Naming convention for a C# Dictionary - c#

How do we name a dictionary variable?
Say in my method I have Dictionary<string, List<string>> dictionary;, where the keys of the dictionary are country names and the values are lists of province/state names. How should I rename dictionary?
I know we can create a Country class for this example. But please don't mention this alternative because I'm thinking of good naming convention here.

ProvincesByCountry

I use mostly one of these:
CountryToStatesDictionary
CountryToStatesMap
CountryToStatesMapping

ProvincesByCountry is not explicit enough, as it sounds like mapping countries to provinces one to one. When accessing ProvincesByCountry["Germany"] I'd expectedly assume one value is an object rather than a list of objects.
My personal pattern is similar:
[Plural of a noun describing the value]By[Singular of a noun describing the key]
However, if a noun describing the value is plural by its nature, then I use the postfix arrays, or lists, as in English you can't really "pluralise" a plural. I personally always stick to arrays, regardless of the actual implementation of IEnumerable or IEnumerable< T> I'm using, be that List, or Array or whatever.
In your case it turns to:
ProvinceArraysByCountry
Tells what it is with scientific precision.
I apply this rule recursively if there are dictionaries as values. The order of accessing then goes in reverse to the order of words in the name. Imagine you add planets:
ProvinceArraysByCountryByPlanet["Earth"]["Germany"][0] = "Bavaria"
ProvinceArraysByCountryByPlanet["Earth"]["Germany"][1] = "Rhineland-Palatinate"
And finally the last little stroke here. If such dictionary maps object properties and the objects themselves, then I leave out the word describing the object in the key section. Here is what I mean:
NodesByIndex[node.Index] = node; // - Do
NodesByNodeIndex[node.Index] = node; // - Don't
I use this pattern unconditionally which is good as it leaves absolutely no room for guess. Con is it generates fairly long names sometimes. But I have no idea how to have always explicit but always short names. You always have to compromise. And it's of course a matter of taste.
This pattern doesn't work (or at least you'd break your brain) when keys are also dictionaries or when you have list of dictionaries of lists of dictionaries or some other crazy exotic stuff. But I don't remember having that many levels of nesting, so I'm happy with it.

I like XtoYMap or YFromX.

Naming is always contextual. So in this specific case some name specifying the country to state mapping is appropriate.
if this was just a device for a loop within a larger context and then discarded, I normally just go with a short temp type var like...
var dict = GetCountryStateMapping();
foreach(var item in dict)
{
//Something....
}

provinces, provinceMap, provinceDictionary
All come to mind. I like provinceMap my self. If it's a member field I would add an "m_" prefix, as in "m_provinceMap".

Related

Modeling data in C#

I'm trying to come up with a good model for (what I would consider) a somewhat complicated class. There are two entities which are independent of each other (in that they have use, by themselves, elsewhere in the app), a User Group and an Event Type.
A User Group has a list of event types to which they are entitled. And beyond that, there is a list of strings which applies to each combination of User Group and Event which tells the app strings to replace later on (so a Dictionary where they key is a field to replace and the value is the value to replace it with).
I come from more of a SQL background, so it's easy for me to think in terms of tables and primary keys. This structure would be keyed off UserGroupID, EventTypeID, NameToReplace, but when I try to come up with a way to do it in C#, I end up with ugly (or at least I think they're ugly) structures like Dictionary<int, Dictionary<tuple<int, string>, string>> or Dictionary<int, Dictionary<int, Dictionary<string, string>>>
I could also do away with the dictionary concept and just make a list of tuples, or a list of custom classes which tie all the logical "keys" together.
My question(s) come down to, is this sort of nested collection structure common and/or a good idea? Are there best practices when modeling data like this anyone can point me to?
Thanks in advance!
So, this might be controversial, but I wanted to put this version out for discussion, to know how appropriate this approach would be.
Assuming you don't have millions of groups and event types, you could pack those into kind of Composite Key in db terms with bits shift :
public static int Combine(int value1, int value2)
{
return value1 | (value2 << 8);
}
Then Dictionary<int, string> would be work fine.
dictionary.Add(Combine(UserGroupID, EventTypeID), NameToReplace)
And to get the value :
dictionary[Combine(UserGroupID, EventTypeID)] // Or TryGetValue()

Is it OK to use dictionaries when I don't need to quickly access their values?

Normally, I use a dictionary like a list, but with a key of a different type. I like the ability to quickly access individual items in the dictionary without having to loop through it until I find the item with the right property (because the property I'm looking for is in the Key).
But there is another possible use of a dictionary. I could just use the Key to store property A and the Value to store property B without ever using the dictionary's special functionality. For example, I could store a list of persons just by storing the forename in the key and the family name in the value (let's assume, for the sake of simplicity, that there won't ever be two people with the same forename, because I just couldn't come up with an better example). I would only use that dictionary to loop through it in a foreach loop and add items to it (no removing, sorting or accessing individual items). There would actually be no difference to using a List<KeyValuePair<string, string>> from using a Dictionary<string, string> (at least not in the example that I gave - I know that I could e. g. store multiple items wiht the same key in the list).
So, to sum it up, what should I do when I don't need to use the special functionalities a dictionary provides and just use it to store something that has exactly two properties:
use a Dictionary<,>
use a List<KeyValuePair<,>
use a List<MyType> with MyType being a custom class that contains the two properties and a constructor.
Don't use dictionaries for that.
If you don't want to create a class for this purpose, use something like List<Tuple<T1,T2>>. But keep in mind a custom class will be both more readable and more flexible.
Here's the reason: it will be much more easy to read your code if you use proper data structures. Using a dictionary will only confuse the reader, and you'll have problems the day a duplicate key shows up.
If someone reads your code and sees a Dictionary being used, he will assume you really mean to use a map-like structure. Your code should be clear and your intent should be obvious when reading it.
If you're concerned with performance you should probably store the data in a List. A Dictionary has lots of internal overhead. Both memory as well as CPU.
If you're just concerned with readability, chose the data structure that best captures your intent. If you are storing key-value pairs (for example, custom fields in a bug tracker issue) then use a Dictionary. If you are just storing items without them having some kind of logical key, use a List.
It takes little work to create a custom class to use as an item in a List. Using a Dictionary just because it gives you a Key property for each item is a misuse of that data structure. It is easy to create a custom class that also has a Key property.
Use List<MyType> where MyType includes all the values.
The problem with the dictionary approach is that it's not flexible. If you later decide to add middle names, you'll need to redesign your whole data structure, rather than just adding another field to MyType.

List of Dictionaries (Search optimization, C#)

I have a List of Dictionaries, List<Dictionary<String,Object>>. The key is an identifier of some abstract record. These Dictionaries come from various places. The size of each Dictionary is in the range [0, 1000].
All Dictionaries contain unique keys. After accumulating some Dictionaries I must make a search by key. It could be done by iterating the List and calling search method on every Dictionary or it could be done by copying all Dictionaries into one. These approaches do not offer very good performance. I am interested in ways to optimize this task.
Edit:
Thank you guys! Maybe I'll change the accumulation method and as result eliminate the problem itself!
Are you expecting there to be lots of key fetches after an initial population phase? If so, amalgamate everything into a single dictionary. If you'll only be doing a few fetches, I can't see any way you could get better than asking every dictionary.
Of course you could create a hybrid approach: create a new (initially empty) dictionary for the amalgamated results, and populate it as you're asked for keys - by searching through all the rest each time you're asked for a key which isn't already in your "big" dictionary.
Is there no way of predicting which dictionary would have a particular key?
If there is any way to localize a dictionary of interest by specifying a key, you can try, naturaly, to create a cross association table where you can try to match the key to dictionary.
If not, imho, don't see any other option that just iterate over collection and ask for the key , may be using standart for and not nicer linq coding.
Adding to what Jon said, there is an API called as PowerCollections which contains MultiDictionary. If my memory is not corrupted, I believe, you can use this for the purpose mentioned.
http://powercollections.codeplex.com/discussions/242163
It sounds like you have lots of dictionaries to "speed up" (assumption of motive) searches that are limited to certain "abstract record" types.
You can get away with one single dictionary, but on limited searches check the result is required abstract record type after finding it. Rather than maintaining a single dictionary for each and every abstract record type as at present.

Is there a way to search a Dictionary in multiple ways using key searching?

Basically I have a Dictionary<Guid, Movie> Movies collection and search for movies using Guid, which is basically movie.Guid. It works great, but I also want to be able to search the same dictionary using movie.Name without looping through each element.
Is this possible or do I have to create another Dictionary<K, V> for this?
Just have two Dictionaries, one of them having the guid as its key and the other with the name as its key.
If you don't want to look at every element, you need to index it the other direction. This means another Dictionary to get O(1).
You can iterate across the variables but then you arnt getting the constant-time searching value in a dictionary (because of the way that the keys are hashed.) The answer above regarding using two dictionarys to hash references to your object may be a good solution if you dont have too many objects to reference.
You could search with the Values property:
dictionary.Values.Where(movie => movie.Name == "Some Name")
You'll lose the efficiency of a key based look up, but it will still work.
Since dictionaries are for one-way mapping you can't get keys from values.
You'll need two dictionaries.
There is also a suggestion:
You can use a custom hash function for keys instead of GUIDs and store Movie Names hash as keys. Then you can actually perform two way search in your dictionary.
Rather than using two dictionaries, you'd be much better off using one container class that has two dictionaries inside it.
Some guy named Jon came up with a partial solution to this (which you could easily build upon), leaving his code here: Getting key of value of a generic Dictionary?
You can't use that dictionary to do that search with anything like the same efficiency. But you can easily just run a LINQ query against your dictionary's Values property, which is just collection of the Movie values.
var moviesIWant = From m in movieLookup.Values
Where m.Name == "Star Wars"
Select m
Some thoughts:
When you find your answer though, you would not have the guids, unless they were also a property of movie.
For a small dictionary, this is just fine. For large and repeated searches, you should consider the creation of other dictionaries keyed on the other values you wish to search on. Only in this way would you achieve the speed of a guid lookup comparable to your original dictionary.
You could create another dictionary keyed by Name. Once you've done this, you could search this dictionary by it's key and it would have the same super-efficiency of your original dictionary, even for a very large dictionary.
var moviesByName = movieLookup.Values.ToDictionary(m => m.Name, m => m)
No I don't believe it is possible. You'll have to use another dictionary.
If you are going to want to search on more movie attributes you may be better off moving the data down to a database and use that for querying. That is what databases are good for after all.

unique key value collection in C#

I want to build 2-dimentional collection where i need unique combination of key value pairs. For example Domain "Company" (Id: 1) can have MachineName "Machine1" and "Machine2", but cannot add another MachineName "Machine1" again. Another Domain "Corporate" (Id:2) can have another machineName "Machine1".
here my collection will be like this 1-Machine1, 1-Machine2, 2-Machine1.
Adding 1-Machine1 or 2-Machine1 should be invalid entry.
Please suggest datatype or approach for this.
I cannot use Dict> datatype, because it may hamper performance if size grows.
So you need some kind of collection with a unique key, and each item within this collection is unique.
So really, you're talking about a dictionary where the value within the dictionary is a unique collection.
Assuming you're only talking about strings, I'd be using something like:
Dictionary<string, HashSet<string>>
Someone correct me if I'm wrong, but I think the advantage of using these generic structures is you can (right off the bat), do this:
Dictionary<string, HashSet<string>> domains = new Dictionary<string, HashSet<string>>();
domains["Domain1"].Add("Machine1");
I'm sorry, but from your description it still sounds like a Dictionary implementation would be a good fit.
If and when the performance of the application suffers due to the speed of the dictionary, then you can revisit the problem and roll your own specifically tailored solution.
You could do something like this:
Dictionary<String, List<String>> mapping = new Dictionary<string, List<string>>();
mapping.Add("1",new List<string>());
mapping["1"].Add("Machine1");
mapping["1"].Add("Machine2");
This will give you a one to many mapping between domain and machines.
or the NameValueCollection class would do the same.
Do you need to be able to look up the list of domains with given machine name efficiently? Otherwise a Hashtable<String, HashSet<String>> seems like a good fit.
There also seems to be something called NameValueCollection which might be a good fit if you change the defaults so that it isn't case- or culture-sensitive.
You didn't state this as a requirement, but my guess is that you also need to be able to query the data structure for all of the machines for a specific "domain". Ex. list the machines belonging to Company 1. This is the only reason I can think of where the performance of using a Dictionary might be unacceptable (since you would have to traverse the entire list to find all of the matching entries).
In that case you might consider representing the data as a tree.
Edit:
Based on your comment above, you could just concatenate your keys as a string and use a HashSet to check if you've already stored that key.

Categories

Resources