1. // This file was generated AUTOMATICALLY from a template file Sun Feb 22 22:52:18 PST 2004

  2. 

  3. /* @(#)CharacterDataLatin1.java.template	1.2 02/11/11

  4.  *

  5.  * Copyright 1994-2002 Sun Microsystems, Inc. All Rights Reserved.

  6.  *

  7.  * This software is the proprietary information of Sun Microsystems, Inc.

  8.  * Use is subject to license terms.

  9.  *

  10.  */

  11. 

  12. package java.lang;

  13. 

  14. /** The CharacterData class encapsulates the large tables found in

  15.     Java.lang.Character. */

  16. 

  17. class CharacterDataLatin1 {

  18. 

  19.     /* The character properties are currently encoded into 32 bits in the following manner:

  20.         1 bit   mirrored property

  21.         4 bits  directionality property

  22.         9 bits  signed offset used for converting case

  23.         1 bit   if 1, adding the signed offset converts the character to lowercase

  24.         1 bit   if 1, subtracting the signed offset converts the character to uppercase

  25.         1 bit   if 1, this character has a titlecase equivalent (possibly itself)

  26.         3 bits  0  may not be part of an identifier

  27.                 1  ignorable control; may continue a Unicode identifier or Java identifier

  28.                 2  may continue a Java identifier but not a Unicode identifier (unused)

  29.                 3  may continue a Unicode identifier or Java identifier

  30.                 4  is a Java whitespace character

  31.                 5  may start or continue a Java identifier;

  32.                    may continue but not start a Unicode identifier (underscores)

  33.                 6  may start or continue a Java identifier but not a Unicode identifier ($)

  34.                 7  may start or continue a Unicode identifier or Java identifier

  35.                 Thus:

  36.                    5, 6, 7 may start a Java identifier

  37.                    1, 2, 3, 5, 6, 7 may continue a Java identifier

  38.                    7 may start a Unicode identifier

  39.                    1, 3, 5, 7 may continue a Unicode identifier

  40.                    1 is ignorable within an identifier

  41.                    4 is Java whitespace

  42.         2 bits  0  this character has no numeric property

  43.                 1  adding the digit offset to the character code and then

  44.                    masking with 0x1F will produce the desired numeric value

  45.                 2  this character has a "strange" numeric value

  46.                 3  a Java supradecimal digit: adding the digit offset to the

  47.                    character code, then masking with 0x1F, then adding 10

  48.                    will produce the desired numeric value

  49.         5 bits  digit offset

  50.         5 bits  character type

  51. 

  52.         The encoding of character properties is subject to change at any time.

  53.      */

  54. 

  55.     static int getProperties(char ch) {

  56.         return A[ch];

  57.     }

  58. 

  59.     static int getType(char ch) {

  60.         return getProperties(ch) & 0x1F;

  61.     }

  62. 

  63.     static boolean isLowerCase(char ch) {

  64.         return getType(ch) == Character.LOWERCASE_LETTER;

  65.     }

  66. 

  67.     static boolean isUpperCase(char ch) {

  68.         return getType(ch) == Character.UPPERCASE_LETTER;

  69.     }

  70. 

  71.     static boolean isTitleCase(char ch) {

  72.         return false;

  73.     }

  74. 

  75.     static boolean isDigit(char ch) {

  76.         return getType(ch) == Character.DECIMAL_DIGIT_NUMBER;

  77.     }

  78. 

  79.     static boolean isDefined(char ch) {

  80.         return getType(ch) != Character.UNASSIGNED;

  81.     }

  82. 

  83.     static boolean isLetter(char ch) {

  84.         return (((((1 << Character.UPPERCASE_LETTER) |

  85.             (1 << Character.LOWERCASE_LETTER) |

  86.             (1 << Character.TITLECASE_LETTER) |

  87.             (1 << Character.MODIFIER_LETTER) |

  88.             (1 << Character.OTHER_LETTER)) >> getType(ch)) & 1) != 0);

  89.     }

  90. 

  91.     static boolean isLetterOrDigit(char ch) {

  92.         return (((((1 << Character.UPPERCASE_LETTER) |

  93.             (1 << Character.LOWERCASE_LETTER) |

  94.             (1 << Character.TITLECASE_LETTER) |

  95.             (1 << Character.MODIFIER_LETTER) |

  96.             (1 << Character.OTHER_LETTER) |

  97.             (1 << Character.DECIMAL_DIGIT_NUMBER)) >> getType(ch)) & 1) != 0);

  98.     }

  99. 

  100.     static boolean isSpaceChar(char ch) {

  101.         return (((((1 << Character.SPACE_SEPARATOR) |

  102.                    (1 << Character.LINE_SEPARATOR) |

  103.                    (1 << Character.PARAGRAPH_SEPARATOR))

  104.                 >> getType(ch)) & 1) != 0);

  105.     }

  106. 

  107. 

  108.     static boolean isJavaIdentifierStart(char ch) {

  109.         return (getProperties(ch) & 0x00007000) >= 0x00005000;

  110.     }

  111. 

  112.     static boolean isJavaIdentifierPart(char ch) {

  113.         return (getProperties(ch) & 0x00003000) != 0;

  114.     }

  115. 

  116.     static boolean isUnicodeIdentifierStart(char ch) {

  117.         return (getProperties(ch) & 0x00007000) == 0x00007000;

  118.     }

  119. 

  120.     static boolean isUnicodeIdentifierPart(char ch) {

  121.         return (getProperties(ch)& 0x00001000) != 0;

  122.     }

  123. 

  124.     static boolean isIdentifierIgnorable(char ch) {

  125.         return (getProperties(ch) & 0x00007000) == 0x00001000;

  126.     }

  127. 

  128.     static char toLowerCase(char ch) {

  129.         char mapChar = ch;

  130.         int val = getProperties(ch);

  131. 

  132.         if (((val & 0x00020000) != 0) && 

  133.                 ((val & 0x07FC0000) != 0x07FC0000)) { 

  134.             int offset = val << 5 >> (5+18);

  135.             mapChar = (char)(ch + offset);

  136.         }

  137.         return mapChar;

  138.     }

  139. 

  140.     static char toUpperCase(char ch) {

  141.         char mapChar = ch;

  142.         int val = getProperties(ch);

  143. 

  144.         if ((val & 0x00010000) != 0) {

  145.             if ((val & 0x07FC0000) != 0x07FC0000) {

  146.                 int offset = val  << 5 >> (5+18);

  147.                 mapChar =  (char)(ch - offset);

  148.             } else if (ch == '\u00B5') {

  149.                 mapChar = '\u039C';

  150.             }

  151.         }

  152.         return mapChar;

  153.     }

  154. 

  155.     static char toTitleCase(char ch) {

  156.         return toUpperCase(ch);

  157.     }

  158. 

  159.     static int digit(char ch, int radix) {

  160.         int value = -1;

  161.         if (radix >= Character.MIN_RADIX && radix <= Character.MAX_RADIX) {

  162.             int val = getProperties(ch);

  163.             int kind = val & 0x1F;

  164.             if (kind == Character.DECIMAL_DIGIT_NUMBER) {

  165.                 value = ch + ((val & 0x3E0) >> 5) & 0x1F;

  166.             }

  167.             else if ((val & 0xC00) == 0x00000C00) {

  168.                 // Java supradecimal digit

  169.                 value = (ch + ((val & 0x3E0) >> 5) & 0x1F) + 10;

  170.             }

  171.         }

  172.         return (value < radix) ? value : -1;

  173.     }

  174. 

  175.     static int getNumericValue(char ch) {

  176.         int val = getProperties(ch);

  177.         int retval = -1;

  178. 

  179.         switch (val & 0xC00) {

  180.             default: // cannot occur

  181.             case (0x00000000):         // not numeric

  182.                 retval = -1;

  183.                 break;

  184.             case (0x00000400):              // simple numeric

  185.                 retval = ch + ((val & 0x3E0) >> 5) & 0x1F;

  186.                 break;

  187.             case (0x00000800)      :       // "strange" numeric

  188.                  retval = -2; 

  189.                  break;

  190.             case (0x00000C00):           // Java supradecimal

  191.                 retval = (ch + ((val & 0x3E0) >> 5) & 0x1F) + 10;

  192.                 break;

  193.         }

  194.         return retval;

  195.     }

  196. 

  197.     static boolean isWhitespace(char ch) {

  198.         return (getProperties(ch) & 0x00007000) == 0x00004000;

  199.     }

  200. 

  201.     static byte getDirectionality(char ch) {

  202.         int val = getProperties(ch);

  203.         byte directionality = (byte)((val & 0x78000000) >> 27);

  204. 

  205.         if (directionality == 0xF ) {

  206.             directionality = -1;

  207.         }

  208.         return directionality;

  209.     }

  210. 

  211.     static boolean isMirrored(char ch) {

  212.         return (getProperties(ch) & 0x80000000) != 0;

  213.     }

  214. 

  215.     static char toUpperCaseEx(char ch) {

  216.         char mapChar = ch;

  217.         int val = getProperties(ch);

  218. 

  219.         if ((val & 0x00010000) != 0) {

  220.             if ((val & 0x07FC0000) != 0x07FC0000) {

  221.                 int offset = val  << 5 >> (5+18);

  222.                 mapChar =  (char)(ch - offset);

  223.             }

  224.             else {

  225.                 switch(ch) {

  226.                     // map overflow characters

  227.                     case '\u00B5' : mapChar = '\u039C'; break;

  228.                     default       : mapChar = Character.CHAR_ERROR; break;

  229.                 }

  230.             }

  231.         }

  232.         return mapChar;

  233.     }

  234. 

  235.     // The following tables and code generated using:

  236.   // java GenerateCharacter -template ../../tools/GenerateCharacter/CharacterDataLatin1.java.template -spec ../../tools/GenerateCharacter/UnicodeData.txt -specialcasing ../../tools/GenerateCharacter/SpecialCasing.txt -o D:/BUILD_AREA/jdk142-update/ws/fcs/control/build/windows-i586/gensrc/java/lang/CharacterDataLatin1.java -string -usecharforbyte -latin1 8

  237.   // The A table has 256 entries for a total of 1024 bytes.

  238. 

  239.   static final int A[] = new int[256];

  240.   static final String A_DATA =

  241.     "\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800"+

  242.     "\u100F\u4800\u100F\u4800\u100F\u5800\u400F\u5000\u400F\u5800\u400F\u6000\u400F"+

  243.     "\u5000\u400F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800"+

  244.     "\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F"+

  245.     "\u4800\u100F\u4800\u100F\u5000\u400F\u5000\u400F\u5000\u400F\u5800\u400F\u6000"+

  246.     "\u400C\u6800\030\u6800\030\u2800\030\u2800\u601A\u2800\030\u6800\030\u6800"+

  247.     "\030\uE800\025\uE800\026\u6800\030\u2800\031\u3800\030\u2800\024\u3800\030"+

  248.     "\u2000\030\u1800\u3609\u1800\u3609\u1800\u3609\u1800\u3609\u1800\u3609\u1800"+

  249.     "\u3609\u1800\u3609\u1800\u3609\u1800\u3609\u1800\u3609\u3800\030\u6800\030"+

  250.     "\uE800\031\u6800\031\uE800\031\u6800\030\u6800\030\202\u7FE1\202\u7FE1\202"+

  251.     "\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1"+

  252.     "\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202"+

  253.     "\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1"+

  254.     "\202\u7FE1\uE800\025\u6800\030\uE800\026\u6800\033\u6800\u5017\u6800\033\201"+

  255.     "\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2"+

  256.     "\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201"+

  257.     "\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2"+

  258.     "\201\u7FE2\201\u7FE2\201\u7FE2\uE800\025\u6800\031\uE800\026\u6800\031\u4800"+

  259.     "\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u5000\u100F"+

  260.     "\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800"+

  261.     "\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F"+

  262.     "\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800"+

  263.     "\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F"+

  264.     "\u3800\014\u6800\030\u2800\u601A\u2800\u601A\u2800\u601A\u2800\u601A\u6800"+

  265.     "\034\u6800\034\u6800\033\u6800\034\000\u7002\uE800\035\u6800\031\u6800\024"+

  266.     "\u6800\034\u6800\033\u2800\034\u2800\031\u1800\u060B\u1800\u060B\u6800\033"+

  267.     "\u07FD\u7002\u6800\034\u6800\030\u6800\033\u1800\u050B\000\u7002\uE800\036"+

  268.     "\u6800\u080B\u6800\u080B\u6800\u080B\u6800\030\202\u7001\202\u7001\202\u7001"+

  269.     "\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202"+

  270.     "\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001"+

  271.     "\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\u6800\031\202\u7001\202"+

  272.     "\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\u07FD\u7002\201\u7002"+

  273.     "\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201"+

  274.     "\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002"+

  275.     "\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\u6800"+

  276.     "\031\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002"+

  277.     "\u061D\u7002";

  278. 

  279.   // In all, the character property tables require 1024 bytes.

  280. 

  281.     static {

  282.                 { // THIS CODE WAS AUTOMATICALLY CREATED BY GenerateCharacter:

  283.             char[] data = A_DATA.toCharArray();

  284.             assert (data.length == (256 * 2));

  285.             int i = 0, j = 0;

  286.             while (i < (256 * 2)) {

  287.                 int entry = data[i++] << 16;

  288.                 A[j++] = entry | data[i++];

  289.             }

  290.         }

  291. 

  292.     }        

  293. }

  294.