String Maps and String Sets

In This Topic

Nevron Open Vision provides two special classes for working with string sets - the NStringMap and the NStringSet. Although similar to standard maps and sets on the outside, they are implemented differently and provide and optimize way for various string search methods such as:

GetStringsStartingWith / GetItemsStartingWith - gets of all strings/items that start with a given prefix. This method is commonly used for implementing of auto complete functionlity.
WildcardSearch - gets all strings that match a given wildcard search pattern. The popular MS DOS symbology is used for the pattern, i.e. you can use the special character '*' to denote zero or more characters and '?' to denote exactly one character.
GetSimilarStrings / GetSimilarItems - gets all strings/items that are similar to a given string based on specific criteria. The criteria is specified through an instance of the NStringSimilarityContext class. This method is commonly used for implementing of spell check and correction functionality.

The results of these methods depends on whether the string comprisons are case sensitive or not, which is specified by passing true or false to the constructor respectively. The following code example demonstrates how to create and initialize a string set and perform a wildcard serch on it:

String Set Example	Copy Code
string[] Words = new string[] { "bar", "bat", "car", "cat", "far", "fat", "ward", "warda", "wa", "w", "wra", "warhead" }; NStringSet stringSet = new NStringSet(true); for (int i = 0; i < Words.Length; i++) { stringSet.Add(Words[i]); } INSet<string> matches = m_Words.WildcardSearch("*ar");

String Set Example

Copy Code

string[] Words = new string[] { "bar", "bat", "car", "cat", "far", "fat", "ward", "warda", "wa", "w", "wra", "warhead" };

NStringSet stringSet = new NStringSet(true);
for (int i = 0; i < Words.Length; i++)
{
    stringSet.Add(Words[i]);
}

INSet<string> matches = m_Words.WildcardSearch("*ar");

The set returned by this WildcardSearch method in this case will contain the strings "bar", "car" and "far".

Note that the optimization of the NStringMap class for the string search methods above comes at the cost of slower access by key. For the NStringMap it is not O(1) as in a hash table based map, but O(log N). That is why you should use the generic NMap class with a string type parameter, when you don't need the special string search methods that the NStringMap offers.