YamlDotNet : Order serialized properties based on inheritance depth - c#

In a tool project, I generate meta files for Unity3D assets. As all unity asset meta start with the same properties, I created a base type :
public class MetaBase {
public int fileFormatVersion { get; set; }
public Guid guid { get; set; }
public long timeCreated { get; set; }
public string licenseType { get; set; }
}
Serializing an instance of a type inheriting from MetaBase generates a file that unity accepts, but the fileFormatVersion, guid, timeCreated and licenseType are written at the bottom of the file whereas Unity writes them at the top. As I said, it works, but whenever Unity decides to overwrite the meta file, it generates some differences that need to be committed on my project repository, and I would prefer to avoid that if possible.
So, my first idea (after upgrading to YamlDotNet 4.0.0) was to add a TypeInspectorSkeleton which would sort the _IPropertyDescriptor_s depending on the declaring types of the property. But the IPropertyDescriptor does not provide access to the actual Property, nor does it provide access to its baseDescriptor property (for PropertyDescriptor/OverridePropertyDescriptor).
Shouldn't there be some way to access a Property owner type ? Or maybe there is a better may to achieve what I'm trying to do ?

Related

What is the design pattern to use for extending a base class with additional properties?

I have "business entities" and their counterpart for saving them to Azure Storage Table, which requires a few additional properties.
// MyData is the business entity with a few properties
public record MyData_AzureTable : MyData, ITableEntity
{
// Required properties for storing data to Azure Storage Table
public string PartitionKey { get; set; } = "";
public string RowKey { get; set; } = "";
public DateTimeOffset? Timestamp { get; set; }
public ETag ETag { get; set; } = new ETag();
}
I am getting tired of having to duplicate each business entity with its AzureTable counterpart but I can't find the correct pattern to use. Something like that, except it's illegal to inherit from a type parameter.
public record AzureTable<T> : T, ITableEntity
{
public string PartitionKey { get; set; } = "";
public string RowKey { get; set; } = "";
public DateTimeOffset? Timestamp { get; set; }
public ETag ETag { get; set; } = new ETag();
}
What pattern should be used for adding properties to a base class?
The object saved to Azure Table Storage needs to be "flat" (tabular data as property values, no hierarchical data or encapsulation)
Not necessarily a pattern but abstract classes may fit well for your need here. Check out the docs: https://learn.microsoft.com/en-us/dotnet/csharp/language-reference/keywords/abstract
In conclusion, those classes have a base one (which would be your always default AZ properties) and all the other classes than inherit from the abstract one will contain those properties as well without needing implementation (like an interface would) but you can extend the children and add more custom properties to each one.
There is no such pattern, currently C# does not support something like traits (though in some cases similar can be achieved with default interface member implementations) and/or multiple inheritance (basically MyData_AzureTable should inherit from both MyData and AzureTable). If you are really tired of writing "duplicates" for data - you should consider using source generators - you can write quite a simple one which will generate azure tables classes for all required classes (for example marked with special attribute like GenerateAzureTable). Potentially it can generate also some useful methods for mapping, copy constructors and so on.

What is best approach for data inheritation in elasticsearch?

I have an parent class and two child like these:
public class Parent {
public Guid Id { get; set; }
public string Name { get; set; }
}
public class FirstChild {
public string IdentityCode { get; set; }
}
public class OtherChild {
public string RegistrationCode { get; set; }
}
There is a question: Is it a good approach to store these two inherited classes in the same Index inside ElasticSearch?
I see there is a _type property that is added to my docs after they are stored in DB but it has always "doc" value.
I test this code to fill it but it seems it is not working this way.
await ElasticClient.IndexAsync<FirstChild>(child, m => m.Index(IndexName));
And Also, I found this question on SO for retrieving my entries from DB but it is outdated and the API is changed and no more accessible.
I want to know if it is a good approach to store sibling data in the same index how can I do this properly.
As of ES 6.0, it is not possible anymore to store multiple types inside the same index, i.e. the _type field you're referring to will always be either doc or _doc. In ES 8.0, the _type field will be removed altogether.
However, if it makes sense for your use case, you can still decide to store several types inside a single index using a custom type field that is present in your document.
You should strive to only store in the same index data that share the same (or very similar) mapping, which doesn't seem to be the case for Parent, FirstChild and SecondChild, but if you add a public string type property to your classes you can still do it.

Unique attribute for each property

I have written an attribute class which I later used for sorting properties.
[AttributeUsage(AttributeTargets.Property)]
class OrderAttribute : Attribute
{
internal OrderAttribute(int order)
{
Order = order;
}
public int Order { get; private set; }
}
I want this to be unique for class properties, e.g.
Valid scenario
[Order(1)]
public string Tier3 { get; set; }
[Order(2)]
public string Tier4 { get; set; }
Invalid Scenario since value "1" is repeated.
[Order(1)]
public string Tier3 { get; set; }
[Order(1)]
public string Tier4 { get; set; }
PS: Attribute values can be repeated for different class properties but not in same. How can I achieve this?
Although attribute values can be repeated, there's no easy way supported to make sure they are unique (they are run time only), you would need to check when actually sorting as Johnathan has already mentioned. There are ways around this, but is it worth it in the long run? A few options I can think of are:
Hook into the actual build process, create a build task that uses reflection to check and fail if needed.
Run a post-build (post build event) step that loads your dll and reflects on those attribute types.
Create a possible rule that will check for uniqueness.
There may be other ways, but these are the one's I could think of at the moment.

Adding Property to Object in WCF Service Breaks Client

We consume a WCF service using C# code. The client was generated in Visual Studio by right-clicking "Add Service Reference" and pointing it at the WSDL.
Recently, the WCF provider adding some properties to one of the objects they serialize. The class went from
public class MyClass
{
public string Foo { get; set; }
public string Baz { get; set; }
public string Zed {get; set; }
}
to this:
public class MyClass
{
public string Foo { get; set; }
public string Bar { get; set; } //<= New Property
public string Baz { get; set; }
public string Zed {get; set; }
}
On our end, this caused Baz and Zed to suddenly start being null when deserialized, until we updated the service reference. In fact, the real object had some ~20 properties alphabetically after Bar, and they were all null (or 0 for ints, false for bools, etc).
It seems an odd way for the deserialization to fail. It didn't throw an exception or ignore the new properties it didn't know anything about.... it just made every property that appeared alphabetically after the new one deserialize to the default value.
So my question is, what's going on here and how do I prevent it? Preferably, I'd like some kind of setting for the client to tell it to "ignore new properties," but telling the service provider how they can prevent future breaking changes would be fine too.
MSDN has an article which lists the serialization ordering of the datamembers. One key point from that document:
current type’s data members that do not have the Order property of the
DataMemberAttribute attribute set, in alphabetical order.
So if you add a new property, without the Order-property of the DataMemberAttribute, the property is alphabetically ordered.
Based on discussion here, your only options are:
Change the serializer to something else
Make sure that the order of the elements in XML matches the order of your properties. Maybe you can always use the Order-property of the DataMemberAttribute?
Make sure that your dll's line up, I've seen some pretty funky issues in the past where one side of a service was pointing to an outdated dll
also remember the fundamentals of data contracts

How to make a Attribute aware of the Name of the Proprty it is on?

I want to implement a simple attribute that is used to map Database Columns to Properties.
So what i have so far is something that attached like so:
[DataField("ID")]
public int ID { get; set; }
[DataField("Name")]
public String Name { get; set; }
[DataField("BirD8")]
public DateTime BirthDay { get; set; }
Is there a way that I can make the attribute "aware" of the field it is on, so that for the properties where the name is the same as the ColumnName I can just apply the attribute without the name parameter, or would I have to deal with that at the point where I reflect the properties. I want to end up doing just this:
[DataField]
public int ID { get; set; }
[DataField]
public String Name { get; set; }
[DataField("BirD8")]
public DateTime BirthDay { get; set; }
The attribute itself won't be aware of what it's applied to, but the code processing the attributes is likely to be running through PropertyInfo values etc and finding the attributes associated with them. That code can then use both the property and the attribute appropriately.
To make things simpler, you might want to write a method on the attribute to allow it to merge its information with the information from the property, so you'd call:
DataFieldAttribute dfa = propertyInfo.GetCustomAttributes(...); // As normal
dfa = dfa.MergeWith(propertyInfo);
Note that for the sake of sanity this should create a new instance of the attribute, rather than changing the existing one. Alternatively, you might want a whole separate class to represent "the information about a data field":
DataFieldAttribute dfa = propertyInfo.GetCustomAttributes(...); // As normal
DataFieldInfo info = dfa.MergeWith(propertyInfo);
That way you could also construct DataFieldInfo objects without any reference to attributes, which might be a nice conceptual separation - allowing you to easily load the config from an XML file or something similar if you wanted to.
If you don't mind using postsharp you can look Here, at a previous question I have asked which was close. I ended up using the compile time validate to do what I wanted, although there are other options, like CompileTimeInitalize.
public override void CompileTimeInitialize(object element)
{
PropertyInfo info = element as PropertyInfo;
//....
}

Categories