I have a strange doubt regarding list and dictionary in c#
In a list we add items to list by using the following method
using System.Collections.Gen
IDictionary
is for key->value maps, ICollection
is for sets of similar objects.
ICollection
is an interface for collections of similar objects: the controls on a form, the elements in a list, the attributes in an XML tag, and so on. As of .NET 2.0, there's a generic version, so you can refer to a collection of integers as ICollection<int>
.
IDictionary is an interface for mapping one type of object or value to another. It works like a real dictionary, or a phone book: you have a "key" in mind like a person’s name, and when you look it up, you get some information that’s identified by that key, like an address or phone number. Each key can only be listed once, although two different keys are still allowed to have the same value. This is also generic in .NET 2.0, so a dictionary whose keys are strings and whose values are integers would be IDictionary<string,int>
.
A dictionary is actually a collection of key/value pairs: you can use an IDictionary<int,string>
as an ICollection<KeyValuePair<int,string>>
, and you can access the keys and values as separate collections with the Keys and Values properties.
Both ICollection
and IDictionary
are unordered, meaning that although you can retrieve the elements in some order with the CopyTo
method or a foreach loop, that order has no special meaning, and it might change for no apparent reason. That’s the main difference between ICollection
and IList
: a list lets you put items in specific positions, just like an array, and they stay there until you move them.
List<>
and Dictionary<,>
- pretty different data structures which used for different purposes, List is simply a set of items and Dictionary is a set of key-value pairs.
Dictionary is pretty useful when you have a set of complex objects and want to have fast access by let's say ObjectName/ObjectId, in this case you create IDictionary<string, TObject>
where key would be ObjectId and Value would be an object itself.
Some differences:
O(1)
time complexity to access an item (value) by a keyI have a class library which accesses a variety of T-sql sprocs; each sproc returns a single row but varying columns. I needed a general purpose solution for an retrieving the values and Dictionary<> provided a much cleaner solution than List<>.
The class common to all of the wrappers declares
public Dictionary<string, String> datadict = new Dictionary<string, string>();
and
public Dictionary<string, String> LoadData(string sproc, string paramName, string paramValue)
Invoking a Reader, the datadict is loaded with
for (int i = Reader.FieldCount; i != 0; i--)
{
datadict.Add(Reader.GetName(i - 1).Trim(), Reader.GetString(i - 1).Trim());
}
and returns datadict to the calling class which can then retrieve the data much like Reader does; E.g.:
datadict = myData.LoadData("spGetSSN", "", "");
ssn1 = datadict["SSN1"];
ssn2 = datadict["SSN2"];
ssn3 = datadict["SSN3"];
Much cleaner for me that List<>.
Dictionary<K,V>
is an associative array, or map. It is a container that can be indexed by values of any type.List<T>
is an integer indexed array. It is a container that is indexed by contiguous integers.The essential difference therefore is in how the containers are indexed.
Don't fall into the trap of believing that Dictionary<int,T>
is semantically equivalent to List<T>
. The difference is that the indexing of List<T>
is contiguous whereas there can be gaps in the indexing for Dictionary<int,T>
.
Elements in a list have the following characteristics:
They maintain their ordering unless explicitly re-ordered (for example, by sorting the list). They can be of any type, and types can be mixed. They are accessed via numeric (zero based) indices. Elements in a Dictionary have the following characteristics:
Every entry has a key and a value Ordering is not guaranteed Elements are accessed using key values Key values can be of any hashtable type (i.e. not a dict) and types can be mixed Values can be of any type (including other dict’s), and types can be mixed