/interpreter/tags/at2-build060407/src/edu/vub/util/regexp/RETokenWordBoundary.java

  1/* gnu/regexp/RETokenWordBoundary.java
  2   Copyright (C) 2001, 2004 Free Software Foundation, Inc.
  3
  4This file is part of GNU Classpath.
  5
  6GNU Classpath is free software; you can redistribute it and/or modify
  7it under the terms of the GNU General Public License as published by
  8the Free Software Foundation; either version 2, or (at your option)
  9any later version.
 10
 11GNU Classpath is distributed in the hope that it will be useful, but
 12WITHOUT ANY WARRANTY; without even the implied warranty of
 13MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 14General Public License for more details.
 15
 16You should have received a copy of the GNU General Public License
 17along with GNU Classpath; see the file COPYING.  If not, write to the
 18Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
 1902110-1301 USA.
 20
 21Linking this library statically or dynamically with other modules is
 22making a combined work based on this library.  Thus, the terms and
 23conditions of the GNU General Public License cover the whole
 24combination.
 25
 26As a special exception, the copyright holders of this library give you
 27permission to link this library with independent modules to produce an
 28executable, regardless of the license terms of these independent
 29modules, and to copy and distribute the resulting executable under
 30terms of your choice, provided that you also meet, for each linked
 31independent module, the terms and conditions of the license of that
 32module.  An independent module is a module which is not derived from
 33or based on this library.  If you modify this library, you may extend
 34this exception to your version of the library, but you are not
 35obligated to do so.  If you do not wish to do so, delete this
 36exception statement from your version. */
 37
 38
 39package edu.vub.util.regexp;
 40
 41/**
 42 * Represents a combination lookahead/lookbehind for POSIX [:alnum:].
 43 */
 44final class RETokenWordBoundary extends REToken {
 45    private boolean negated;
 46    private int where;
 47    static final int BEGIN = 1;
 48    static final int END = 2;
 49
 50    RETokenWordBoundary(int subIndex, int where, boolean negated) {
 51	super(subIndex);
 52	this.where = where;
 53	this.negated = negated;
 54    }
 55
 56    int getMaximumLength() {
 57        return 0;
 58    }
 59
 60    
 61    boolean match(CharIndexed input, REMatch mymatch) {
 62	// Word boundary means input[index-1] was a word character
 63	// and input[index] is not, or input[index] is a word character
 64	// and input[index-1] was not
 65	//  In the string "one two three", these positions match:
 66	//  |o|n|e| |t|w|o| |t|h|r|e|e|
 67	//  ^     ^ ^     ^ ^         ^
 68	boolean after = false;  // is current character a letter or digit?
 69	boolean before = false; // is previous character a letter or digit?
 70	char ch;
 71
 72	// TODO: Also check REG_ANCHORINDEX vs. anchor
 73	if (((mymatch.eflags & RE.REG_ANCHORINDEX) != RE.REG_ANCHORINDEX) 
 74	    || (mymatch.offset + mymatch.index > mymatch.anchor)) {
 75	    if ((ch = input.charAt(mymatch.index - 1)) != CharIndexed.OUT_OF_BOUNDS) {
 76		before = Character.isLetterOrDigit(ch) || (ch == '_');
 77	    }
 78	}
 79
 80	if ((ch = input.charAt(mymatch.index)) != CharIndexed.OUT_OF_BOUNDS) {
 81	    after = Character.isLetterOrDigit(ch) || (ch == '_');
 82	}
 83
 84	// if (before) and (!after), we're at end (\>)
 85	// if (after) and (!before), we're at beginning (\<)
 86	boolean doNext = false;
 87
 88	if ((where & BEGIN) == BEGIN) {
 89	    doNext = after && !before;
 90	}
 91	if ((where & END) == END) {
 92	    doNext ^= before && !after;
 93	}
 94
 95	if (negated) doNext = !doNext;
 96
 97	return (doNext ? next(input, mymatch) : false);
 98    }
 99    
100    void dump(StringBuffer os) {
101	if (where == (BEGIN | END)) {
102	    os.append( negated ? "\\B" : "\\b" );
103	} else if (where == BEGIN) {
104	    os.append("\\<");
105	} else {
106	    os.append("\\>");
107	}
108    }
109}