htmlparser/jericho/RowColumnVector.java

// Jericho HTML Parser - Java based library for analysing and manipulating HTML
// Version 3.2
// Copyright (C) 2004-2009 Martin Jericho
// http://jericho.htmlparser.net/
//
// This library is free software; you can redistribute it and/or
// modify it under the terms of either one of the following licences:
//
// 1. The Eclipse Public License (EPL) version 1.0,
// included in this distribution in the file licence-epl-1.0.html
// or available at http://www.eclipse.org/legal/epl-v10.html
//
// 2. The GNU Lesser General Public License (LGPL) version 2.1 or later,
// included in this distribution in the file licence-lgpl-2.1.txt
// or available at http://www.gnu.org/licenses/lgpl.txt
//
// This library is distributed on an "AS IS" basis,
// WITHOUT WARRANTY OF ANY KIND, either express or implied.
// See the individual licence texts for more details.

package net.htmlparser.jericho;

import java.util.*;

/**
 * Represents the row and column number of a character position in the source document.
 * <p>
 * Obtained using the {@link Source#getRowColumnVector(int pos)} method.
 */
public final class RowColumnVector {
        private final int row;
        private final int column;
        private final int pos;
        
        private static final RowColumnVector FIRST=new RowColumnVector(1,1,0);
        private static final RowColumnVector[] STREAMED=new RowColumnVector[0];

        private RowColumnVector(final int row, final int column, final int pos) {
                this.row=row;
                this.column=column;
                this.pos=pos;
        }

        private RowColumnVector(final int pos) {
                // used in Streamed source where row and column aren't available.
                this(-1,-1,pos);
        }

        /**
         * Returns the row number of this character position in the source document.
         * <p>
         * If a {@link StreamedSource} is in use, this method always returns <code>-1</code>.
         *
         * @return the row number of this character position in the source document.
         */
        public int getRow() {
                return row;
        }
        
        /**
         * Returns the column number of this character position in the source document.
         * <p>
         * If a {@link StreamedSource} is in use, this method always returns <code>-1</code>.
         *
         * @return the column number of this character position in the source document.
         */
        public int getColumn() {
                return column;
        }

        /**
         * Returns the character position in the source document.
         * @return the character position in the source document.
         */
        public int getPos() {
                return pos;
        }
        
        /**
         * Returns a string representation of this character position.
         * <p>
         * The returned string has the format "<code>(<var>row</var>,<var>column</var>:<var>pos</var>)</code>".
         *
         * @return a string representation of this character position.
         */
        public String toString() {
                return appendTo(new StringBuilder(20)).toString();
        }

        StringBuilder appendTo(final StringBuilder sb) {
                if (row!=-1) return sb.append("(r").append(row).append(",c").append(column).append(",p").append(pos).append(')');
                return sb.append("(p").append(pos).append(')');
        }
        
        static RowColumnVector[] getCacheArray(final Source source) {
                if (source.isStreamed()) return STREAMED;
                final int lastSourcePos=source.end-1;
                final ArrayList<RowColumnVector> list=new ArrayList<RowColumnVector>();
                int pos=0;
                list.add(FIRST);
                int row=1;
                while (pos<=lastSourcePos) {
                        final char ch=source.charAt(pos);
                        if (ch=='\n' || (ch=='\r' && (pos==lastSourcePos || source.charAt(pos+1)!='\n'))) list.add(new RowColumnVector(++row,1,pos+1));
                        pos++;
                }
                return list.toArray(new RowColumnVector[list.size()]);
        }

        static RowColumnVector get(final RowColumnVector[] cacheArray, final int pos) {
                if (cacheArray==STREAMED) return new RowColumnVector(pos);
                int low=0;
                int high=cacheArray.length-1;
                while (true) {
                        int mid=(low+high) >> 1;
                        final RowColumnVector rowColumnVector=cacheArray[mid];
                        if (rowColumnVector.pos<pos) {
                                if (mid==high) return new RowColumnVector(rowColumnVector.row,pos-rowColumnVector.pos+1,pos);
                                low=mid+1;
                        } else if (rowColumnVector.pos>pos) {
                                high=mid-1;
                        } else {
                                return rowColumnVector;
                        }
                }
        }
}
1	// Jericho HTML Parser - Java based library for analysing and manipulating HTML
2	// Version 3.2
3	// Copyright (C) 2004-2009 Martin Jericho
4	// http://jericho.htmlparser.net/
5	//
6	// This library is free software; you can redistribute it and/or
7	// modify it under the terms of either one of the following licences:
8	//
9	// 1. The Eclipse Public License (EPL) version 1.0,
10	// included in this distribution in the file licence-epl-1.0.html
11	// or available at http://www.eclipse.org/legal/epl-v10.html
12	//
13	// 2. The GNU Lesser General Public License (LGPL) version 2.1 or later,
14	// included in this distribution in the file licence-lgpl-2.1.txt
15	// or available at http://www.gnu.org/licenses/lgpl.txt
16	//
17	// This library is distributed on an "AS IS" basis,
18	// WITHOUT WARRANTY OF ANY KIND, either express or implied.
19	// See the individual licence texts for more details.
20
21	package net.htmlparser.jericho;
22
23	import java.util.*;
24
25	/**
26	* Represents the row and column number of a character position in the source document.
27	* <p>
28	* Obtained using the {@link Source#getRowColumnVector(int pos)} method.
29	*/
30	public final class RowColumnVector {
31	private final int row;
32	private final int column;
33	private final int pos;
34
35	private static final RowColumnVector FIRST=new RowColumnVector(1,1,0);
36	private static final RowColumnVector[] STREAMED=new RowColumnVector[0];
37
38	private RowColumnVector(final int row, final int column, final int pos) {
39	this.row=row;
40	this.column=column;
41	this.pos=pos;
42	}
43
44	private RowColumnVector(final int pos) {
45	// used in Streamed source where row and column aren't available.
46	this(-1,-1,pos);
47	}
48
49	/**
50	* Returns the row number of this character position in the source document.
51	* <p>
52	* If a {@link StreamedSource} is in use, this method always returns <code>-1</code>.
53	*
54	* @return the row number of this character position in the source document.
55	*/
56	public int getRow() {
57	return row;
58	}
59
60	/**
61	* Returns the column number of this character position in the source document.
62	* <p>
63	* If a {@link StreamedSource} is in use, this method always returns <code>-1</code>.
64	*
65	* @return the column number of this character position in the source document.
66	*/
67	public int getColumn() {
68	return column;
69	}
70
71	/**
72	* Returns the character position in the source document.
73	* @return the character position in the source document.
74	*/
75	public int getPos() {
76	return pos;
77	}
78
79	/**
80	* Returns a string representation of this character position.
81	* <p>
82	* The returned string has the format "<code>(<var>row</var>,<var>column</var>:<var>pos</var>)</code>".
83	*
84	* @return a string representation of this character position.
85	*/
86	public String toString() {
87	return appendTo(new StringBuilder(20)).toString();
88	}
89
90	StringBuilder appendTo(final StringBuilder sb) {
91	if (row!=-1) return sb.append("(r").append(row).append(",c").append(column).append(",p").append(pos).append(')');
92	return sb.append("(p").append(pos).append(')');
93	}
94
95	static RowColumnVector[] getCacheArray(final Source source) {
96	if (source.isStreamed()) return STREAMED;
97	final int lastSourcePos=source.end-1;
98	final ArrayList<RowColumnVector> list=new ArrayList<RowColumnVector>();
99	int pos=0;
100	list.add(FIRST);
101	int row=1;
102	while (pos<=lastSourcePos) {
103	final char ch=source.charAt(pos);
104	if (ch=='\n' \|\| (ch=='\r' && (pos==lastSourcePos \|\| source.charAt(pos+1)!='\n'))) list.add(new RowColumnVector(++row,1,pos+1));
105	pos++;
106	}
107	return list.toArray(new RowColumnVector[list.size()]);
108	}
109
110	static RowColumnVector get(final RowColumnVector[] cacheArray, final int pos) {
111	if (cacheArray==STREAMED) return new RowColumnVector(pos);
112	int low=0;
113	int high=cacheArray.length-1;
114	while (true) {
115	int mid=(low+high) >> 1;
116	final RowColumnVector rowColumnVector=cacheArray[mid];
117	if (rowColumnVector.pos<pos) {
118	if (mid==high) return new RowColumnVector(rowColumnVector.row,pos-rowColumnVector.pos+1,pos);
119	low=mid+1;
120	} else if (rowColumnVector.pos>pos) {
121	high=mid-1;
122	} else {
123	return rowColumnVector;
124	}
125	}
126	}
127	}