Share via


Attribute Data Translation Dialog Box (Analysis Services - Multidimensional Data)

Use the Attribute Data Translation dialog box to set the column that contains the translation caption data, as well as the collation and sort order to be used with the translated data. You can display the Attribute Data Translation dialog box by:

  • Clicking New caption column or Edit caption column from the Toolbar pane on the Translations tab of Dimension Designer.

  • Right-clicking the Translation Details pane on the Translations tab of Dimension Designer and selecting New caption column or Edit caption column.

Options

  • Attribute
    Displays the selected attribute.

  • Language
    Displays the selected language.

  • Translated caption
    Sets the current translated caption for the selected attribute.

  • Translation columns
    Sets the column where the translation caption data is stored. Only one column can be selected. All related tables that are referred to by the primary dimension table will be displayed.

  • Collation designator
    Sets the collation designator for the selected attribute. By default, the current Windows collation is selected. Click the down arrow to select from the available collations.

  • Binary
    Select this option to sort and compare data based on the bit patterns defined for each character. Binary sort order is case-sensitive, that is, lowercase precedes uppercase, and accent-sensitive. This is the fastest sorting order.

    If this option is not selected, Analysis Services follows sorting and comparison rules as defined in dictionaries for the associated language or alphabet.

    Note

    If this option is selected, the Case sensitive, Accent sensitive, Kana sensitive, and Width sensitive options are disabled.

  • Case sensitive
    Select this option to sort and compare data based on the dictionary rules provided for the associated language or alphabet, and to distinguish between uppercase and lowercase letters.

    If not selected, Analysis Services considers the uppercase and lowercase versions of letters to be equal. Analysis Services does not define whether lowercase letters sort lower or higher in relation to uppercase letters when Case-sensitive is not selected.

  • Accent sensitive
    Select this option to sort and compare data based on the dictionary rules provided for the associated language or alphabet, and to distinguish between accented and unaccented characters. For example, 'a' is not equal to 'รก'.

    If not selected, Analysis Services considers the accented and unaccented versions of letters to be equal.

  • Kana sensitive
    Select this option to sort and compare data based on the dictionary rules provided for the associated language or alphabet, and to distinguish between the two types of Japanese kana characters: Hiragana and Katakana.

    If not selected, Analysis Services considers Hiragana and Katakana characters to be equal.

  • Width sensitive
    Select this option to sort and compare data based on the dictionary rules provided for the associated language or alphabet, and to distinguish between a single-byte character (half-width) and the same character when represented as a double-byte character (full-width).

    If not selected, Analysis Services considers the single-byte and double-byte representation of the same character to be equal.

See Also

Reference

Analysis Services Designers and Dialog Boxes (Multidimensional Data)

Translation Details (Translations Tab, Dimension Designer) (Analysis Services - Multidimensional Data)