org.marc4j.MarcSplitStreamWriter Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of freelib-marc4j Show documentation
An easy to use Application Programming Interface (API) for working with MARC and MARCXML in Java.
There is a newer version: 2.6.12

package org.marc4j;

import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.io.OutputStream;

import org.marc4j.marc.ControlField;
import org.marc4j.marc.DataField;
import org.marc4j.marc.Leader;
import org.marc4j.marc.Record;
import org.marc4j.marc.Subfield;

public class MarcSplitStreamWriter extends MarcStreamWriter {

    private final int recordThreshold;

    private final String fieldsToSplit;

    /**
     * Creates a MarcSplitStreamWriter with a record threshold and fields to split from the supplied
     * {@link OutputStream}.
     *
     * @param out
     * @param threshold
     * @param fieldsToSplit
     */
    public MarcSplitStreamWriter(final OutputStream out, final int threshold, final String fieldsToSplit) {
        super(out, false);
        recordThreshold = threshold;
        this.fieldsToSplit = fieldsToSplit;
    }

    /**
     * Creates a MarcSplitStreamWriter with an encoding, a record threshold, and fields to split from the supplied
     * {@link OutputStream}.
     *
     * @param out
     * @param encoding
     * @param threshold
     * @param fieldsToSplit
     */
    public MarcSplitStreamWriter(final OutputStream out, final String encoding, final int threshold,
            final String fieldsToSplit) {
        super(out, encoding, false);
        recordThreshold = threshold;
        this.fieldsToSplit = fieldsToSplit;
    }

    /**
     * Writes a Record object to the writer.
     *
     * @param record - the Record object
     */
    @Override
    public void write(final Record record) {
        boolean doneWithRec = false;
        for (final DataField df : record.getDataFields()) {
            if (!df.getTag().matches(fieldsToSplit)) {
                continue;
            }
            df.setId(null);
        }

        while (!doneWithRec) {
            try {
                int previous = 0;
                final ByteArrayOutputStream data = new ByteArrayOutputStream();
                final ByteArrayOutputStream dir = new ByteArrayOutputStream();

                // control fields
                for (final ControlField cf : record.getControlFields()) {
                    data.write(getDataElement(cf.getData()));
                    data.write(Constants.FT);
                    dir.write(getEntry(cf.getTag(), data.size() - previous, previous));
                    previous = data.size();
                }

                // data fields
                for (final DataField df : record.getDataFields()) {
                    if (df.getTag().matches(fieldsToSplit)) {
                        continue;
                    }
                    data.write(df.getIndicator1());
                    data.write(df.getIndicator2());
                    for (final Subfield sf : df.getSubfields()) {
                        data.write(Constants.US);
                        data.write(sf.getCode());
                        data.write(getDataElement(sf.getData()));
                    }
                    data.write(Constants.FT);
                    dir.write(getEntry(df.getTag(), data.size() - previous, previous));
                    previous = data.size();
                }
                // data fields
                doneWithRec = true;
                for (final DataField df : record.getDataFields()) {
                    if (previous >= recordThreshold) {
                        doneWithRec = false;
                        break;
                    }
                    if (!df.getTag().matches(fieldsToSplit)) {
                        continue;
                    }
                    if (!(df.getId() == null || df.getId().intValue() != 0)) {
                        continue;
                    }
                    df.setId(new Long(0));
                    data.write(df.getIndicator1());
                    data.write(df.getIndicator2());
                    for (final Subfield sf : df.getSubfields()) {
                        data.write(Constants.US);
                        data.write(sf.getCode());
                        data.write(getDataElement(sf.getData()));
                    }
                    data.write(Constants.FT);
                    dir.write(getEntry(df.getTag(), data.size() - previous, previous));
                    previous = data.size();
                }
                dir.write(Constants.FT);

                // base address of data and logical record length
                final Leader ldr = record.getLeader();

                final int baseAddress = 24 + dir.size();
                ldr.setBaseAddressOfData(baseAddress);
                final int recordLength = ldr.getBaseAddressOfData() + data.size() + 1;
                ldr.setRecordLength(recordLength);

                // write record to output stream
                dir.close();
                data.close();

                if (!allowOversizeEntry && (hasOversizeLength)) {
                    throw new MarcException("Record has field that is too long to be a valid MARC binary record. "
                            + "The maximum length for a field counting all of the sub-fields is 9999 bytes.");
                }
                writeLeader(ldr);
                out.write(dir.toByteArray());
                out.write(data.toByteArray());
                out.write(Constants.RT);

            } catch (final IOException e) {
                throw new MarcException("IO Error occured while writing record", e);
            } catch (final MarcException e) {
                throw e;
            }
        }
    }

}