[Missing <summary> documentation for "N:Iesi.Collections"]
Classes
| Class | Description | |
|---|---|---|
 | DictionarySet | DictionarySet is an abstract class that supports the creation of new Set types where the underlying data store is an IDictionary instance. You can use any object that implements the IDictionary interface to hold set data. You can define your own, or you can use one of the objects provided in the Framework. The type of IDictionary you choose will affect both the performance and the behavior of the Set using it. To make a Set typed based on your own IDictionary, simply derive a new class with a constructor that takes no parameters. Some Set implmentations cannot be defined with a default constructor. If this is the case for your class, you will need to override Clone() as well. It is also standard practice that at least one of your constructors takes an ICollection or an ISet as an argument. |
| HashedSet |
Implements a Set based on a hash table. This will give the best lookup, add, and remove
performance for very large data-sets, but iteration will occur in no particular order.
|
| HybridSet |
Implements a Set that automatically changes from a list to a hash table
when the size reaches a certain threshold. This is good if you are unsure about
whether you data-set will be tiny or huge. Because this uses a dual implementation,
iteration order is not guaranteed!
|
| ImmutableSet | Implements an immutable (read-only) Set wrapper. Although this is advertised as immutable, it really isn't. Anyone with access to the basisSet can still change the data-set. So GetHashCode() is not implemented for this Set, as is the case for all Set implementations in this library. This design decision was based on the efficiency of not having to Clone() the basisSet every time you wrap a mutable Set. |
| ListSet |
Implements a Set based on a list. Performance is much better for very small lists
than either HashedSet or SortedSet. However, performance degrades rapidly as
the data-set gets bigger. Use a HybridSet instead if you are not sure your data-set
will always remain very small. Iteration produces elements in the order they were added.
However, element order is not guaranteed to be maintained by the various Set
mathematical operators.
|
| Set | A collection that contains no duplicate elements. |
| SortedSet |
Implements a set based on a sorted tree. This gives good performance for operations on very
large data-sets, though not as good - asymptotically - as a HashedSet.
However, iteration occurs in order. Elements that you put into this type of collection must
implement IComparable, and they must actually be comparable. You can't mix
String and Int32 values, for example.
|
| SynchronizedSet |
Implements a thread-safe ISet wrapper.
|
Interfaces
| Interface | Description | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
 | ISet | A collection that contains no duplicate elements. This interface models the mathematical Set abstraction. The order of elements in a set is dependant on (a)the data-structure implementation, and (b)the implementation of the various Set methods, and thus is not guaranteed. None of the Set implementations in this library are guranteed to be thread-safe in any way unless wrapped in a SynchronizedSet. The following table summarizes the binary operators that are supported by the Set class.
|
