Recognize duplicate items in a list

I could just post the link to SO.en , but I find it interesting to have the answer here.

var list = new List<string> { "a", "a", "b", "b", "r", "t" };

var distinct = new HashSet<string>();    
var duplicates = new HashSet<string>();

foreach (var s in list)
    if (!distinct.Add(s))
        duplicates.Add(s);

// distinct == { "a", "b", "r", "t" }
// duplicates == { "a", "b" }

Only duplicates

List<string> list = new List<string>() { "a", "a", "b", "b", "r", "t" };

var dups = list.GroupBy(x => x)
    .Where(x => x.Count() > 1)
    .Select(x => x.Key)
    .ToList();

First you can group all the items given a grouping criterion ... and then we put each grouping inside a dictionary:

var dicionarioDeArrays = lista
    .GroupBy(x => x)
    .ToDictionary(x => x.Key, x => x.ToArray());

Then we can filter the created dictionary, to return only groups that have an X number of elements, in > 1 :

var listaDeDuplicatasAgrupadasEmArrays = dicionarioDeArrays
    .Where(x => x.Value.Length > 1)
    .Select(x => x.Value)
    .ToList();

The advantage of this approach is that it is extremely flexible ... you can for example:

• group using a property:

  // agrupar pelo nome das pessoas (caso os itens sejam pessoas)
  .GroupBy(x => x.NomeDaPessoa)
  

• filter by different amounts:

  // quero pegar os trigêmeos (ou seja, 3 irmãos com o mesmo pai)
  .GroupBy(x => x.Pai)
  
  // ... e depois na hora de verificar, quero apenas os grupos com 3
  .Where(x => x.Value.Length == 3)
  

The MoreLinq package has the optimal list extension called DistinctBy :

var distinctList = Itens.DistinctBy(x => x.UmaPropriedadeQualquer).ToList();

Now, if you just want to know who it was, it would look something like:

var repetidos = Itens
                 .GroupBy(x => x.UmaPropriedadeQualquer)
                 .Where(g => g.Count() > 1)
                 .Select(g => g.Key)
                 .ToList();