FileDocCategorySizeDatePackage
SetOfIntegerSyntax.javaAPI DocJava SE 5 API17254Fri Aug 26 14:57:44 BST 2005javax.print.attribute

SetOfIntegerSyntax

public abstract class SetOfIntegerSyntax extends Object implements Serializable, Cloneable
Class SetOfIntegerSyntax is an abstract base class providing the common implementation of all attributes whose value is a set of nonnegative integers. This includes attributes whose value is a single range of integers and attributes whose value is a set of ranges of integers.

You can construct an instance of SetOfIntegerSyntax by giving it in "string form." The string consists of zero or more comma-separated integer groups. Each integer group consists of either one integer, two integers separated by a hyphen (-), or two integers separated by a colon (:). Each integer consists of one or more decimal digits (0 through 9). Whitespace characters cannot appear within an integer but are otherwise ignored. For example: "", "1", "5-10", "1:2, 4".

You can also construct an instance of SetOfIntegerSyntax by giving it in "array form." Array form consists of an array of zero or more integer groups where each integer group is a length-1 or length-2 array of ints; for example, int[0][], int[][]{{1}}, int[][]{{5,10}}, int[][]{{1,2},{4}}.

In both string form and array form, each successive integer group gives a range of integers to be included in the set. The first integer in each group gives the lower bound of the range; the second integer in each group gives the upper bound of the range; if there is only one integer in the group, the upper bound is the same as the lower bound. If the upper bound is less than the lower bound, it denotes a null range (no values). If the upper bound is equal to the lower bound, it denotes a range consisting of a single value. If the upper bound is greater than the lower bound, it denotes a range consisting of more than one value. The ranges may appear in any order and are allowed to overlap. The union of all the ranges gives the set's contents. Once a SetOfIntegerSyntax instance is constructed, its value is immutable.

The SetOfIntegerSyntax object's value is actually stored in "canonical array form." This is the same as array form, except there are no null ranges; the members of the set are represented in as few ranges as possible (i.e., overlapping ranges are coalesced); the ranges appear in ascending order; and each range is always represented as a length-two array of ints in the form {lower bound, upper bound}. An empty set is represented as a zero-length array.

Class SetOfIntegerSyntax has operations to return the set's members in canonical array form, to test whether a given integer is a member of the set, and to iterate through the members of the set.

author
David Mendenhall
author
Alan Kaminsky

Fields Summary
private static final long
serialVersionUID
private int[]
members
This set's members in canonical array form.
Constructors Summary
protected SetOfIntegerSyntax(String members)
Construct a new set-of-integer attribute with the given members in string form.

param
members Set members in string form. If null, an empty set is constructed.
exception
IllegalArgumentException (Unchecked exception) Thrown if members does not obey the proper syntax.



                                                                                  
       
	this.members = parse (members);
    
protected SetOfIntegerSyntax(int[] members)
Construct a new set-of-integer attribute with the given members in array form.

param
members Set members in array form. If null, an empty set is constructed.
exception
NullPointerException (Unchecked exception) Thrown if any element of members is null.
exception
IllegalArgumentException (Unchecked exception) Thrown if any element of members is not a length-one or length-two array or if any non-null range in members has a lower bound less than zero.

	this.members = parse (members);
    
protected SetOfIntegerSyntax(int member)
Construct a new set-of-integer attribute containing a single integer.

param
member Set member.
exception
IllegalArgumentException (Unchecked exception) Thrown if member is less than zero.

	if (member < 0) {
	    throw new IllegalArgumentException();
	}
	members = new int[][] {{member, member}};
    
protected SetOfIntegerSyntax(int lowerBound, int upperBound)
Construct a new set-of-integer attribute containing a single range of integers. If the lower bound is greater than the upper bound (a null range), an empty set is constructed.

param
lowerBound Lower bound of the range.
param
upperBound Upper bound of the range.
exception
IllegalArgumentException (Unchecked exception) Thrown if the range is non-null and lowerBound is less than zero.

	if (lowerBound <= upperBound && lowerBound < 0) {
	    throw new IllegalArgumentException();
	}
	members = lowerBound <=upperBound ?
	    new int[][] {{lowerBound, upperBound}} :
	    new int[0][];
    
Methods Summary
private static voidaccumulate(java.util.Vector ranges, int lb, int ub)
Accumulate the given range (lb .. ub) into the canonical array form into the given vector of int[] objects.

	// Make sure range is non-null.
	if (lb <= ub) {
	    // Stick range at the back of the vector.
	    ranges.add(new int[] {lb, ub});

	    // Work towards the front of the vector to integrate the new range 
	    // with the existing ranges. 
	    for (int j = ranges.size()-2; j >= 0; -- j) {
	    // Get lower and upper bounds of the two ranges being compared.
		int[] rangea = (int[]) ranges.elementAt (j);
		int lba = rangea[0];
		int uba = rangea[1];
		int[] rangeb = (int[]) ranges.elementAt (j+1);
		int lbb = rangeb[0];
		int ubb = rangeb[1];

		/* If the two ranges overlap or are adjacent, coalesce them. 
		 * The two ranges overlap if the larger lower bound is less 
		 * than or equal to the smaller upper bound. The two ranges 
		 * are adjacent if the larger lower bound is one greater
		 * than the smaller upper bound. 
		 */
		if (Math.max(lba, lbb) - Math.min(uba, ubb) <= 1) {
		    // The coalesced range is from the smaller lower bound to 
		    // the larger upper bound.
		    ranges.setElementAt(new int[]
					   {Math.min(lba, lbb),
						Math.max(uba, ubb)}, j);
		    ranges.remove (j+1);
		} else if (lba > lbb) {

		    /* If the two ranges don't overlap and aren't adjacent but 
		     * are out of order, swap them. 
		     */
		    ranges.setElementAt (rangeb, j);
		    ranges.setElementAt (rangea, j+1);
		} else {
		/* If the two ranges don't overlap and aren't adjacent and 
		 * aren't out of order, we're done early. 
		 */
		    break;
		}
	    }
	}
    
private static int[][]canonicalArrayForm(java.util.Vector ranges)
Convert the given vector of int[] objects to canonical array form.

	return (int[][]) ranges.toArray (new int[ranges.size()][]);
    
public booleancontains(int x)
Determine if this set-of-integer attribute contains the given value.

param
x Integer value.
return
True if this set-of-integer attribute contains the value x, false otherwise.

	// Do a linear search to find the range that contains x, if any. 
	int n = members.length;
	for (int i = 0; i < n; ++ i) {
	    if (x < members[i][0]) {
		return false;
	    } else if (x <= members[i][1]) {
		return true;
	    }
	}
	return false;
    
public booleancontains(javax.print.attribute.IntegerSyntax attribute)
Determine if this set-of-integer attribute contains the given integer attribute's value.

param
attribute Integer attribute.
return
True if this set-of-integer attribute contains theAttribute's value, false otherwise.

	return contains (attribute.getValue());
    
public booleanequals(java.lang.Object object)
Returns whether this set-of-integer attribute is equivalent to the passed in object. To be equivalent, all of the following conditions must be true:
  1. object is not null.
  2. object is an instance of class SetOfIntegerSyntax.
  3. This set-of-integer attribute's members and object's members are the same.

param
object Object to compare to.
return
True if object is equivalent to this set-of-integer attribute, false otherwise.

	if (object != null && object instanceof SetOfIntegerSyntax) {
	    int[][] myMembers = this.members;
	    int[][] otherMembers = ((SetOfIntegerSyntax) object).members;
	    int m = myMembers.length;
	    int n = otherMembers.length;
	    if (m == n) { 
		for (int i = 0; i < m; ++ i) {
		    if (myMembers[i][0] != otherMembers[i][0] ||
			myMembers[i][1] != otherMembers[i][1]) {
			return false;
		    }
		}
		return true;
	    } else {
		return false;
	    }
	} else {
	    return false;
	}
    
public int[][]getMembers()
Obtain this set-of-integer attribute's members in canonical array form. The returned array is "safe;" the client may alter it without affecting this set-of-integer attribute.

return
This set-of-integer attribute's members in canonical array form.

	int n = members.length;
	int[][] result = new int[n][];
	for (int i = 0; i < n; ++ i) {
	    result[i] = new int[] {members[i][0], members[i][1]};
	}
	return result;
    
public inthashCode()
Returns a hash code value for this set-of-integer attribute. The hash code is the sum of the lower and upper bounds of the ranges in the canonical array form, or 0 for an empty set.

	int result = 0;
	int n = members.length;
	for (int i = 0; i < n; ++ i) {
	    result += members[i][0] + members[i][1];
	}
	return result;
    
public intnext(int x)
Determine the smallest integer in this set-of-integer attribute that is greater than the given value. If there are no integers in this set-of-integer attribute greater than the given value, -1 is returned. (Since a set-of-integer attribute can only contain nonnegative values, -1 will never appear in the set.) You can use the next() method to iterate through the integer values in a set-of-integer attribute in ascending order, like this:
SetOfIntegerSyntax attribute = . . .;
int i = -1;
while ((i = attribute.next (i)) != -1)
{
foo (i);
}

param
x Integer value.
return
The smallest integer in this set-of-integer attribute that is greater than x, or -1 if no integer in this set-of-integer attribute is greater than x.

	// Do a linear search to find the range that contains x, if any. 
	int n = members.length;
	for (int i = 0; i < n; ++ i) {
	    if (x < members[i][0]) {
		return members[i][0];
	    } else if (x < members[i][1]) {
		return x + 1;
	    }
	}
	return -1;
    
private static int[][]parse(java.lang.String members)
Parse the given string, returning canonical array form.

	// Create vector to hold int[] elements, each element being one range 
	// parsed out of members. 
	Vector theRanges = new Vector();
	
	// Run state machine over members.
	int n = (members == null ? 0 : members.length());
	int i = 0;
	int state = 0;
	int lb = 0;
	int ub = 0;
	char c;
	int digit;
	while (i < n) {
	    c = members.charAt(i ++);
	    switch (state) {

	    case 0: // Before first integer in first group
		if (Character.isWhitespace(c)) {
		    state = 0;
		}
		else if ((digit = Character.digit(c, 10)) != -1) {
		    lb = digit;
		    state = 1;
		} else {
		    throw new IllegalArgumentException();
		}
		break;
		
	    case 1: // In first integer in a group
		if (Character.isWhitespace(c)){
			state = 2;
		} else if ((digit = Character.digit(c, 10)) != -1) {
		    lb = 10 * lb + digit;
		    state = 1;
		} else if (c == '-" || c == ':") {
		    state = 3;
		} else if (c == ',") {
		    accumulate (theRanges, lb, lb);
		    state = 6;
		} else {
		    throw new IllegalArgumentException();
		}
		break;

	    case 2: // After first integer in a group
		if (Character.isWhitespace(c)) {
		    state = 2;
		} 
		else if (c == '-" || c == ':") {
		    state = 3;
		}
		else if (c == ',") {
		    accumulate(theRanges, lb, lb);
		    state = 6;
		} else {
		    throw new IllegalArgumentException();
		}
		break;

	    case 3: // Before second integer in a group
		if (Character.isWhitespace(c)) {
		    state = 3;
		} else if ((digit = Character.digit(c, 10)) != -1) {
		    ub = digit;
		    state = 4;
		} else {
		    throw new IllegalArgumentException();
		}
		break;

	    case 4: // In second integer in a group
		if (Character.isWhitespace(c)) {
		    state = 5;
		} else if ((digit = Character.digit(c, 10)) != -1) {
		    ub = 10 * ub + digit;
		    state = 4;
		} else if (c == ',") {
		    accumulate(theRanges, lb, ub);
		    state = 6;
		} else {
		    throw new IllegalArgumentException();
		}
		break;
		
	    case 5: // After second integer in a group
		if (Character.isWhitespace(c)) {
		    state = 5;
		} else if (c == ',") {
		    accumulate(theRanges, lb, ub);
		    state = 6;
		} else {
		    throw new IllegalArgumentException();
		}
		break;

	    case 6: // Before first integer in second or later group
		if (Character.isWhitespace(c)) {
		    state = 6;
		} else if ((digit = Character.digit(c, 10)) != -1) {
		    lb = digit;
		    state = 1;
		} else {
		    throw new IllegalArgumentException();
		}
		break;
	    }
	}

	// Finish off the state machine.
	switch (state) {
	case 0: // Before first integer in first group
	    break;
	case 1: // In first integer in a group
	case 2: // After first integer in a group
	    accumulate(theRanges, lb, lb);
	    break;
	case 4: // In second integer in a group
	case 5: // After second integer in a group
	    accumulate(theRanges, lb, ub);
	    break;
	case 3: // Before second integer in a group
	case 6: // Before first integer in second or later group
	    throw new IllegalArgumentException();
	}

	// Return canonical array form.
	return canonicalArrayForm (theRanges);
    
private static int[][]parse(int[][] members)
Parse the given array form, returning canonical array form.

	// Create vector to hold int[] elements, each element being one range 
	// parsed out of members. 
	Vector ranges = new Vector();

	// Process all integer groups in members.
	int n = (members == null ? 0 : members.length);
	for (int i = 0; i < n; ++ i) {
	    // Get lower and upper bounds of the range.
	    int lb, ub;
	    if (members[i].length == 1) {
		lb = ub = members[i][0];
	    } else if (members[i].length == 2) {
		lb = members[i][0];
		ub = members[i][1];
	    } else {
		throw new IllegalArgumentException();
	    }

	    // Verify valid bounds.
	    if (lb <= ub && lb < 0) {
		throw new IllegalArgumentException();
	    }

	    // Accumulate the range.
	    accumulate(ranges, lb, ub);
	}

		// Return canonical array form.
		return canonicalArrayForm (ranges);
		
public java.lang.StringtoString()
Returns a string value corresponding to this set-of-integer attribute. The string value is a zero-length string if this set is empty. Otherwise, the string value is a comma-separated list of the ranges in the canonical array form, where each range is represented as "i" if the lower bound equals the upper bound or "i-j" otherwise.

	StringBuffer result = new StringBuffer();
	int n = members.length;
	for (int i = 0; i < n; i++) {
	    if (i > 0) {
		result.append (',");
	    }
	    result.append (members[i][0]);
	    if (members[i][0] != members[i][1]) {
		result.append ('-");
		result.append (members[i][1]);
	    }
	}
	return result.toString();