1 | /**
|
---|
2 | *#########################################################################
|
---|
3 | *
|
---|
4 | * A component of the Gatherer application, part of the Greenstone digital
|
---|
5 | * library suite from the New Zealand Digital Library Project at the
|
---|
6 | * University of Waikato, New Zealand.
|
---|
7 | *
|
---|
8 | * <BR><BR>
|
---|
9 | *
|
---|
10 | * Author: John Thompson, Greenstone Digital Library, University of Waikato
|
---|
11 | *
|
---|
12 | * <BR><BR>
|
---|
13 | *
|
---|
14 | * Copyright (C) 1999 New Zealand Digital Library Project
|
---|
15 | *
|
---|
16 | * <BR><BR>
|
---|
17 | *
|
---|
18 | * This program is free software; you can redistribute it and/or modify
|
---|
19 | * it under the terms of the GNU General Public License as published by
|
---|
20 | * the Free Software Foundation; either version 2 of the License, or
|
---|
21 | * (at your option) any later version.
|
---|
22 | *
|
---|
23 | * <BR><BR>
|
---|
24 | *
|
---|
25 | * This program is distributed in the hope that it will be useful,
|
---|
26 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
27 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
28 | * GNU General Public License for more details.
|
---|
29 | *
|
---|
30 | * <BR><BR>
|
---|
31 | *
|
---|
32 | * You should have received a copy of the GNU General Public License
|
---|
33 | * along with this program; if not, write to the Free Software
|
---|
34 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
35 | *########################################################################
|
---|
36 | */
|
---|
37 | package org.greenstone.gatherer.util;
|
---|
38 |
|
---|
39 | /**
|
---|
40 | * Title: The Gatherer<br>
|
---|
41 | * Description: The Gatherer: a tool for gathering and enriching digital collections.<br>
|
---|
42 | * Copyright: Copyright (c) 2001<br>
|
---|
43 | * Company: The University of Waikato<br>
|
---|
44 | * @author John Thompson, Greenstone Digital Libraries
|
---|
45 | * @version 2.1
|
---|
46 | */
|
---|
47 | import org.greenstone.gatherer.util.Utility;
|
---|
48 |
|
---|
49 | /** This class functions much like a <strong>StringTokenizer</strong> in that it tokenizes a long string into tokens, however this tokenizer cleverly notices HTML formatting tags. */
|
---|
50 | public class HTMLStringTokenizer {
|
---|
51 | /** The current position in the source string. */
|
---|
52 | private int pos = 0;
|
---|
53 | /** The current token, usually created by the last nextToken call. */
|
---|
54 | private String current = null;
|
---|
55 | /** The previous token. */
|
---|
56 | private String previous = null;
|
---|
57 | /** The string to be tokenized, including any HTML markup. */
|
---|
58 | private String source = null;
|
---|
59 | /** Constructor.
|
---|
60 | * @param source The source <strong>String</strong> to be tokenized.
|
---|
61 | */
|
---|
62 | public HTMLStringTokenizer(String source) {
|
---|
63 | this.source = source;
|
---|
64 | // Parse the first token.
|
---|
65 | parseToken();
|
---|
66 | }
|
---|
67 | /** Determines if there are still tokens remaining unparsed in the source.
|
---|
68 | * @return A <strong>boolean</strong> which is <i>true</i> if there are more tokens.
|
---|
69 | */
|
---|
70 | public boolean hasMoreTokens() {
|
---|
71 | if(current != null && current.length() > 0) {
|
---|
72 | return true;
|
---|
73 | }
|
---|
74 | return false;
|
---|
75 | }
|
---|
76 | /** Determines if the tag currently being returned by sameToken is a tag.
|
---|
77 | * @return A <strong>boolean</strong> indicating if the token is a tag.
|
---|
78 | */
|
---|
79 | public boolean isTag() {
|
---|
80 | if(previous.startsWith("<") && previous.endsWith(">")) {
|
---|
81 | return true;
|
---|
82 | }
|
---|
83 | return false;
|
---|
84 | }
|
---|
85 | /** Retrieves the next token.
|
---|
86 | * @return A <strong>String</strong> representing the token.
|
---|
87 | */
|
---|
88 | public String nextToken() {
|
---|
89 | previous = current;
|
---|
90 | // Get the next token.
|
---|
91 | parseToken();
|
---|
92 | // Return previous.
|
---|
93 | return previous;
|
---|
94 | }
|
---|
95 | /** Repeats the result of the last <i>nextToken()</i>.
|
---|
96 | * @return A <strong>String</strong> representing the token.
|
---|
97 | */
|
---|
98 | /* private String sameToken() {
|
---|
99 | return previous;
|
---|
100 | } */
|
---|
101 | /** Parses the next token and stores it in current.
|
---|
102 | */
|
---|
103 | private void parseToken() {
|
---|
104 | boolean found = false;
|
---|
105 | boolean tag = false;
|
---|
106 | boolean text = false;
|
---|
107 | // Reset current
|
---|
108 | current = "";
|
---|
109 | // Parse away
|
---|
110 | dumpWhiteSpace();
|
---|
111 | while(pos < source.length() && !found) {
|
---|
112 | char c = (char)source.charAt(pos);
|
---|
113 | if(!tag && !text) {
|
---|
114 | if(c == '<') {
|
---|
115 | tag = true;
|
---|
116 | }
|
---|
117 | else {
|
---|
118 | text = true;
|
---|
119 | }
|
---|
120 | current = current + c;
|
---|
121 | }
|
---|
122 | // Reading a tag. Watch only for '>'.
|
---|
123 | else if(tag) {
|
---|
124 | if(c == '>') {
|
---|
125 | found = true;
|
---|
126 | }
|
---|
127 | current = current + c;
|
---|
128 | }
|
---|
129 | // Reading text. Watch for ' ' and '<'. Rollback '<'.
|
---|
130 | else if(text) {
|
---|
131 | if(c == ' ') {
|
---|
132 | found = true;
|
---|
133 | }
|
---|
134 | else if(c == '<') {
|
---|
135 | found = true;
|
---|
136 | pos--;
|
---|
137 | }
|
---|
138 | else {
|
---|
139 | current = current + c;
|
---|
140 | }
|
---|
141 | }
|
---|
142 | pos++;
|
---|
143 | }
|
---|
144 | }
|
---|
145 | /** Method to ignore whitespace in the source.
|
---|
146 | */
|
---|
147 | private void dumpWhiteSpace() {
|
---|
148 | while(pos < source.length() && source.charAt(pos) == ' ') {
|
---|
149 | pos++;
|
---|
150 | }
|
---|
151 | }
|
---|
152 |
|
---|
153 | static public void main(String args[]) {
|
---|
154 | String init = "<HTML>Where material to be imported is found. Defaults to <i>GSDLHOME/collection/col_name/gimport</i></HTML>";
|
---|
155 | ///ystem.err.println("Before: " + init);
|
---|
156 | String result = Utility.formatHTMLWidth(init, 40);
|
---|
157 | ///ystem.err.println("After: " + result);
|
---|
158 | }
|
---|
159 | }
|
---|