1 /*
2 * Licensed to the Apache Software Foundation (ASF) under one or more
3 * contributor license agreements. See the NOTICE file distributed with
4 * this work for additional information regarding copyright ownership.
5 * The ASF licenses this file to You under the Apache License, Version 2.0
6 * (the "License"); you may not use this file except in compliance with
7 * the License. You may obtain a copy of the License at
8 *
9 * http://www.apache.org/licenses/LICENSE-2.0
10 *
11 * Unless required by applicable law or agreed to in writing, software
12 * distributed under the License is distributed on an "AS IS" BASIS,
13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 * See the License for the specific language governing permissions and
15 * limitations under the License.
16 */
17
18 package xni.parser;
19
20 import java.io.BufferedReader;
21 import java.io.IOException;
22 import java.io.InputStream;
23 import java.io.InputStreamReader;
24 import java.io.Reader;
25 import java.util.StringTokenizer;
26
27 import org.apache.xerces.util.NamespaceSupport;
28 import org.apache.xerces.util.XMLAttributesImpl;
29 import org.apache.xerces.util.XMLStringBuffer;
30 import org.apache.xerces.xni.QName;
31 import org.apache.xerces.xni.XMLAttributes;
32 import org.apache.xerces.xni.XMLDTDContentModelHandler;
33 import org.apache.xerces.xni.XMLString;
34 import org.apache.xerces.xni.XNIException;
35 import org.apache.xerces.xni.parser.XMLInputSource;
36
37 /**
38 * This example is a very simple parser configuration that can
39 * parse files with comma-separated values (CSV) to generate XML
40 * events. For example, the following CSV document:
41 * <pre>
42 * Andy Clark,16 Jan 1973,Cincinnati
43 * </pre>
44 * produces the following XML "document" as represented by the
45 * XNI streaming document information:
46 * <pre>
47 * <?xml version='1.0' encoding='UTF-8' standalone='true'?>
48 * <!DOCTYPE csv [
49 * <!ELEMENT csv (row)*>
50 * <!ELEMENT row (col)*>
51 * <!ELEMENT col (#PCDATA)>
52 * ]>
53 * <csv>
54 * <row>
55 * <col>Andy Clark</col>
56 * <col>16 Jan 1973</col>
57 * <col>Cincinnati</col>
58 * </row>
59 * </csv>
60 * </pre>
61 *
62 * @author Andy Clark, IBM
63 *
64 * @version $Id: CSVConfiguration.java 447690 2006-09-19 02:41:53Z mrglavas $
65 */
66 public class CSVConfiguration
67 extends AbstractConfiguration {
68
69 //
70 // Constants
71 //
72
73 /** A QName for the <csv> element name. */
74 protected static final QName CSV = new QName(null, null, "csv", null);
75
76 /** A QName for the <row> element name. */
77 protected static final QName ROW = new QName(null, null, "row", null);
78
79 /** A QName for the <col> element name. */
80 protected static final QName COL = new QName(null, null, "col", null);
81
82 /** An empty list of attributes. */
83 protected static final XMLAttributes EMPTY_ATTRS = new XMLAttributesImpl();
84
85 /** A newline XMLString. */
86 private final XMLString NEWLINE = new XMLStringBuffer("\n");
87
88 /** A newline + one space XMLString. */
89 private final XMLString NEWLINE_ONE_SPACE = new XMLStringBuffer("\n ");
90
91 /** A newline + two spaces XMLString. */
92 private final XMLString NEWLINE_TWO_SPACES = new XMLStringBuffer("\n ");
93
94 //
95 // Data
96 //
97
98 /**
99 * A string buffer for use in copying string into an XMLString
100 * object for passing to the characters method.
101 */
102 private final XMLStringBuffer fStringBuffer = new XMLStringBuffer();
103
104 //
105 // XMLParserConfiguration methods
106 //
107
108 /**
109 * Parse an XML document.
110 * <p>
111 * The parser can use this method to instruct this configuration
112 * to begin parsing an XML document from any valid input source
113 * (a character stream, a byte stream, or a URI).
114 * <p>
115 * Parsers may not invoke this method while a parse is in progress.
116 * Once a parse is complete, the parser may then parse another XML
117 * document.
118 * <p>
119 * This method is synchronous: it will not return until parsing
120 * has ended. If a client application wants to terminate
121 * parsing early, it should throw an exception.
122 *
123 * @param source The input source for the top-level of the
124 * XML document.
125 *
126 * @exception XNIException Any XNI exception, possibly wrapping
127 * another exception.
128 * @exception IOException An IO exception from the parser, possibly
129 * from a byte stream or character stream
130 * supplied by the parser.
131 */
132 public void parse(XMLInputSource source)
133 throws IOException, XNIException {
134
135 // get reader
136 openInputSourceStream(source);
137 Reader reader = source.getCharacterStream();
138 if (reader == null) {
139 InputStream stream = source.getByteStream();
140 reader = new InputStreamReader(stream);
141 }
142 BufferedReader bufferedReader = new BufferedReader(reader);
143
144 // start document
145 if (fDocumentHandler != null) {
146 fDocumentHandler.startDocument(null, "UTF-8", new NamespaceSupport(), null);
147 fDocumentHandler.xmlDecl("1.0", "UTF-8", "true", null);
148 fDocumentHandler.doctypeDecl("csv", null, null, null);
149 }
150 if (fDTDHandler != null) {
151 fDTDHandler.startDTD(null, null);
152 fDTDHandler.elementDecl("csv", "(row)*", null);
153 fDTDHandler.elementDecl("row", "(col)*", null);
154 fDTDHandler.elementDecl("col", "(#PCDATA)", null);
155 }
156 if (fDTDContentModelHandler != null) {
157 fDTDContentModelHandler.startContentModel("csv", null);
158 fDTDContentModelHandler.startGroup(null);
159 fDTDContentModelHandler.element("row", null);
160 fDTDContentModelHandler.endGroup(null);
161 short csvOccurs = XMLDTDContentModelHandler.OCCURS_ZERO_OR_MORE;
162 fDTDContentModelHandler.occurrence(csvOccurs, null);
163 fDTDContentModelHandler.endContentModel(null);
164
165 fDTDContentModelHandler.startContentModel("row", null);
166 fDTDContentModelHandler.startGroup(null);
167 fDTDContentModelHandler.element("col", null);
168 fDTDContentModelHandler.endGroup(null);
169 short rowOccurs = XMLDTDContentModelHandler.OCCURS_ZERO_OR_MORE;
170 fDTDContentModelHandler.occurrence(rowOccurs, null);
171 fDTDContentModelHandler.endContentModel(null);
172
173 fDTDContentModelHandler.startContentModel("col", null);
174 fDTDContentModelHandler.startGroup(null);
175 fDTDContentModelHandler.pcdata(null);
176 fDTDContentModelHandler.endGroup(null);
177 fDTDContentModelHandler.endContentModel(null);
178 }
179 if (fDTDHandler != null) {
180 fDTDHandler.endDTD(null);
181 }
182 if (fDocumentHandler != null) {
183 fDocumentHandler.startElement(CSV, EMPTY_ATTRS, null);
184 }
185
186 // read lines
187 String line;
188 while ((line = bufferedReader.readLine()) != null) {
189 if (fDocumentHandler != null) {
190 fDocumentHandler.ignorableWhitespace(NEWLINE_ONE_SPACE, null);
191 fDocumentHandler.startElement(ROW, EMPTY_ATTRS, null);
192 StringTokenizer tokenizer = new StringTokenizer(line, ",");
193 while (tokenizer.hasMoreTokens()) {
194 fDocumentHandler.ignorableWhitespace(NEWLINE_TWO_SPACES, null);
195 fDocumentHandler.startElement(COL, EMPTY_ATTRS, null);
196 String token = tokenizer.nextToken();
197 fStringBuffer.clear();
198 fStringBuffer.append(token);
199 fDocumentHandler.characters(fStringBuffer, null);
200 fDocumentHandler.endElement(COL, null);
201 }
202 fDocumentHandler.ignorableWhitespace(NEWLINE_ONE_SPACE, null);
203 fDocumentHandler.endElement(ROW, null);
204 }
205 }
206 bufferedReader.close();
207
208 // end document
209 if (fDocumentHandler != null) {
210 fDocumentHandler.ignorableWhitespace(NEWLINE, null);
211 fDocumentHandler.endElement(CSV, null);
212 fDocumentHandler.endDocument(null);
213 }
214
215 } // parse(XMLInputSource)
216
217 // NOTE: The following methods are overloaded to ignore setting
218 // of parser state so that this configuration does not
219 // throw configuration exceptions for features and properties
220 // that it doesn't care about.
221
222 public void setFeature(String featureId, boolean state) {}
223 public boolean getFeature(String featureId) { return false; }
224 public void setProperty(String propertyId, Object value) {}
225 public Object getProperty(String propertyId) { return null; }
226
227 } // class CSVConfiguration