001: /*
002: * The Apache Software License, Version 1.1
003: *
004: *
005: * Copyright (c) 1999 The Apache Software Foundation. All rights
006: * reserved.
007: *
008: * Redistribution and use in source and binary forms, with or without
009: * modification, are permitted provided that the following conditions
010: * are met:
011: *
012: * 1. Redistributions of source code must retain the above copyright
013: * notice, this list of conditions and the following disclaimer.
014: *
015: * 2. Redistributions in binary form must reproduce the above copyright
016: * notice, this list of conditions and the following disclaimer in
017: * the documentation and/or other materials provided with the
018: * distribution.
019: *
020: * 3. The end-user documentation included with the redistribution,
021: * if any, must include the following acknowledgment:
022: * "This product includes software developed by the
023: * Apache Software Foundation (http://www.apache.org/)."
024: * Alternately, this acknowledgment may appear in the software itself,
025: * if and wherever such third-party acknowledgments normally appear.
026: *
027: * 4. The names "Xerces" and "Apache Software Foundation" must
028: * not be used to endorse or promote products derived from this
029: * software without prior written permission. For written
030: * permission, please contact apache@apache.org.
031: *
032: * 5. Products derived from this software may not be called "Apache",
033: * nor may "Apache" appear in their name, without prior written
034: * permission of the Apache Software Foundation.
035: *
036: * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED
037: * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
038: * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
039: * DISCLAIMED. IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR
040: * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
041: * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
042: * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
043: * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
044: * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
045: * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
046: * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
047: * SUCH DAMAGE.
048: * ====================================================================
049: *
050: * This software consists of voluntary contributions made by many
051: * individuals on behalf of the Apache Software Foundation and was
052: * originally based on software copyright (c) 1999, International
053: * Business Machines, Inc., http://www.apache.org. For more
054: * information on the Apache Software Foundation, please see
055: * <http://www.apache.org/>.
056: */
057:
058: package org.apache.xerces.readers;
059:
060: import org.apache.xerces.framework.XMLErrorReporter;
061: import org.apache.xerces.utils.ChunkyByteArray;
062: import org.apache.xerces.utils.StringPool;
063:
064: import java.io.InputStreamReader;
065: import java.io.IOException;
066: import java.io.UnsupportedEncodingException;
067:
068: /**
069: *
070: * @version
071: */
072: final class EBCDICRecognizer extends XMLDeclRecognizer {
073: //
074: //
075: //
076: public XMLEntityHandler.EntityReader recognize(
077: XMLEntityReaderFactory readerFactory,
078: XMLEntityHandler entityHandler,
079: XMLErrorReporter errorReporter,
080: boolean sendCharDataAsCharArray, StringPool stringPool,
081: ChunkyByteArray data, boolean xmlDecl,
082: boolean allowJavaEncodingName) throws Exception {
083: XMLEntityHandler.EntityReader reader = null;
084: byte b0 = data.byteAt(0);
085: byte b1 = data.byteAt(1);
086: byte b2 = data.byteAt(2);
087: byte b3 = data.byteAt(3);
088: boolean debug = false;
089:
090: if (b0 != 0x4c || b1 != 0x6f || b2 != (byte) 0xa7
091: || b3 != (byte) 0x94)
092: return reader;
093: XMLEntityHandler.EntityReader declReader = readerFactory
094: .createCharReader(entityHandler, errorReporter,
095: sendCharDataAsCharArray, new InputStreamReader(
096: data, "CP037"), stringPool);
097: int encoding = prescanXMLDeclOrTextDecl(declReader, xmlDecl);
098: if (encoding == -1) {
099: data.rewind();
100: // REVISIT - The document is not well-formed. There is no encoding, yet the file is
101: // clearly not UTF8.
102: throw new UnsupportedEncodingException(null);
103: }
104: String enc = stringPool.orphanString(encoding).toUpperCase();
105: if ("ISO-10646-UCS-2".equals(enc))
106: throw new UnsupportedEncodingException(enc);
107: if ("ISO-10646-UCS-4".equals(enc))
108: throw new UnsupportedEncodingException(enc);
109: if ("UTF-16".equals(enc))
110: throw new UnsupportedEncodingException(enc);
111: String javaencname = MIME2Java.convert(enc);
112: if (null == javaencname) {
113: if (allowJavaEncodingName) {
114: javaencname = enc;
115: } else {
116: throw new UnsupportedEncodingException(enc);
117: }
118: }
119: try {
120: data.rewind();
121: reader = readerFactory.createCharReader(entityHandler,
122: errorReporter, sendCharDataAsCharArray,
123: new InputStreamReader(data, javaencname),
124: stringPool);
125: } catch (UnsupportedEncodingException e) {
126: throw e;
127: } catch (Exception e) {
128: if (debug == true)
129: e.printStackTrace(); // Internal Error
130: }
131: return reader;
132: }
133: }
|