Exercise - Enumerate items using language-integrated query (LINQ)

• 7 minutes

Although the application already meets all of the requirements, there's still one more change we can make. C# developers use the language-integrated query (LINQ) syntax to perform queries over collections of data. The .NET SDK for Azure Cosmos DB for NoSQL comes with a built-in mechanism to build your query using the LINQ method syntax.

There are two key requirements at this time:

1. Build a new query using LINQ syntax
2. Convert the query to a feed iterator to get results

After you complete this exercise, your queries will now use the LINQ syntax to make it easier for developers to maintain your application moving forward.

Query using LINQ expressions

Your team wants a unique query that returns all products within the container regardless of category. Recall that we used the type property to separate our category items from the individual product items. Here, we're going to create a cross-partition query using the LINQ method syntax.

1. Open the Program.cs file one last time.

2. Create a new LINQ queryable by using the GetItemLinqQueryable method.

   IOrderedQueryable<Product> queryable = container.GetItemLinqQueryable<Product>();
   

3. Build a LINQ expression using the Where and OrderBy methods and store the expression in a new variable.

   var matches = queryable
       .Where(p => p.Type == nameof(Product))
       .Where(p => !p.Archived)
       .OrderBy(p => p.Price);
   

4. Use the IOrderedQueryable<>.ToFeedIterator method to get a feed iterator from the LINQ expression.

   using FeedIterator<Product> linqFeed = matches.ToFeedIterator();
   

5. Output a message to the console.

   Console.WriteLine($"[Start LINQ query]");
   

6. Save the Program.cs file.

Paginate LINQ query results

Because we're querying across multiple logical partitions, we need to make sure that we return all results even if a logical partition doesn't have any matching results. For example, the helmets category doesn't have any matching products. If we didn't return all possible pages, our application could inadvertently halt when it sees an empty page of results for the helmets category. Here, you again use a while and foreach loop in C# to iterate over all result pages. The code should be similar to how you used a feed iterator previously.

1. In Program.cs, create a while loop that iterates until the feed iterator doesn't have any more pages.

   while (linqFeed.HasMoreResults)
   {    
   }
   

2. Inside of the while loop, get a new page of results.

   FeedResponse<Product> page = await linqFeed.ReadNextAsync();
   

3. Inside of the while loop, output the current page's request charge.

   Console.WriteLine($"[Page RU charge]:\t{page.RequestCharge}");
   

4. Inside of the while loop, create a new foreach loop to iterate over the page's items.

   foreach (Product item in page)
   {
   }
   

5. Inside of the foreach loop, write the individual items to the console.

   Console.WriteLine($"[Returned item]:\t{item}");
   

6. Save the Program.cs file.

Check your work

Your app now uses LINQ to build a cross-partition query that your developer team natively understands. Here, you run the application one last time to ensure that all queries return the results you expected.