The last record in a CSV file may or may not have a newline ... how to implement this in DFDL?

[Costello, Roger L.]

Hi Folks,

I am creating a DFDL schema for the CSV format as defined in RFC 4180.

I have learned that a CSV file may or may not have a header record.

I have learned that each record is separated by a newline and the last record 
may or may not have a new line.

My DFDL Schema is working great except when the last record contains a newline.

With this input:

[cid:image003.png@01D5915E.1A9E6840]

I get this XML:

<csv>
  <record>
    <field>1997</field>
    <field>Ford</field>
    <field>E350</field>
    <field>ac, abs, moon</field>
    <field>2999.99</field>
  </record>
  <record>
    <field>1999</field>
    <field>Chevy</field>
    <field>Venture Extended Edition</field>
    <field>4900.00</field>
  </record>
  <record>
    <field>1999</field>
    <field>Chevy</field>
    <field>Venture Extended Edition</field>
    <field>Very Large</field>
    <field>5000.00</field>
  </record>
  <record>
    <field>1996</field>
    <field>Jeep</field>
    <field>Grand Cherokee</field>
    <field>MUST SELL! air, moon roof, loaded</field>
    <field>4799.00</field>
  </record>
  <record>
    <field></field>
  </record>
</csv>

Notice the last <record> element has just one <field> element and it is empty. 
I don't want that last record.

I tried specifying in the declaration for the field element that 
dfdl:indexOccurs must be greater than 0 but that didn't help. See below. What's 
the solution, please? I've attached the TDML file.  /Roger

<xs:element name="field" maxOccurs="unbounded" type="xs:string"
    dfdl:escapeSchemeRef="Quotes"
    dfdl:occursCountKind="implicit">
    <xs:annotation>
        <xs:appinfo source="http://www.ogf.org/dfdl/";>
            <dfdl:assert test="{ dfdl:occursIndex() gt 0 }"
                message="{'If the last record is empty, then no fields should 
be generated'}" />
        </xs:appinfo>
    </xs:annotation>
</xs:element>

csv.tdml.xml
Description: csv.tdml.xml