See Also: SortKey Members
A culture-sensitive comparison of two strings depends on each character in the strings having several categories of sort weights, including script, alphabetic, case, and diacritic weights. A sort key serves as the repository of these weights for a particular string. In particular, the value of a System.Globalization.SortKey object is its key data, which is a series of bytes that encode the string, culture-specific sorting rules, and user-specified compare options. A comparison using sort keys consists of a bitwise comparison of the corresponding key data in each sort key.
In effect, the erload:System.Globalization.CompareInfo.Compare method generates the sort key for each string, performs the comparison, then discards the sort key and returns the result of the comparison. In fact, the erload:System.Globalization.CompareInfo.Compare method does not generate an entire sort key and then perform the comparison. Instead, the method generates the key data for each text element, that is, base character, surrogate pair, or combining character sequence, in each string. The method then compares the key data for the corresponding text elements. The operation terminates as soon as the ultimate result of the comparison is determined. Sort key information is computed, but no System.Globalization.SortKey object is created. This strategy is economical in terms of performance if both strings are compared once, but becomes expensive if the same strings are compared many times.
The SortKey.Compare(SortKey, SortKey) method requires generation of a System.Globalization.SortKey object for each string before performing the comparison. This strategy is expensive in terms of performance for the first comparison because of the time and memory invested to generate the System.Globalization.SortKey objects. However, it becomes economical if the same sort keys are compared many times.
For example, suppose you write an application that searches a database table for the row in which the string-based index column matches a specified search string. The table contains thousands of rows, and comparing the search string to the index in each row will take a long time. Therefore, when the application stores a row and its index column, it also generates and stores the sort key for the index in a column dedicated to improving search performance. When the application searches for a target row, it compares the sort key for the search string to the sort key for the index string, instead of comparing the search string to the index string.
The CompareInfo.GetSortKey(string, CompareOptions) method returns a System.Globalization.SortKey object with the value based on a specified string and System.Globalization.CompareOptions value, and the culture associated with the underlying System.Globalization.CompareInfo object. If a security decision depends on a string comparison or case change, the application should use the CompareInfo.GetSortKey(string, CompareOptions) method of the invariant culture to ensure that the behavior of the operation is consistent, regardless of the culture settings of the operating system.
The application should use the following steps to obtain the CompareInfo.GetSortKey(string, CompareOptions) method:[The 'ordered' type of list has not been implemented in the ECMA stylesheet.]
Working with the value of a System.Globalization.SortKey object is equivalent to calling the Windows API LCMapString method with the LCMAP_SORTKEY value specified. However, for the SortKey object, the sort keys for English characters precede the sort keys for Korean characters.
System.Globalization.SortKey objects can be serialized, but only so that they can cross AppDomain objects. If an application serializes a System.Globalization.SortKey object, the application must regenerate all of the sort keys when there is a new version of the .NET Framework.
For more information about sort keys, see Unicode Technical Standard #10, "Unicode Collation Algorithm" at the tp://go.microsoft.com/fwlink/?linkid=37123.