Array Performance/Optimization

Question

I am reading a CSV file and I would like to cache the results in an array.

This is my getter/setter:

private RedirectionRule[] RedirectionRules
{
    get
    {
        if (_redirectionRules == null)
        {
            return new RedirectionRule[MAXLENGTH];
        }

        return _redirectionRules;
    }
    set
    {
        _redirectionRules = value;
    }
}

Is this the right approach to an optimal way of caching the results?

Answer 1

I don't think there's much point in returning a new array in your getter when _redirectionRules is null. If you set the property in your code that parses the CSV, then it will be cached.

---

In other words, somewhere you should have a function like this to parse the CSV data (as an example, I've put it in the RedirectionRule class, but you could have a RedirectionRuleParser class or something like that depending on your needs):

class RedirectionRule {
    public static RedirectionRule Parse(string text) {
        // some code here to parse text for your RedirectionRule object
    }

    public static RedirectionRule[] ParseCsv(string csv) {
        string[] values = csv.Split(',');
        RedirectionRule[] rules = new RedirectionRule[values.Length];

        for (int i = 0; i < values.Length; i++) {
            rules[i] = RedirectionRule.Parse(values[i]);
        }
    }
}

Then, if you have code like this somewhere, you are caching the data:

string csv = "RuleType1,RuleType1,RuleType1";
RedirectionRules = RedirectionRule.ParseCsv(csv);

Elsewhere, where you want to access the data you have cached:

if (RedirectionRules != null) {
    // do something with your cached data
} else {
    // I don't know, throw an exception or something
}

The only thing your example code would accomplish by creating a new RedirectionRule[MAXLENGTH] array in your property's getter would be to sneak past the RedirectionRules != null check above, thereby opening up the possibility of accessing data that looks like it's been cached but really came out of thin air.

Answer 2

Hard to say in general, it depends on what you need exactly.

Just another thought: When you create a full sized array in the getter, why do you need a setter anymore? There should probably be only one place where the array is created, just to keep it clean.

Then - when you create the array once, and this for sure, why not creating it in the constructor?

Answer 3

I'm not sure I understand the question really, as I am not sure which of the following questions you are asking

1. Are there any optimisations possible to the shown code?
2. Is an array the optimal container for caching the results?
3. What is the optimal way of loading the csv into this container?

Perhaps you were asking all 3, so I will do my best to answer all 3

1. It is not optimal to check for null in the getter each time. I would suggest you initialise it in the constructor instead.
2. Yep, array sounds good to me
3. I think you would have to provide more info in order to answer this question fully. String.Split will get your csv into a string[] which is what I would normally do.