This topic provides details on completing the Data Reuse page in the "Generate Parameterized Messages from Traffic" wizard. These settings determine how new data from the traffic file will extend and/or update existing repository data sets and data records.
The left panel displays data sets. Any data sets that are already present in the specified repository will be marked with the (existing) label.
The right panel displays the record types and fields for a group. Any record types or fields that match existing data repository records will be marked with the (matching) label.
A record identity is the subset of a record type’s fields, which uniquely identifies that record type. For example, a bank customer record type might have 15 different fields, but its identity might use only social security number and account number. Or, a book record might have ISBN as its identity.
Identities enable you to correlate imported traffic data with existing data repository records. This matching helps determine which data from the traffic file is new and which match existing data. When matching data is detected, record import settings determine whether the existing record will be referenced/shared or whether it will be updated.
For more details on record identities, see Specifying Record Identities.
If record identities have not already been specified for the data set associated with this traffic file, they can be set in the record type tree shown in the wizard.
Fields that are already part of an identity are marked with green icons. Existing record types (annotated with (existing) labels in the tree) cannot have their identities adjusted in the wizard (you cannot remove identities or add new ones here). If you want to adjust identity settings on an existing data set then you need to do it using the Data Repository editor (as described in Specifying Record Identities).
If there are no existing identities and you want to specify that you want a field to be used as the identity:
To remove a field from the identity:
For greater control over the schema of the data being imported, you can rename record types and fields that will be created. This lets you customize names, as well as indicate that items marked with different names are actually the same.
Existing records and fields cannot be renamed, but new duplicate ones can be created.
To rename a record type of field:
The new name will be indicated in the tree (to the right of the arrow) and the icon will be marked with a red asterisk (*).
If you chose to rename an existing record type or field, note that the original entry will remain unchanged and a new one will be added immediately below it. Data from the traffic file will use the new record type or field.
Assume that the recorded traffic references multiple elements (e.g., address, work_address, home_address) that are actually of the same address type. By default, the wizard will treat them as different types (since they have different names). By renaming them, you can indicate that they should be treated as the same type.
If you right-click work_address, and rename it to address, then address child nodes that were not in work_address will be added to work_address. Additionally, work_address child nodes that were not in address will be added to address.
If you rename home_address to address, then home_address, work_address, and address are all merged together. The node text indicates which types are being merged. Any renaming and identity settings applied after the merge will be applied to all three types.
When you rename a merged type, you will be asked whether the renaming should apply to all the merged types, or only to the selected type. For example, renaming any of the merged types to addr and applying it to all the merged types would result in the following:
If you then renamed work_address to secondary_address and applied that change to only the selected node, it would be "unmerged" from the others—reverting its child list to the original—and then renamed. The wizard will treat it as a different type than addr.
Two record types cannot be merged if one type is an ancestor of the other anywhere in the traffic file.
You can choose among the following options to control how new data from the traffic file will extend and/or update existing repository data sets:
You can control whether matching data (data that matches existing record types, as determined by the identity) reuses existing record types or updates an existing record. Available options are:
For example, assume an identity for SocialSec and existing records with FirstName, LastName, SocialSec, and Email.
With the Update option...
With the Reuse option...
If you prefer to create a duplicate record type with a different name, right-click the name of the matching record type, choose Rename option, then enter the new record type name.
The tree will then show both the original record type and the new (renamed) record type.