遍历数组列表的更有效方法

Question

My program takes in thousands of PL/SQL functions, procedures and views, saves them as objects and then adds them to an array list. My array list stores objects with the following format:

ArrayList<PLSQLItemStore> storedList = new ArrayList<>(); 
storedList.add(new PLSQLItemStore(String, String, String,   Long            ));
storedList.add(new PLSQLItemStore(Name,   Type,   FileName, DatelastModified));

What I wanted to do is remove duplicate objects from the array-list based on their Name. The older object would be removed based on its dateLastModified variable. The approach i took was to have an outer loop and an inner loop with each object comparing themselves to every other object and then changing the name to "remove" if it was considered to be older. The program then does one final pass backwards through the array-list removing any objects whose name is set as "remove". While this works fine it seems extremely inefficient. 1000 objects will mean 1,000,000 passes need to be made. I was wondering if someone could help me make it more efficient? Thanks.

Sample Input:

storedList.add(new PLSQLItemStore("a", "function", "players.sql", 1234));
storedList.add(new PLSQLItemStore("a", "function", "team.sql", 2345));
storedList.add(new PLSQLItemStore("b", "function", "toon.sql", 1111));
storedList.add(new PLSQLItemStore("c", "function", "toon.sql", 2222));
storedList.add(new PLSQLItemStore("c", "function", "toon.sql", 1243));
storedList.add(new PLSQLItemStore("d", "function", "toon.sql", 3333));

ArrayList Iterator:

for(int i = 0; i < storedList.size();i++)
{
    for(int k = 0; k < storedList.size();k++)
    {
        if (storedList.get(i).getName().equalsIgnoreCase("remove"))
        {
            System.out.println("This was already removed");
            break;
        }

        if (storedList.get(i).getName().equalsIgnoreCase(storedList.get(k).getName()) &&  // checks to see if it is valid to be removed
           !storedList.get(k).getName().equalsIgnoreCase("remove") &&
           i != k )
        {
            if(storedList.get(i).getLastModified() >= storedList.get(k).getLastModified())
            {
                storedList.get(k).setName("remove");
                System.out.println("Set To Remove");
            }
            else
            {
                System.out.println("Not Older");
            }
        }
    } 
}

Final Pass to remove Objects:

System.out.println("size: " + storedList.size());
for (int i= storedList.size() - 1; i >= 0; i--)
{
    if (storedList.get(i).getName().equalsIgnoreCase("remove"))
    {
        System.out.println("removed: " + storedList.get(i).getName());

        storedList.remove(i);                
    }
}
System.out.println("size: " + storedList.size());

Answer 1

You need to make PLSQLItemStore implement hashCode and equals methods and then you can use Set to remove the duplicates.

public class PLSQLItemStore {

    private String name;   

    @Override
    public int hashCode() {
        int hash = 7;
        hash = 47 * hash + (this.name != null ? this.name.hashCode() : 0);
        return hash;
    }

    @Override
    public boolean equals(Object obj) {
        if (obj == null) {
            return false;
        }
        if (getClass() != obj.getClass()) {
            return false;
        }
        final PLSQLItemStore other = (PLSQLItemStore) obj;
        if ((this.name == null) ? (other.name != null) : !this.name.equals(other.name))     {
            return false;
        }
        return true;
    }
}

And then just do Set<PLSQLItemStore> withoutDups = new HashSet<>(storedList);

PS equals and hashCode are generated by NetBeans IDE.

Answer 2

Put them in a Guava ArrayListMultimap<String,PLSQLItemStore> .

Add each PLSQLItemStore using name as the key.

When you're done adding, loop through the multimap, sort each List with a Comparator<PLSQLItemStore> which sorts by dateLastModified , and pull out the last entry of each List - it will be the latest PLSQLItemStore .

Put these entries in another Map<String,PLSQLItemStore> (or List<PLSQLItemStore> , if you no longer care about the name) and throw away the ArrayListMultimap .

Answer 3

Building off of Petr Mensik's answer, you should implement equals and hashCode . From there, you can put items into the map. If you come across a duplicate, you can decide then what to do:

Map<String, PLSQLItemStore> storeMap = new HashMap<String, PLSQLItemStore>();
for(PLSQLItemStore currentStore : storedList) {

    // See if an item exists in the map with this name
    PLSQLItemStore buffStore = storeMap.get(currentStore.getName();

    // If this value was never in the map, put it in the map and move on
    if(buffStore == null) {
        storeMap.put(currentStore.getName(), currentStore);
        continue;
    }

    // If we've gotten here, then something is in buffStore.
    // If buffStore is newer, put it in the map.  Otherwise, do nothing
    // (this might be backwards --  I didn't quite follow your logic.  
    // Feel free to correct me
    if(buffStore.getLastModified() > currentStore.getLastModified())
        storeMap.put(currentStore.getName(), currentStore);

}

Your map is dup-free. Because Map is a Collection , you can iterate through it later in your code:

for(PLSQLItemStore currentStore : storeMap) {
    // Do whatever you want with your items
}