Visokio website     Downloads     Video tutorials     KnowledgeBase  
Sources: XML data import wizard? - Visokio Forums
Sources: XML data import wizard?
  • VoteVote Up1Vote Down     chris November 24, 2011 5:06AM
    At the moment when you try and open an XML file in Omniscope you need to specify a XSL document to allow Omniscope convert the XML to the correct format. It would be much more user friendly if Omniscope allowed you to configure the XML import inside the dialog, by allowing you to specify the row element, columns etc. This would only work for simple XML files but would mean you wouldn't have to create an XSL file for each distinct XML document you wanted to open.
  • 6 Comments
  •     chris November 24, 2011 5:28AM
    Another note on this idea, not all XML files are tabular, so an XML file may not contain a clearly defined set of rows/columns.
  •     steve November 24, 2011 5:51AM
    You can think of this as a visual way of building up the equivalent XSL translation used to convert from one schema to the Visokio XML data schema.
  •     sherriff November 25, 2011 5:26AM
    We would like to submit IATA industry standard schema file for use in testing
  •     sherriff January 27, 2012 10:48PM
    I would like to ask if there is any prospect of an XML import wizard?
  •     steve January 28, 2012 2:39AM
    The difficulty is XML is a very flexible and hierarchical format, and one document's solution might not work for another's.

    If you're still talking about the invoice XML above, for example, this is highly hierarchical. In this case what node in the XML would constitute a record in the output? How would attributes and body text of child nodes be handled?

    We're considering a solution which can be entirely automatic without configuration, where each terminating XML node (a node without children) results in a single record, and all attributes of all parent nodes are repeated for every record.

    In this case that wouldn't work very well, since InvoiceIDVATRegistrationNumber and InvoiceIDName1 would result in separate records.

    You could then configure what node level constituted a record, e.g. Invoice. But this would then need to deal with flattening multiple children (InvoiceHeader, InvoiceIDDetails) into a single record cell for each invoice field.

    But the principle of our current planned automatic solution is to convert the data into tabular form in a consistently reproducible way; you could then use other operations to restructure the data appropriately.

    Currently you have to write an XSL transformation file, which is the most flexible way of dealing with an arbitrary schema.
  •     sherriff January 28, 2012 11:48AM
    Thanks for explaining this.

    It would still be exceptionally helpful if an XML wizard were available to assist in defining the record start/end nodes and other hierarchy nodes. In effect to build an XSL file with the assistance of a wizard. I have no idea whether this might be possible since I am not really familiar with the complexities of XML schema.

Welcome!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In Apply for Membership

Tagged