1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
|
/*
* Copyright 2001,2002,2004,2005 The Apache Software Foundation.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package xni.parser;
import java.io.BufferedReader;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.io.Reader;
import java.util.StringTokenizer;
import org.apache.xerces.util.NamespaceSupport;
import org.apache.xerces.util.XMLAttributesImpl;
import org.apache.xerces.util.XMLStringBuffer;
import org.apache.xerces.xni.QName;
import org.apache.xerces.xni.XMLAttributes;
import org.apache.xerces.xni.XMLDTDContentModelHandler;
import org.apache.xerces.xni.XMLString;
import org.apache.xerces.xni.XNIException;
import org.apache.xerces.xni.parser.XMLInputSource;
/**
* This example is a very simple parser configuration that can
* parse files with comma-separated values (CSV) to generate XML
* events. For example, the following CSV document:
* <pre>
* Andy Clark,16 Jan 1973,Cincinnati
* </pre>
* produces the following XML "document" as represented by the
* XNI streaming document information:
* <pre>
* <?xml version='1.0' encoding='UTF-8' standalone='true'?>
* <!DOCTYPE csv [
* <!ELEMENT csv (row)*>
* <!ELEMENT row (col)*>
* <!ELEMENT col (#PCDATA)>
* ]>
* <csv>
* <row>
* <col>Andy Clark</col>
* <col>16 Jan 1973</col>
* <col>Cincinnati</col>
* </row>
* </csv>
* </pre>
*
* @author Andy Clark, IBM
*
* @version $Id: CSVConfiguration.java 320203 2005-01-11 13:47:23Z mrglavas $
*/
public class CSVConfiguration
extends AbstractConfiguration {
//
// Constants
//
/** A QName for the <csv> element name. */
protected static final QName CSV = new QName(null, null, "csv", null);
/** A QName for the <row> element name. */
protected static final QName ROW = new QName(null, null, "row", null);
/** A QName for the <col> element name. */
protected static final QName COL = new QName(null, null, "col", null);
/** An empty list of attributes. */
protected static final XMLAttributes EMPTY_ATTRS = new XMLAttributesImpl();
/** A newline XMLString. */
private final XMLString NEWLINE = new XMLStringBuffer("\n");
/** A newline + one space XMLString. */
private final XMLString NEWLINE_ONE_SPACE = new XMLStringBuffer("\n ");
/** A newline + two spaces XMLString. */
private final XMLString NEWLINE_TWO_SPACES = new XMLStringBuffer("\n ");
//
// Data
//
/**
* A string buffer for use in copying string into an XMLString
* object for passing to the characters method.
*/
private final XMLStringBuffer fStringBuffer = new XMLStringBuffer();
//
// XMLParserConfiguration methods
//
/**
* Parse an XML document.
* <p>
* The parser can use this method to instruct this configuration
* to begin parsing an XML document from any valid input source
* (a character stream, a byte stream, or a URI).
* <p>
* Parsers may not invoke this method while a parse is in progress.
* Once a parse is complete, the parser may then parse another XML
* document.
* <p>
* This method is synchronous: it will not return until parsing
* has ended. If a client application wants to terminate
* parsing early, it should throw an exception.
*
* @param source The input source for the top-level of the
* XML document.
*
* @exception XNIException Any XNI exception, possibly wrapping
* another exception.
* @exception IOException An IO exception from the parser, possibly
* from a byte stream or character stream
* supplied by the parser.
*/
public void parse(XMLInputSource source)
throws IOException, XNIException {
// get reader
openInputSourceStream(source);
Reader reader = source.getCharacterStream();
if (reader == null) {
InputStream stream = source.getByteStream();
reader = new InputStreamReader(stream);
}
BufferedReader bufferedReader = new BufferedReader(reader);
// start document
if (fDocumentHandler != null) {
fDocumentHandler.startDocument(null, "UTF-8", new NamespaceSupport(), null);
fDocumentHandler.xmlDecl("1.0", "UTF-8", "true", null);
fDocumentHandler.doctypeDecl("csv", null, null, null);
}
if (fDTDHandler != null) {
fDTDHandler.startDTD(null, null);
fDTDHandler.elementDecl("csv", "(row)*", null);
fDTDHandler.elementDecl("row", "(col)*", null);
fDTDHandler.elementDecl("col", "(#PCDATA)", null);
}
if (fDTDContentModelHandler != null) {
fDTDContentModelHandler.startContentModel("csv", null);
fDTDContentModelHandler.startGroup(null);
fDTDContentModelHandler.element("row", null);
fDTDContentModelHandler.endGroup(null);
short csvOccurs = XMLDTDContentModelHandler.OCCURS_ZERO_OR_MORE;
fDTDContentModelHandler.occurrence(csvOccurs, null);
fDTDContentModelHandler.endContentModel(null);
fDTDContentModelHandler.startContentModel("row", null);
fDTDContentModelHandler.startGroup(null);
fDTDContentModelHandler.element("col", null);
fDTDContentModelHandler.endGroup(null);
short rowOccurs = XMLDTDContentModelHandler.OCCURS_ZERO_OR_MORE;
fDTDContentModelHandler.occurrence(rowOccurs, null);
fDTDContentModelHandler.endContentModel(null);
fDTDContentModelHandler.startContentModel("col", null);
fDTDContentModelHandler.startGroup(null);
fDTDContentModelHandler.pcdata(null);
fDTDContentModelHandler.endGroup(null);
fDTDContentModelHandler.endContentModel(null);
}
if (fDTDHandler != null) {
fDTDHandler.endDTD(null);
}
if (fDocumentHandler != null) {
fDocumentHandler.startElement(CSV, EMPTY_ATTRS, null);
}
// read lines
String line;
while ((line = bufferedReader.readLine()) != null) {
if (fDocumentHandler != null) {
fDocumentHandler.ignorableWhitespace(NEWLINE_ONE_SPACE, null);
fDocumentHandler.startElement(ROW, EMPTY_ATTRS, null);
StringTokenizer tokenizer = new StringTokenizer(line, ",");
while (tokenizer.hasMoreTokens()) {
fDocumentHandler.ignorableWhitespace(NEWLINE_TWO_SPACES, null);
fDocumentHandler.startElement(COL, EMPTY_ATTRS, null);
String token = tokenizer.nextToken();
fStringBuffer.clear();
fStringBuffer.append(token);
fDocumentHandler.characters(fStringBuffer, null);
fDocumentHandler.endElement(COL, null);
}
fDocumentHandler.ignorableWhitespace(NEWLINE_ONE_SPACE, null);
fDocumentHandler.endElement(ROW, null);
}
}
bufferedReader.close();
// end document
if (fDocumentHandler != null) {
fDocumentHandler.ignorableWhitespace(NEWLINE, null);
fDocumentHandler.endElement(CSV, null);
fDocumentHandler.endDocument(null);
}
} // parse(XMLInputSource)
// NOTE: The following methods are overloaded to ignore setting
// of parser state so that this configuration does not
// throw configuration exceptions for features and properties
// that it doesn't care about.
public void setFeature(String featureId, boolean state) {}
public boolean getFeature(String featureId) { return false; }
public void setProperty(String propertyId, Object value) {}
public Object getProperty(String propertyId) { return null; }
} // class CSVConfiguration
|