We Keep Coding

How to process a DynamoDBEvent in a C# Lambda Function?

I created a C# Lambda function that is being triggered with a DynamoDB stream. It gets excecuted just fine. However, the StreamRecord of the NewImage value returns no values. The count is 0. What am I doing wrong? I have checked all the AWS documention but this seems more and more like a bug. My below lambda function should work and it should return at least 1 StreamRecord in my example. attributeMap.Count always returns 0 but it should return 1.
public void FunctionHandler(DynamoDBEvent dynamoDbEvent, ILambdaContext context)
{
Console.WriteLine($"Beginning to process {dynamoDbEvent.Records.Count} records...");
foreach (var record in dynamoDbEvent.Records)
{
Console.WriteLine($"Event ID: {record.EventID}");
Console.WriteLine($"Event Name: {record.EventName}");
var attributeMap = record.Dynamodb.NewImage;
if (attributeMap.Count > 0) // If item does not exist, attributeMap.Count will be 0
{
Console.WriteLine(attributeMap["AccountId"].S);
}
}
Console.WriteLine("Stream processing complete.");
}

UPDATE: October, 4th, 2018. I no longer use an admin app. I now use CloudFormation exclusively to create and maintain everyting including full CI/CD pipelines with CodePipelines. This includes all Serverless lamba functions in .NET Core 2.1.
I figured it out. The sucky thing is that AWS Docs do not say anything about this. This was a pain in the ass to find out. For anyone else who might need this information here it is: You have to set the stream view type when you create the DynamoDB stream for the table. Here is a picture for the AWS Console:
However, since I setup all tables via an admin console (in C# Core 2.0), here is how I setup the setup the table including the stream specification and the event source mapping request to the lambda function:
var request = new CreateTableRequest
{
TableName = TABLE_CREATE_ACCOUNT,
AttributeDefinitions = new List<AttributeDefinition>()
{
new AttributeDefinition
{
AttributeName = "CommandId",
AttributeType = ScalarAttributeType.S
}
},
KeySchema = new List<KeySchemaElement>()
{
new KeySchemaElement
{
AttributeName = "CommandId",
KeyType = KeyType.HASH
}
},
ProvisionedThroughput = new ProvisionedThroughput
{
ReadCapacityUnits = 1,
WriteCapacityUnits = 1
},
StreamSpecification = new StreamSpecification
{
StreamEnabled = true,
StreamViewType = StreamViewType.NEW_IMAGE
}
};
try
{
var response = _db.CreateTableAsync(request);
var tableDescription = response.Result.TableDescription;
Console.WriteLine("{1}: {0} ReadCapacityUnits: {2} WriteCapacityUnits: {3}",
tableDescription.TableStatus,
tableDescription.TableName,
tableDescription.ProvisionedThroughput.ReadCapacityUnits,
tableDescription.ProvisionedThroughput.WriteCapacityUnits);
string status = tableDescription.TableStatus;
Console.WriteLine(TABLE_CREATE_ACCOUNT + " - " + status);
WaitUntilTableReady(TABLE_CREATE_ACCOUNT);
// This connects the DynamoDB stream to a lambda function
Console.WriteLine("Creating event source mapping between table stream '"+ TABLE_CREATE_ACCOUNT + "' and lambda 'ProcessCreateAccount'");
var req = new CreateEventSourceMappingRequest
{
BatchSize = 100,
Enabled = true,
EventSourceArn = tableDescription.LatestStreamArn,
FunctionName = "ProcessCreateAccount",
StartingPosition = EventSourcePosition.LATEST
};
var reqResponse =_lambda.CreateEventSourceMappingAsync(req);
Console.WriteLine("Event source mapping state: " + reqResponse.Result.State);
}
catch (AmazonDynamoDBException e)
{
Console.WriteLine("Error creating table '" + TABLE_CREATE_ACCOUNT + "'");
Console.WriteLine("Amazon error code: {0}", string.IsNullOrEmpty(e.ErrorCode) ? "None" : e.ErrorCode);
Console.WriteLine("Exception message: {0}", e.Message);
}
catch (Exception e)
{
Console.WriteLine("Error creating table '" + TABLE_CREATE_ACCOUNT + "'");
Console.WriteLine("Exception message: {0}", e.Message);
}
The key is
StreamViewType = StreamViewType.NEW_IMAGE
That was it.

Mismatch in .NET metadata for SQL DateTime2 type

I'm getting a thrown exception when attempting to import a SqlDataRecord with a DateTime2 data type. There is no exception thrown in this snippet, but the cause of the later exception is evident in the following code.
using System;
using Microsoft.SqlServer.Server;
using System.Data.SqlClient;
using System.Data;
using System.Web.UI.WebControls;
using System.Linq;
using System.Collections;
using System.Collections.Generic;
using System.Reflection;
using System.Diagnostics;
namespace Testing123
{
public class Program
{
public static object SQLTParser(SqlMetaData smd, string val)
{
TypeCode systc = Parameter.ConvertDbTypeToTypeCode(smd.DbType);
try
{
return Convert.ChangeType(val, systc);
}
catch (Exception ex)
{
if (ex is InvalidCastException || ex is FormatException || ex is OverflowException)
{
Console.WriteLine("Exception reached casting " + val + " to " + Type.GetType("System." + Enum.GetName(typeof(TypeCode), systc)) + ": " + ex.Message + ex.ToString()); //smd.GetType().Name
return null;
}
else
{
Console.WriteLine("Null value exception");
return null;
}
}
}
public static void Main()
{
SqlMetaData sqmd = new SqlMetaData("dt", SqlDbType.DateTime2, 27, 7);
SqlDataRecord sdr = new SqlDataRecord(sqmd);
sdr.SetValue(0, SQLTParser(sqmd, "2017-01-12 01:23:12.3456789"));
//set BreakPoint
//sdr -> Non-Public members -> _columnMetaData[0] -> Precision = 27
//sdr -> Non-Public members -> _columnSmiMetaData[0] -> Non-Public members -> Precision = 0
}
}
}
As noted, if you set the breakpoint and watches as indicated in MS Visual Studio*, Precision does not match between columnMetaData and columnSmiMetaData, which by the second such entry (!!) throws an exception:
Metadata for field 'dt' of record '2' did not match the original record's metadata.
which matches the exception thrown by
line 3755 of ValueUtilsSmi
due to the return of MetadataUtilsSmi.IsCompatible
Essentially, precision in the field's metadata of record 2 doesn't match that which is in the SmiMetaData of record 1. In record 1, the MD and SMD don't match either, but based on the logic Microsoft was using for an IEnumerator'd SqlDataRecord, it doesn't become an issue until the 2nd record.
Is this a MS bug? Or is there a way to force the precision value of the SmiMetaData? Or ignore that particular field check in ValueUtilsSmi? Specifying SqlMetaData sqmd = new SqlMetaData("dt", SqlDbType.DateTime2, 0, 7); allows the parsing to proceed, but drops sub-second precision from the data.
Here is, in a nutshell, how I'm attempting to send the data to a database. Apologies if this portion is not a complete example.
public class FullStreamingDataRecord : IEnumerable<SqlDataRecord>
{
private string _filePath;
public bool _hasHeader { get; private set; }
private ParserDict _pd; //notably has colStructure which is an array of SqlMetaData[]
public FullStreamingDataRecord(string FilePath, ParserDict pd)
{
_filePath = FilePath;
_pd = pd;
_hasHeader = true;
}
public static object SQLTParser(SqlMetaData smd, string val)
{
TypeCode systc = Parameter.ConvertDbTypeToTypeCode(smd.DbType);
try
{
return Convert.ChangeType(val, systc);
}
catch (Exception ex)
{
if (ex is InvalidCastException || ex is FormatException || ex is OverflowException)
{
Console.WriteLine("Exception reached casting " + val + " to " + Type.GetType("System."+Enum.GetName(typeof(TypeCode),systc)) + ": " + ex.Message + ex.ToString()); //smd.GetType().Name
return null; //smd.TParse(val);
}
else
{
Console.WriteLine("Null value exception casting...attempting a different method");
return null; smd.TParse(val);
}
}
}
public IEnumerator<SqlDataRecord> GetEnumerator()
{
int len = this._pd.colStructure.Length;
StreamReader fileReader = null;
try
{
using (fileReader = new StreamReader(this._filePath))
{
string inputRow = "";
string[] inputColumns = new string[len];
if (_hasHeader && !fileReader.EndOfStream)
{
inputRow = fileReader.ReadLine(); //and ignore
}
while (!fileReader.EndOfStream)
{
string temp = "";
inputRow = fileReader.ReadLine();
inputColumns = inputRow.Split(new char[]{','},len);
SqlDataRecord dataRecord = this._pd.colStructure
for (int j = 0; j < len; j++) { // i = counter for input columns
string currentKey = this._pd.colStructure[j].SqlMetaData.Name;
string curval = inputColumns[j];
var ty = this._pd.colStructure[j].SqlMetaData; //.DbType;
dataRecord.SetValue(j, SQLTParser(ty, curval));
// dataRecord.GetSqlMetaData(j).Adjust(dataRecord.GetValue(j));
}
yield return dataRecord;
}
}
}
// no catch block allowed due to the "yield" command
finally
{
fileReader.Close();
}
}
IEnumerator IEnumerable.GetEnumerator()
{
return GetEnumerator();
}
Which I call with this in my Main()
using (SqlConnection conn = new DBConnect().conn)
{
conn.Open();
SqlCommand importProc = new SqlCommand("tvp_"+pd.tableDestination, conn);
importProc.CommandType = CommandType.StoredProcedure;
importProc.CommandTimeout = 300;
SqlParameter importTable = new SqlParameter();
importTable.ParameterName = "#ImportTable";
importTable.TypeName = "dbo.tt_"+pd.tableDestination;
importTable.SqlDbType = SqlDbType.Structured;
importTable.Value = new FullStreamingDataRecord(fn, pd);
importProc.Parameters.Add(importTable);
importProc.ExecuteNonQuery(); //this line throws the exception
}
Apologies for not Console.WriteLineing the values of these watches.
Unfortunately, it seems that these parameters are protected against examination using Reflection's sdr.GetType().GetProperties(), even with the appropriate BindingFlags set. But at least you can see the values in debug mode!

C# - Visual Studio - 'object' does not contain a definition for 'ExecuteQuery'

I am trying to add code to an existing project that will check for existence of a device in SCCM and delete it if it does exist. I seem to be missing something, in that particular block of code. I get an error - 'object' does not contain a definition for 'ExecuteQuery' and no extension method 'ExecuteQuery' accepting a first argument of type 'object' could be found.
Here is the C# code:
using System;
using System.Collections.Generic;
using System.Linq;
using System.Web;
using System.Web.Mvc;
using System.Net.NetworkInformation;
using SupportToolkit.Models;
using SupportToolkit.WindowsAutomationServices;
using NLog;
using System.Text;
using System.DirectoryServices;
using System.DirectoryServices.AccountManagement;
using System.Management;
namespace SupportToolkit.Controllers
{
public class TPOSRecordDeletionController : Controller
{
private static Logger recordDeletionLogger = LogManager.GetCurrentClassLogger();
[Authorize(Roles = #"nor\NST_RIT_Users,nor\NST_STM_Users,nor\NST_Admin_Users,nor\NST_CORP_Users")]
// GET: TPOSRecordDeletion
public ActionResult Index(TPOSRecordDeletionModel model)
{
if (model.ComputerName != null)
{
}
if (ModelState.IsValid)
{
if (!(string.IsNullOrEmpty(model.btnDeleteRecord)))
{
InvokeRecordDeletion(model);
}
}
return View(model);
}
[Authorize(Roles = #"nor\NST_RIT_Users,nor\NST_STM_Users,nor\NST_Admin_Users,nor\NST_CORP_Users")]
public string InvokeRecordDeletion(TPOSRecordDeletionModel model)
{
model.Status = "Running Service";
var windowsAutomationService = new WindowsAutomationServicesClient();
string shortServiceOutput;
var serviceAction = "Remove-TPOSRecords";
var SCCMServer = "server.nor.net";
var siteCode = "PO60";
string[] recordDeletionArguments = new string[1];
recordDeletionArguments[0] = model.ComputerName;
model.Status = "Processing" + model.ComputerName;
Ping pingSender = new Ping();
PingOptions options = new PingOptions();
// Use the default Ttl value which is 128,
// but change the fragmentation behavior.
options.DontFragment = true;
// Create a buffer of 32 bytes of data to be transmitted.
string data = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa";
byte[] buffer = Encoding.ASCII.GetBytes(data);
int timeout = 120;
PingReply reply = pingSender.Send(model.ComputerName, timeout, buffer, options);
if (reply.Status == IPStatus.Success)
{
model.Status = model.ComputerName + "is currently online and will not be removed!";
}
else
{
// set up domain context
using (var ctx = new System.DirectoryServices.AccountManagement.PrincipalContext(System.DirectoryServices.AccountManagement.ContextType.Domain))
{
// find a computer
ComputerPrincipal computer = ComputerPrincipal.FindByIdentity(ctx, model.ComputerName);
if (computer == null)
{
model.Status = model.ComputerName + "does not exist in Active Directory.";
}
else
{
computer.Delete();
model.Status = model.ComputerName + "successfully removed from Active Directory!";
}
//insert code here for checking for existence of computer in SCCM and removing from SCCM if exists
SmsNamedValuesDictionary namedValues = new SmsNamedValuesDictionary();
WqlConnectionManager connection = new WqlConnectionManager(namedValues);
connection.Connect("s0319p60.nordstrom.net");
foreach (IResultObject computerobject in connection.QueryProcessor.ExecuteQuery("Select ResourceID From SMS_R_System Where Name ='" + model.ComputerName + "'"))
{
if (computerobject == null)
{
model.Status = model.ComputerName + "does not exist in SCCM.";
}
else
{
computerobject.Delete();
model.Status = model.ComputerName + "successfully removed from SCCM!";
}
}
}
var userName = User.Identity.Name;
var serviceOutput = windowsAutomationService.RunAutomationService(serviceAction, userName, recordDeletionArguments);
recordDeletionLogger.Info(userName + " is attempting to remove the record " + model.ComputerName);
if (serviceOutput.Length >= 7)
{
shortServiceOutput = serviceOutput.Substring(0, 7);
shortServiceOutput = shortServiceOutput.ToLower();
}
else
{
shortServiceOutput = serviceOutput;
shortServiceOutput = shortServiceOutput.ToLower();
}
if (shortServiceOutput == "success")
{
model.Status = "Successfully removed " + model.ComputerName + " from SCCM and Active Directory";
recordDeletionLogger.Info(userName + " successfully removed " + model.ComputerName + " from SCCM and Active Directory");
return "Success";
}
else
{
model.Status = "Failure removing " + model.ComputerName + " from SCCM and Active Directory. Unknown Error";
recordDeletionLogger.Info(userName + " failed to remove " + model.ComputerName + " from SCCM and Active Directory");
return "Failure";
}
}
}
}
internal interface IResultObject
{
//void Delete();
void Delete();
}
internal class WqlConnectionManager
{
private SmsNamedValuesDictionary namedValues;
public WqlConnectionManager(SmsNamedValuesDictionary namedValues)
{
this.namedValues = namedValues;
}
public object QueryProcessor { get; internal set; }
internal void Connect(string v)
{
throw new NotImplementedException();
}
public object ExecuteQuery { get; internal set; }
}
internal class SmsNamedValuesDictionary
{
public SmsNamedValuesDictionary()
{
}
}

Well - after a few days of searching, I FINALLY figured out the issue.
The entire block of code at the end, was not necessary. The issue was that there were missing assembly references and using statements in the code. Specifically:
using Microsoft.ConfigurationManagement.ManagementProvider;
using Microsoft.ConfigurationManagement.ManagementProvider.WqlQueryEngine;
The corresponding DLLs also needed to be added to the project as references.
I hope this helps someone who runs into similar issues and is not versed in C# coding. I myself am only versed in PowerShell, so figuring this out took a lot of work for me. Thanks.

You create a custom "connection" object:
WqlConnectionManager connection = new WqlConnectionManager(namedValues);
Then call a method on one of its properties:
connection.QueryProcessor.ExecuteQuery("...")
But what is that QueryProcessor property?...
public object QueryProcessor { get; internal set; }
It's an object. As the error states, object doesn't have a method called ExecuteQuery. (It doesn't have very many methods or properties at all, really.)
I can't really tell from this code (maybe I'm missing something?) what specific type you're expecting QueryProcessor to be, but it should definitely be something more specific than object. Something analogous to a SQLCommand object, perhaps? Essentially, whatever type would have that ExecuteQuery method.
If there's a compelling reason in the existing codebase for this to be of type object, you'd need to determine what that reason is. There seems to be a lot of use of object here, which smells of some bad design choices from before you got there.

ProcessEvent not called in TFS plugin

Recently i have decided to create a plugin for the TFS for tracking work item changes based on ISubscriber interface.
So the workflow would be the following:
1) Work item state changes
2) Plugin catches the WorkItemChangedEvent
3) Send an email to the person specified in the Requester
As the base for my project, i used the following project from CodePlex - The Mail Alert
After i adopted it to my needs and saved compiled binaries in %TFS-DIR%\Microsoft Team Foundation Server 14.0\Application Tier\Web Services\bin\Plugins the TFS restarted the tier and... that's it. On work item change the ProcessEvent method is not called, but it should be.
using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Text.RegularExpressions;
using Microsoft.TeamFoundation.Client;
using Microsoft.TeamFoundation.Common;
using Microsoft.TeamFoundation.Framework.Client;
using Microsoft.TeamFoundation.Framework.Common;
using Microsoft.TeamFoundation.Framework.Server;
using Microsoft.TeamFoundation.VersionControl.Client;
using Microsoft.TeamFoundation.WorkItemTracking.Client;
using Microsoft.TeamFoundation.WorkItemTracking.Server;
using System.Diagnostics;
using System.Xml.Linq;
using System.Xml;
using System.DirectoryServices;
using System.Net.Mail;
using System.Xml.Xsl;
using System.Configuration;
using System.Reflection;
namespace MailAlert
{
public class WorkItemChangedEventHandler : ISubscriber
{
static string serverPath = "";
static string ExternalURL = "";
static string MailAddressFrom = "";
static string SMTPHost = "";
static string Password = "";
static int Port = 25;
static int index = 0;
static string projectCollectionFolder;
static Uri projectCollectionUri;
static WorkItemStore wiStore;
static WorkItem wItem;
static WorkItemChangedEvent workItemChangedEvent;
static string teamProjectPath = "";
static VersionControlServer versionControlServer;
static TfsTeamProjectCollection projectCollection;
static Dictionary<IdentityDescriptor, TeamFoundationIdentity> m_identities = new Dictionary<IdentityDescriptor, TeamFoundationIdentity>(IdentityDescriptorComparer.Instance);
public Type[] SubscribedTypes()
{
return new Type[1] { typeof(WorkItemChangedEvent) };
}
public WorkItemChangedEventHandler()
{
TeamFoundationApplicationCore.Log("WorkItemChangedEvent Started", index++, EventLogEntryType.Information);
}
public EventNotificationStatus ProcessEvent(TeamFoundationRequestContext requestContext, NotificationType notificationType,
object notificationEventArgs, out int statusCode, out string statusMessage, out ExceptionPropertyCollection properties)
{
TeamFoundationApplicationCore.Log("WorkItemChangedEventHandler: ProcessEvent entered", index++, EventLogEntryType.Information);
statusCode = 0;
properties = null;
statusMessage = String.Empty;
GetTfsServerName();
projectCollection = TfsTeamProjectCollectionFactory.GetTeamProjectCollection(TfsTeamProjectCollection.GetFullyQualifiedUriForName(serverPath));
try
{
if (notificationType == NotificationType.Notification && notificationEventArgs is WorkItemChangedEvent)
{
workItemChangedEvent = notificationEventArgs as WorkItemChangedEvent;
TeamFoundationApplicationCore.Log("WorkItemChangedEventHandler: WorkItem " + workItemChangedEvent.WorkItemTitle + " was modified", index++, EventLogEntryType.Information);
TeamFoundationApplicationCore.Log("WorkItemChangedEventHandler: serverPath - " + serverPath, index++, EventLogEntryType.Information);
projectCollectionFolder = requestContext.ServiceHost.VirtualDirectory.ToString();
projectCollectionUri = new Uri(serverPath + projectCollectionFolder);
projectCollection = TfsTeamProjectCollectionFactory.GetTeamProjectCollection(projectCollectionUri);
wiStore = projectCollection.GetService<WorkItemStore>();
versionControlServer = projectCollection.GetService<VersionControlServer>();
TeamFoundationApplicationCore.Log("WorkItemChangedEventHandler: Before process workitem", index++, EventLogEntryType.Information);
ProcessWorkItem();
TeamFoundationApplicationCore.Log("WorkItemChangedEventHandler: After process workitem", index++, EventLogEntryType.Information);
}
}
catch (Exception ex)
{
TeamFoundationApplicationCore.Log("WorkItemChangedEventHandler: FUCKING EXCEPTION! =>\n" + ex.Message, index++, EventLogEntryType.Error);
}
return EventNotificationStatus.ActionPermitted;
}
private static void GetTfsServerName()
{
try
{
string assemblyFolder = Path.GetDirectoryName(Assembly.GetExecutingAssembly().CodeBase);
XmlDocument XmlDoc = new XmlDocument();
XmlDoc.Load(assemblyFolder + #"\Settings.xml");
// Declare the xpath for finding objects inside the XML file
XmlNodeList XmlDocNodes = XmlDoc.SelectNodes("/configuration/tfssettings");
XmlNodeList XmlDocExt = XmlDoc.SelectNodes("/configuration/Externaltfssettings");
// Define a new List, to store the objects we pull out of the XML
serverPath = XmlDocNodes[0].InnerText;
ExternalURL = XmlDocExt[0].InnerText;
XmlNodeList XmlDocNodes2 = XmlDoc.SelectNodes("/configuration/appSettings");
foreach (XmlNode mailNode in XmlDocNodes2)
{
foreach (XmlNode varElement in mailNode.ChildNodes)
{
switch (varElement.Attributes["key"].Value)
{
case "MailAddressFrom":
MailAddressFrom = varElement.Attributes["value"].Value;
break;
case "SMTPHost":
SMTPHost = varElement.Attributes["value"].Value;
break;
case "Password":
Password = varElement.Attributes["value"].Value;
break;
case "Port":
Port = Convert.ToInt32(varElement.Attributes["value"].Value);
break;
}
}
}
}
catch (Exception ex)
{
EventLog.WriteEntry("WorkItemChangedEventHandler", ex.Message);
}
}
public string Name
{
get { return "WorkItemChangedEventHandler"; }
}
public SubscriberPriority Priority
{
get { return SubscriberPriority.High; }
}
private static void ProcessWorkItem()
{
var teamProjects = versionControlServer.GetAllTeamProjects(false);
for (int i = 0; i < teamProjects.Length; i++)
{
string teamProjectName = teamProjects[i].Name;
var teamProject = teamProjects[i];
Project teamProjectWI = wiStore.Projects[i];
teamProjectPath = projectCollectionUri + teamProject.Name;
if (workItemChangedEvent.PortfolioProject == teamProjectName)
{
//get the workitem by ID ( CoreFields.IntegerFields[0] == ID ?!)
//check if any of String changed fields
foreach(StringField sf in workItemChangedEvent.ChangedFields.StringFields)
{
//is the State field
if (sf.Name.Equals("State"))
{
//then notify Reuqester
wItem = wiStore.GetWorkItem(workItemChangedEvent.CoreFields.IntegerFields[0].NewValue);
string CollGuid = projectCollection.InstanceId.ToString();
string Requester = wItem.Fields["Requester"].Value.ToString();
string WorkItemId = wItem.Id.ToString();
string mail = GetEmailAddress(Requester);
SendMail(CollGuid, WorkItemId, mail);
}
}
}
}
}
private static string GetEmailAddress(string userDisplayName)
{
DirectorySearcher ds = new DirectorySearcher();
ds.PropertiesToLoad.Add("mail");
ds.Filter = String.Format("(&(displayName={0})(objectCategory=person)((objectClass=user)))", userDisplayName);
SearchResultCollection results = ds.FindAll();
if (results.Count == 0)
{
return string.Empty;
}
ResultPropertyValueCollection values = results[0].Properties["mail"];
if (values.Count == 0)
{
return string.Empty;
}
return values[0].ToString();
}
private static void SendMail(string collID,string workItemId,string tomailAddrees)
{
MailMessage objeto_mail = new MailMessage();
SmtpClient client = new SmtpClient();
client.Port = Port;
client.Host = SMTPHost;
client.Timeout = 200000;
client.DeliveryMethod = SmtpDeliveryMethod.Network;
client.UseDefaultCredentials = true;
client.EnableSsl = true;
client.Credentials = new System.Net.NetworkCredential(MailAddressFrom, Password);
objeto_mail.From = new MailAddress(MailAddressFrom);
objeto_mail.To.Add(new MailAddress(tomailAddrees));
//objeto_mail.CC.Add(new MailAddress("nagarajb#hotmail.com"));
objeto_mail.Subject = "Work Item Changed:"+workItemId;
string mailbody = serverPath+"/tfs/web/wi.aspx?pcguid=" + collID + "&id=" + workItemId;
string mailbody2 = "";
if (ExternalURL.Length > 0)
{
mailbody2 = ExternalURL + "/tfs/web/wi.aspx?pcguid=" + collID + "&id=" + workItemId;
}
string tables = "<table border=1><tr><td>Work Item ID</td><td>" + wItem.Id.ToString() + "</td></tr><tr><td>Title</td><td>" + wItem.Title + "</td></tr><tr><td>State</td><td>" + wItem.State + "</td></tr><tr><td>Assigned To</td><td>" + wItem.Fields["Assigned to"].Value.ToString() + "</td></tr><tr><td>Internal URL</td><td>" + mailbody + "</td></tr><tr><td>External URL</td><td>" + mailbody2 + "</td></tr></table>";
objeto_mail.IsBodyHtml = true;
objeto_mail.Body = "<i>Hi " + wItem.Fields["Requester"].Value.ToString() + ","+"</i></br></br></br>" + tables + " </br></br> Best regards; </br></br>Configuration Management Team</br></br></br>";
client.Send(objeto_mail);
EventLog.WriteEntry("WorkItemChangedEventHandler", "Email Sent");
}
}
}
No errors or exceptions are thrown in the Event Log either. Tracing TFS (trace=true property in web.config) also was of no help.
Maybe someone could help or shed light on this mysterious case?
UPDATE:
Thanks for a reply Giulio Vian!
Here how it goes:
1) I haven't seen the dependencies broken, plus the constructor WorkItemChangedEventHandler is successfully called. That is seen in the Windows Event Log - WorkItemChangedEvent Started message is written.
2) I am not sure how to register the event handler.... i'll look that up
3) I am not sure how this works. I thought just copy-pasting dlls in the appropriate folder will do the trick, and there is no need for an account for the plugin. Mind giving a bit more info on this?
4) yes. Web.config in the main Web config in Application Tier\Web Services
5) Yes. Using an account with Administer rights. if i set a break point in the constructor, it is reached. Any other place in the code is not reached.

Many things can be wrong.
If any dependency is broken, you should see an error in the Event log
If you do not register the "WorkItemChangedEventHandler" event source, no message is written
Can the user account running the plugin access TFS?
How did you enabled the tracing (you pasted a bunch of code, but no configuration)?
Have you attached the debugger to the TFS process and set a breakpoint (do not this on a production TFS)?
We have collected similar suggestions in the documentation for our plugin at
https://github.com/tfsaggregator/tfsaggregator/wiki.

Compile C# Roslyn

I'm using Roslyn to try and compile and run code at runtime. I've ysed some code I found online and have it somewhat working.
public Type EvalTableScript(string Script, CRMMobileFramework.EnbuUtils EnbuUtils, CRMMobileFramework.Includes.DBAdapter dbConn)
{
var syntaxTree = SyntaxTree.ParseText(Script);
var compilation = Compilation.Create("EnbuScript.dll",
options: new CompilationOptions(outputKind: OutputKind.DynamicallyLinkedLibrary),
references: new[]
{
new MetadataFileReference(typeof(object).Assembly.Location),
new MetadataFileReference(typeof(EnbuUtils).Assembly.Location),
new MetadataFileReference(typeof(DBAdapter).Assembly.Location),
MetadataFileReference.CreateAssemblyReference("System.Data"),
MetadataFileReference.CreateAssemblyReference("System.Linq"),
MetadataFileReference.CreateAssemblyReference("System"),
MetadataFileReference.CreateAssemblyReference("System.XML")
},
syntaxTrees: new[] { syntaxTree });
var diagnostics = compilation.GetDiagnostics();
foreach (var diagnostic in diagnostics)
{
Console.WriteLine("Error: {0}", diagnostic.Info.GetMessage());
}
Assembly assembly;
using (var stream = new MemoryStream())
{
EmitResult emitResult = compilation.Emit(stream);
assembly = Assembly.Load(stream.GetBuffer());
}
Type ScriptClass = assembly.GetType("EnbuScript");
// Pass back the entire class so we can call it at the appropriate time.
return ScriptClass;
}
Then I'm trying to call this:
string Script = #"
using System;
using System.Data;
using System.IO;
using System.Linq;
public class EnbuScript
{
public string PostInsertRecord(CRMMobileFramework.EnbuUtils EnbuUtils,CRMMobileFramework.Includes.DBAdapter dbConn)
{
string ScriptTable = ""QuoteItems"";
DataSet EntityRecord = dbConn.FindRecord(""*"", ScriptTable, ""QuIt_LineItemID='"" + EnbuUtils.GetContextInfo(ScriptTable) + ""'"", """", 1, 1, false);
string OrderId = EntityRecord.Tables[""item""].Rows[0][""QuIt_orderquoteid""].ToString();
string UpdateOrderTotalCommand = ""UPDATE Quotes SET Quot_nettamt = (select SUM(QuIt_listprice * quit_quantity) from QuoteItems where quit_orderquoteid = "" + OrderId + "" ) where Quot_OrderQuoteID = "" + OrderId;
dbConn.ExecSql(UpdateOrderTotalCommand);
return ""Complete"";
}
}";
Type EnbuScript = EnbuUtils.EvalTableScript(Script, EnbuUtils, dbConn);
MethodInfo methodInfo = EnbuScript.GetMethod("InsertRecord");
object[] parameters = { EnbuUtils, dbConn };
string InsertRecordResult = methodInfo.Invoke(null, parameters).ToString();
As you can see I've been messing around with trying to pass parameters to the compilation.
Basically I've got 4 functions I need to support, that will come in as a string. What I'm trying to do is create a class for these 4 functions and compile and run them. This part works.
What I now need to be able to do is pass class instances to this. In the code you'll see a dbConn which is basically my database connection. I need to pass the instance of this to the method I'm calling at runtime so it has it's correct context.
I have another implementation of this where I'm using the Roslyn session. I originally tried to use this and override my function at runtime but that didn't work either. See below what I tried:
public static void EvalTableScript(ref EnbuUtils EnbuUtils, DBAdapter dbConn, string EvaluateString)
{
ScriptEngine roslynEngine = new ScriptEngine();
Roslyn.Scripting.Session Session = roslynEngine.CreateSession(EnbuUtils);
Session.AddReference(EnbuUtils.GetType().Assembly);
Session.AddReference(dbConn.GetType().Assembly);
Session.AddReference("System.Web");
Session.AddReference("System.Data");
Session.AddReference("System");
Session.AddReference("System.XML");
Session.ImportNamespace("System");
Session.ImportNamespace("System.Web");
Session.ImportNamespace("System.Data");
Session.ImportNamespace("CRMMobileFramework");
Session.ImportNamespace("CRMMobileFramework.Includes");
try
{
var result = (string)Session.Execute(EvaluateString);
}
catch (Exception ex)
{
}
}
I tried to call this using:
string PostInsertRecord = "" +
" public override void PostInsertRecord() " +
"{ " +
" string ScriptTable = \"QuoteItems\"; " +
"DataSet EntityRecord = dbConn.FindRecord(\"*\", ScriptTable, \"QuIt_LineItemID='\" + EnbuUtils.GetContextInfo(ScriptTable) + \"'\", \"\", 1, 1, false); " +
"string OrderId = EntityRecord.Tables[\"item\"].Rows[0][\"QuIt_orderquoteid\"].ToString(); " +
"string UpdateOrderTotalCommand = \"UPDATE Quotes SET Quot_nettamt = (select SUM(QuIt_listprice * quit_quantity) from QuoteItems where quit_orderquoteid = \" + OrderId + \" ) where Quot_OrderQuoteID = \" + OrderId; " +
"dbConn.ExecSql(UpdateOrderTotalCommand); " +
"} ";
The function is declared as a public virtual void in the EnbuUtils class but it says it doesn't have a suitable method to override.
Safe to say, I'm stumped!
Any help appreciated!
Thanks

I got this in the end - this first method was very close to what I actually needed. Changed the method to static and had to add a few references including the full namespace.