Creating the Metadata Import File

Before you begin loading, you must create an import file for each artifact that you want to import (dimensions, smart lists, and exchange rate tables). The import file must contain a header record, and then below the header record, a list of metadata records that you want to import or update. The file format can be .csv (comma-delimited), or .txt (tab-delimited or other delimiter characters).

The file consists of these sections:

  1. A header record, the first line in the file:

    • Lists the dimension and any member properties used by subsequent metadata records. The header record and subsequent records do not need to include all properties. Properties that are not included are inherited from the corresponding parent’s default property value.

    • Is case-sensitive

    • Can list properties in any order, as long as the subsequent metadata records are in the same order

    • Can use comma or tab delimiters. Other delimiter characters can be used if they are supported and the same delimiter character is used throughout the file. For a list of supported delimiter characters and exceptions, see Other Supported Delimiter Characters.

  2. After the header record, include a list of metadata records that you want to import or update. Each metadata record contains a delimited list (comma, tab, or other) of property values that matches the order designated in the header record. A metadata record can skip a property that is specified in the header record because the default property is assumed.

    Note:

    Only characters that are included in the UTF-8 character set are supported in import or export files. Characters that aren't included in the UTF-8 character set are considered invalid.

    To delete UDAs and attributes in metadata load files, replace the UDAs and attributes with <none> instead of leaving the property value blank.

    Note that properties that are not included in the header or metadata records are inherited from the corresponding parent’s default properties. If your application has multiple cubes enabled, the metadata that you load specifically for one cube may result in errors in other cubes. In the metadata load file, to disable members for any cube, set the column "Plan Type(Cubename)" to FALSE. For example, if you have cubes Consol2 and Consol3 enabled, but want to import metadata only for Consol3, set the PlanType(Consol3) to TRUE, and the PlanType(Consol) and PlanType(Consol2) to FALSE.

For an example import file, see Example: Entity Dimension Import File.