ch - '\uFE70']; 1638 } 1639 } 1640 return lacount; 1641 } 1642 1643 /* 1644 * Name : shapeUnicode 1645 * Function: Converts an Arabic Unicode buffer in 06xx Range into a shaped 1646 * arabic Unicode buffer in FExx Range 1647 */ shapeUnicode(char[] dest, int start, int length, int destSize, int tashkeelFlag)1648 private int shapeUnicode(char[] dest, 1649 int start, 1650 int length, 1651 int destSize, 1652 int tashkeelFlag)throws ArabicShapingException { 1653 1654 int lamalef_count = normalize(dest, start, length); 1655 1656 // resolve the link between the characters. 1657 // Arabic characters have four forms: Isolated, Initial, Medial and Final. 1658 // Tashkeel characters have two, isolated or medial, and sometimes only isolated. 1659 // tashkeelFlag == 0: shape normally, 1: shape isolated, 2: don't shape 1660 1661 boolean lamalef_found = false, seenfam_found = false; 1662 boolean yehhamza_found = false, tashkeel_found = false; 1663 int i = start + length - 1; 1664 int currLink = getLink(dest[i]); 1665 int nextLink = 0; 1666 int prevLink = 0; 1667 int lastLink = 0; 1668 //int prevPos = i; 1669 int lastPos = i; 1670 int nx = -2; 1671 int nw = 0; 1672 1673 while (i >= 0) { 1674 // If high byte of currLink > 0 then there might be more than one shape 1675 if ((currLink & '\uFF00') > 0 || isTashkeelChar(dest[i])) { 1676 nw = i - 1; 1677 nx = -2; 1678 while (nx < 0) { // we need to know about next char 1679 if (nw == -1) { 1680 nextLink = 0; 1681 nx = Integer.MAX_VALUE; 1682 } else { 1683 nextLink = getLink(dest[nw]); 1684 if ((nextLink & IRRELEVANT) == 0) { 1685 nx = nw; 1686 } else { 1687 --nw; 1688 } 1689 } 1690 } 1691 1692 if (((currLink & ALEFTYPE) > 0) && ((lastLink & LAMTYPE) > 0)) { 1693 lamalef_found = true; 1694 char wLamalef = changeLamAlef(dest[i]); // get from 0x065C-0x065f 1695 if (wLamalef != '\u0000') { 1696 // replace alef by marker, it will be removed later 1697 dest[i] = '\uffff'; 1698 dest[lastPos] = wLamalef; 1699 i = lastPos; 1700 } 1701 1702 lastLink = prevLink; 1703 currLink = getLink(wLamalef); // requires '\u0000', unfortunately 1704 } 1705 if ((i > 0) && (dest[i-1] == SPACE_CHAR)) 1706 { 1707 if ( isSeenFamilyChar(dest[i]) == 1){ 1708 seenfam_found = true; 1709 } else if (dest[i] == YEH_HAMZA_CHAR) { 1710 yehhamza_found = true; 1711 } 1712 } 1713 else if(i==0){ 1714 if ( isSeenFamilyChar(dest[i]) == 1){ 1715 seenfam_found = true; 1716 } else if (dest[i] == YEH_HAMZA_CHAR) { 1717 yehhamza_found = true; 1718 } 1719 } 1720 1721 1722 // get the proper shape according to link ability of neighbors 1723 // and of character; depends on the order of the shapes 1724 // (isolated, initial, middle, final) in the compatibility area 1725 1726 int flag = specialChar(dest[i]); 1727 1728 int shape = shapeTable[nextLink & LINK_MASK] 1729 [lastLink & LINK_MASK] 1730 [currLink & LINK_MASK]; 1731 1732 if (flag == 1) { 1733 shape &= 0x1; 1734 } else if (flag == 2) { 1735 if (tashkeelFlag == 0 && 1736 ((lastLink & LINKL) != 0) && 1737 ((nextLink & LINKR) != 0) && 1738 dest[i] != '\u064C' && 1739 dest[i] != '\u064D' && 1740 !((nextLink & ALEFTYPE) == ALEFTYPE && 1741 (lastLink & LAMTYPE) == LAMTYPE)) { 1742 1743 shape = 1; 1744 } else { 1745 shape = 0; 1746 } 1747 } 1748 if (flag == 2) { 1749 if (tashkeelFlag == 2) { 1750 dest[i] = TASHKEEL_SPACE_SUB; 1751 tashkeel_found = true; 1752 } 1753 else{ 1754 dest[i] = (char)('\uFE70' + irrelevantPos[dest[i] - '\u064B'] + shape); 1755 } 1756 // else leave tashkeel alone 1757 } else { 1758 dest[i] = (char)('\uFE70' + (currLink >> 8) + shape); 1759 } 1760 } 1761 1762 // move one notch forward 1763 if ((currLink & IRRELEVANT) == 0) { 1764 prevLink = lastLink; 1765 lastLink = currLink; 1766 //prevPos = lastPos; 1767 lastPos = i; 1768 } 1769 1770 --i; 1771 if (i == nx) { 1772 currLink = nextLink; 1773 nx = -2; 1774 } else if (i != -1) { 1775 currLink = getLink(dest[i]); 1776 } 1777 } 1778 1779 // If we found a lam/alef pair in the buffer 1780 // call handleGeneratedSpaces to remove the spaces that were added 1781 1782 destSize = length; 1783 if (lamalef_found || tashkeel_found) { 1784 destSize = handleGeneratedSpaces(dest, start, length); 1785 } 1786 if (seenfam_found || yehhamza_found){ 1787 destSize = expandCompositChar(dest, start, destSize, lamalef_count, SHAPE_MODE); 1788 } 1789 return destSize; 1790 } 1791 1792 /* 1793 * Name : deShapeUnicode 1794 * Function: Converts an Arabic Unicode buffer in FExx Range into unshaped 1795 * arabic Unicode buffer in 06xx Range 1796 */ deShapeUnicode(char[] dest, int start, int length, int destSize)1797 private int deShapeUnicode(char[] dest, 1798 int start, 1799 int length, 1800 int destSize) throws ArabicShapingException { 1801 1802 int lamalef_count = deshapeNormalize(dest, start, length); 1803 1804 // If there was a lamalef in the buffer call expandLamAlef 1805 if (lamalef_count != 0) { 1806 // need to adjust dest to fit expanded buffer... !!! 1807 destSize = expandCompositChar(dest, start, length, lamalef_count,DESHAPE_MODE); 1808 } else { 1809 destSize = length; 1810 } 1811 1812 return destSize; 1813 } 1814 internalShape(char[] source, int sourceStart, int sourceLength, char[] dest, int destStart, int destSize)1815 private int internalShape(char[] source, 1816 int sourceStart, 1817 int sourceLength, 1818 char[] dest, 1819 int destStart, 1820 int destSize) throws ArabicShapingException { 1821 1822 if (sourceLength == 0) { 1823 return 0; 1824 } 1825 1826 if (destSize == 0) { 1827 if (((options & LETTERS_MASK) != LETTERS_NOOP) && 1828 ((options & LAMALEF_MASK) == LAMALEF_RESIZE)) { 1829 1830 return calculateSize(source, sourceStart, sourceLength); 1831 } else { 1832 return sourceLength; // by definition 1833 } 1834 } 1835 1836 // always use temp buffer 1837 char[] temp = new char[sourceLength * 2]; // all lamalefs requiring expansion 1838 System.arraycopy(source, sourceStart, temp, 0, sourceLength); 1839 1840 if (isLogical) { 1841 invertBuffer(temp, 0, sourceLength); 1842 } 1843 1844 int outputSize = sourceLength; 1845 1846 switch (options & LETTERS_MASK) { 1847 case LETTERS_SHAPE_TASHKEEL_ISOLATED: 1848 outputSize = shapeUnicode(temp, 0, sourceLength, destSize, 1); 1849 break; 1850 1851 case LETTERS_SHAPE: 1852 if( ((options&TASHKEEL_MASK)> 0) && 1853 ((options&TASHKEEL_MASK) !=TASHKEEL_REPLACE_BY_TATWEEL)) { 1854 /* Call the shaping function with tashkeel flag == 2 for removal of tashkeel */ 1855 outputSize = shapeUnicode(temp, 0, sourceLength, destSize, 2); 1856 }else { 1857 //default Call the shaping function with tashkeel flag == 1 */ 1858 outputSize = shapeUnicode(temp, 0, sourceLength, destSize, 0); 1859 1860 /*After shaping text check if user wants to remove tashkeel and replace it with tatweel*/ 1861 if( (options&TASHKEEL_MASK) == TASHKEEL_REPLACE_BY_TATWEEL){ 1862 outputSize = handleTashkeelWithTatweel(temp,sourceLength); 1863 } 1864 } 1865 break; 1866 1867 case LETTERS_UNSHAPE: 1868 outputSize = deShapeUnicode(temp, 0, sourceLength, destSize); 1869 break; 1870 1871 default: 1872 break; 1873 } 1874 1875 if (outputSize > destSize) { 1876 throw new ArabicShapingException("not enough room for result data"); 1877 } 1878 1879 if ((options & DIGITS_MASK) != DIGITS_NOOP) { 1880 char digitBase = '\u0030'; // European digits 1881 switch (options & DIGIT_TYPE_MASK) { 1882 case DIGIT_TYPE_AN: 1883 digitBase = '\u0660'; // Arabic-Indic digits 1884 break; 1885 1886 case DIGIT_TYPE_AN_EXTENDED: 1887 digitBase = '\u06f0'; // Eastern Arabic-Indic digits (Persian and Urdu) 1888 break; 1889 1890 default: 1891 break; 1892 } 1893 1894 switch (options & DIGITS_MASK) { 1895 case DIGITS_EN2AN: 1896 { 1897 int digitDelta = digitBase - '\u0030'; 1898 for (int i = 0; i < outputSize; ++i) { 1899 char ch = temp[i]; 1900 if (ch <= '\u0039' && ch >= '\u0030') { 1901 temp[i] += digitDelta; 1902 } 1903 } 1904 } 1905 break; 1906 1907 case DIGITS_AN2EN: 1908 { 1909 char digitTop = (char)(digitBase + 9); 1910 int digitDelta = '\u0030' - digitBase; 1911 for (int i = 0; i < outputSize; ++i) { 1912 char ch = temp[i]; 1913 if (ch <= digitTop && ch >= digitBase) { 1914 temp[i] += digitDelta; 1915 } 1916 } 1917 } 1918 break; 1919 1920 case DIGITS_EN2AN_INIT_LR: 1921 shapeToArabicDigitsWithContext(temp, 0, outputSize, digitBase, false); 1922 break; 1923 1924 case DIGITS_EN2AN_INIT_AL: 1925 shapeToArabicDigitsWithContext(temp, 0, outputSize, digitBase, true); 1926 break; 1927 1928 default: 1929 break; 1930 } 1931 } 1932 1933 if (isLogical) { 1934 invertBuffer(temp, 0, outputSize); 1935 } 1936 1937 System.arraycopy(temp, 0, dest, destStart, outputSize); 1938 1939 return outputSize; 1940 } 1941 1942 private static class ArabicShapingException extends RuntimeException { ArabicShapingException(String msg)1943 ArabicShapingException(String msg) { 1944 super(msg); 1945 } 1946 } 1947 } 1948