Release 10.1A: OpenEdge Development:
Programming Interfaces

Inferring definitions from XML data

If the temp-table or ProDataSet has no definition, and the READ-XML() method does not specify a schema-location, and the XML document does not contain references to XML Schema, READ-XML() will attempt to construct schema from the XML data. This is known as inferring schema.

If an element in the instance data contains the "xsi:type" attribute, the field is assigned a Progress data type based on XML Schema mapping rules. See Appendix D "XML Reference of Data Type and Code Page mappings." If the element is not typed, the field will be CHARACTER. All attributes and text elements will become CHARACTER fields.

Inferring schema for a temp-table

Suppose you have an XML document for a single temp-table, but you have no XML Schema for it. You can either call the READ-XML() method on a temp-table handle and create a dynamic temp-table object in Progress; or, you can call the READ-XML() method on a ProDataSet handle and create a dynamic ProDataSet object with a single temp-table buffer in Progress.

For example, examine this XML document:

<?xml version="1.0"?> <Rows> <ttFam RelativeName="Ken Koberlein"> <Relation>Spouse</Relation> <IncludedOnBenefits>true</IncludedOnBenefits> <EmpNum>1</EmpNum> </ttFam> <ttFam RelativeName="Kelly Smith"> <Relation>Daughter</Relation> <IncludedOnBenefits>true</IncludedOnBenefits> <EmpNum>3</EmpNum> </ttFam> <ttFam RelativeName="Mark Smith"> <Relation>Spouse</Relation> <IncludedOnBenefits>true</IncludedOnBenefits> <EmpNum>3</EmpNum> </ttFam> <ttFam RelativeName="Sam Smith"> <Relation>Son</Relation> <IncludedOnBenefits>false</IncludedOnBenefits> <EmpNum>3</EmpNum> </ttFam> </Rows>

Using READ-XML() on a temp-table object with this XML document will create a temp-table named ttFam, with four CHARACTER fields: RelativeName, Relation, IncludedOnBenefits and EmpNum.

Using READ-XML() on a ProDataSet object with this XML will create a ProDataSet named Rows, with one temp-table buffer, ttFam, with all its buffer-fields.

Inferring schema for a ProDataSet

Suppose you have an XML document for a dataset, but you have no XML Schema for it. You can call the READ-XML() method on a ProDataSet handle and create a dynamic ProDataSet object in Progress.

For example, examine this XML document:

<?xml version="1.0"?> <myDataSet> <Department> <DeptCode>400</DeptCode> <DeptName>Sales</DeptName> </Department> <Employee EmpNum="3"> <DeptCode>400</DeptCode> <LastName>Smith</LastName> <FirstName>Lee</FirstName> <Address>1342 Atlantic Ave</Address> <City>Boston</City> </Employee> <Family RelativeName="Ken Koberlein"> <Relation>Spouse</Relation> <IncludedOnBenefits>true</IncludedOnBenefits> <EmpNum>3</EmpNum> </Family> </myDataSet>

Using READ-XML() on a ProDataSet object will create a ProDataSet named myDataSet with three temp-table buffers: Department, Employee and Family. The Department temp-table has two CHARACTER fields: DeptCode and DeptName. The Employee temp-table has six CHARACTER fields: EmpNum, DeptCode, LastName, FirstName, Address and City. The Family temp-table has four CHARACTER fields: RelativeName, Relation, IncludedOnBenefits, and EmpNum.

Using READ-XML() on a temp-table object will fail, since there is more than one temp-table represented in the XML document.