define dataset variables for Dataset Metadata Check against Define XML

[SFJohnson24]

currently for rules that are Dataset Metadata Check against Define XML, the ContentsDefineDatasetBuilder does not report define_dataset_variables. Having this produced as one of the pieces of metadata would be useful for rule development

I think we should use: define_dataset_variable_order
This will eliminate the need for #784 get_column_order_from_define operation

Links to related JIRA Tickets: CORERULES-199

[SFJohnson24]

see also: #784

[mhungria]

I look forward to this enhancement.

[mhungria]

I think there may be confusion on the functionality I requested with get_column_order_from_define ("Returns list of dataset columns in order" as per the Define-XML document; i.e., list of ItemGroupDef/ItemRef ordered by ItemGroupDef/ItemRef/@OrderNumber. It should be similar to get_column_order_from_dataset ("Returns list of dataset columns in order" as included in the actual dataset).
Note the metadata variable define_variable_order_number (ItemGroupDef.ItemRef.OrderNumber) is only one value, not a list.

[gerrycampion]

@mhungria, I think we are on the same page. My idea is that given the following XML:

<ItemGroupDef OID="IG.SE" Name="SE" Domain="SE">
    <ItemRef ItemOID="IT.SE.STUDYID" OrderNumber="1"/>
    <ItemRef ItemOID="IT.SE.DOMAIN" OrderNumber="2"/>
    <ItemRef ItemOID="IT.SE.USUBJID" OrderNumber="3"/>
    <ItemRef ItemOID="IT.SE.ETCD" OrderNumber="5"/>
    <ItemRef ItemOID="IT.SE.SESEQ" OrderNumber="4"/>
</ItemGroupDef>

define_dataset_variable_order will return a list with:

• STUDYID
• DOMAIN
• USUBJID
• SESEQ*
• ETCD*

*Note that ETCD and SESEQ are switched because OrderNumber attribute exists

given the following XML:

<ItemGroupDef OID="IG.SE" Name="SE" Domain="SE">
    <ItemRef ItemOID="IT.SE.STUDYID"/>
    <ItemRef ItemOID="IT.SE.DOMAIN"/>
    <ItemRef ItemOID="IT.SE.USUBJID"/>
    <ItemRef ItemOID="IT.SE.ETCD"/>
    <ItemRef ItemOID="IT.SE.SESEQ"/>
</ItemGroupDef>

define_dataset_variable_order will return a list with:

• STUDYID
• DOMAIN
• USUBJID
• ETCD
• SESEQ

Since there is no OrderNumber attribute, it follows the element order.

Is there anything this is missing?

[mhungria]

The examples for the two use cases where a user wants to retrieve the list of variables in order look ok - thanks!

However, regarding my note "Note the metadata variable define_variable_order_number (ItemGroupDef.ItemRef.OrderNumber) is only one value, not a list.", I would expect to be able to retrieve that value for other use cases that a user needs a variable's order number value as indicated explicitly in the Define-XML document, or implicitly if ItemGroupDef/ItemRef/@OrderNumber is not provided.

[gerrycampion]

Are you saying define_variable_order_number provides the explicit ordering, but does not provide the implicit order? So we need something like define_variable_element_order?