Sort Order

[sukishan]

Sort Order
A sort order specifies the rules used by SQL Server to interpret, collate, compare, and present character data. For example, a sort order defines whether 'a' is less than, equal to, or greater than 'b'. A sort order defines whether the collation is case-sensitive, for example whether 'm' is equal or not equal to 'M'. It also defines if the collation is accent-sensitive, for example whether 'á' is equal or not equal to 'ä'.

SQL Server 2000 uses two sort orders with each collation, one for Unicode data and another for the character code page.

Many SQL Server collations use the same code page, but have a different sort order for the code page. This allows sites to choose:

Whether characters will simply be sorted based on the numeric value represented by their bit patterns. Binary sorting is fastest because SQL Server does not have to make any adjustments and can use fast, simple sorting algorithms. Binary sort orders are always case-sensitive. Because the bit patterns in a code page may not be arranged in the same sequence as defined by the dictionary rules for a specific language, binary sorting sometimes does not sort characters in a sequence users who speak that language might expect.

Between case-sensitive or case-insensitive behavior.
Between accent-sensitive or accent-insensitive behavior.