Table of Contents

Search

  1. Preface
  2. Components
  3. Business services
  4. File listeners
  5. Fixed-width file formats
  6. Hierarchical schemas
  7. Intelligent structure models
  8. Mapplets
  9. Saved queries
  10. Shared sequences
  11. User-defined functions

Components

Components

Editing the nodes

Editing the nodes

You can change the output by editing the
intelligent structure
nodes. The available options depend on the type of sample file that you base the model on and on the output data type of the node.
You perform actions nodes on the
Visual Model
tab.
Perform the following actions to edit the nodes:
Change the root node.
To change the root, right-click the new root node and select
Mark As Root
. The brown-colored node shows the current root node. The green-colored node shows the selected root node.
Changes to root nodes only affect the model if you use it in a
Data Integration
elastic mapping or in a data engineering mapping.
Intelligent Structure Discovery
doesn't apply the changes if you use the model in a Structure Parser transformation in a
Data Integration
mapping.
Rename a node.
To rename a node, right-click a node, select
Rename
, and then type in a name, for example
Street Address
. Node names are not case sensitive. Names must be unique across sibling nodes, that is, nodes that reside under the same node.
Apply unique names to nodes
To apply unique naming across the structure, right-click the root node and select
Apply Unique Naming
. When you apply unique naming to a model, if the data in the sample files contains fields with identical names in different groups,
Intelligent Structure Discovery
uses suffixes to rename identical names to unique names.
Restore original node names.
You can restore the original node names and remove the suffixes that
Intelligent Structure Discovery
adds to identical names in different hierarchy levels.
To restore the original node names in a structure, right-click the root node of the structure and select
Restore Original Names
.
Don't restore original node names for models that you plan to use in a
Data Integration
mapping with a relational output. Under these conditions, node names must be unique across the structure.
Combine data in two nodes.
To join two nodes, click and drag a node to the other node. For example, in the following model, the
time
node is selected and dragged to the
date
node:
The date node is combined with the time node.
The following image shows the combined node:
After you combine the date node with the time node, the model includes one node named datetime that contains the data from both nodes.
Combine all child node data into a parent node.
To flatten a parent node and merge its child nodes, right-click the node and select
Flatten
. When you flatten a node, the data from the child nodes is merged into one output field for the parent node. The child nodes are no longer separate fields in the output.
Exclude a node from the intelligent structure and re-include an excluded node.
To exclude a node from the structure, right-click the node and select
Exclude from Structure
. When you exclude a node from the structure, the data from the node is not part of the output of the model and is not parsed during run time. To re-include an excluded node in the structure, right-click the node and select
Include in Structure
.
Determine how
Intelligent Structure Discovery
parses data from Microsoft Excel files.
You can determine whether
Intelligent Structure Discovery
parses the formatted data that Microsoft Excel shows or the underlying, unformatted data when it parses fields in a Microsoft Excel file.
To determine how
Intelligent Structure Discovery
parses the data, right-click the node and select either of the following options:
  • Extract Formatted Data.
    Intelligent Structure Discovery
    parses the fields as formatted data.
  • Extract Unformatted Data.
    Intelligent Structure Discovery
    parses the fields as unformatted data.
Split a node.
To split the data in a node into two nodes, for example, for a credit card number for which you only want to store the last four digits, perform the following steps:
  1. Right-click the node and select
    Split
    . The
    Split
    dialog box appears.
  2. In the
    Split
    dialog box, highlight the data in the display field to indicate where to split the data. The data is split and displayed as two fields, with one field containing the highlighted data, and the second field containing the non-highlighted data.
  3. To confirm the split, click
    OK
    .
  4. To undo the split, click
    Reset
    .
For example, you want to split a node with year-month-date-hour-minute data in one field into two fields. You want to create a year-month-date field and an hour-minute field. The following image shows the highlighted section that you want to split:
In the Split Field dialog box, you highlight the part of the field that you want to split from the rest.
Treat JSON model as repeating
By default, when the sample JSON model does not contain repetitions, the Structure Parser transformation does not parse repeating groups. If you want to parse repeating groups, right-click the JSON root node and select
Treat as Recurring
.
Treat an element as repeating
By default, when an element in the sample JSON or XML model does not contain repetitions, the Structure Parser transformation does not parse repetitions of the element. If you want to parse repetitions of the element, right-click the node and select
Treat as a List
. To undo a
Treat as a List
action, right-click the node and select
Treat as Single
.
Define table headers in Microsoft Excel files.
You can define rows and columns in Microsoft Excel files as table headers. To define rows or columns as table headers, click the Edit button in the sample pane, select rows or columns that show within the table boundaries, and click the Apply icon. To clear the selection of table headers, click the Remove all headers icon.


Updated January 22, 2021