15 Sorting Algorithms in 6 Minutes

Deploy code first DB based web app to Windows Azure

If you created a ASP.NET MVC4 app using EF5 code first model, and you want to deploy the app onto Windows Azure, basically there are 2 steps you need to follow:

First, enable code first migration so server side knows how to re-create the database. You can achieve this in Package Management Console by executing “enable-migrations”. If succeeded you will see a new folder name Migrations created in the project file root (see screenshot), click here for more details of code first migrations.

Then create a web site on Azure server and remember to create a database at the same time by following the online guide. Once website created you can get the publish profile which contains the database connection string so later when you use the publish profile to deploy your website those values can be used to replace your LocalDB connection string value (see screenshot below), and remember tick the Execute Code First Migrations option before click Next.

Launch an app from another app: cross platform review

Android platform allows launch any app from inside of other app as long as we know the full package name of the target app. Developer can even specify the activity to be opened initially.

iOS and Windows Phone don’t allow launch arbitrary app on the device, instead, they have different approach, the URL scheme. The targe app can register a specific URL scheme (for example “com.facebook:” for Facebook app), later when user launches the same URL from either another app or inside a webpage the target app will be launched automatically. Developer can also pass parameters in the URL just like we do in a normal URL.

Please be aware, the URL scheme introduces a security flaw because any third party app can register the same URL scheme. If more than one third-party app registers to handle the same URL scheme: for iOS, there is currently no process for determining which app will be given that scheme (see doc); for Windows Phone, each time the user opens the file, they will be asked which app they want to use.

Besides, Windows Phone provides a way that launching apps from the same publisher who published the current app. See below the code sample:

IEnumerable<Package> apps = Windows.Phone.Management.Deployment
                                                    .InstallationManager
                                                    .FindPackagesForCurrentPublisher();
apps.First().Launch(string.Empty);

HttpWebRequest Connection limitation

As you may know, by default, the System.Net implementation of C# is using connection pool to improving performance which respects restrictions stated in http 1.1 - one user agent SHOULD NOT create more than 2 connection to a server. This causes max 2 live connections can be created by HttpWebRequest in the same app domain at any time. So the wordaround can be just host the code in different app domains and fire the method in multi threads.

Or, you can change app configuration to rewrite the default connection limit:
<system.net>
     <connectionManagement>
         <add address="*" maxcoonection="2000" />
     </connectionManagement>
</system.net>

Or, you can rewrite the limitation in your code before initialize the first HttpWebRequest:
ServicePointManager.DefaultConnectionLimit = 100;

C# language Evolution Matrix by Raymond Tang

ASP.NET Web Forms Application Paths Related

This article is going to list all ASP.NET Web Forms application paths related properties and their values on the Request object. Below the test result are based on this scenario: local hosted IIS site named VirtualSite, first we access to http://localhost/VirtualSite/Redirect/dummy.aspx and there is a HttpModule to redirect the request to http://localhost/VirtualSite/Redirect/Redirect.aspx. The request has parameters "a=1&b=2&c=3". So here it is:

Request Property	Value
ApplicationPath	/VirtualSite
AppRelativeCurrentExecutionFilePath	~/Redirect/Redirect.aspx
CurrentExecutionFilePath	/VirtualSite/Redirect/Redirect.aspx
FilePath	/VirtualSite/Redirect/Redirect.aspx
Path	/VirtualSite/Redirect/Redirect.aspx
QueryString	{a=1&b=2&c=3}
RawUrl	/VirtualSite/Redirect/Dummy.aspx?a=1&b=2&c=3
Url.AbsolutePath	/VirtualSite/Redirect/Redirect.aspx
Url.AbsoluteUrl	http://localhost:25739/VitualSite/Redirect/Redirect.aspx?a=1&b=2&c=3
Url.Authority	localhost:25739
Url.DnsSafeHost	localhost
Url.Host	localhost
Url.LocalPath	/VitualSite/Redirect/Redirect.aspx
Url.OriginalString	http://localhost:25739/VitualSite/Redirect/Redirect.aspx?a=1&b=2&c=3
Url.PathAndQuery	/VitualSite/Redirect/Redirect.aspx?a=1&b=2&c=3
Url.Port	25739
Url.Query	?a=1&b=2&c=3
Url.Scheme	http

Since there is a redirect, as you can see, only the RawUrl contains the very original request, all other values are for the redirected Url, event the Request.Url.OriginalString

Do not mix using HtmlTextWriter and WebControl itself to set its attributes

Have a look the simplified code below first. The code is inside a ButtonControl class (assuming there is no null reference exception for CssClass property).

What does it do? First in the Render method we set the class to "Disabled" if Enabled property is false. Later when AddAttributesToRender method fired during base.Render() it will attach more CSS classes on the web control. All look good, right? No. It won’t work when Enabled == false.

Why? Check what the base.AddAttributesToRender() in line 16 does first. The AddAttributesToRender method will copy every single attribute from the web control into the HtmlTextWriter for page rendering later. If the attribute name has already existed in the HtmlTextWriter then the value of this attribute is abandoned to avoid value overwritten. This rule doesn't apply for Style attribute. For Style attribute, the value will be appended at the end of existing style string.

Ok, in our case when Enabled == false, since we have already set the Class attribute equal to "Disabled" then no matter how many other CssClass values assign to the web control directly those values will be abandoned when base. AddAttributesToRender() called.

So here is my simple advice: see the title of this article. Since we can always set attributes value on web control level it is unnecessary to touch the HtmlTextWriter under any condition.

Uhuru Cloud - My First App

I recently had a chance to test a newly release cloud service provided by Uhuru. The service is named Uhuru .NET Services for Cloud Foundry and you can find all information about it by accessing Products page of Uhuru website.

Uhuru Cloud support JAVA, .NET, php and other web applications but I'd like to start with the simplest one, my company website, a single html/js webpage.

To deploy app to Uhuru Cloud you will need either install a MMC (Microsoft Management Console) or Uhuru Visual Studio plugin. I choose MMC because the multi page user guide for installation of Visual Studio plugin is quite scary. Uhuru gives a very detailed PDF doc to help deploying app through MMC and the whole process is very straight forward.

One thing worth mentioning, when pushing your app to Uhuru CloudFoundry, it will automatically identify the suitable framework and runtime version for your app. In my case, since it is just a html/js site, CloudFoundry treat it as a php app (fair enough, compare with JAVA and IIS, php host cost less).  While after couple of mouse clicking and filling one small windows form, my first app is officially online now. See here: wandj.uhurucloud.net

Cheat Sheets for Web Development you must have

The original article is from: Cheat Sheets for Web Development you must have. Here I only copy those I am familiar with as a back up.

Tips of securing your ASP.NET MVC3 application

Tip 1, Always encode user input content before showing it on web page. You can do this simply by using Html.Encode() for innerHTML, Html.AttributeEncode() for attribute values and Url.Encode() for href link values.

You should know code expressions in Razor are implicitly HTML encoded, therefore in cases you do need to render some HTML value that should not be HTML encoded you can use the Html.Raw() method to ensure that the value is not encoded. BUT by doing so you have thrown open the door to evil.

Tip 2, Always JavaScript Encode user input values in your JavaScript code. You can use Ajax.JavaScriptStringEncode() to do this. The reason behind this is the html encode methods mentioned in Tip 1 cannot fully stop the XSS attacks because the html encoding cannot handle hex escape string but DOM does. Therefore the hacker may use your JS code to pass hex escape string into DOM object and the DOM object will convert the string back to JavaScript code and hurt you. Have a look the code below:

   1:  <div id="test">Text here is suppose to be replaced</div>

   2:  <script type="text/javascript">

   3:      $(function () {

   4:          $("#test").html(

   5:              '@Html.Encode("\\x3cscript\\x3e alert(\\x27oops\\x27) \\x3c/script\\x3e")');

   6:      });

   7:  </script>

Copy the code into your razor page and run it you will see an alert message says “oops”.

Tip 3, Follow the listed principles below to prevent CSRF attacks (if you don’t know what CSRF is please google it).

• First thing first, you need follow Tip1 and 2 to make you site XSS proof
• Always use POST instead of GET to submit data changes
• Validate HTTPReferrer by checking the HttpContext.Request.UrlReferrer to see if the client that posted the data was indeed from your site page
• Put token Verification in your form. You can use Html.AntiForgeryToken() to create a hidden token value in your form and in the related controller use ValidateAntiforgeryToken attribute to check if the token is matched

Tip 4, Protect your open redirection URL. The ASP.NET MVC 3 template includes code to protect against open redirection attacks in its AccountController after user successfully log on. You need also do similar thing in other places where a potential URL redirection may occur.

Tip 5, Use the AuthrizeAttribute to require a login access or role membership access. AuthrizeAttribute can be used on Action level or Controller level. You can use FormsAuthentication.SetAuthCookie() to keep the user login info into http cookie and it can be read by calling User.Identity.IsAuthenticated() in the requested Controller.

Tip 6, Use ValidationAttibute to do view-model level data value validation automatically. Most of the ValidationAttributes support client side validation if related jQuery lib file included which is a huge benefit for us developers.

Tip 7, Prevent over posting data through ASP.NET MVC3 model binding feature which can allow your attacker an opportunity to populate model properties you didn’t even put on your input forms. You can use BindAttribute to specify the name of properties which allow auto model binding. Alternatively you can do a manual data binding by calling UpdateModel() or TryUpdateModel() methods in the controller.

Still another way to deal with over-posting is to avoid binding directly to the data model. You can
do this by define a View-Model class that holds only the properties you want to allow the user to set.

Dynamic HttpModule registration in .NET Framework 4.0

This article could be useful for the scenario that you are developing HttpModule handling related functionality for a web application and you do not have the right to or not intend to modify the Global.asax file to register your HttpModule handler.

So here we go. Step 1, before register a HttpModule handler dynamically the first thing is to create the handler. See code below:

public class MyDynamicModule : IHttpModule

{

    void IHttpModule.Dispose()

    {

    }

    void IHttpModule.Init(HttpApplication context)

    {

        context.BeginRequest += new EventHandler(context_BeginRequest);

        context.EndRequest += new EventHandler(context_EndRequest);

    }

    //... more code ignored

}

Step 2, to register this HttpModule handler, we need the reference to Microsoft.Web.Infrastructure assemble and use the contained method DynamicModuleUtility.RegisterModule. See sample code below:

DynamicModuleUtility.RegisterModule(typeof(MyDynamicModule));

But where we call this line of code? As you know HttpModule registration has to be done before the instance of HttpAppliction start. To achieve that we need the newly introduced attribute class System.Web.PreApplicationStartMethodAttribute in .NET framework 4.0. Note this attribute is assembly level attribute. See the code below, as step 3 of our example, of how to use the attribute:

[assembly: PreApplicationStartMethod(typeof(DynamicHttpModule), "RegisterMyModule")]

The code above basically says I want the web application to do something in its pre-start event, which is, to execute the static method DynamicHttpModule.RegisterMyModule, see below the code of the DynamicHttpModule class:

public static class DynamicHttpModule

{

    public static void RegisterMyModule()

    {

        DynamicModuleUtility.RegisterModule(typeof(MyDynamicModule)); //code from step 2

    }

}

And that is it.

Controlling XML Serialization Using Attributes

The article is mainly focus on mapping the object type name with the xml file node name when De-serializing the xml file and when the node name is not a proper C# type name

Long story short, here are some brief ideas:

• Mapping xml root element name with class name: XmlRootAttribute
• Mapping element name other than the root with class property/field name: XmlElementAttribute
• Mapping List of elements underneath the parent node, normally to a ArrayList (or generic List<T>): XmlArrayAttribute and XmlArrayItemAttribute
• Mapping a xml attribute to a C# class property/field: XmlAttribute

More details you can check the Microsoft article: Controlling XML Serialization Using Attributes

Control.BeginInvoke() vs Control.Invoke()

The BeginInvoke() method has a IAsyncResult type return value and the Invoke() method simply return a object. This makes a big difference between them which is the BeginInvoke() is making a asynchronous callback to the thread created the handle of the control.

Assuming the main thread is A and currently executing code is in thread B, if call Invoke() from thread B then thread B has to wait for the callback to thread A being finished and then continue the following process.

NLog configuration file locations

Ways of passing data between WP7 pages

The book Programming Windows Phone 7 (MS Press) introduced 4 ways:

1. add a public propery on App class to retain the data during current lifecycle;
2. add a public propery on the target page, and then in the prev page's OnNavigatedFrom event set the value of the property for the next page;
3. use the property State from the application's current PhoneApplicationService instance, which is a type of generic dictionary. But the object been stored has to be serializable (the reason is when app goes to deactivated it will serialize the objects to isolated storage);
4. isolate storage. BUT obviously it is not wise to frequently access isolate storage during application running. Here is a great article of compairing the performance of using isolate storage http://blogs.claritycon.com/kevinmarshall/2010/11/03/wp7-serialization-comparison/

How Secure Are Query Strings Over HTTPS?

The original article is from HttpWatch's blog, a very good post. So, to make a long story short, the conclusions are: At the network level, HTTPS URL parameters are secure, but there are some other ways in which URL based data can leak:

1. URLs are stored in web server logs,
2. URLs are stored in the browser history,
3. URLs are passed in Referrer headers

Developing on SharePoint 2010 (Day 2)

a) SharePoint Server Architecture

b) SharePoint Object Hierarchy

So as you can see the SharePoint Object model is relatively straightforward. Here we have a very simple slip of code which start from reaching all services of the Farm and navigate through to the item and file level objects.

   1:  foreach (SPService svc in SPFarm.Local.Services)

   2:  {

   3:      if (svc is SPWebService)

   4:      {

   5:          foreach (SPWebApplication webApp in ((SPWebService)svc).WebApplications)

   6:          {

   7:              foreach (SPSite site in webApp.Sites)

   8:              {

   9:                  foreach (SPWeb web in site.AllWebs)

  10:                  {

  11:                      foreach (SPList list in web.Lists)

  12:                      {

  13:                          foreach (SPListItem i in list.Items)

  14:                          {

  15:                              // do something

  16:                          }

  17:                      }

  18:   

  19:                      foreach (SPFile f in web.Files)

  20:                      {

  21:                          // do something

  22:                      }

  23:                  }

  24:              }

  25:          }

  26:      }

  27:  }

Oracle Procedures (version 11.1)

General Related Data Dictionary Objects error$ source$   DBA ALL USER dba_arguments all_arguments user_arguments dba_errors all_errors user_errors dba_object_size all_object_size user_object_size dba_procedures all_procedures user_procedures dba_source all_source user_source System Privileges Related To Procedures create procedure alter any procedure   create any procedure   debug any procedure   drop any procedure   execute any procedure Object Privileges GRANT execute ON <procedure_name>; Privileges to tables and views granted through roles may not be valid within a procedure. See the section on AUTHID. GRANT execute ON testproc TO uwclass;   Stored Procedure No Parameters CREATE OR REPLACE PROCEDURE <procedure_name> IS BEGIN   <code here> END <procedure_name>; / CREATE OR REPLACE PROCEDURE no_param IS BEGIN   dbms_output.put_line('No Params'); END no_param; / set serveroutput on exec no_param; Single IN Parameter CREATE OR REPLACE PROCEDURE <procedure_name> ( <parameter_name> IN <data_type>) IS BEGIN   <code here> END <procedure_name>; / CREATE OR REPLACE PROCEDURE in_param (mesg VARCHAR2) IS BEGIN   dbms_output.put_line(mesg); END in_param; / set serveroutput on exec in_param('Single IN Parameter'); OUT Parameter CREATE OR REPLACE PROCEDURE <procedure_name> ( <parameter_name> OUT <data_type>) IS BEGIN   <code here> END <procedure_name>; / CREATE OR REPLACE PROCEDURE out_param(mesg OUT VARCHAR2) IS BEGIN   mesg := 'Single OUT Parameter'; END out_param; / set serveroutput on DECLARE s VARCHAR2(50); BEGIN   out_param(s);   dbms_output.put_line(s); END; / CREATE OR REPLACE PROCEDURE out_param (mesg OUT VARCHAR2) IS BEGIN   mesg := 'Single OUT Parameter'; END out_param; / SQL> var x VARCHAR2(30) SQL> exec out_param(:x) SQL> print x IN OUT Parameter CREATE OR REPLACE PROCEDURE <procedure_name> ( <parameter_name> IN OUT <data_type>) IS BEGIN   <code here> END <procedure_name>; / CREATE OR REPLACE PROCEDURE inout_param (mesg IN OUT VARCHAR2) IS BEGIN   mesg := mesg || ' an IN OUT Parameter'; END inout_param; / set serveroutput on DECLARE s VARCHAR2(50) := 'This procedure uses'; BEGIN   inout_param(s);   dbms_output.put_line(s); END; / Multiple Parameters CREATE OR REPLACE PROCEDURE <procedure_name> ( <parameter_name> IN <data_type>, <parameter_name> OUT <data_type>, <parameter_name> IN OUT <data_type>) IS BEGIN   <code here> END <procedure_name>; / CREATE OR REPLACE PROCEDURE many_params ( mesg1 IN     VARCHAR2, mesg2 OUT   VARCHAR2, mesg3 IN OUT VARCHAR2) IS BEGIN   mesg2 := mesg1 || 'Parameter As The OUT';   mesg3 := mesg3 || 'Returned'; END many_params; / set serveroutput on DECLARE iparm  VARCHAR2(50) := 'This is the IN '; oparm  VARCHAR2(50); ioparm VARCHAR2(50) := 'And This is the IN OUT '; BEGIN   many_params(iparm, oparm, ioparm);   dbms_output.put_line(oparm || ' ' || ioparm); END; /   Parameter DEFAULT Procedure Without Default CREATE OR REPLACE PROCEDURE <procedure_name> ( <parameter_name> <data_type> <default_value>) IS BEGIN   <code> END; / CREATE OR REPLACE PROCEDURE no_default(num_rows PLS_INTEGER) IS BEGIN   FOR r IN (SELECT object_name FROM all_objects             WHERE rownum < num_rows+1)   LOOP     dbms_output.put_line(r.object_name);   END LOOP; END no_default; / set serveroutput on exec no_default exec no_default(5); Procedure With Default CREATE OR REPLACE PROCEDURE <procedure_name> ( <parameter_name> <data_type> DEFAULT <default_value>) IS BEGIN   <code> END; / CREATE OR REPLACE PROCEDURE with_default ( num_rows PLS_INTEGER DEFAULT 20) IS BEGIN   FOR r IN (SELECT object_name FROM all_objects             WHERE rownum < num_rows+1)   LOOP     dbms_output.put_line(r.object_name);   END LOOP; END with_default; / set serveroutput on exec with_default exec with_default(5);   Parameter Notations Positional Notation exec <procedure_name> (<parameter>,<parameter>); CREATE OR REPLACE PROCEDURE positional ( min_nr PLS_INTEGER DEFAULT 100, max_nr PLS_INTEGER DEFAULT 1000) IS BEGIN   FOR r IN (SELECT table_name FROM user_tables             WHERE num_rows BETWEEN min_nr AND max_nr)   LOOP     dbms_output.put_line(r.table_name);   END LOOP; END positional; / set serveroutput on exec positional; exec positional(1); exec positional(1000, 500000); Named Notation exec <procedure_name> (<parameter_name> => <parameter>); exec positional; exec positional(min_nr => 1); exec positional(max_nr => 500); exec positional(max_nr => 5000); exec positional(max_nr => 10000); exec positional(max_nr => 50000); exec positional(max_nr => 999999); Mixed Notation exec <procedure_name> (<parameter>, <parameter_name> => <parameter>); exec positional(10, max_nr => 1000); exec dbms_stats(USER, cascade=>TRUE);   Parameter NOCOPY Note: NOCOPY is a hint ... not a directive ... and may be ignored. NOCOPY Performance Demo CREATE OR REPLACE PROCEDURE <procedure_name> ( <parameter_name> NOCOPY <data_type> <default_value>) IS BEGIN   <code> END; / -- Note: This first demo shows an example of NOCOPY not working conn / as sysdba GRANT execute ON dbms_crypto TO uwclass; conn uwclass/uwclass CREATE OR REPLACE PROCEDURE default_out (retval OUT VARCHAR2) AS BEGIN   retval := dbms_crypto.randombytes(32); END default_out; / CREATE OR REPLACE PROCEDURE nocopy_out (retval OUT NOCOPY VARCHAR2) AS BEGIN   retval := dbms_crypto.randombytes(32); END nocopy_out; / CREATE OR REPLACE PROCEDURE nocopy_test (reps IN PLS_INTEGER) AS bt1 TIMESTAMP(9); et1 TIMESTAMP(9); bt2 TIMESTAMP(9); et2 TIMESTAMP(9); outval VARCHAR2(64); BEGIN   bt1 := SYSTIMESTAMP;   FOR i IN 1..reps LOOP     default_out(outval);   END LOOP;   et1 := SYSTIMESTAMP;   bt2 := SYSTIMESTAMP;   FOR i IN 1..reps LOOP     nocopy_out(outval);   END LOOP;   et2 := SYSTIMESTAMP;   dbms_output.put_line('Default: ' || TO_CHAR(et1-bt1));   dbms_output.put_line('No Copy: ' || TO_CHAR(et2-bt2)); END nocopy_test; / -- Note: Tried this with REF CURSORS returning 40K rows with, again,  -- no measurable difference. But now watch what happens when used in a  -- demo received from Tom Kyte on 2/9/2007. CREATE OR REPLACE PROCEDURE p3(x IN OUT dbms_sql.varchar2s) AS BEGIN   FOR i IN 1 .. 2000000 LOOP     x(i) := RPAD('*', 255, '*');   END LOOP;   RAISE PROGRAM_ERROR; END; / CREATE OR REPLACE PROCEDURE p4(x IN OUT NOCOPY dbms_sql.varchar2s) AS BEGIN   FOR i IN 1 .. 2000000 LOOP     x(i) := RPAD( '*', 255, '*' );   END LOOP;   RAISE PROGRAM_ERROR; END; / set serveroutput on set timing on DECLARE l_x dbms_sql.varchar2s; BEGIN   FOR i IN 1 .. 1000000 LOOP     l_x(i) := RPAD('*', 250, '*');   END LOOP;   p3(l_x); EXCEPTION   WHEN OTHERS THEN     dbms_output.put_line('count = ' || l_x.count); END; / DECLARE   l_x dbms_sql.varchar2s; BEGIN   FOR i IN 1 .. 1000000 LOOP     l_x(i) := RPAD( '*', 250, '*' );   END LOOP;   p4(l_x); EXCEPTION   WHEN OTHERS THEN     dbms_output.put_line('count = ' || l_x.count); END; / Note: If a subprogram exists with an unhandled exception, the value assigned to its OUT and IN OUT formal parameters are not copied into the corresponding actual parameters, the changes appear to roll back. However, when you specify NOCOPY, assignments to the formal parameters immediately affect the actual parameters. So, if the subprogram exits with an unhandled exception, the (possibly unfinished) changes are not "rolled back." Error Handling With NOCOPY CREATE OR REPLACE PROCEDURE raise_error ( p_Raise BOOLEAN, p_ParameterA OUT NOCOPY NUMBER) AS BEGIN   p_ParameterA := 7;   IF p_Raise THEN     RAISE DUP_VAL_ON_INDEX;   ELSE     RETURN;   END IF; EXCEPTION   WHEN OTHERS THEN     RETURN; END raise_error; / set serveroutput on DECLARE p_B BOOLEAN := TRUE; n NUMBER; BEGIN   raise_error(p_B, n);   dbms_output.put_line(n); END; / CREATE OR REPLACE PROCEDURE raise_error ( p_Raise      BOOLEAN, p_ParameterA IN OUT NOCOPY NUMBER) AS BEGIN   IF p_Raise THEN     RAISE DUP_VAL_ON_INDEX;   ELSE     p_ParameterA := 999;   END IF; EXCEPTION   WHEN OTHERS THEN     NULL; END raise_error; / DECLARE p_B BOOLEAN := FALSE; n NUMBER := 100; BEGIN   raise_error(p_B, n);   dbms_output.put_line(n); END; / DECLARE   p_B BOOLEAN := TRUE;   n NUMBER := 100; BEGIN   raise_error(p_B, n);   dbms_output.put_line(n); END; /   AUTHID Note: For DEFINER RIGHTS objects execute privileges must be granted directly to the user; not to a role. With CURRENT USER rights can be granted to a role. AUTHID DEFINER (default) stored procedures are executed in environment equivalent to the one you get after SET ROLE NONE. In other words, roles are disabled for PL/SQL and any privileges granted via roles do not apply unless you created the procedure with AUTHID CURRENT_USER, in which case role privileges do apply (but executing such procedures is a bit more expensive because Oracle has to evaluate the privileges on every call). AUTHID Demo AUTHID DEFINER (the default) run the procedure with the rights of the procedure's owner. AUTHID CURRENT_USER run the procedure with the rights of the executing schema. conn / as sysdba CREATE USER abc IDENTIFIED BY abc DEFAULT TABLESPACE uwdata TEMPORARY TABLESPACE temp QUOTA 10M ON uwdata; GRANT create session, create table, create procedure to abc; conn abc/abc CREATE TABLE t1 ( mycol VARCHAR2(20)); CREATE TABLE t2 ( yourcol NUMBER(10,2)); CREATE TABLE t3 ( ourcol DATE); conn uwclass/uwclass CREATE OR REPLACE PROCEDURE definer_test AUTHID DEFINER IS BEGIN   FOR rec IN (SELECT table_name FROM user_tables)   LOOP     dbms_output.put_line(rec.table_name);   END LOOP; END definer_test; / CREATE OR REPLACE PROCEDURE cu_test AUTHID CURRENT_USER IS BEGIN   FOR rec IN (SELECT table_name FROM user_tables)   LOOP     dbms_output.put_line(rec.table_name);   END LOOP; END cu_test; / set serveroutput on exec definer_test; exec cu_test; GRANT execute on definer_test TO abc; GRANT execute on cu_test TO abc; conn abc/abc set serveroutput on exec uwclass.definer_test; exec uwclass.cu_test; --====================================== conn abc/abc CREATE OR REPLACE PROCEDURE definer_test AUTHID DEFINER IS BEGIN   FOR rec IN (SELECT srvr_id FROM servers)   LOOP     dbms_output.put_line(rec.srvr_id);   END LOOP; END definer_test; / CREATE OR REPLACE PROCEDURE cu_test AUTHID CURRENT_USER IS BEGIN   FOR rec IN (SELECT srvr_id FROM servers)   LOOP     dbms_output.put_line(rec.srvr_id);   END LOOP; END cu_test; / set serveroutput on exec definer_test; exec cu_test; GRANT execute on definer_test TO uwclass; GRANT execute on cu_test TO uwclass; conn uwclass/uwclass set serveroutput on exec abc.definer_test; exec abc.cu_test;   Procedure Demos Calculate Business Days CREATE TABLE daterange ( beg_date DATE, end_date DATE, biz_days NUMBER(5)); INSERT INTO daterange VALUES (SYSDATE-10, SYSDATE+10, NULL); INSERT INTO daterange VALUES (SYSDATE-17, SYSDATE+10, NULL); INSERT INTO daterange VALUES (SYSDATE-22, SYSDATE+12, NULL); CREATE OR REPLACE PROCEDURE bizdays AUTHID CURRENT_USER IS CURSOR bd_cur IS SELECT beg_date, end_date, 0 FROM daterange; TYPE bDate_tab IS TABLE OF DATE INDEX BY BINARY_INTEGER; bdate bDate_tab; TYPE eDate_tab IS TABLE OF DATE INDEX BY BINARY_INTEGER; edate eDate_tab; TYPE bDay_tab IS TABLE OF NUMBER INDEX BY BINARY_INTEGER; bDays bDay_tab; NumDays PLS_INTEGER; TestDate DATE; BEGIN   OPEN bd_cur;   LOOP     FETCH bd_cur BULK COLLECT INTO bDate, eDate, bDays LIMIT 100;     EXIT WHEN bd_cur%NOTFOUND;     FOR i IN bdate.FIRST .. bdate.LAST LOOP       TestDate := bdate(i);       NumDays := 0;       FOR j IN 1 .. (edate(i) - bdate(i) + 1) LOOP         IF TO_CHAR(TestDate, 'D') BETWEEN '2' AND '6' THEN           NumDays := NumDays+1;         END IF;         TestDate := TestDate + 1;       END LOOP;       bDays(i) := NumDays;       UPDATE daterange       SET biz_days = bDays(i)       WHERE beg_date = bDate(i)       AND end_date = eDate(i);     END LOOP;   END LOOP;   COMMIT;   CLOSE bd_cur; END bizdays; /

Temporary Tables vs. Table Variables

	Temp Table	Table Variable
record transaction logs?	Yes	No
inside transaction scope?	Yes	No
cross execution batch?	Global temp table	No way
support index?	Yes	No
where to store?	Tempdb	Memory

Developing on SharePoint 2010 (Day 1)

This article is talking about developing web parts for SharePoint 2010 using Visual Studio 2010. If you are looking for articles for older version of SharePoint or Visual Studio I recommend this URL: http://www.aspfree.com/c/a/Windows-Scripting/Beginning-SharePoint-Web-Part-Development/

First of all, there are two types of web parts VS 2010 supports: Standard Web Part (which in VS 2010 Item Template list named Web Part) and Visual Web Part.

A standard web part is just a class inherited from abstract class WebPart. Below is the Inheritance Hierarchy of WebPart class.

   1:  WebControl

   2:  |---Panel

   3:      |---Part

   4:          |---WebPart

To add contents into standard web part we need override the CreateChildControls method and inside put code logic to create sub controls manually and add them into the Controls collection. See the code below:

   1:  protected override void CreateChildControls()

   2:  {

   3:      LiteralControl msg = new LiteralControl("test");

   4:      this.Controls.Add(msg);

   5:  }

The visual web part, in the other hand, give you a visual design UI to drag existed web control in. But if you check the source code, a visual web part is just a standard web part plus a Web.UI.UserControl. And in the overrided CreateChildControls method used a trick like this:

this.Controls.Add(Page.LoadControl(_controlPath));

One more thing for visual web part: the value _controlPath is the physical URL of the user control and the value of it is maintained by VS 2010.