-
Notifications
You must be signed in to change notification settings - Fork 636
Coding Standards
- All system usings should go before custom usings
- No unused usings should be introduced to the code base
- All conditional blocks should be enclosed in curly brackets. If a block is a one word statement, such as
return
, it can be placed on the same line.
// Acceptable...
if(something) return;
if(something)
{
return;
}
//Not acceptable...
if(something)
return;
-
Magic numbers should be avoided. Mark values as
const
and name them according to the MSDN Capitalization Conventions
Class property getters and setters should minimize side effects. Although possible, it's particularly dangerous to mutate other properties in a property setter.
Namespaces should adhere to the Namespace Naming Guidelines provided by Microsoft. Project sub-folder organization in Visual Studio should reflect the namespace structure.
Where not specified otherwise in this document, you should adhere to the C# Coding Conventions.
Use common sense. If the block of code you're writing is abstruse or does a lot in one line, consider commenting it. A good example are dense LINQ calls.
For each method, provide the following:
- A summary describing the purpose of the class and basic implementation details.
Here's an example
/// <summary>
/// Manages instantiation of custom nodes. All custom nodes known to Dynamo should be stored
/// with this type. This object implements late initialization of custom nodes by providing a
/// single interface to initialize custom nodes.
/// </summary>
public class CustomNodeLoader {
...
}
For properties, provide the following:
- A summary what the property does.
Here's an example:
/// <summary>
/// A dictionary of node ids, keyed by the node names.
/// </summary>
public ObservableDictionary<string, Guid> NodeNames
{
get;
private set;
}
For each method, provide the following:
- A summary written in the imperative mood.
- A descriptions for each parameter
- A description of what the method returns if applicable.
Here's an example
/// <summary>
/// Get a node header (guid, name, category) from a given path. This path points
/// usually to a .dyf file.
/// </summary>
/// <param name="path">The path from which to get the header</param>
/// <param name="guid">A reference to the guid (OUT) Guid.Empty if function returns false. </param>
/// <param name="name">A reference to the node name (OUT) Empty string if function returns false </param>
/// <param name="category">A reference to the category name (OUT) Empty string if function returns false </param>
/// <returns>Whether we successfully obtained the header or not. </returns>
public static bool GetHeaderFromPath(string path, out Guid guid, out string name, out string category ) {
...
}
For zero-touch methods, it is highly recommended that you include the names of the output ports. This is discussed in more depth here.
For all new pieces of functionality, provide unit tests.
The Dynamo project uses NUnit for testing. Using ReSharper allows you to run unit tests directly from Visual Studio. If you don't have ReSharper, you probably should get it. Otherwise, get yourself a copy of NUnit - it's free.
Unit tests should be placed in a separate assembly from the one you are testing and named by appending "Tests" to the project name. For example, for the "DynamoCore" project the unit test assembly is called "DynamoCoreTests." Putting unit tests in a separate assemblies allows us to omit unit tests when shipping and naming them as such allows us to easily identify those assemblies.
If applicable, design your classes such that they can be unit-tested without needing to instantiate a large amount of other infrastructure (particularly user interfaces). This is all obvious good OOP. If you are passing DynamoModel
as a reference to a method or class, you're probably doing something bad.
Looking for help with using the Dynamo application? Try dynamobim.org.
- Dynamo 2.0 Language Changes Explained
- How Replication and Replication Guide work: Part 1
- How Replication and Replication Guide work: Part 2
- How Replication and Replication Guide work: Part 3