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 an optimized way for various string search methods such as:

FindStrings, FindItems - finds all strings/items that match a given key using a specified string match mode. The available match modes are:

Value	Description
StartsWith	Matches all strings that start with a given substring. This is the default value.
WordStartsWith	Matches all strings that contain a word starting with a given substring.
MultiWordStartsWithCount	Matches all strings that contain a word starting with one of the words of a given substring and orders the matched strings by the number of matched words.
Contains	Matches all strings that contain a given substring.
EndsWith	Matches all strings that end with a given substring.
WordEndsWith	Matches all strings that contains a word ending with a given substring.
Wildcard	Matches all strings that match a given wildcard 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. For example, the pattern "B?l" in a list of countries matches "Belarus", "Belgium", "Belize", "Bolivia" and "Bulgaria".

FindSimilarStrings, FindSimilarItems - finds 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 spell check and correction functionality.

The results of these methods depend on whether the string comparisons are case sensitive or not, which is specified by passing true or false to the constructor. 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.FindStrings("*ar", ENStringMatchMode.Wildcard);

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.FindStrings("*ar", ENStringMatchMode.Wildcard);

The set returned by wildcard pattern search in this case will contain the strings "bar", "car" and "far".

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