清除清单 <string> 从字符串字典列表中删除所有值，C＃

Question

Trying to read a csv file, and take the first word in the stream, throw it in to a dictionary while the following words get added to a list in that dictionary.

However, I find that (during debugging) when, inside my loop I decide to clear my list, all of the values it had added to the dictionary previously also get cleared. I guess I am mistaken in assuming it makes a copy of the list, it is actually just referencing that same list? Should I be creating a new list with every iteration? Code below:

public class TestScript : MonoBehaviour {

// Use this for initialization
void Start() {

    Dictionary<string, List<string>> theDatabase = new Dictionary<string, List<string>>();
    string word;
    string delimStr = ",.:";
    char[] delimiter = delimStr.ToCharArray();
    List<string> theList = new List<string>();


    using (StreamReader reader = new StreamReader("testComma.csv")) {
        while (true) {
            //Begin reading lines
            string line = reader.ReadLine();
            if (line == null) {
                break;
            }
            //Begin splitting lines, adding to array.
            string[] split2 = line.Split(delimiter, StringSplitOptions.RemoveEmptyEntries);

            //Loop to hold the first word in the stream
            for(int i = 0; i <= 0; i++) {
                word = split2[i];

                //loop to hold the following words in to list.
                for (int y = 1; y < split2.Length; y++) {
                    theList.Add(split2[y]);
                }

                //Add word/list combo in to the database
                theDatabase.Add(word, theList);

                //clear the list.
                theList.Clear();
            }
        }
    }

    foreach (KeyValuePair<string, List<string>> pair in theDatabase) {
        string keys;
        List<string> values;

        keys = pair.Key;
        values = pair.Value;
        print(keys + " = " + values);

    }
  }
}

The bottom foreach loop is just so I can see the results. Also, any critique is welcome in regards to how this is written, as I'm a beginner.

Answer 1

Yes, you're adding the same object to your dictionary.

You can just change :

theDatabase.Add(word, theList);

To :

theDatabase.Add(word, theList.ToList());

Method ToList() makes shallow copy of your List<T>

Answer 2

C# is pass by reference.
So, theList and the list in your Dictionary are the same object.

The simplest solution is to stop clearing your List and create a new one every time instead:

for(int i = 0; i <= 0; i++) {
    List<string> theList = new List<string>(); // it is in a loop now

    word = split2[i];

    //loop to hold the following words in to list.
    for (int y = 1; y < split2.Length; y++) {
        theList.Add(split2[y]);
    }

    //Add word/list combo in to the database
    theDatabase.Add(word, theList);

    //clear the list.
    //theList.Clear(); - not required anymore
}

It is more readable and clear solution: create a list, insert items, paste a list into the dictionary, continue the iteration.
It is also much more performant since there is no List clearing - List<T>.Clear() is a linear operation, which takes O(n) operations.

Answer 3

Yes, as everyone says, lists are reference types. You need to make a copy to avoid the .Clear() clearing all the lists.

You could always write your code like this:

void Start()
{
    string delimStr = ",.:";
    Dictionary<string, List<string>> theDatabase = 
        File
            .ReadAllLines("testComma.csv")
            .Select(line => line.Split(delimStr.ToCharArray(), StringSplitOptions.RemoveEmptyEntries))
            .ToDictionary(x => x[0], x => x.Skip(1).ToList());

    /* foreach here */
}

}

This doesn't have the problem with the list references.