All Downloads are FREE. Search and download functionalities are using the official Maven repository.

src.com.ibm.as400.access.ConvTableReader Maven / Gradle / Ivy

There is a newer version: 11.1
Show newest version
///////////////////////////////////////////////////////////////////////////////
//
// JTOpen (IBM Toolbox for Java - OSS version)
//
// Filename:  ConvTableReader.java
//
// The source code contained herein is licensed under the IBM Public License
// Version 1.0, which has been approved by the Open Source Initiative.
// Copyright (C) 1997-2004 International Business Machines Corporation and
// others.  All rights reserved.
//
///////////////////////////////////////////////////////////////////////////////

package com.ibm.as400.access;

import java.io.BufferedInputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.io.UnsupportedEncodingException;
import java.sql.SQLException;

/**
 Represents a Toolbox converter that uses stateful character conversion.  That is, it wraps an underlying InputStream and reads/caches the appropriate number of bytes to return the requested number of Unicode characters.  This is especially useful for mixed-byte tables where the number of converted Unicode characters is almost never the same as the number of underlying EBCDIC bytes.  This class exists primarily for use with the IFSTextFile... classes, but other components are free to use it as well.
 @see  com.ibm.as400.access.ConvTableWriter
 @see  com.ibm.as400.access.ReaderInputStream
 **/
public class ConvTableReader extends InputStreamReader
{
    private BufferedInputStream is_ = null;

    private int ccsid_ = -1;
    private ConvTable table_ = null;

    private BidiConversionProperties properties_ = new BidiConversionProperties();

    // The mode is used for mixed-byte tables only.
    static final int DB_MODE = 1;
    static final int SB_MODE = 2;
    private int mode_ = SB_MODE;  // default to single-byte mode unless we receive a shift-out

    // The different table types, based on the instance of the ConvTable.
    private static final int SB_TABLE = 10;
    private static final int DB_TABLE = 11;
    private static final int MB_TABLE = 12;
    private static final int JV_TABLE = 13;
    private static final int UTF8_TABLE = 14;
    private int tableType_ = SB_TABLE;

    private char[] cache_ = new char[1024];  // The character cache.
    private byte[] b_cache_ = new byte[2562];  // ((1024*5)+3)/2 == worst case mixed-byte array size +1 for extra shift byte, just in case.
    private boolean isCachedByte_ = false;  // Used for double-byte tables.
    private byte cachedByte_ = 0;  // Used for double-byte tables.

    private byte[] leftovers = new byte[3];  // Used for portions of utf-8 characters at the cache_ boundary.
    private int  leftoverCount = 0;  // How many bytes are in the leftovers.

    private int nextRead_ = 0;  // Cache needs to be filled when nextRead_ >= nextWrite_
    private int nextWrite_ = 0;
    boolean isXML_ = false;      //@xml3 true if this data originated from a native XML column type //@xml4 @ascii
    private boolean isFirstRead_ = true; //@xml3

    /**
     Creates a ConvTableReader that uses the default character encoding.  The CCSID this reader uses may be set if a known mapping exists for this platform's default character encoding.
     @param  in  The InputStream from which to read characters.
     @exception  UnsupportedEncodingException  If the default character encoding or its associated CCSID is not supported.
     **/
    public ConvTableReader(InputStream in) throws UnsupportedEncodingException
    {
        super(in);
        is_ = new BufferedInputStream(in);
        initializeCcsid();
        initializeTable();
    }

    /**
     Creates a ConvTableReader that uses the specified character encoding.  The CCSID this reader uses may be set if a known mapping exists for the given encoding.
     @param  in  The InputStream from which to read characters.
     @param  encoding  The name of a supported character encoding.
     @exception  UnsupportedEncodingException  If the specified character encoding or its associated CCSID is not supported.
     **/
    public ConvTableReader(InputStream in, String encoding) throws UnsupportedEncodingException
    {
        super(in, encoding);
        is_ = new BufferedInputStream(in);
        initializeCcsid();
        initializeTable();
    }

    /**
     Creates a ConvTableReader that uses the specified CCSID.
     @param  in  The InputStream from which to read characters.
     @param  ccsid  The CCSID.
     @exception  UnsupportedEncodingException  If the specified CCSID or its corresponding character encoding is not supported.
     **/
    public ConvTableReader(InputStream in, int ccsid) throws UnsupportedEncodingException
    {
        super(in);
        if (ccsid < 0 || ccsid > 65535)
        {
            Trace.log(Trace.ERROR, "Value of parameter 'ccsid' is not valid:", ccsid);
            throw new ExtendedIllegalArgumentException("ccsid", ExtendedIllegalArgumentException.RANGE_NOT_VALID);
        }
        is_ = new BufferedInputStream(in);
        ccsid_ = ccsid;
        initializeTable();
    }

    //@xml3 new constructor
    /**
    Creates a ConvTableReader that uses the specified CCSID.
    @param  in  The InputStream from which to read characters.
    @param  ccsid  The CCSID.
    @param  isXML  flag that stream is from an XML column type (needed to strip xml declaration)
    @exception  UnsupportedEncodingException  If the specified CCSID or its corresponding character encoding is not supported.
    **/
   public ConvTableReader(InputStream in, int ccsid, boolean isXML) throws UnsupportedEncodingException
   {
        this(in, ccsid);
        isXML_ = isXML;
   }
   
    /**
     Creates a ConvTableReader that uses the specified CCSID and bi-directional string type.
     @param  in  The InputStream from which to read characters.
     @param  ccsid  The CCSID.
     @param  bidiStringType  The {@link com.ibm.as400.access.BidiStringType bi-directional string type}.
     @exception  UnsupportedEncodingException  If the specified CCSID or its corresponding character encoding is not supported.
     **/
    public ConvTableReader(InputStream in, int ccsid, int bidiStringType) throws UnsupportedEncodingException
    {
        this(in, ccsid, new BidiConversionProperties(bidiStringType));
    }

    //@xml3 new constructor
    /**
    Creates a ConvTableReader that uses the specified CCSID and bi-directional string type.
    @param  in  The InputStream from which to read characters.
    @param  ccsid  The CCSID.
    @param  bidiStringType  The {@link com.ibm.as400.access.BidiStringType bi-directional string type}.
    @param  isXML  flag that stream is from an XML column type (needed to strip xml declaration)
    @exception  UnsupportedEncodingException  If the specified CCSID or its corresponding character encoding is not supported.
    **/
   public ConvTableReader(InputStream in, int ccsid, int bidiStringType, boolean isXML) throws UnsupportedEncodingException
   {
       this(in, ccsid, new BidiConversionProperties(bidiStringType));
       isXML_ = isXML;
   }

   
    /**
     Creates a ConvTableReader that uses the specified CCSID and bi-directional string type.
     @param  in  The InputStream from which to read characters.
     @param  ccsid  The CCSID.
     @param  properties  The bidi conversion properties.
     @exception  UnsupportedEncodingException  If the specified CCSID or its corresponding character encoding is not supported.
     **/
    public ConvTableReader(InputStream in, int ccsid, BidiConversionProperties properties) throws UnsupportedEncodingException
    {
        super(in);
        if (ccsid < 0 || ccsid > 65535)
        {
            Trace.log(Trace.ERROR, "Value of parameter 'ccsid' is not valid:", ccsid);
            throw new ExtendedIllegalArgumentException("ccsid", ExtendedIllegalArgumentException.RANGE_NOT_VALID);
        }

        is_ = new BufferedInputStream(in);
        ccsid_ = ccsid;
        properties_ = properties;
        initializeTable();
    }

    /**
     Creates a ConvTableReader that uses the specified CCSID, bi-directional string type, and internal cache size.
     @param  in  The InputStream from which to read characters.
     @param  ccsid  The CCSID.
     @param  bidiStringType  The {@link com.ibm.as400.access.BidiStringType bi-directional string type}.
     @param  cacheSize  The number of characters to store in the internal buffer.  The default is 1024.  This number must be greater than zero.
     @exception  UnsupportedEncodingException  If the specified CCSID or its corresponding character encoding is not supported.
     **/
    public ConvTableReader(InputStream in, int ccsid, int bidiStringType, int cacheSize) throws UnsupportedEncodingException
    {
        this(in, ccsid, bidiStringType);
        if (cacheSize < 1)
        {
            Trace.log(Trace.ERROR, "Value of parameter 'cacheSize' is not valid:", cacheSize);
            throw new ExtendedIllegalArgumentException("cacheSize", ExtendedIllegalArgumentException.RANGE_NOT_VALID);
        }

        cache_ = new char[cacheSize]; // The character cache.
        b_cache_ = new byte[((cacheSize*5)+3)/2]; // ((1024*5)+3)/2 == worst case mixed-byte array size +1 for extra shift byte, just in case.
    }

    /**
     Closes this ConvTableReader and its underlying input stream.  Calling close() multiple times will not throw an exception.
     @exception  IOException  If an I/O exception occurs.
     **/
    public void close() throws IOException
    {
        synchronized (lock)
        {
            if (table_ == null) return; //we are already closed
            table_ = null;
            cache_ = null;
            b_cache_ = null;
            super.close();
            is_.close();
        }
    }

    private boolean fillCache() throws IOException
    {
        synchronized (lock)
        {
            checkOpen();
            if (nextRead_ >= nextWrite_)
            {
                int numRead = 0;
                if (Trace.traceOn_) Trace.log(Trace.CONVERSION, "Filling cache for reader " + ccsid_ + "/" + tableType_ + " [" + toString() + "]: " + nextRead_ + "," + nextWrite_ + "," + cache_.length);
                if (tableType_ == SB_TABLE || tableType_ == JV_TABLE)
                {
                    numRead = is_.read(b_cache_, 0, cache_.length);
                }
                else if (tableType_ == DB_TABLE)
                {
                    if (isCachedByte_)
                    {
                        numRead = is_.read(b_cache_, 1, cache_.length-1);
                        if (numRead == -1)
                        {
                            if (Trace.traceOn_) Trace.log(Trace.CONVERSION, "Cache not filled, end of stream reached.");
                            return false;
                        }
                        b_cache_[0] = cachedByte_;
                        if (numRead % 2 == 0)  // if we've read an even number of chars, need to proliferate the last byte.
                        {
                            cachedByte_ = b_cache_[numRead];
                            isCachedByte_ = true;
                        }
                        else
                        {
                            isCachedByte_ = false;
                        }
                    }
                    else
                    {
                        numRead = is_.read(b_cache_, 0, cache_.length);
                        if (numRead > 0 && numRead % 2 != 0)  // did we read an odd number of chars
                        {
                            cachedByte_ = b_cache_[numRead-1];
                            isCachedByte_ = true;
                            --numRead;
                        }
                    }
                }
                else if (tableType_ == MB_TABLE)
                {
                    // Max number of bytes for worst-case mixed-byte data scenario is (5x+3)/2.
                    int c = 0;
                    if (mode_ == DB_MODE)
                    {
                        // Begin with a shift-out since we left off in DB_MODE last time.
                        b_cache_[numRead++] = ConvTableMixedMap.shiftOut_;
                        if (isCachedByte_)  // Note that we don't ever cache a shift byte or a single-byte char - we only cache half of a double-byte char.
                        {
                            b_cache_[numRead++] = cachedByte_;
                            isCachedByte_ = false;
                        }
                    }
                    // Don't read too much, we only want to read enough that will fit in our character cache after conversion.
                    int curRead = is_.read(b_cache_, numRead, cache_.length - 1);
                    if (curRead == -1 && numRead == 0)
                    {
                        if (Trace.traceOn_) Trace.log(Trace.CONVERSION, "Cache not filled, end of stream reached.");
                        return false;  // End-of-stream.
                    }
                    if (curRead > -1) numRead += curRead;

                    // Find out which mode we are in when we stopped reading.
                    boolean needToCache = false;
                    for (int i = 0; i < numRead; ++i)
                    {
                        if (mode_ == SB_MODE)
                        {
                            if (b_cache_[i] == ConvTableMixedMap.shiftOut_) mode_ = DB_MODE;
                        }
                        else
                        {
                            // In DB_MODE.
                            if (b_cache_[i] == ConvTableMixedMap.shiftIn_)
                            {
                                mode_ = SB_MODE;
                                needToCache = false;
                            }
                            else
                            {
                                needToCache = !needToCache;
                            }
                        }
                    }
                    if (mode_ == DB_MODE)
                    {
                        if (needToCache)
                        {
                            cachedByte_ = b_cache_[--numRead];
                            isCachedByte_ = true;
                        }

                        // Need to finish with a shift-in.
                        b_cache_[numRead++] = ConvTableMixedMap.shiftIn_;
                    }
                }
                else if (tableType_ == UTF8_TABLE)
                {
                    // Were there leftovers from the previous read?
                    if (leftoverCount > 0)
                    {
                        // Move the leftovers into the cache prior to reading in any more.
                        System.arraycopy(leftovers, 0, b_cache_, 0, leftoverCount);
                        // Fill in the rest of the cache bytes read from the stream.
                        numRead = is_.read(b_cache_, leftoverCount, cache_.length-leftoverCount);

                        // It's possible that the numRead is -1, we still have to pretend to read the number of bytes indicated by leftoverCount.
                        numRead = numRead == -1 ? leftoverCount : numRead + leftoverCount;
                    }
                    else
                    {
                        // No leftovers, try to fill the entire cache with bytes from the stream.
                        numRead = is_.read(b_cache_, 0, cache_.length);
                    }

                    // If no bytes were read, then we have reached the end.
                    if (numRead == -1)
                    {
                        if (Trace.traceOn_) Trace.log(Trace.CONVERSION, "Cache not filled, end of stream reached.");
                        return false;
                    }

                    // If fewer bytes were read than requested, can we assume there are no characaters remaining to be read?
                    if (numRead < cache_.length) 
                    {
                        leftoverCount = 0;
                    }
                    else
                    {
                        // This is where we figure out if a utf-8 character (1-4 bytes) is straddling the cache boundary (|)
                        // The leftoverCount is how many bytes we need to carryover to the next read.
                        // case 1:    0xxxxxxx |
                        // case 2:    110xxxxx 10xxxxxx |
                        // case 3:    1110xxxx 10xxxxxx 10xxxxxx |
                        // case 4:    11110xxx 10xxxxxx 10xxxxxx 10xxxxxx |
                        // leftoverCount = 0
                        //
                        // case 5:    110xxxxx |
                        // case 6:    1110xxxx |
                        // case 7:    11110xxx |
                        // leftoverCount = 1
                        //
                        // case 8:    1110xxxx 10xxxxxx |
                        // case 9:    11110xxx 10xxxxxx |
                        // leftoverCount = 2
                        //
                        // case 10:   11110xxx 10xxxxxx 10xxxxxx |
                        // leftoverCount = 3

                        int n = cache_.length - 1;
                        if ((b_cache_[n] & 0x80) == 0 ) 
                        {
                            leftoverCount = 0;  // Case 1.
                        }
                        else if ((b_cache_[n] & 0xC0) == 0xC0)
                        {
                            leftoverCount = 1;  // Case 5, 6, and 7.
                            leftovers[0] = b_cache_[n];
                        }
                        else if (n>0 && ((b_cache_[n-1] & 0xE0) == 0xE0))
                        {
                            leftoverCount = 2;  // Case 8, and 9.
                            System.arraycopy(b_cache_, n-1, leftovers, 0, leftoverCount);
                        }
                        else if (n>1 && ((b_cache_[n-2] & 0xF0) == 0xF0))
                        {
                            leftoverCount = 3;  // Case 10.
                            System.arraycopy(b_cache_, n-2, leftovers, 0, leftoverCount);
                        }
                        else
                        {
                            leftoverCount = 0;  // Case 2, 3, and 4.
                        }
                        // Adjust the numRead, so it appears the the leftovers aren't in the cache yet.
                        numRead -= leftoverCount;
                    }
                }
                else
                {
                    if (Trace.traceOn_) Trace.log(Trace.ERROR, "Unknown table type during conversion: " + tableType_);
                    throw new InternalErrorException(InternalErrorException.UNKNOWN);
                }

                if (numRead == -1)
                {
                    if (Trace.traceOn_) Trace.log(Trace.CONVERSION, "Cache not filled, end of stream reached.");
                    return false;
                }

                String s = table_.byteArrayToString(b_cache_, 0, numRead, properties_);
                nextWrite_ = s.length();
                s.getChars(0, nextWrite_, cache_, 0);
                nextRead_ = 0;

                if (Trace.traceOn_) Trace.log(Trace.CONVERSION, "Filled cache for reader: " + nextRead_ + "," + nextWrite_ + "," + cache_.length, ConvTable.dumpCharArray(cache_, nextWrite_));
            }

            if (nextRead_ >= nextWrite_) // Still didn't read enough, so try again.
            {
                // This should never happen, but the javadoc for InputStream is unclear if the read(byte[],int,int) method will sometimes return 0 or always read at least 1 byte.
                return fillCache();
            }
            //@xml3 trim xml declaration if needed ""
            //read from stream until declaration is matched.
            if(isXML_ && isFirstRead_)
            {
                String s = String.copyValueOf(cache_,0, nextWrite_);
                int origLen = s.length();
                try{
                    s = JDUtilities.stripXMLDeclaration(s);
                    //here, we know s has no declaration (to begin with or after strip)
                    //next shift chars after declaration to index 0
                    s.getChars(0, s.length(), cache_, 0);
                    int trimmed = origLen - s.length();
                    nextWrite_ -= trimmed;
                    isFirstRead_ = false;
                }catch(SQLException e){
                    //here we know that start was found, but no ending
                    //get some more data and remove any chars after " (nextWrite_ - nextRead_) ? (nextWrite_ - nextRead_) : buffer.length;
                System.arraycopy(cache_, nextRead_, buffer, 0, max);
                nextRead_ += max;
                return max;
            }
        }
        return -1;
    }

    /**
     Reads characters into a portion of the specified array.  If close() is called prior to calling this method, an exception will be thrown.
     @param  buffer  The destination buffer.
     @param  offset  The offset into the buffer at which to begin storing data.
     @param  length  The maximum number of characters to store.  If a zero length is specified, this method does nothing and returns 0.
     @return  The number of characters read, or -1 if the end of the stream has been reached.
     @exception  IOException  If an I/O exception occurs.
     **/
    public int read(char[] buffer, int offset, int length) throws IOException
    {
        if (buffer == null)
        {
            Trace.log(Trace.ERROR, "Parameter 'buffer' is null.");
            throw new NullPointerException("buffer");
        }
        if (length == 0) return 0; // The JDK doesn't throw exceptions when the length is 0.
        if (offset < 0 || offset >= buffer.length)
        {
            Trace.log(Trace.ERROR, "Value of parameter 'offset' is not valid:", offset);
            throw new ExtendedIllegalArgumentException("offset", ExtendedIllegalArgumentException.RANGE_NOT_VALID);
        }
        if (length < 0 || (offset + length) > buffer.length)
        {
            // Note: InputStreamReader throws IndexOutOfBoundsException if length overflow buffer.  We try to be a little nicer.
            Trace.log(Trace.ERROR, "Specified length would overflow buffer:", length);
            throw new ExtendedIllegalArgumentException("length", ExtendedIllegalArgumentException.RANGE_NOT_VALID);
        }
        synchronized (lock)
        {
            if (fillCache())
            {
                int max = length > (nextWrite_ - nextRead_) ? (nextWrite_ - nextRead_) : length;
                System.arraycopy(cache_, nextRead_, buffer, offset, max);
                nextRead_ += max;
                return max;
            }
        }
        return -1;
    }

    /**
     Reads up to length characters out of the underlying stream.  If close() is called prior to calling this method, an exception will be thrown.
     @param  length  The maximum number of Unicode characters to return as a String.  Note that the number of bytes read from the underlying InputStream could be greater than length.  If a zero length is specified, this method does nothing and returns "".
     @return  A String of up to length Unicode characters, or null if the end of the stream has been reached.  The actual number of characters returned may be less than the specified length if the end of the underlying InputStream is reached while reading.
     @exception  IOException  If an I/O exception occurs.
     **/
    public String read(int length) throws IOException
    {
        if (length < 0)
        {
            Trace.log(Trace.ERROR, "Value of parameter 'length' is not valid:", length);
            throw new ExtendedIllegalArgumentException("length", ExtendedIllegalArgumentException.RANGE_NOT_VALID);
        }
        if (length == 0) return "";
        synchronized (lock)
        {
            StringBuffer buf = new StringBuffer();
            if (fillCache())
            {
                while (fillCache() && buf.length() < length)
                {
                    buf.append(cache_, nextRead_++, 1);
                }
                return buf.toString();
            }
        }
        return null;
    }

    /**
     Tells whether this ConvTableReader is ready to be read.  A ConvTableReader is ready if its input buffer is not empty or if bytes are available to be read from the underlying input stream.  If close() is called, a call to ready() will always return false.
     @return  true if the ConvTableReader is ready to read characters; false otherwise.
     @exception  IOException  If an I/O exception occurs.
     **/
    public boolean ready() throws IOException
    {
        synchronized (lock)
        {
            if (table_ == null) // We are closed.
            {
                return super.ready();  // This should throw an IOException.
            }
            return (nextRead_ < nextWrite_) || is_.available() > 0;
        }
    }

    /**
     Skips the specified number of characters in the underlying stream.  If close() is called prior to calling this method, an exception will be thrown.
     @param  length  The number of characters to skip.
     @return  The number of characters actually skipped.
     @exception  IOException  If an I/O exception occurs.
     **/
    public long skip(long length) throws IOException
    {
        if (length < 0)
        {
            Trace.log(Trace.ERROR, "Value of parameter 'length' is not valid: " + length);
            throw new ExtendedIllegalArgumentException("length", ExtendedIllegalArgumentException.RANGE_NOT_VALID);
        }
        if (length == 0) return 0;
        long total = 0;
        synchronized (lock)
        {
            // Note: We can't just call is_.skip(length), since that would just skip 'length' bytes, but we need to skip 'length' characters.
            checkOpen();
            char[] buf = new char[length < cache_.length ? (int)length : cache_.length];
            int r = read(buf);
            total += r;
            while (r > 0 && total < length)
            {
                r = read(buf);
                if (r > 0) total += r;
            }

            if (isXML_)
            {
              if((r!=total) && (length > (total-r))) //@xml5 (if length is > than the total read from previous reads (skips))
              {
                nextRead_ = (int)(length - (total-(long)r));//@xml5 (set nextRead_ to offset that could be after reading in first buffer of chars)
              }
            }
        }
        return total;
    }
}




© 2015 - 2024 Weber Informatics LLC | Privacy Policy