1. // This file was generated AUTOMATICALLY from a template file Wed Sep 15 03:12:50 PDT 2004

  2. 

  3. /* @(#)CharacterDataLatin1.java.template	1.5 03/07/26

  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(int ch) {

  56. 		char offset = (char)ch;

  57.         int props = A[offset];

  58.         return props;

  59.     }

  60. 

  61.     static int getType(int ch) {

  62.         int props = getProperties(ch);

  63.         return (props & 0x1F);

  64.     }

  65. 

  66.     static boolean isLowerCase(int ch) {

  67.         int type = getType(ch);

  68.         return (type == Character.LOWERCASE_LETTER);

  69.     }

  70. 

  71.     static boolean isUpperCase(int ch) {

  72.         int type = getType(ch);

  73.         return (type == Character.UPPERCASE_LETTER);

  74.     }

  75. 

  76.     static boolean isTitleCase(int ch) {

  77.         return false;

  78.     }

  79. 

  80.     static boolean isDigit(int ch) {

  81.         int type = getType(ch);

  82.         return (type == Character.DECIMAL_DIGIT_NUMBER);

  83.     }

  84. 

  85.     static boolean isDefined(int ch) {

  86.         int type = getType(ch);

  87.         return (type != Character.UNASSIGNED);

  88.     }

  89. 

  90.     static boolean isLetter(int ch) {

  91.         int type = getType(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)) >> type) & 1) != 0);

  97.     }

  98. 

  99.     static boolean isLetterOrDigit(int ch) {

  100.         int type = getType(ch);

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

  102.             (1 << Character.LOWERCASE_LETTER) |

  103.             (1 << Character.TITLECASE_LETTER) |

  104.             (1 << Character.MODIFIER_LETTER) |

  105.             (1 << Character.OTHER_LETTER) |

  106.             (1 << Character.DECIMAL_DIGIT_NUMBER)) >> type) & 1) != 0);

  107.     }

  108. 

  109.     static boolean isSpaceChar(int ch) {

  110.         int type = getType(ch);

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

  112.             (1 << Character.LINE_SEPARATOR) |

  113.             (1 << Character.PARAGRAPH_SEPARATOR)) >> type) & 1) != 0);

  114.     }

  115. 

  116. 

  117.     static boolean isJavaIdentifierStart(int ch) {

  118.         int props = getProperties(ch);

  119.         return ((props & 0x00007000) >= 0x00005000);

  120.     }

  121. 

  122.     static boolean isJavaIdentifierPart(int ch) {

  123.         int props = getProperties(ch);

  124.         return ((props & 0x00003000) != 0);

  125.     }

  126. 

  127.     static boolean isUnicodeIdentifierStart(int ch) {

  128.         int props = getProperties(ch);

  129.         return ((props & 0x00007000) == 0x00007000);

  130.     }

  131. 

  132.     static boolean isUnicodeIdentifierPart(int ch) {

  133.         int props = getProperties(ch);

  134.         return ((props & 0x00001000) != 0);

  135.     }

  136. 

  137.     static boolean isIdentifierIgnorable(int ch) {

  138.         int props = getProperties(ch);

  139.         return ((props & 0x00007000) == 0x00001000);

  140.     }

  141. 

  142.     static int toLowerCase(int ch) {

  143.         int mapChar = ch;

  144.         int val = getProperties(ch);

  145. 

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

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

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

  149.             mapChar = ch + offset;

  150.         }

  151.         return mapChar;

  152.     }

  153. 

  154.     static int toUpperCase(int ch) {

  155.         int mapChar = ch;

  156.         int val = getProperties(ch);

  157. 

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

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

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

  161.                 mapChar =  ch - offset;

  162.             } else if (ch == 0x00B5) {

  163.                 mapChar = 0x039C;

  164.             }

  165.         }

  166.         return mapChar;

  167.     }

  168. 

  169.     static int toTitleCase(int ch) {

  170.         return toUpperCase(ch);

  171.     }

  172. 

  173.     static int digit(int ch, int radix) {

  174.         int value = -1;

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

  176.             int val = getProperties(ch);

  177.             int kind = val & 0x1F;

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

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

  180.             }

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

  182.                 // Java supradecimal digit

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

  184.             }

  185.         }

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

  187.     }

  188. 

  189.     static int getNumericValue(int ch) {

  190.         int val = getProperties(ch);

  191.         int retval = -1;

  192. 

  193.         switch (val & 0xC00) {

  194.             default: // cannot occur

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

  196.                 retval = -1;

  197.                 break;

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

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

  200.                 break;

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

  202.                  retval = -2; 

  203.                  break;

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

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

  206.                 break;

  207.         }

  208.         return retval;

  209.     }

  210. 

  211.     static boolean isWhitespace(int ch) {

  212.         int props = getProperties(ch);

  213.         return ((props & 0x00007000) == 0x00004000);

  214.     }

  215. 

  216.     static byte getDirectionality(int ch) {

  217.         int val = getProperties(ch);

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

  219. 

  220.         if (directionality == 0xF ) {

  221.             directionality = -1;

  222.         }

  223.         return directionality;

  224.     }

  225. 

  226.     static boolean isMirrored(int ch) {

  227.         int props = getProperties(ch);

  228.         return ((props & 0x80000000) != 0);

  229.     }

  230. 

  231.     static int toUpperCaseEx(int ch) {

  232.         int mapChar = ch;

  233.         int val = getProperties(ch);

  234. 

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

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

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

  238.                 mapChar =  ch - offset;

  239.             }

  240.             else {

  241.                 switch(ch) {

  242.                     // map overflow characters

  243.                     case 0x00B5 : mapChar = 0x039C; break;

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

  245.                 }

  246.             }

  247.         }

  248.         return mapChar;

  249.     }

  250. 

  251.     static char[] sharpsMap = new char[] {'S', 'S'};

  252. 

  253.     static char[] toUpperCaseCharArray(int ch) {

  254.         char[] upperMap = {(char)ch};

  255.         if (ch == 0x00DF) {

  256.             upperMap = sharpsMap;

  257.         }

  258.         return upperMap;

  259.     }

  260. 

  261. 

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

  263.   // java GenerateCharacter -template ../../tools/GenerateCharacter/CharacterDataLatin1.java.template -spec ../../tools/GenerateCharacter/UnicodeData.txt -specialcasing ../../tools/GenerateCharacter/SpecialCasing.txt -o C:/BUILD_AREA/jdk1.5.0/control/build/windows-i586/gensrc/java/lang/CharacterDataLatin1.java -string -usecharforbyte -latin1 8

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

  265. 

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

  267.   static final String A_DATA =

  268.     "\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800"+

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

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

  271.     "\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F"+

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

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

  274.     "\030\uE800\025\uE800\026\u6800\030\u2800\031\u3800\030\u2800\024\u3800\030"+

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

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

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

  278.     "\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1"+

  279.     "\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202"+

  280.     "\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1\202\u7FE1"+

  281.     "\202\u7FE1\uE800\025\u6800\030\uE800\026\u6800\033\u6800\u5017\u6800\033\201"+

  282.     "\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2"+

  283.     "\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201"+

  284.     "\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2\201\u7FE2"+

  285.     "\201\u7FE2\201\u7FE2\201\u7FE2\uE800\025\u6800\031\uE800\026\u6800\031\u4800"+

  286.     "\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u5000\u100F"+

  287.     "\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800"+

  288.     "\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F"+

  289.     "\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800"+

  290.     "\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F\u4800\u100F"+

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

  292.     "\034\u6800\034\u6800\033\u6800\034\000\u7002\uE800\035\u6800\031\u6800\u1010"+

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

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

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

  296.     "\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202"+

  297.     "\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001"+

  298.     "\202\u7001\202\u7001\202\u7001\202\u7001\202\u7001\u6800\031\202\u7001\202"+

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

  300.     "\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201"+

  301.     "\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002"+

  302.     "\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\u6800"+

  303.     "\031\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002\201\u7002"+

  304.     "\u061D\u7002";

  305. 

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

  307. 

  308.     static {

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

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

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

  312.             int i = 0, j = 0;

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

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

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

  316.             }

  317.         }

  318. 

  319.     }        

  320. }

  321.