A simple indexing workflow that uses syntax similar to the superbly documented and inspiring IndexTankDotNet project

        var stockItem = new StockItem
        {
            Id = 1,
            Title = "Edam",
            Description = "Tasty spherical cheese from Holland",
            Price = 12.34M,
            Categories = new Dictionary<string, string> {
                                                            {"Country", "Holland"},
                                                            {"PriceRange", "11-20"}
                                                        }
        };

        var tankTopClient = new TankTopClient("http://:begyhuzatybu@vehehu.api.indexden.com");
        var index = tankTopClient.CreateIndex("Stock");

        var document = new Document(stockItem.Id.ToString());
        document.AddField("title", stockItem.Title);
        document.AddField("description", stockItem.Description);
        document.AddVariable(0, stockItem.Price);
        document.Categories = stockItem.Categories;

        index.AddDocument(document);

        var query = new Query("title:Edam").WithAllFields().WithVariables();
        var searchResult = index.Search(query);
        var resultDocument = searchResult.Results.First();

        var resultStockItem = new StockItem {
                                                Id = int.Parse(resultDocument.DocId),
                                                Title = resultDocument.Fields["title"],
                                                Description = resultDocument.Fields["description"],
                                                Price = (decimal)resultDocument.Variables[0],
                                                Categories = resultDocument.Categories
                                            };

        Assert.AreEqual("Edam", resultStockItem.Title);

The same workflow using TankTops generic syntax

        var stockItem = new StockItem
        {
            Id = 1,
            Title = "Edam",
            Description = "Tasty spherical cheese from Holland",
            Price = 12.34M,
            Categories = new Dictionary<string, string> {
                                                            {"Country", "Holland"},
                                                            {"PriceRange", "11-20"}
                                                        }
        };

        var tankTopClient = new TankTopClient("http://:begyhuzatybu@vehehu.api.indexden.com");
        var index = tankTopClient.CreateIndex("Stock");

        var document = new Document<StockItem>(stockItem.Id.ToString(), stockItem)
            .AddFields(x => x.Title, x => x.Description)
            .AddVariable(0, stockItem.Price)
            .AddCategories(stockItem.Categories);

        index.AddDocument(document);

        var query = new Query<StockItem>("title:Edam");
        var searchResult = index.Search(query);

        var resultStockItem = searchResult.Results.First().Obj;

        Assert.AreEqual("Edam", resultStockItem.Title);
        Assert.AreEqual(12.34, resultStockItem.Price);

As you can see, it cuts out all the object mapping and gives a more ORM like experience

The limitations I discovered with IndexTankDotNet were unit testing and mocking, so TankTop has a public interface, ITankTopClient, that can be mocked. Here I'm using NSubstitute:

        var tankTopClient = Substitute.For<ITankTopClient>();
        var index = new Index { Name = "MyIndex", TankTopClient = tankTopClient };
        var document = new Document("id").AddField("key", "value");
        index.AddDocument(document);
        tankTopClient.Received().AddDocument("MyIndex", document);