File Splitters

To configure new Marc Exchange file splitters, specify the following class name in the File Splitters mapping table:

• com.exlibris.primo.publish.platform.harvest.splitters.marc_exchange.MarcExchangeSplitter

The following table lists the supported parameters:

This splitter is helpful for ILS systems that can extract MARC XML but may not be able to expand holdings information into the xml and may not be able to add the OAI header.
Customers will need extract their own data which will include the ID, status and holdings info – outside the MARC record and also the MARC metadata – in one line.
The splitter will take each bib according to the bib ID and merge the holdings/location/availability information into the MARC XML as 999 subfields.
In cases of multiple location of the same bib the splitter output will be MARC xml with multiple 999 fields.

Format – The splitter can deal with records with the following format:

• ID | Status | MARC XML | Holdings information in multiple fields delimited (no specific limit to number of fields)

Input data should be consistent with the following requirements:

• The extract output must have 1 line per each bib + holdings.
• The extract output file needs to be sorted according to the record ID (control no. – 1st field).
• The extract output may be at library level or item level.
• Records with multiple holdings/items should be extracted multiple time – each holdings records will have the full ID, Status, MARC and will only differ in the holdings info. The number of lines written to the file should be equal the number of items that bibliographic record has. (i.e. the bib information will be repeated and only the holdings information will vary from line to line).
• The 1st field must always be the record ID.
• The 2nd field must always be the record status – the delete flag.
• The 3rd field must always be the MARC xml (legal xml).
• Fields 4 and onwards are for holdings, location and availability information.

To configure new MARC XML with character delimited holdings splitter, specify the following class name in the File Splitters mapping table:

• com.exlibris.primo.publish.platform.harvest.splitters.marc_delimited.MarcDelimitedFileSplitter

The following table lists the supported parameters:

Input example (one line, pipe delimited between fields):

477222 | n | <?xml version="1.0" encoding="utf-8"?>

<collection>

    <record>

        <leader>00397cas a2200157 a 4500</leader>

        <controlfield tag="001">0000477222</controlfield>

        <controlfield tag="005">20060227110415</controlfield>

        <controlfield tag="008">970721c19759999caumr1p       0uuua0eng  </controlfield>

        <datafield tag="022" ind1=" " ind2=" ">

            <subfield code="a">0730-0158</subfield>

        </datafield>

        <datafield tag="040" ind1=" " ind2=" ">

            <subfield code="a">211006</subfield>

            <subfield code="c">211006</subfield>

        </datafield>

        <datafield tag="090" ind1=" " ind2=" ">

            <subfield code="a">780.5</subfield>

            <subfield code="b">K441</subfield>

            <subfield code="x">P</subfield>

        </datafield>

        <datafield tag="245" ind1="0" ind2="0">

            <subfield code="a">Keyboard.</subfield>

        </datafield>

        <datafield tag="260" ind1="0" ind2="0">

            <subfield code="a">San Francisco, CA. :</subfield>

            <subfield code="b">Miller Freeman, Inc.,</subfield>

            <subfield code="c">1975-</subfield>

        </datafield>

        <datafield tag="650" ind1=" " ind2="0">

            <subfield code="a">Music</subfield>

        </datafield>

    </record>

</collection> | MED | W0187164 | available | P 780.5 K441 v.35 no.1-12 2009 | 5462 |

Output example:

Record ID: 477222

<?xml version="1.0" encoding="utf-8"?>

<record>

    <leader>00397cas a2200157 a 4500</leader>

    <controlfield tag="001">0000477222</controlfield>

    <controlfield tag="005">20060227110415</controlfield>

    <controlfield tag="008">970721c19759999caumr1p 0uuua0eng </controlfield>

    <datafield tag="999">

        <subfield code="a">MED</subfield>

        <subfield code="b"> W0187164</subfield>

        <subfield code="c">available</subfield>

        <subfield code="d"> P 780.5 K441 v.35 no.1-12 2009</subfield>

        <subfield code="e">5462</subfield>

    </datafield>

    <datafield tag="022" ind1=" " ind2=" ">

        <subfield code="a">0730-0158</subfield>

    </datafield>

    <datafield tag="040" ind1=" " ind2=" ">

        <subfield code="a">211006</subfield>

        <subfield code="c">211006</subfield>

    </datafield>

    <datafield tag="090" ind1=" " ind2=" ">

        <subfield code="a">780.5</subfield>

        <subfield code="b">K441</subfield>

        <subfield code="x">P</subfield>

    </datafield>

    <datafield tag="245" ind1="0" ind2="0">

        <subfield code="a">Keyboard.</subfield>

    </datafield>

    <datafield tag="260" ind1="0" ind2="0">

        <subfield code="a">San Francisco, CA. :</subfield>

        <subfield code="b">Miller Freeman, Inc.,</subfield>

        <subfield code="c">1975-</subfield>

    </datafield>

    <datafield tag="650" ind1=" " ind2="0">

        <subfield code="a">Music</subfield>

    </datafield>

</record>

//parse a record from the file and create a RecordData object for it
RecordData record = getNextParsedRecord();
 
//call the recordSaver to save the parsed record into the dataabse
recordSaver.save(record);

//parse a record from the file and create a RecordData object for it
RecordData record = getNextParsedRecord();
 
//adding 2 different full texts
record.addExtension(new ExtensionData("FULLTEXT", "blah blah balh…"));
record.addExtension(new ExtensionData("FULLTEXT", "kuku kuku kuku…"));
 
//call the recordSaver to save the parsed record into the dataabse
recordSaver.save(record);

RecordData rd = getNextRecord();

recordSaver.save(rd);

public interface IRecordSaverWrapper {


    //will be called before recordSaver.save(rd) is called.

    //returning false will cause the record to not be saved

    public boolean doBeforeSave(RecordData rd);


    //will be called once, after the class is created

    public void init(IPrimoLogger logger, Map<String, Object> params);

}

//call the file splitters doneParsing function

splitter.doneParsing();


//call end parsing interface if defined

doEndParsing(recordSaver, params);


//save to the database the remaining records

recordSaver.saveLeftOvers(config);

public interface IEndParsingHandler {


    public void init(IRecordSaver recordSaver,

            IPrimoLogger logger, Map<String, Object> params);


    public void execute();

}