This project is read-only.

OData Aggregation Extension on V4

Topics: ASP.NET Web API
Apr 4, 2014 at 10:57 AM
Edited Apr 6, 2014 at 4:54 PM
I would like to extend the Web API with some aggregation function. It would be great if it is already on the roadmap (could not find that). But I may also be able to contribute in this area.

Some background:

What I am looking for is some more understanding of where to start extending this framework. From the source code alone I could not directly see how to add a queryoption. If someone could point me in the right direction.. this would be great.

I can extend the ODataQueryOption class to start supporting new odata uri options, but the aggregations would require an anonymous type to be returned. Any ideas on this?


/Blogs/$apply=groupby(Author/LastName, aggregate($count as BlogCount))

This should result in a count of blogs grouped by the LastName of the Author of the BlogEntry

"@odata.context": "$metadata#Blogs(Author(LastName),BlogCount",
"value": [
{ "": null, "Author": { "LastName" : "Clancy" }, "BlogCount": 23 },
{ "": null, "Author": { "LastName" : "Kirk" }, "BlogCount": 2 },
{ "": null, "Author": { "LastName" : "George" }, "BlogCount": 12 }
May 17, 2014 at 12:07 PM
I am trying to do the same. Were you able to make any headway?
May 27, 2014 at 9:54 PM
I developed a LINQ extension that allows to do this:
Using the AdaptiveLINQ .QueryByCube() method, the use of a projection leads to an aggregation effect.
Thus, the following C# code:
        .Select(x => new {
          LastName = x.AuthorLastName,
          BlogCount = x.TotalCount
will execute the query:
Blogs.GroupBy(x => x.Author.LastName)
        .Select(group => new {
          LastName = group.Key,
          BlogCount = group.Count()
Exposing the IQueryable expression Blogs.QueryByCube(myBlogAnalysisCube) as an OData feed, you can query:
Of Course, this projection can be combined with a filter or any other OData operator:
BlogsAnalysis/$select=AuthorLastName,TotalCount/$filter=AuthorLastName eq Clancy
Jun 4, 2014 at 9:11 AM
Could someone from the OData team please answer Powdor's questions? I have the exactly same interests on OData V4 Aggregation Extension.

Basically 2 questions:
(1) Is OData Aggregation on the todo list of OData team? What is the plan for it?
(2) Will it be open for contribution? If yes, how to get started?

Thanks a lot,
Jul 15, 2014 at 10:17 PM
would like to know to.
Jul 16, 2014 at 6:26 AM
  1. aggregation definitely is one of the key features we want to support. the feature is ducumented here. Most likely we would want to start this work sometime late this year. it'd be refected on codeplex once we prioritize this feature.
  2. we are now in the process of making all odata projects truly "open source", which means all feature planning, bug triage, release are open to the community including taking contributions. we should open source at least one of our projects in Aug and gradually open all the others. we will announce this on odata team blog please watch it to get the latest status.