Sample: Generic virtual table data provider plug-in

Demonstrates

This sample shows a minimal implementation for a generic Microsoft Dataverse virtual table data provider plug-in, DropboxRetrieveMultiplePlugin, for the Dropbox file-sharing service. It uses the "bare metal" approach, translating the QueryExpression through the creation of the custom visitor class, DropBoxExpressionVisitor. It returns a collection of the files that satisfy the search criteria as an EntityCollection.

Getting started

In order to build this sample, you must first install the Dropbox.Api and Microsoft.CrmSdk.Data NuGet packages in your solution. You'll also need a DropBox account and pass a real access token when creating an instance of the DropboxClient.

Add the following using statements to your code:

using Microsoft.Xrm.Sdk;
using Microsoft.Xrm.Sdk.Query;
using Dropbox.Api;
using Dropbox.Api.Files;

Sample code

public class DropBoxExpressionVisitor : QueryExpressionVisitorBase
{
    public string SearchKeyWords { get; private set; }

    public override QueryExpression Visit(QueryExpression query)
    {
        // Very simple visitor that extracts search keywords
        var filter = query.Criteria;
        if (filter.Conditions.Count > 0)
        {
            foreach (ConditionExpression condition in filter.Conditions)
            {
                if (condition.Operator == ConditionOperator.Like && condition.Values.Count > 0)
                {
                    string exprVal = (string)condition.Values[0];

                    if (exprVal.Length > 2)
                    {
                        this.SearchKeyWords += " " + exprVal.Substring(1, exprVal.Length - 2);
                    }
                }
            }
            return query;
        }
    }
}

public class DropboxRetrieveMultiplePlugin : IPlugin
{
    public void Execute(IServiceProvider serviceProvider)
    {
        var context = (IPluginExecutionContext)serviceProvider.GetService(typeof(IPluginExecutionContext));
        var qe = (QueryExpression)context.InputParameters["Query"];
        if (qe != null)
        {
            var visitor = new DropBoxExpressionVisitor();
            qe.Accept(visitor);
            using (var dbx = new DropboxClient(AccessToken))
            {
                if (visitor.SearchKeyWords != string.Empty)
                {
                    var searchCriteria = new SearchArg(string.Empty, visitor.SearchKeyWords);
                    var task = Task.Run(() => this.SearchFile(dbx, searchCriteria));
                    context.OutputParameters["BusinessEntityCollection"] = task.Result;
                }
            }
        }
    }

    public async Task<EntityCollection> SearchFile(DropboxClient dbx, SearchArg arg)
    {
        EntityCollection ec = new EntityCollection();
        var list = await dbx.Files.SearchAsync(arg);
        foreach (var item in list.Matches)
        {
            if (item.Metadata.IsFile)
            {
                Entity e = new Entity("new_dropbox");
                e.Attributes.Add("new_dropboxid", Guid.NewGuid());
                e.Attributes.Add("new_filename", item.Metadata.AsFile.Name);
                e.Attributes.Add("new_filesize", item.Metadata.AsFile.Size);
                e.Attributes.Add("new_modifiedon", item.Metadata.AsFile.ServerModified);
                ec.Entities.Add(e);
            }
        }
        return ec;
    }
}

See also

Get started with virtual tables
API considerations of virtual tables
Custom virtual table data providers