Encrypting config sections on the fly in .Net

Question:

In a asp.net application I want to keep some config sections encrypted. How do I do this?

Put config sections into a separate file (.Net)

Question:

How do I put DotNet config sections into separate file?

Useful queries for optimizing Sql Server 2005

Question:

How can I figure out the queries that take most resources on a server.

Answer:

We use this query.
It lists the worst queries (I/Os) on your database since the last restart.

select top 5 (total_logical_reads/execution_count) as avg_logical_reads,
(total_logical_writes/execution_count) as avg_logical_writes,
(total_physical_reads/execution_count) as avg_physical_reads,
Execution_count, statement_start_offset, p.query_plan, q.text
from sys.dm_exec_query_stats
cross apply sys.dm_exec_query_plan(plan_handle) p
cross apply sys.dm_exec_sql_text(plan_handle) as q
order by (total_logical_reads + total_logical_writes)/execution_count Desc

I grabbed it somewhere form TechNet: Monitoring SQL Server health

How to use the HttpContext cache in a non-web (test) environment

Question:

How to test a class that relies on HttpContext.Current.Cache (or so) in order to work.
If you test this in NUnit, for example, there will be no current HttpContext around.
So abstract the access to the Context away and in the property you'll provide another cache if need is.

Attach Debugger macro in VS 2005/ 2008

Question:

How to use a macro in Visual Studio 2005/2008 to attach a debugger to a process?

Call private methods of an object from a test assembly. (NUnit)

I have to test a class of Foo from my test assembly Foo.Test.
Unfortunately, many methods I should test are not public.

This class uses reflection to call private methods on any type or object for testing.

using System;
using System.Reflection;

namespace TestUtilities
{
    public class TestHelper
    {
        public static object InvokeStaticMethod(System.Type type, string methodName, object[] parameters)
        {
            BindingFlags bindFlags = BindingFlags.Static | BindingFlags.Public | BindingFlags.NonPublic;
            return InvokeMethod(type, methodName, null, parameters, bindFlags);
        }
        public static object InvokeInstanceMethod(System.Type type, object instance, string methodName, object[] parameters)
        {
            BindingFlags bindFlags = BindingFlags.Instance | BindingFlags.Public | BindingFlags.NonPublic;
            return InvokeMethod(type, methodName, instance, parameters, bindFlags);
        }
        private static object InvokeMethod(System.Type type, string methodName, object instance, object[] parameters, BindingFlags bindFlags)
        {
            MethodInfo methInfo;
            methInfo = type.GetMethod(methodName, bindFlags);
            if (methInfo == null)
            {
                throw new ArgumentException(String.Format("There is no method with name {0} on type {1}", methodName, type.Name));
            }

            object result = methInfo.Invoke(instance, parameters);
            return result;
        }

    }
}

Get the date portion of a datetime for sql server 2000

Using T-SQL in SQL Server 2000 you sometimes want compare a given datetime value with date by ignoring the time portion of the datetime value.

Here are three ways to strip off the time which gives the following:
2007-06-21 11:23:41.321 => 2007-06-21 00:00:00.000

1) convert to string in format (ISO yymmdd), then cast to datetime
- select on 500000 items takes 1703 ms

SELECT CAST( CONVERT( CHAR(8), GetDate(), 112) AS DATETIME)

2) convert to number of days in float, cut all that is smaller than a day, then cast to datetime
- select on 500000 items takes 228 ms

SELECT CAST( FLOOR( CAST( GETDATE() AS FLOAT ) ) AS DATETIME )

3) get the difference form zero date as number of days, then add this number to zero date.
- select on 500000 items takes 184 ms

SELECT DATEADD(d, 0, DATEDIFF(d, 0, GETDATE()))

---

Performance tested like this.

SET DATEFORMAT dmy
SET STATISTICS IO ON
SET STATISTICS TIME ON
select myDateTime
from #dummy
where CAST( FLOOR( CAST( myDateTime AS FLOAT ) ) AS DATETIME ) = '02/07/2007'
SET STATISTICS TIME OFF
SET STATISTICS IO OFF

Write to different files with Log4Net

Logging in dot.Net using Log4Net is rather easy.
We often want to write all standard infos, errors and warnings to one rolling file.
To configure this you do the following
- configure one or more appenders
- add a root logger
- configure the root logger to use your appenders

Now, if you want to write in certain cases to another file, all you need to do is to create an extra named logger and attach the appender of your choice.

Here is the config

<log4net>
  <!-- create all appenders you need here -->
  <appender name="Console" type="log4net.Appender.ConsoleAppender">
  <layout type="log4net.Layout.PatternLayout">
    <!-- Pattern to output the caller's file name and line number -->
    <conversionPattern value="%5level [%thread] (%file:%line) - %message%newline"/>
  </layout>
</appender>

<appender name="RollingFile" type="log4net.Appender.RollingFileAppender">
<file value="MyProject.log"/>
<appendToFile value="true"/>
<rollingStyle value="composite"/>
<datePattern value="yyyyMMdd"/>
<maximumFileSize value="5MB"/>
<maxSizeRollBackups value="20"/>
<layout type="log4net.Layout.PatternLayout">
<conversionPattern value="%date %-5level [%thread] %logger - %message%newline"/>
</layout>
</appender>

<appender name="SpecialAuditFile" type="log4net.Appender.RollingFileAppender">
   <file value="SpecialAudit.log"/>
   <appendToFile value="true"/>
   <rollingStyle value="composite"/>
   <datePattern value="yyyyMMdd"/>
   <maximumFileSize value="5MB"/>
   <maxSizeRollBackups value="20"/>
   <layout type="log4net.Layout.PatternLayout">
     <conversionPattern value="%date %-5level [%thread] %logger - %message%newline"/>
   </layout>
</appender>

<!-- These are your loggers: one root which is always there and one named logger -->

<root>
<level value="ERROR"/>
<appender-ref ref="Console"/>
<appender-ref ref="RollingFile"/>
</root>

<logger name="SpecialAudit" additivity="false">
  <level value="ALL" />
  <appender-ref ref= "SpecialAuditFile" />
</logger>

</log4net>

Don't forget to add the section in the configSections region

<section name="log4net" type="log4net.Config.Log4NetConfigurationSectionHandler, log4net" />

And here is how you get the one and the other:

// logger for class hierarchy
private static readonly ILog logger = LogManager.GetLogger(typeof(MyClass));
// named logger
private static readonly ILog logAudit = LogManager.GetLogger("SpecialAudit");

In some cases (like Test in NUnit) an explicite call to the conifuration manager is a good idea.

XmlConfigurator.Configure();
...

And again for some test scenarios you should indicate the config filename.
You can do this with an assembly attribute in the AssemblyInfo.cs

...
[assembly: log4net.Config.XmlConfigurator(ConfigFile = "ues.myproject.core.test.dll.config", Watch = true)]
...

The basic intro from Apache is here

How to disable all foreign key constraints on all tables in SQL Server?

The other day I wanted to import some table data from Excel using the import data facility. But, as it goes, SQL Server insisted to import some tables having FKs before their corresponding PK tables. So we have to disable/enable FK constraints on some or all tables.
Here is a very usefull script from decipherinfosys
I tried this out on SQL server 2000. Its working fine.

/************** DISABLE ALL TABLE CONSTRAINTS *********************************

This script will disable all constraints on all tables within the database
that it is run in.

******************************************************************************/

SET NOCOUNT ON
SET ROWCOUNT 0

DECLARE @Count int
DECLARE @String nvarchar (1000)
DECLARE @ConstraintName varchar(128)
DECLARE @TableName varchar(128)

--Find all constraints and their respective tables from the sysobjects table and place into a temp table.
--Primary Key and Unique Constraints via Unique Indexes are not disabled through this command
--You should use the ALTER INDEX...DISABLE command in SQL Server 2005
SELECT
name AS constraintname,
object_name(parent_obj)  AS tablename
INTO #Const_Table
FROM sysobjects s
where xtype in ('F')

SELECT @Count = Count(*) FROM #Const_Table

--Setting the rowcount to one allows for one row from the temp table to be picked off at a time.
--Used as an alternative to a cursor.
SET ROWCOUNT 1

--Loop until all rows in temp table have been processed.
WHILE @Count > 0
BEGIN
--The rowcount of one ensures that only one tablename and constraint name is picked.
SELECT @TableName = TableName, @ConstraintName = ConstraintName
FROM #Const_Table

--Build execution string to disable constraint.
SET @String = 'ALTER TABLE ['+ @tablename + '] NOCHECK CONSTRAINT [' + @constraintname +']'

--Execute the SQL
exec sp_executesql @string

--Remove this row from the temp table, since it has now been processed.
DELETE FROM #Const_Table WHERE ConstraintName = @ConstraintName and TableName = @TableName

SET @Count = @Count - 1
END -- Loop

set rowcount 0

After having re-enabled the constrains you can check constraints using the following command:

DBCC CHECKCONSTRAINTS

And look at your constraints using this command:

SELECT (CASE
WHEN OBJECTPROPERTY(CONSTID, 'CNSTISDISABLED') = 0 THEN 'ENABLED'
ELSE 'DISABLED'
END) AS STATUS,
OBJECT_NAME(CONSTID) AS CONSTRAINT_NAME,
OBJECT_NAME(FKEYID) AS TABLE_NAME,
COL_NAME(FKEYID, FKEY) AS COLUMN_NAME,
OBJECT_NAME(RKEYID) AS REFERENCED_TABLE_NAME,
COL_NAME(RKEYID, RKEY) AS REFERENCED_COLUMN_NAME
FROM SYSFOREIGNKEYS
ORDER BY TABLE_NAME, CONSTRAINT_NAME,REFERENCED_TABLE_NAME, KEYNO

IEnumerable, IEnumerators, IEnumerable<string> and yield

If you want to use custom data collection in the foreach loop, you have to make sure that the collection imprements a GetEnumerator() method that returns an IEnumerator interface.
In order to indicate that your list implements the GetEnumerator() method you can implement the IEnumerable interface as follows:

public class MyEnumerableList : IEnumberable
{
   public IEnumerator GetEnumerator(){...}

}

Here are the framework definitions for .NET Framework 1

public interface IEnumerable 
{
   IEnumerator GetEnumerator();
}
public interface IEnumerator 
{
   object Current{get;}
   bool MoveNext();
   void Reset();
}

Here is a custom class that implements an iterator in .Net 1.

public class MyNumbers : IEnumerable
{
    string[] _numberNames = { "Eins", "Zwei", "Drei", "Vier" };

    IEnumerator IEnumerable.GetEnumerator()
    {
        return new MyEnumerator(this);
    }

    IEnumerator IEnumerable.GetEnumerator()
    {
        return ((IEnumerable)this).GetEnumerator();
    }

    //Inner class for the iteration
    class MyEnumerator : IEnumerator
    {
        // inner collection
        MyNumbers _myNumbers;
        // keep the index in mind
        int _current;
        // constructor
        public MyEnumerator(MyNumbers collection)
        {
            _myNumbers = collection;
            _current = -1;
        }
        void IEnumerator.Reset()
        {
            _current = -1;
        }
        bool IEnumerator.MoveNext()
        {
            _current++;
            return (_current < _myNumbers._numberNames.Length);
        }
        string IEnumerator.Current
        {
            get
            {
                if (_current == -1) throw new InvalidOperationException();
                return _myNumbers._numberNames[_current];
            }
        }
        object IEnumerator.Current
        {
            get
            {
                return ((IEnumerator)this).Current;
            }
        }
        public void Dispose() { }
    }
}

In the .NET Framework 2.0 they added a generic interface that inherits from the older one.

public interface IEnumerable<T> : IEnumerable
{   
   IEnumerator<T> GetEnumerator();
} 
public interface IEnumerator<T> : IEnumerator,IDisposable
{
   T Current{get;}
}

Add now thanks to the new yield statement the compiler generates the inner iterator class itself in order to keep the state.

public class MyNumbers2 : IEnumerable<string>
{
    string[] _numberNames = { "Eins", "Zwei", "Drei", "Vier" };
    IEnumerator<string> IEnumerable<string>.GetEnumerator()
    {
        for (int i = 0; i < _numberNames.Length; i++)
            yield return _numberNames[i];
    }
    IEnumerator IEnumerable.GetEnumerator()
    {
        return ((IEnumerable<string>)this).GetEnumerator();
    }
}

Just created this blog

I've just created this new blog about things I found on the web or made up and like to put somewhere if not in mind.

This will be mainly about IT, programming, philosophy, language learning etc.

"Donkey bridge" is not really English - the correct word would be "Mnemonic Device". But I prefer the direct and literal translation "DONKEY BRIDGE" of the expression we use in German: "Eselsbrücke".