/*
 *  gnu/regexp/REMatch.java
 *  Copyright (C) 1998-2001 Wes Biggs
 *
 *  This library is free software; you can redistribute it and/or modify
 *  it under the terms of the GNU Lesser General Public License as published
 *  by the Free Software Foundation; either version 2.1 of the License, or
 *  (at your option) any later version.
 *
 *  This library is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU Lesser General Public License for more details.
 *
 *  You should have received a copy of the GNU Lesser General Public License
 *  along with this program; if not, write to the Free Software
 *  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */

package gnu.regexp;
import java.io.Serializable;

/**
 * An instance of this class represents a match
 * completed by a gnu.regexp matching function. It can be used
 * to obtain relevant information about the location of a match
 * or submatch.
 *
 * @author <A HREF="mailto:wes@cacas.org">Wes Biggs</A>
 */
public final class REMatch implements Serializable, Cloneable {
    private String m_match;

    // These variables are package scope for fast access within the engine
    int eflags; // execution flags this match was made using
    int offset; // offset in source text where match was tried
    int anchor; // anchor position, for ANCHORINDEX option

    int[] start; // start positions (relative to offset) for each (sub)exp.
    int[] end;   // end positions for the same
    int index; // used while matching to mark current match position in input
    REMatch next; // other possibility (to avoid having to use arrays)

    public Object clone() {
	try {
	    REMatch copy = (REMatch) super.clone();
	    copy.next = null;

	    copy.start = new int [start.length];
	    System.arraycopy(start,0,copy.start,0,start.length);

	    copy.end = new int [end.length];
	    System.arraycopy(end,0,copy.end,0,end.length);

	    return copy;
	} catch (CloneNotSupportedException e) {
	    throw new Error(); // doesn't happen
	}
    }

    void assignFrom(REMatch other) {
	start = other.start;
	end = other.end;
	index = other.index;
	next = other.next;
    }

    REMatch(int f_subs, int f_index, int f_eflags) {
	start = new int[f_subs+1];
	end = new int[f_subs+1];
	anchor = f_index;
	eflags = f_eflags;
	clear(f_index);
    }

    void finish(CharIndexed text) {
	start[0] = 0;
	StringBuffer sb = new StringBuffer();
	int i;
	for (i = 0; i < end[0]; i++)
	    sb.append(text.charAt(i));
	m_match = sb.toString();
	for (i = 0; i < start.length; i++) {
	    // If any subexpressions didn't terminate, they don't count
	    // TODO check if this code ever gets hit
	    if ((start[i] == -1) ^ (end[i] == -1)) {
		start[i] = -1;
		end[i] = -1;
	    }
	}
	next = null; // cut off alternates
    }
    
    /** Clears the current match and moves the offset to the new index. */
    void clear(int f_index) {
	offset = f_index;
	index = 0;
	for (int i = 0; i < start.length; i++) {
	    start[i] = end[i] = -1;
	}
	next = null; // cut off alternates
    }
    
    /**
     * Returns the string matching the pattern.  This makes it convenient
     * to write code like the following:
     * <P>
     * <code> 
     * REMatch myMatch = myExpression.getMatch(myString);<br>
     * if (myMatch != null) System.out.println("Regexp found: "+myMatch);
     * </code>
     */
    public String toString() {
	return m_match;
    }
    
    /**
     * Returns the index within the input text where the match in its entirety
     * began.
     */
    public int getStartIndex() {
	return offset + start[0];
    }
    
    /**
     * Returns the index within the input string where the match in
     * its entirety ends.  The return value is the next position after
     * the end of the string; therefore, a match created by the
     * following call:
     *
     * <P>
     * <code>REMatch myMatch = myExpression.getMatch(myString);</code>
     * <P>
     * can be viewed (given that myMatch is not null) by creating
     * <P>
     * <code>String theMatch = myString.substring(myMatch.getStartIndex(),
     * myMatch.getEndIndex());</code>
     * <P>
     * But you can save yourself that work, since the <code>toString()</code>
     * method (above) does exactly that for you.  
     */
    public int getEndIndex() {
	return offset + end[0];
    }
  
    /**
     * Returns the string matching the given subexpression.  The subexpressions
     * are indexed starting with one, not zero.  That is, the subexpression
     * identified by the first set of parentheses in a regular expression
     * could be retrieved from an REMatch by calling match.toString(1).
     *
     * @param sub Index of the subexpression.
     */
    public String toString(int sub) {
	if ((sub >= start.length) || (start[sub] == -1)) return "";
	return (m_match.substring(start[sub],end[sub]));
    }
    
    /** 
     * Returns the index within the input string used to generate this match
     * where subexpression number <i>sub</i> begins, or <code>-1</code> if
     * the subexpression does not exist.  The initial position is zero.
     *
     * @param sub Subexpression index
     * @deprecated Use getStartIndex(int) instead.
     */
    public int getSubStartIndex(int sub) {
	if (sub >= start.length) return -1;
	int x = start[sub];
	return (x == -1) ? x : offset + x;
    }
    
    /** 
     * Returns the index within the input string used to generate this match
     * where subexpression number <i>sub</i> begins, or <code>-1</code> if
     * the subexpression does not exist.  The initial position is zero.
     *
     * @param sub Subexpression index
     * @since gnu.regexp 1.1.0
     */
    public int getStartIndex(int sub) {
	if (sub >= start.length) return -1;
	int x = start[sub];
	return (x == -1) ? x : offset + x;
    }
  
    /** 
     * Returns the index within the input string used to generate this match
     * where subexpression number <i>sub</i> ends, or <code>-1</code> if
     * the subexpression does not exist.  The initial position is zero.
     *
     * @param sub Subexpression index
     * @deprecated Use getEndIndex(int) instead
     */
    public int getSubEndIndex(int sub) {
	if (sub >= start.length) return -1;
	int x = end[sub];
	return (x == -1) ? x : offset + x;
    }
    
    /** 
     * Returns the index within the input string used to generate this match
     * where subexpression number <i>sub</i> ends, or <code>-1</code> if
     * the subexpression does not exist.  The initial position is zero.
     *
     * @param sub Subexpression index
     */
    public int getEndIndex(int sub) {
	if (sub >= start.length) return -1;
	int x = end[sub];
	return (x == -1) ? x : offset + x;
    }
    
    /**
     * Substitute the results of this match to create a new string.
     * This is patterned after PERL, so the tokens to watch out for are
     * <code>$0</code> through <code>$9</code>.  <code>$0</code> matches
     * the full substring matched; <code>$<i>n</i></code> matches
     * subexpression number <i>n</i>.
     *
     * @param input A string consisting of literals and <code>$<i>n</i></code> tokens.
     */
    public String substituteInto(String input) {
	// a la Perl, $0 is whole thing, $1 - $9 are subexpressions
	StringBuffer output = new StringBuffer();
	int pos;
	for (pos = 0; pos < input.length()-1; pos++) {
	    if ((input.charAt(pos) == '$') && (Character.isDigit(input.charAt(pos+1)))) {
		int val = Character.digit(input.charAt(++pos),10);
		if (val < start.length) {
		    output.append(toString(val));
		} 
	    } else output.append(input.charAt(pos));
	}
	if (pos < input.length()) output.append(input.charAt(pos));
	return output.toString();
    }
}