Gaurav http://allyourcodearebelongto.me/blog All About Everything Sun, 23 Nov 2008 19:44:48 +0000 http://wordpress.org/?v=2.9.1 en hourly 1 Autocompleting ssh, rake, cap command parameters using PowerShell http://allyourcodearebelongto.me/blog/2008/11/23/autocompleting-ssh-rake-cap-command-parameters-using-powershell/ http://allyourcodearebelongto.me/blog/2008/11/23/autocompleting-ssh-rake-cap-command-parameters-using-powershell/#comments Sun, 23 Nov 2008 19:42:54 +0000 Gaurav http://allyourcodearebelongto.me/blog/2008/11/23/autocompleting-ssh-rake-cap-command-parameters-using-powershell/ I have mentioned in my previous posts that PowerShell has excellent and customizabme autocompletion support.
And some awesome plugins like powertab exist to take this functionality to another level.
Here is one easy way you can customize autocompletion easily by writing a couple of lines of code.

Open your PowerShell profile file (for help see here) and create a function with name "global:TabExpansion" in it.
This function basically overrides default autocompletion and sends your values to the prompt. For example for it to read your ssh files and display hosts in it the function would look something like this: 

# customized tab expansion
function global:TabExpansion {
  param($line, $lastWord)
  
  if ($line -match 'ssh'){
    return cat C:\Users\Gaurav\.ssh\config | Select-String '^Host ' | % { if($_.Line -match "^Host ($lastword\S*)" ){ $matches[1] }}
  }
}

Here it is just parsing the config file and running it through a regular-expression and returning the results as an array. Here is what a typical ssh config file looks like:

Host blog
  User username
  HostName sitename.com
  IdentityFile ~/.ssh/id_dsa
  
Host blog_2
  User username
  HostName sitename.com
  IdentityFile ~/.ssh/id_dsa

 

Here is what the output looks like if I write “ssh ” and the press tab key twice:

image

image

 

Also this same function can easily be modded to autocomplete for ruby/rails commands like rake and cap.

Here is is my complete function:

# customized tab expansion
function global:TabExpansion {
  param($line, $lastWord)
  
  if ($line -match 'ssh'){
    return cat C:\Users\Gaurav\.ssh\config | Select-String '^Host ' | % { if($_.Line -match "^Host ($lastword\S*)" ){ $matches[1] }}
  }
  elseif ($line -match 'rake'){
    rake -T | % { if($_ -match "^rake ($lastword\S*)" ){ $matches[1] }}
  } 
  elseif ($line -match 'cap'){
    cap -T | % { if($_ -match "^cap ($lastword\S*)" ){ $matches[1] }}
  }
}

]]> http://allyourcodearebelongto.me/blog/2008/11/23/autocompleting-ssh-rake-cap-command-parameters-using-powershell/feed/ 0 Using RubyAmf for creating a CRUD application in Rails http://allyourcodearebelongto.me/blog/2008/11/12/using-rubyamf-for-creating-a-crud-application-in-rails/ http://allyourcodearebelongto.me/blog/2008/11/12/using-rubyamf-for-creating-a-crud-application-in-rails/#comments Wed, 12 Nov 2008 11:35:45 +0000 Gaurav http://allyourcodearebelongto.me/blog/2008/11/12/using-rubyamf-for-creating-a-crud-application-in-rails/ CRUD applications can be easily created using Ruby on Rails as the backend and Flex as the frontend app using the XML format as demonstrated here.

Another way to create such applications is using the AMF protocol which is optimized for network communications and stores objects in binary format so consumes less bandwidth. Another compelling reason for using this library is that it doesn’t feel like a hack as it integrates very nicely with Rails :)

The Ruby port of this protocol known as RubyAmf can be downloaded from from here. More information about this protocol can be found here.

Lets start making the application.

 

Create a Rails Application:

   1: > rails amf_demo

This will create a rails application. Please note that I am using Rails 2.1.2 for this example and it is not tested on any other Rails version.

Generate a scaffold:

   1: > ruby script/generate scaffold blog_post title:string body:text

Create Tables and run Migrations:

   1: > rake db:migrate

   1: > rake db:create

Install the RubyAmf Plugin:

 

 

Step 1: Lets try if everything is working fine by creating a new Flex Project

Open Flex Builder and create a new Flex Project

image

 

Select all the default options and click Finish.

Now right click on the project you just created in Flex Navigator and select Properties. The Flex Build Path should have the following settings:

image

The Output folder will point to the public folder of the amf_demo Rails application and the Output folder URL will point to the Rails server which in  my case is localhost on port 80.

 

Start the Rails server:

   1: > ruby script/server -p 80

And start the project in Debugging mode. The Web Browser will show up a blank page.

image

 

Some helper files can be downloaded from the RubyAmf project site here

These are located in the “flash9_helloworld\flash9_helloworld\app\flash\org” of the archive.

Or just download the extracted archive from here. Put these files in the “src” directory of your Flex project. So the “src” folder should now have a folder with name “org” inside it.

Fetching Data off the server:

Firstly we will need to update the “BlogPostsController” so that it can generate amf content. It can be done by modifying the the index action to following:

   1: def index
   2:   @blog_posts = BlogPost.find(:all)
   3:  
   4:   respond_to do |format|
   5:     format.html # index.html.erb
   6:     format.amf  { render :amf => @blog_posts }
   7:     format.xml  { render :xml => @blog_posts }
   8:   end
   9: end

 

Now change the main mxml file of the Flex project to include code that fetches data off the server. In the current project the name of the file should be “amf_demo.mxml”

   1: <?xml version="1.0" encoding="utf-8"?>
   2: <mx:Application xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical"
   3:     creationComplete="creationCompleteEvent(event)" xmlns:ssr="org.rubyamf.remoting.ssr.*">
   4:     <mx:Script>
   5:         <![CDATA[
   6:             import org.rubyamf.remoting.ssr.FaultEvent;
   7:             import org.rubyamf.remoting.ssr.ResultEvent;
   8:             import org.rubyamf.remoting.ssr.RemotingService;
   9:             private function creationCompleteEvent(evemt:Event):void{
  10:                 var service:RemotingService = new RemotingService("/rubyamf/gateway", "BlogPostsController");
  11:                 service.index([], onIndexResult, onFault);
  12:             }
  13:             private function onIndexResult(event:ResultEvent):void{
  14:                 trace(event.result);
  15:             }
  16:             private function onFault(event:FaultEvent):void{
  17:                 trace(event.fault);
  18:             }
  19:         ]]>
  20:     </mx:Script>
  21: </mx:Application>

 

The main component of this code is the “RemoteService” class and in the above given example it is invoking the index action on the server. On running this example the onIndexResult should be triggered and the debug window should show all the BlogPosts in the db. You can also put  a break-pointer to analyze the results.

image

One thing you will notice here is that the output result are plain ActionScript Objects. In the next step I will show how to map these objects to an ActionScript class.

Download code for upto this point:

  1. Rails Code
  2. Flex Code

 

Step 2: Mapping results to Action Script classes

Generate Mappings:

In the project root directory run the following command.

   1: > ruby script/generate rubyamf_mappings

Copy the output of this command to config/rubyamf_config.rb in the Rails project and now this file should look like this after some modifications:

   1: require 'app/configuration'
   2: module RubyAMF
   3:   module Configuration
   4:     #set the service path used in all requests
   5:     # RubyAMF::App::RequestStore.service_path = File.expand_path(RAILS_ROOT) + '/app/controllers'
   6:  
   7:     # => CLASS MAPPING CONFIGURATION
   8:     
   9:     # => Global Property Ignoring
  10:     # By putting attribute names into this array, you opt in to globally ignore these properties on incoming objects.
  11:     # If you want to ignore specific properties on certain objects, use the :ignore_fields property in a 
  12:     # Class Mapping definition (see CLASS MAPPING DEFINITIONS)
  13:     # ClassMappings.ignore_fields = ['created_at','created_on','updated_at','updated_on']
  14:   
  15:     # => Case Translations
  16:     # Most actionscript uses camelCase instead of snake_case. Set ClassMappings.translate_case to true if want translations to occur.
  17:     # The translations only occur on object properties
  18:     # An incoming property like: myProperty gets turned into my_property
  19:     # An outgoing property like my_property gets turned into myProperty
  20:     ClassMappings.translate_case = true
  21:   
  22:     # => Force Active Record Ids
  23:     # includes the id field for activerecord objects even if you don't specify it when using custom attributes. This is important for deserialization
  24:     # where ids are needed to keep active record association integrity.
  25:     # ClassMappings.force_active_record_ids = true
  26:     
  27:     # => Hash key access
  28:     # You can choose how keys in hashes are created. As :string, :symbol, or :indifferent. 
  29:     # :string creates keys as hash['key']
  30:     # :symbol creates keys as hash[:key]
  31:     # :indifferent uses rails HashWithIndifferentAccess so you can use hash[:key] of hash['key']
  32:     # There are performance issues with HashWithIndifferentAccess. Use :symbol or :string for best performance.
  33:     # The default is :symbol
  34:     ClassMappings.hash_key_access = :string
  35:     
  36:     # => Assume Class Types
  37:     # This tells RubyAMF to assume class type transfers. So when you register a class Alias from Flash or Flex like this:
  38:     # Flash::   fl.net.registerClassAlias('User',User)
  39:     # Flex::    [RemoteClass(alias='User')]
  40:     # RubyAMF will automagically convert it to a User active record without you having to create a class mapping.
  41:     # This also works with non active record class mappings. See the wiki on the google code page for a downloadable example.
  42:     # ClassMappings.assume_types = false
  43:     
  44:     # => Class Mapping Definitions
  45:     # A Class Mapping definition conists of at least these two properties:
  46:     # :actionscript   # The incoming action script class to watch for
  47:     # :ruby           # The ruby class to turn it into
  48:     #
  49:     # => Optional value object properties:
  50:     # :type           # Used to spectify the type of VO, valid options are 'active_record', 'custom',  (or don't specify at all)
  51:     # :associations   # Specify which associations to read on the active record (only applies to active records)
  52:     # :attributes     # Specifically which attributes to include in the serialization
  53:     # :methods        # An array of methods to call and place values in a similarly named attribute on the Actionscript Object (outgoing, or Rails => Actionscript only)
  54:     # :ignore_fields  # An array of field names you want to ignore on incoming classes
  55:     #
  56:     # If you are using ActiveRecord VO's you do not need to specify a fully qualified class path to the model, you can just define the class name, 
  57:     # EX: ClassMappings.register(:actionscript => 'vo.Person', :ruby => 'Person', :type => 'active_record')
  58:     #
  59:     # If you are using custom VO's you would need to specify the fully qualified class path to the file
  60:     # EX: ClassMappings.register(:actionscript => 'vo.Person', :ruby => 'org.mypackage.Person')
  61:     #
  62:     # ClassMappings.register(:actionscript => 'Person', :ruby => 'Person', :type => 'active_record', :attributes => ["id", "address_id"])
  63:     # ClassMappings.register(:actionscript => 'User', :ruby => 'User', :type => 'active_record', :associations=> ["addresses", "credit_cards"])
  64:     # ClassMappings.register(:actionscript => 'Address', :ruby => 'Address', :type => 'active_record')
  65:     # ClassMappings.register(:actionscript => 'User', :ruby => 'User', :type => 'active_record', :associations=> ["addresses", "credit_cards"], :methods => ["friends"])
  66:     #
  67:     # => Class Mapping Scope (Advanced Usage)
  68:     # You can also specify a class mapping scope if you want. For example, lets say you need certain attributes for a book when you are viewing a book
  69:     # in flex as opposed to editing a book (where you would need more parameters). You can define a scope mapping parameter for ":attributes"
  70:     # or for ":associations." You're mapping would look something like this.
  71:     # ClassMappings.register(
  72:     #   :actionscript  => 'com.mixbook.vo.books.BookVO',      
  73:     #   :ruby          => 'Book',      
  74:     #   :type          => 'active_record',
  75:     #   :associations  => ["access_info", "pages", "page_ratio"],  
  76:     #   :attributes    => {:viewing => ["description", "title"], :editing => ["id","published_at","theme_id"] }    <=== notice the hash instead of an array
  77:     #
  78:     # Now, to call the class mapping scope of editing (you are sending objects to the editing application), your controller call would look like this:
  79:     # EX: render :amf => book, :class_mapping_scope => :editing
  80:     #
  81:     # You can also specify a default scope to use. If you don't set this and you don't specify a class mapping scope on an attribute or association, then 
  82:     # it will not have a scope to use and will not add any attributes or associations (whichever it cant match) to that association.
  83:     # ClassMappings.default_mapping_scope = :viewing
  84:   
  85:     # => Date Conversion
  86:     # Incoming dates from Flash by default are Time objects, this can conver to DateTime if needed
  87:     # ClassMappings.use_ruby_date_time = false
  88:   
  89:     # => Use Array Collection
  90:     # By setting this to true, you opt in to using array collections for all the arrays generated by the body of your request.
  91:     # Note: This only works for amf3 with Remote Object, NOT with Net Connection.
  92:     # ClassMappings.use_array_collection = false
  93:   
  94:     # => Check for Associations
  95:     # Enabling this will automagically pick up eager loaded association data on objects returned through RubyAMF.
  96:     # If this is disabled, you will need to specify any associations you DO want picked up in the ClassMapping
  97:     # ClassMappings.check_for_associations = true
  98:     
  99:     # => NAMED PARAMETER MAPPING CONFIGURATION
 100:     
 101:     #=> Always Put Remoting Parameters into the "params" hash
 102:     # If set to true, arguments from Flash/Flex will come in to the controllers as params[0], params[1], etc.. This is especally useful if you are sending huge objects
 103:     # from Flex into Ruby so it doesnt eat up all your output window with outputting the params in the controller/action header information while in dev mode.
 104:     # Even if its set to false, if you specify specific ParameterMappings, those will still get entered as the param keys you specify. Likewise, you
 105:     # always have access to the parameters from rubyamf in your controller by calling rubyamf_params[0], rubyamf_params[1], etc regardless of
 106:     # if it this is set or not.
 107:     # ParameterMappings.always_add_to_params = true
 108:      
 109:     # => Return Top Level Hash
 110:     # For those scaffolding users out there, who want the top-level object to come as a hash so scaffolding works out of the box.
 111:     ParameterMappings.scaffolding = true
 112:   
 113:     # => Incoming Remoting Parameter Mappings
 114:     # Incoming Remoting Parameter mappings allow you to map an incoming requests parameters into rails' params hash
 115:     #
 116:     # Here's an example:
 117:     # ParameterMappings.register(:controller => :UserController, :action => :find_friend, :params => { :friend => "[0]['friend']" })
 118:     ClassMappings.register(
 119:       :actionscript  => 'BlogPost',
 120:       :ruby          => 'BlogPost',
 121:       :type          => 'active_record',
 122:       :attributes    => ["id", "title", "body", "created_at", "updated_at"])
 123:   end
 124: end

 

Create a new ActionScript Class with name “BlogPost” and it should have the following code:

   1: package
   2: {
   3:     [RemoteClass(alias='BlogPost')]
   4:     public class BlogPost
   5:     {
   6:         public var id:int;
   7:         public var title:String; 
   8:         public var body:String;
   9:     }
  10: }

 

And modify the “amf_demo.mxml” file to include this file. Also change the “service.index” method to to use this file:

   1: service.index([], onIndexResult, onFault, BlogPost);

 

Now when you debug and check the output in the “onIndexResult” it should look like:

image

 

Download code for upto this point:

  1. Rails Code
  2. Flex Code

 

Step 3: Creating a UI

Change the amf_demo.mxml to the following code.

   1: <?xml version="1.0" encoding="utf-8"?>
   2: <mx:Application xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical"
   3:     creationComplete="creationCompleteEvent(event)" xmlns:ssr="org.rubyamf.remoting.ssr.*">
   4:     <mx:Script>
   5:         <![CDATA[
   6:             import mx.collections.ArrayCollection;
   7:             import org.rubyamf.remoting.ssr.FaultEvent;
   8:             import org.rubyamf.remoting.ssr.ResultEvent;
   9:             import org.rubyamf.remoting.ssr.RemotingService;
  10:             [Bindable]
  11:             private var blogPosts:ArrayCollection;
  12:             
  13:             private function creationCompleteEvent(evemt:Event):void{
  14:                 var service:RemotingService = new RemotingService("/rubyamf/gateway", "BlogPostsController");
  15:                 service.index([], onIndexResult, onFault, BlogPost);
  16:             }
  17:             private function onIndexResult(event:ResultEvent):void{
  18:                 blogPosts = new ArrayCollection(event.result as Array);
  19:             }
  20:             private function onFault(event:FaultEvent):void{
  21:                 trace(event.fault);
  22:             }
  23:         ]]>
  24:     </mx:Script>
  25:     
  26:     <mx:List width="300" dataProvider="{blogPosts}" labelField="title" />
  27: </mx:Application>

Here the result of the onIndexResult event is bound to a list to Display all the blog posts in the database. This is probably the simplest way to do this. The output will look like:

image

Download code for upto this point:

  1. Rails Code (no changes)
  2. Flex Code

 

Step 4: Refactoring Completing CRUD

As a single page can display more than one list, it would make more sense if we move the code into separate files for each list. So I will be moving all the BlogPost related code in the “BlogPost” class.

So the “BlogPost.as” will look like:

   1: package
   2: {
   3:     import mx.collections.ArrayCollection;
   4:     import org.rubyamf.remoting.ssr.FaultEvent;
   5:     import org.rubyamf.remoting.ssr.ResultEvent;
   6:     import org.rubyamf.remoting.ssr.RemotingService;
   7:             
   8:     [RemoteClass(alias='BlogPost')]
   9:     public class BlogPost
  10:     {
  11:         public var id:int;
  12:         public var title:String; 
  13:         public var body:String;
  14:         
  15:         [Bindable]
  16:         public static var blogPosts:ArrayCollection;
  17:         private var service:RemotingService = new RemotingService("/rubyamf/gateway", "BlogPostsController");
  18:         
  19:         public static function getBlogPosts():void{
  20:             (new BlogPost()).getBlogPosts();
  21:         }
  22:         private function getBlogPosts():void{
  23:             service.index([], onIndexResult, onFault, BlogPost);
  24:         }
  25:         private function onIndexResult(event:ResultEvent):void{
  26:             BlogPost.blogPosts = new ArrayCollection(event.result as Array);
  27:         }
  28:         private function onFault(event:FaultEvent):void{
  29:             trace(event.fault);
  30:         }
  31:     }
  32: }

 

And “amf_demo.mxml”:

   1: <?xml version="1.0" encoding="utf-8"?>
   2: <mx:Application xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical"
   3:     creationComplete="creationCompleteEvent(event)" xmlns:ssr="org.rubyamf.remoting.ssr.*">
   4:     <mx:Script>
   5:         <![CDATA[
   6:             private function creationCompleteEvent(evemt:Event):void{
   7:                 BlogPost.getBlogPosts();
   8:             }
   9:         ]]>
  10:     </mx:Script>
  11:     <mx:List width="300" dataProvider="{BlogPost.blogPosts}" labelField="title" />
  12: </mx:Application>

 

Download code for upto this point:

  1. Rails Code (no changes)
  2. Flex Code

 

Step 5: Completing CRUD

After further refactoring the code and adding methods for creating/editing/destroying the code looks like:

  • amf_demo.mxml 
   1: <?xml version="1.0" encoding="utf-8"?>
   2: <mx:Application xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical"
   3:     creationComplete="creationCompleteEvent(event)" xmlns:ssr="org.rubyamf.remoting.ssr.*">
   4:     <mx:Script>
   5:         <![CDATA[
   6:             private function creationCompleteEvent(event:Event):void{
   7:                 BlogPost.parentContainer = this;
   8:                 BlogPost.getBlogPosts();
   9:             }
  10:             private function blogPostCreateEvent(event:Event):void{
  11:                 (new BlogPost(title.text, body.text)).create();
  12:             }
  13:             private function blogPostUpdateEvent(event:Event):void{
  14:                 var blogPost:BlogPost     = currentPost();
  15:                 blogPost.title             = title.text;
  16:                 blogPost.body              = body.text; 
  17:                 blogPost.update();
  18:             }
  19:             private function blogPostdestroyEvent(event:Event):void{
  20:                 currentPost().destroy();
  21:             }
  22:             private function currentPost():BlogPost{
  23:                 return blogPosts.selectedItem as BlogPost;
  24:             }
  25:         ]]>
  26:     </mx:Script>
  27:     <mx:List width="300" dataProvider="{BlogPost.blogPosts}" labelField="title" id="blogPosts"/>
  28:     <mx:Panel layout="vertical" paddingBottom="5" paddingLeft="5" paddingRight="5" paddingTop="5">
  29:         <mx:HBox width="100%">
  30:             <mx:Label text="Post Title" width="60"/>
  31:             <mx:TextInput id="title" text="{blogPosts.selectedItem.title}"/>
  32:         </mx:HBox>
  33:         <mx:HBox width="100%">
  34:             <mx:Label text="Post Body" width="60" />
  35:             <mx:TextArea id="body" text="{blogPosts.selectedItem.body}"/>
  36:         </mx:HBox>
  37:         <mx:ControlBar>
  38:             <mx:Button label="Create" click="blogPostCreateEvent(event)" />
  39:             <mx:Button label="Update" click="blogPostUpdateEvent(event)" />
  40:             <mx:Button label="Destroy" click="blogPostdestroyEvent(event)" />
  41:         </mx:ControlBar>
  42:     </mx:Panel>
  43: </mx:Application>

  • BlogPost.as 
   1: package
   2: {
   3:     import mx.collections.ArrayCollection;
   4:     import mx.core.UIComponent;
   5:     
   6:     import org.rubyamf.remoting.ssr.FaultEvent;
   7:     import org.rubyamf.remoting.ssr.RemotingService;
   8:     import org.rubyamf.remoting.ssr.ResultEvent;
   9:             
  10:     [RemoteClass(alias='BlogPost')]
  11:     public class BlogPost
  12:     {
  13:         public var id:int;
  14:         public var title:String; 
  15:         public var body:String;
  16:         public var simpleErrors:Object;
  17:         
  18:         [Bindable]
  19:         public static var blogPosts:ArrayCollection;
  20:         public static var parentContainer:UIComponent;
  21:         private var service:RemotingService = new RemotingService("/rubyamf/gateway", "BlogPostsController");
  22:         
  23:         public function BlogPost(title:String = "", body:String = ""){
  24:             this.title = title;
  25:             this.body  = body;
  26:         }
  27:         // index methods
  28:         public static function getBlogPosts():void{
  29:             (new BlogPost()).getBlogPosts();
  30:         }
  31:         private function getBlogPosts():void{
  32:             service.index([], onIndexResult, onFault, BlogPost);
  33:         }
  34:         private function onIndexResult(event:ResultEvent):void{
  35:             BlogPost.blogPosts = new ArrayCollection(event.result as Array);
  36:         }
  37:         // create methods
  38:         public function create():void{
  39:             service.create([this], onCreateResult, onFault, BlogPost);
  40:         }
  41:         private function onCreateResult(event:ResultEvent):void{
  42:             var blogPost:BlogPost = event.result as BlogPost;
  43:             if(BlogPost.count(blogPost.simpleErrors) == 0)
  44:                 BlogPost.blogPosts.addItem(blogPost);
  45:             else
  46:                 showErrors(blogPost.simpleErrors);
  47:         }
  48:         // update methods
  49:         public function update():void{
  50:             service.update([this], onUpdateResult, onFault, BlogPost);
  51:         }
  52:         private function onUpdateResult(event:ResultEvent):void{
  53:             var blogPost:BlogPost = event.result as BlogPost;
  54:             if(BlogPost.count(blogPost.simpleErrors) == 0)
  55:                 BlogPost.blogPosts[BlogPost.blogPosts.getItemIndex(this)] = blogPost;
  56:             else
  57:                 showErrors(blogPost.simpleErrors);
  58:         }
  59:         // update methods
  60:         public function destroy():void{
  61:             service.destroy([this], onDestroyResult, onFault, BlogPost);
  62:         }
  63:         private function onDestroyResult(event:ResultEvent):void{
  64:             var blogPost:BlogPost = event.result as BlogPost;
  65:             BlogPost.blogPosts.removeItemAt(BlogPost.blogPosts.getItemIndex(this));
  66:         }
  67:         
  68:         // common methods
  69:         private function showErrors(errors:Object):void{
  70:             for(var error:String in errors){
  71:                 var field:UIComponent =  parentContainer[error] as UIComponent;
  72:                 field.errorString = (errors[error] as Array).join('\r');
  73:             }
  74:         }
  75:         private static function count(object:Object):int{
  76:             var i:int = i;
  77:             for(var obj:* in object) i++;
  78:             return i;
  79:         }
  80:         private function onFault(event:FaultEvent):void{
  81:             trace(event.fault);
  82:         }
  83:     }
  84: }

 

  • BlogPostsController 
   1: class BlogPostsController < ApplicationController
   2:   before_filter :filter_attributes, :only=> [:create, :update]
   3:  
   4:   def filter_attributes
   5:     params[:blog_post] && params[:blog_post][:attributes] && params[:blog_post] = params[:blog_post][:attributes]
   6:   end
   7:   
   8:   # GET /blog_posts
   9:   # GET /blog_posts.xml
  10:   def index
  11:     @blog_posts = BlogPost.find(:all)
  12:  
  13:     respond_to do |format|
  14:       format.html # index.html.erb
  15:       format.amf  { render :amf => @blog_posts }
  16:       format.xml  { render :xml => @blog_posts }
  17:     end
  18:   end
  19:   
  20:   # GET /blog_posts/1
  21:   # GET /blog_posts/1.xml
  22:   def show
  23:     @blog_post = BlogPost.find(params[:id])
  24:  
  25:     respond_to do |format|
  26:       format.html # show.html.erb
  27:       format.xml  { render :xml => @blog_post }
  28:     end
  29:   end
  30:  
  31:   # GET /blog_posts/new
  32:   # GET /blog_posts/new.xml
  33:   def new
  34:     @blog_post = BlogPost.new
  35:  
  36:     respond_to do |format|
  37:       format.html # new.html.erb
  38:       format.xml  { render :xml => @blog_post }
  39:     end
  40:   end
  41:  
  42:   # GET /blog_posts/1/edit
  43:   def edit
  44:     @blog_post = BlogPost.find(params[:id])
  45:   end
  46:  
  47:   # POST /blog_posts
  48:   # POST /blog_posts.xml
  49:   def create
  50:     @blog_post = BlogPost.new(params[:blog_post])
  51:  
  52:     respond_to do |format|
  53:       if @blog_post.save
  54:         flash[:notice] = 'BlogPost was successfully created.'
  55:         format.html { redirect_to(@blog_post) }
  56:         format.amf  { render :amf => @blog_post }
  57:         format.xml  { render :xml => @blog_post, :status => :created, :location => @blog_post }
  58:       else
  59:         format.html { render :action => "new" }
  60:         format.amf  { render :amf => @blog_post }
  61:         format.xml  { render :xml => @blog_post.errors, :status => :unprocessable_entity }
  62:       end
  63:     end
  64:   end
  65:  
  66:   # PUT /blog_posts/1
  67:   # PUT /blog_posts/1.xml
  68:   def update
  69:     @blog_post = BlogPost.find(params[:id])
  70:  
  71:     respond_to do |format|
  72:       if @blog_post.update_attributes(params[:blog_post])
  73:         flash[:notice] = 'BlogPost was successfully updated.'
  74:         format.html { redirect_to(@blog_post) }
  75:         format.amf  { render :amf => @blog_post }
  76:         format.xml  { head :ok }
  77:       else
  78:         format.html { render :action => "edit" }
  79:         format.amf  { render :amf => @blog_post }
  80:         format.xml  { render :xml => @blog_post.errors, :status => :unprocessable_entity }
  81:       end
  82:     end
  83:   end
  84:  
  85:   # DELETE /blog_posts/1
  86:   # DELETE /blog_posts/1.xml
  87:   def destroy
  88:     @blog_post = BlogPost.find(params[:id])
  89:     @blog_post.destroy
  90:  
  91:     respond_to do |format|
  92:       format.html { redirect_to(blog_posts_url) }
  93:       format.amf  { render :amf => @blog_post }
  94:       format.xml  { head :ok }
  95:     end
  96:   end
  97: end

Please note the before filter in the rails controller which is converting the params in appropriate format so that rest of the code continues to work seamlessly.

 

Here is what the output looks like when you run the application

image

 

Download code for upto this point:

  1. Rails Code
  2. Flex Code

 

So the steps taken in completing this application were:

  1. Creating a Rails app
  2. Install RubyAmf plugin
  3. Configure the RubyAmf plugin
  4. Create a Flex app
  5. Install the appropriate libraries in it
  6. Create and refactor code

 

Questions and comments are welcome :)

]]> http://allyourcodearebelongto.me/blog/2008/11/12/using-rubyamf-for-creating-a-crud-application-in-rails/feed/ 8 Multiple ways to open PowerShell in the current Explorer window http://allyourcodearebelongto.me/blog/2008/10/14/multiple-ways-to-open-powershell-in-the-current-explorer-window/ http://allyourcodearebelongto.me/blog/2008/10/14/multiple-ways-to-open-powershell-in-the-current-explorer-window/#comments Tue, 14 Oct 2008 16:42:35 +0000 Gaurav http://techblogging.wordpress.com/2008/10/14/multiple-ways-to-open-powershell-in-the-current-explorer-window/

These techniques are generic and can be used for other items like command prompt as well. A word of caution, everything given below involves fiddling with the registry so please backup your system’s registry or create a restore point before trying this out.

1) Adding it to Vista Explorer’s toolbar:

image

This is the most convenient one in my opinion, but requires most work to enable it.

Firstly, navigate to the following registry key:

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\explorer\FolderTypes\{7d49d726-3c21-4f05-99aa-fdc2c9474656}

Its should have the following items:

image

Note that it has following two items:

  1. TasksItemsSelected: Enabled when an item is selected in the view.
  2. TasksNoItemsSelected: Enabled when nothing is selected in the view.

Each of these keys have multiple items inside them each pointing to separate shortcut inside the toolbar.

Import the following registry file so that the PowerShell icon appears in the toolbar:

Download From Skydrive

Before importing I will also point out that I am assuming the a batch file exists at the following path:

C:\\Users\\Gaurav\\Programs\\shortcuts\\psh_launch.bat %2

The content of this file should be:

   1: @echo off
   2: pushd %1
   3: %SystemRoot%\system32\WindowsPowerShell\v1.0\powershell.exe
   4: @echo

You can search and replace the location of this file to your custom location inside the just downloaded registry file. Its being used in two places.

Also, remember to take a backup of this registry key or make a restore point before importing this registry key.

That’s it. Now a PowerShell Icon should appear inside the every folder.

 

The PowerShell icon still doesn’t appear?

Actually the registry file only modifies the folders of Document Type and the PowerShell icon will appear only inside folders of Document Type.

An easy way to make all the folders inside a Drive Document Type:

right click on drive/partition –> select Properties –> select Customize tab-> and select following options:

image 

So this way we just forced all the folders in the drive to be of Document Type.

The same procedure can be used to add multiple items on the toolbar. Its a lot of work but the end result is worth it :)

 

2) Using AutoHotKey

AutoHotKey is used for creating keyboard macros.

AutoHotKey will use the same batch file that was created to launch PowerShell in the current directory.

Here is the script:

   1: !p::
   2: IfWinActive ahk_class CabinetWClass
   3: {
   4: WinGetClass explorerClass, A
   5: ControlGetText currentPath, Edit1, ahk_class %explorerClass%
   6: run C:\\Users\\Gaurav\\Programs\\shortcuts\\psh_launch.bat "%currentPath%"
   7: }
   8: return

Just fix path to the batch file, save and reload AutoHotKey.

Now press Alt + P to launch PowerShell inside the current explorer window.

 

3) Right Click Menu

image

To add PowerShell to right click menu navigate to following registry entry:

HKEY_CLASSES_ROOT\Directory\Background\shell

and create a key with name powershell inside shell and a key with name command inside it. Edit the default value of command key to:

C:\Users\Gaurav\Programs\shortcuts\psh_launch.bat “%V”

This is same batch file created earlier and the exact path needs to be fixed.

or just import this file into your registry and fix the path to the batch file:

Download from Skydrive

That’s it, now PowerShell should show up in you right-click menu inside any directory.

]]> http://allyourcodearebelongto.me/blog/2008/10/14/multiple-ways-to-open-powershell-in-the-current-explorer-window/feed/ 2 Context sensitive auto-completion using PowerShell, PowerTab and GIT http://allyourcodearebelongto.me/blog/2008/10/13/context-sensitive-auto-completion-using-powershell-powertab-and-git/ http://allyourcodearebelongto.me/blog/2008/10/13/context-sensitive-auto-completion-using-powershell-powertab-and-git/#comments Mon, 13 Oct 2008 19:33:46 +0000 Gaurav http://techblogging.wordpress.com/2008/10/13/context-sensitive-auto-completion-using-powershell-powertab-and-git/ Powertab is an Awesome PowerShell TabExpansion extension. It extends the default PowerShell autocompletion and shows the results like:

image

The best part is that it is easily customizable.

One way to customize it is by editing it Tab Expansion database file which is located in your powershell profile directory:

TabExpansion.xml

To display some simple git commands in autocompletion all the following entries to it:

   1: <Custom>
   2:   <Filter>git</Filter>
   3:   <Text>status</Text>
   4:   <Type>Custom</Type>
   5: </Custom>
   6: <Custom>
   7:   <Filter>git</Filter>
   8:   <Text>add .</Text>
   9:   <Type>Custom</Type>
  10: </Custom>
  11: <Custom>
  12:   <Filter>git</Filter>
  13:   <Text>pull</Text>
  14:   <Type>Custom</Type>
  15: </Custom>
  16: <Custom>
  17:   <Filter>git</Filter>
  18:   <Text>push</Text>
  19:   <Type>Custom</Type>
  20: </Custom>
  21: <Custom>
  22:   <Filter>git</Filter>
  23:   <Text>commit -m ""</Text>
  24:   <Type>Custom</Type>
  25: </Custom>
  26: <Custom>
  27:   <Filter>git</Filter>
  28:   <Text>commit -a -m ""</Text>
  29:   <Type>Custom</Type>
  30: </Custom>

To invoke this list add the following to TabExpansion.ps1 file which should be in the source directory:

   1: 'git(.*)' {
   2:   if(Test-Path .git) {
   3:     $global:dsTabExpansion.Tables['Custom'].select("filter = 'git' AND type = 'Custom'") | % {$_.text} | Invoke-TabItemSelector  -SelectionHandler $SelectionHandler
   4:   }
   5: }

This should be added in the Main tabcompletion block inside the switch statement. The function is basically matching the current command name with git and then invoking autocompletion.

Now when you type git<tab> then the output should look like below:

image

I am sure this is not the best way to do custom autocompletion but i couldn’t find an easier way.

This is all static stuff. To add dynamic autocompletion (like providing branch names) add following to the TabExpansion.ps1 file:

   1: 'git(.*)' {
   2:   if(Test-Path .git) {
   3:     # git checkout
   4:     if($lastBlock -match 'git checkout'){
   5:       git branch | %{ if($_ -match "^\*?\s*(.*)"){ $matches[1] } } | Invoke-TabItemSelector  -SelectionHandler $SelectionHandler
   6:     }
   7:     else{
   8:       $global:dsTabExpansion.Tables['Custom'].select("filter = 'git' AND type = 'Custom'") | % {$_.text} | Invoke-TabItemSelector  -SelectionHandler $SelectionHandler
   9:     }
  10:   }
  11: }

Now when you type git checkout <tab> then it should show the following in the autocomplete list:

image

This function checks if the user has entered git checkout as the command then it runs the git branch command, parses it and shows the result as autocomplete list.

Here are couple more commands that I wrote while trying this out:

   1: 'git(.*)' {
   2:   if(Test-Path .git) {
   3:     # git checkout
   4:     if($lastBlock -match 'git checkout'){
   5:       git branch | %{ if($_ -match "^\*?\s*(.*)"){ $matches[1] } } | Invoke-TabItemSelector  -SelectionHandler $SelectionHandler
   6:     }
   7:     
   8:     # git add
   9:     Elseif($lastBlock -match 'git add'){
  10:       $s = $False ; git status | %{ if($s -eq $True -and $_ -match "^#\s{1,1}(\S.+)$"){ $matches[1] } ElseIf($_ -match 'Untracked files:'){ $s = $True } } | Invoke-TabItemSelector  -SelectionHandler $SelectionHandler
  11:     }
  12:     
  13:     # git rm
  14:     Elseif($lastBlock -match 'git rm'){
  15:       git status | %{ if($_ -match "^#\s+deleted:\s+(\S.+)$"){ $matches[1] } } | Invoke-TabItemSelector  -SelectionHandler $SelectionHandler
  16:     }
  17:     
  18:     else{
  19:       $global:dsTabExpansion.Tables['Custom'].select("filter = 'git' AND type = 'Custom'") | % {$_.text} | Invoke-TabItemSelector  -SelectionHandler $SelectionHandler
  20:     }
  21:   }
  22: }

These will autocomplete parameters for git add and git rm commands by running the git status command and parsing the output.

For example:

image

 

I think a lot more interesting stuff can be created using PowerTab and GIT.

 

Update 1:

Turns out as pointed by MOW that there is a much easier way to add custom expansions instead of editing the TabExpansion.xml file.

Just run this command inside PowerShell:

Invoke-TabExpansionEditor

It will open up a dialog box like:

image

Here custom items can be added by clicking the Custom link and then the database change can by saved by running the command:

Export-tabExpansionDataBase

]]> http://allyourcodearebelongto.me/blog/2008/10/13/context-sensitive-auto-completion-using-powershell-powertab-and-git/feed/ 5 Displaying GIT Branch on your PowerShell prompt http://allyourcodearebelongto.me/blog/2008/10/12/displaying-git-branch-on-your-powershell-prompt/ http://allyourcodearebelongto.me/blog/2008/10/12/displaying-git-branch-on-your-powershell-prompt/#comments Sun, 12 Oct 2008 07:48:28 +0000 Gaurav http://techblogging.wordpress.com/2008/10/12/displaying-git-branch-on-your-powershell-prompt/ PowerShell is a command line shell like the command prompt (cmd.exe) but with a lot more features. It can be downloaded from here.

One of the features is that you can change the default prompt. Prompt can be changed by overriding the function with name “prompt”.

This function can be declared inside the PowerShell profile file which has the path:

C:\Users\Gaurav\Documents\WindowsPowerShell\Microsoft.PowerShell_profile.ps1

If the file does not exist then you will need to create this file. Declaring the following function inside the profile file will change the prompt.

   1: function prompt {
   2:   return "prompt> "
   3: }

to:

image

Now, to display the git branch as a prompt change the function to the following:

   1: function prompt {
   2:   $host.ui.rawui.WindowTitle = $(get-location)
   3:   
   4:   $prompt_string = ""
   5:   
   6:   if(Test-Path .git) {
   7:     $prompt_string = "GIT"
   8:     git branch | foreach {
   9:       if ($_ -match "^\*(.*)"){
  10:         $prompt_string += $matches[1] + "> "
  11:       }
  12:     }
  13:   }
  14:   else{
  15:     $prompt_string = "PS> "
  16:   }
  17:  
  18:   Write-Host ($prompt_string) -nonewline -foregroundcolor yellow
  19:   return " "
  20: }

and the prompt will look like:

image

The code basically is checking if current directory is a git project. This can be done by checking if there is a directory with name “.git” in the current directory. If yes then it runs the “git branch” command and then passes the result through a regular expression to get the current branch.

Here is my full PowerShell prompt function:

   1: $Global:CurrentUser = [System.Security.Principal.WindowsIdentity]::GetCurrent()
   2: $UserType = "User"
   3: $CurrentUser.Groups | foreach { if($_.value -eq "S-1-5-32-544") {$UserType = "Admin"} }
   4:  
   5: function prompt {
   6:   if($UserType -eq "Admin") {
   7:     $host.ui.rawui.WindowTitle = "" + $(get-location) + " : Admin"
   8:     $host.UI.RawUI.ForegroundColor = "Yellow"
   9:   }
  10:   else {
  11:     $host.ui.rawui.WindowTitle = $(get-location)
  12:   }
  13:   
  14:   Write-Host("")
  15:   
  16:   $status_string = ""
  17:   
  18:   if(Test-Path .git) {
  19:     $status_string = "GIT"
  20:     git branch | foreach {
  21:       if ($_ -match "^\*(.*)"){
  22:         $status_string += $matches[1]
  23:       }
  24:     }
  25:     
  26:     $git_create_count = 0
  27:     $git_update_count = 0
  28:     $git_delete_count = 0
  29:     
  30:     git status | foreach {
  31:       if($_ -match "modified:"){ 
  32:         $git_update_count += 1
  33:       }
  34:       elseif($_ -match "deleted:"){ 
  35:         $git_delete_count += 1
  36:       }
  37:       elseif($_ -match "added:"){ 
  38:         $git_create_count += 1
  39:       }
  40:     }
  41:     $status_string += " u:" + $git_update_count + " d:" + $git_delete_count + ">"
  42:   }
  43:   else{
  44:     $status_string = "PS>"
  45:   }
  46:  
  47:   Write-Host ($status_string) -nonewline -foregroundcolor yellow
  48:   return " "
  49: }

This function also displays updated and deleted files in the current git project.

image

 

Scripts execution in PowerShell is disabled?

If you just created the profile file then PowerShell may display the following error message when loading this profile we just created:

image

or

File C:\Users\Gaurav\Documents\WindowsPowerShell\Microsoft.PowerShell_profile.ps1 cannot be loaded because the execution of scripts is disabled on this system. Please see “get-help about_signing” for more details.
At line:1 char:2
+ .  <<<< ‘C:\Users\Gaurav\Documents\WindowsPowerShell\Microsoft.PowerShell_profile.ps1′

This is because PowerShell does not allow the execution of scripts by default. To fix this open PowerShell in Admin mode and set the execution policy to unrestricted by running the following command:

   1: Set-ExecutionPolicy Unrestricted
]]> http://allyourcodearebelongto.me/blog/2008/10/12/displaying-git-branch-on-your-powershell-prompt/feed/ 6 Creating a chat application using Socket in Silverlight http://allyourcodearebelongto.me/blog/2008/04/17/creating-a-chat-application-using-socket-in-silverlight/ http://allyourcodearebelongto.me/blog/2008/04/17/creating-a-chat-application-using-socket-in-silverlight/#comments Thu, 17 Apr 2008 21:12:52 +0000 Gaurav http://techblogging.wordpress.com/2008/04/17/creating-a-chat-application-using-socket-in-silverlight/ Here I will explain how to communicate with the server using the Socket class. The socket class allows Asynchronous communication between the client and server. This application will show how to share data in real-time in two different browser windows using Silverlight.

This application will contain two parts:

  1. Server application using Ruby.
  2. Client application using C# in Silverlight.
Creating the server application:

Making a multi-threaded server application in ruby is really easy and that is the reason I chose Ruby to create the server in least amount of time possible.

   1: require 'socket'
   2:  
   3: HOST = 'localhost'
   4: PORT = 4505
   5:  
   6: server = TCPServer.new(HOST, PORT)
   7:  
   8: # array to store all the active connections
   9: sessions = []
  10: while (session = server.accept)
  11:   # push the current session(socket) in the array
  12:   sessions << session
  13:   # initialize a new thead for each connection
  14:   Thread.new(session) do |local_session|
  15:     # each time a client sends some data send it to all the connections
  16:     while(true)
  17:       data = local_session.gets
  18:       sessions.each do |s| 
  19:         begin
  20:           s.puts data
  21:         rescue Errno::ECONNRESET
  22:           # an exception is raised, that means the connection to the client is broken
  23:           sessions.delete(s)
  24:         end
  25:       end
  26:     end
  27:   end
  28: end

 

Here is the pastie.

Save this file as chat_server.rb

 

This code is really quite simple and all it is doing is creating a new thread for each client connection and rejecting the invalid ones. Whenever a client sends some data it is propagated to all the clients.

 

Creating the client application:

This is the difficult part among the two. As I mentioned earlier I will be using the Socket class. The important instance methods of this class are:

  1. ConnectAsync
  2. SendAyns
  3. ReceiveAync

As their names suggest all of them are asynchronous. I am not sure of the reason. May be its because the communication does not block the UI thread and making it a bad experience for the client or may be its because of browser’s limitations.

Also in the beta-1 implementation the Silverlight Socket class only allows connection to 4502 – 4532 port range. I guess this limitation will be removed in the later versions.

Open up Visual Studio 2008 (with Silverlight tools installed. You can also download a trial version of Visual Studio 2008 from here). And create a new project:

image

Name the application ChatClient and select rest of the default options.

Run the application. Visual Studio will prompt you to enable debugging, click OK to enable debugging and continue.

Add the System.net assembly to add the Socket class to solution.

Now to make a connection to the server we can use the Socket class:

   1: socket = new Socket(AddressFamily.InterNetwork, SocketType.Stream, ProtocolType.Tcp);
   2: SocketAsyncEventArgs connectEventArgs = new SocketAsyncEventArgs();
   3: connectEventArgs.RemoteEndPoint = IP_END_POINT;
   4: connectEventArgs.Completed += new EventHandler<SocketAsyncEventArgs>(connectEventArgs_Completed);
   5: socket.ConnectAsync(connectEventArgs);

Here IP_END_POINT is an object of DnsEndPoint class and basically points to the server. I will be showing the code for it later for simplicity.

In the above code I created an instances of Socket class and SocketAsyncEventArgs class. The latter does the event handling for the former. So when the connection is complete it will just ca a event handler and then we can start sending/receiving data to server.

Code for sending data to the server:

   1: sendEventArgs = new SocketAsyncEventArgs();
   2: sendEventArgs.RemoteEndPoint = IP_END_POINT;
   3:  
   4: List<ArraySegment<byte>> l = new List<ArraySegment<byte>>();
   5: l.Add(new ArraySegment<byte>(Encoding.UTF8.GetBytes(sendBox.Text + "\n")));
   6:  
   7: sendEventArgs.BufferList = l;
   8: socket.SendAsync(sendEventArgs);

 

Socket uses a List to fill the data buffer. I have appended a “\n” in the data so that it is handled properly at the server.

Code to receive data from the server:

   1: void connectEventArgs_Completed(object sender, SocketAsyncEventArgs e)
   2: {
   3:     e.Completed -= new EventHandler<SocketAsyncEventArgs>(connectEventArgs_Completed);
   4:     e.Completed += new EventHandler<SocketAsyncEventArgs>(receiveEventArgs_Completed);
   5:     transferBuffer = new byte[BUFFER_SIZE];
   6:     e.SetBuffer(transferBuffer, 0, transferBuffer.Length);
   7:     socket.ReceiveAsync(e);
   8: }
   9:  
  10: void receiveEventArgs_Completed(object sender, SocketAsyncEventArgs e)
  11: {
  12:     String a = Encoding.UTF8.GetString(transferBuffer, e.Offset, e.BytesTransferred);
  13:     SetText(a);
  14:     socket.ReceiveAsync(e);
  15: }

 

When the connection to the server is complete the client begins to wait for the server to send data. This is done by removing the old event handler an creating a new one to receive the data. Also we need to initialise buffer so that incoming data can be received in it.

That is the basic structure you need to make for Asynchronous data transfer using Socket class.

The full code:

Page.xaml.cs

   1: using System;
   2: using System.Collections.Generic;
   3: using System.Linq;
   4: using System.Windows;
   5: using System.Windows.Controls;
   6: using System.Windows.Documents;
   7: using System.Windows.Input;
   8: using System.Windows.Media;
   9: using System.Windows.Media.Animation;
  10: using System.Windows.Shapes;
  11: using System.Net.Sockets;
  12: using System.Net;
  13: using System.Text;
  14:  
  15: namespace DeleetThis
  16: {
  17:     public partial class Page : UserControl
  18:     {
  19:         private Socket socket;
  20:         private byte[] transferBuffer;
  21:         private SocketAsyncEventArgs sendEventArgs;
  22:  
  23:         // contants
  24:         private static int BUFFER_SIZE = 500;
  25:         private static int PORT = 4505;
  26:         private static DnsEndPoint IP_END_POINT = new DnsEndPoint(Application.Current.Host.Source.DnsSafeHost, PORT);
  27:  
  28:         public delegate void SetTextCallback(String _text);
  29:  
  30:         // initialize stuff
  31:         public Page()
  32:         {
  33:             InitializeComponent();
  34:             socket = new Socket(AddressFamily.InterNetwork, SocketType.Stream, ProtocolType.Tcp);
  35:             SocketAsyncEventArgs connectEventArgs = new SocketAsyncEventArgs();
  36:             connectEventArgs.RemoteEndPoint = IP_END_POINT;
  37:             connectEventArgs.Completed += new EventHandler<SocketAsyncEventArgs>(connectEventArgs_Completed);
  38:             socket.ConnectAsync(connectEventArgs);
  39:         }
  40:  
  41:         // connection complete
  42:         void connectEventArgs_Completed(object sender, SocketAsyncEventArgs e)
  43:         {
  44:             sendEventArgs = new SocketAsyncEventArgs();
  45:             sendEventArgs.RemoteEndPoint = IP_END_POINT;
  46:  
  47:             e.Completed -= new EventHandler<SocketAsyncEventArgs>(connectEventArgs_Completed);
  48:             e.Completed += new EventHandler<SocketAsyncEventArgs>(receiveEventArgs_Completed);
  49:             transferBuffer = new byte[BUFFER_SIZE];
  50:             e.SetBuffer(transferBuffer, 0, transferBuffer.Length);
  51:             socket.ReceiveAsync(e);
  52:         }
  53:  
  54:         // incoming data
  55:         void receiveEventArgs_Completed(object sender, SocketAsyncEventArgs e)
  56:         {
  57:             String a = Encoding.UTF8.GetString(transferBuffer, e.Offset, e.BytesTransferred);
  58:             SetText(a);
  59:             socket.ReceiveAsync(e);
  60:         }
  61:  
  62:         // this process is asynchronous so we need to invoke the ui thread to set the received text
  63:         private void SetText(String text)
  64:         {
  65:             if (this.Dispatcher.CheckAccess())
  66:                 receiveBox.Text += text;
  67:             else
  68:                 this.Dispatcher.BeginInvoke(new SetTextCallback(SetText), text);
  69:         }
  70:  
  71:         // data sending
  72:         private void sendButton_Click(object sender, RoutedEventArgs e)
  73:         {
  74:             List<ArraySegment<byte>> l = new List<ArraySegment<byte>>();
  75:             l.Add(new ArraySegment<byte>(Encoding.UTF8.GetBytes(sendBox.Text + "\n")));
  76:  
  77:             sendEventArgs.BufferList = l;
  78:             socket.SendAsync(sendEventArgs);
  79:         }
  80:     }
  81: }

Here is the pastie.

Page.xaml:

   1: <UserControl x:Class="ChatClient.Page"
   2:     xmlns="http://schemas.microsoft.com/client/2007" 
   3:     xmlns:x="http://schemas.microsoft.com/winfx/2006/xaml" 
   4:     Width="400" Height="230">
   5:     <Grid x:Name="LayoutRoot" Background="White">
   6:         <StackPanel>
   7:             <ScrollViewer Height="200" >
   8:                 <TextBlock x:Name="receiveBox"/>
   9:             </ScrollViewer>
  10:             <StackPanel Orientation="Horizontal">
  11:                 <TextBox x:Name="sendBox" Height="30" Width="365"/>
  12:                 <Button x:Name="sendButton" Content="Send" Click="sendButton_Click"/>
  13:             </StackPanel>
  14:  
  15:         </StackPanel>
  16:     </Grid>
  17: </UserControl>

Here is the pastie.

Running the application:

Now run this application in two different browser windows and whenever you type in one window the content of both windows should be updated :)

image 

Download the source code:

]]> http://allyourcodearebelongto.me/blog/2008/04/17/creating-a-chat-application-using-socket-in-silverlight/feed/ 8 Compiling and Installing GIT on windows under cygwin http://allyourcodearebelongto.me/blog/2008/04/11/compiling-and-installing-git-on-windows-under-cygwin/ http://allyourcodearebelongto.me/blog/2008/04/11/compiling-and-installing-git-on-windows-under-cygwin/#comments Fri, 11 Apr 2008 19:56:10 +0000 Gaurav http://techblogging.wordpress.com/2008/04/11/installing-git-on-windows/ Using GIT was one of the requirements of one of the project that I am doing. Here is how I installed GIT on Windows.

So What is GIT?

The supplied documentation of GIT says that:

Git is a fast, scalable, distributed revision control system with an unusually rich command set that provides both high-level operations and full access to internals.

Here is the link for the wikipedia entry.

Where can I download GIT?

Here is the link.

Before following the given steps you will need to install cygwin on your system, and add its bin directory to the system path (so that all the cygwin commands are available from inside the command prompt).

So after installing cygwin run setup.exe inside the cygwin installation folder and make sure the following modules are installed.

  1. make
  2. gcc c anc c++ compilers
  3. openssl-devel
  4. curl-devel
  5. Any other dependencies that are missing..

image

All these packages are located inside the Devel branch of the package tree. Required dependencies will be selected automatically.

Click next to install selected packages.

Compiling GIT form source

Untar the GIT source you downloaded using the command “tar xvzf git-1.5.5.tar.gz” .

Navigate to the untarred directory using a command prompt. Now run the “make” command inside the GIT source directory.

If it runs successfully then run “make install”.

Before running the command make sure you have Perl (language) installed. You can check it by running the command:

perl -MCPAN -e 'shell'

and then running the following inside the cpan shell:

install blib

it should show no error. I am not sure if this is the best way to do this as I am not much of a Perl guy.

So as I mentioned before you need to run the following two command to make and install Perl:

   1: make
   2: make install

If these two steps complete successfully then your GIT is ready to go :). You can test this by running the “git” command. This command is located inside the “C:/cygwin/home/gaurav/bin”. You might to add this folder to the system path too.

Pre-configured version

Here is a cygwin version pre-configured to compile GIT. It is installed on my system and works fine. Cygwin 1.5.25 and GIT 1.5.5 are preinstalled in this archive. Just decompress this archive under folder “C:\cygwin” and add “C:\cygwin\bin” and “C:/cygwin/home/gaurav/bin” to your system path and GIT should work form any command window.

]]> http://allyourcodearebelongto.me/blog/2008/04/11/compiling-and-installing-git-on-windows-under-cygwin/feed/ 3 New and Improved Sudo for Vista (now remembers credentials) http://allyourcodearebelongto.me/blog/2008/04/02/new-and-improved-sudo-for-vista-now-remembers-credentials/ http://allyourcodearebelongto.me/blog/2008/04/02/new-and-improved-sudo-for-vista-now-remembers-credentials/#comments Wed, 02 Apr 2008 11:02:14 +0000 Gaurav http://techblogging.wordpress.com/2008/04/02/new-and-improved-sudo-for-vista-now-remembers-credentials/ In one of my earlier blog post I shared source code for a simple utility that I had made. It could be used to launch elevated processes from the command line.

So opening a Elevated command prompt was as simple as writing

sudo cmd

Actually the code for this is very simple as it just executes a well documented system function ShellExecute.

I have made some changes to the script and now it remembers the credentials. So once you execute any command, Vista will ask you confirmation only once and any subsequent call won’t ask for the confirmation with the UAC dialog box.

The concept behind this is very simple too:

A process launched from any elevated process will be (by default) automatically elevated at administrative assess. So what this script (client) does is launch a elevated Ruby server first time you run it, so it will Vista will prompt you for confirmation. For any subsequent call it will just see that the server is already running and it will just use that server for elevating launched processes.

 

How it works:

When you write “sudo cmd” in the command prompt (or run it any where else):

  • Check if the “sudo server” process is running. If not then start it. (This will result in a UAC prompt).
  • Connect to the server using TCP and ask it to run the command “sudo cmd”.
  • The server will launch the elevated process.

 

The code for Client and the Server is in Ruby and thus can easily be improved upon. One can easily add code to kill the server after a specific time interval or after running a fixed number of commands. Actually I am not sure if this is the correct way to handle this, so any comments will be welcome.

 Download Here

Just un-rar this file anywhere and add it to system path. To run it just write sudo <command name> in “command prompt” of in the “Run dialog box”. Example to run notepad: “sudo notepad”

This code was written with ruby 1.8.6.

]]> http://allyourcodearebelongto.me/blog/2008/04/02/new-and-improved-sudo-for-vista-now-remembers-credentials/feed/ 0 Creating Smart Playlists in Windows Media Player http://allyourcodearebelongto.me/blog/2008/03/31/creating-smart-playlists-in-windows-media-player/ http://allyourcodearebelongto.me/blog/2008/03/31/creating-smart-playlists-in-windows-media-player/#comments Mon, 31 Mar 2008 12:47:31 +0000 Gaurav http://techblogging.wordpress.com/2008/03/31/creating-smart-playlists-in-windows-media-player/ A “Smart  Playlist” is a type of playlist that changes automatically according to criteria you have specified. In Windows Media Player it is called an “Auto Playlist”.

I know that this feature exists in almost every media player but in Windows Media Player this feature is not clearly visible and many people don’t even know anything like this exists.

So here is how you can create a Smart Playlist in Windows Media Player:

Right click on the Playlist button on the left side of the library in window media player and select Create Auto Playlist.

screen1

Now a dialog box will pop up where you can specify parameters according to which the playlist will be generated:

screen3

So there you have it. This playlist will be automatically updated when you add a new song or rate a new song.

]]> http://allyourcodearebelongto.me/blog/2008/03/31/creating-smart-playlists-in-windows-media-player/feed/ 0 Using Flex with Ruby on Rails http://allyourcodearebelongto.me/blog/2008/03/26/using-flex-with-rubyonrails/ http://allyourcodearebelongto.me/blog/2008/03/26/using-flex-with-rubyonrails/#comments Wed, 26 Mar 2008 19:08:39 +0000 Gaurav http://techblogging.wordpress.com/2008/03/26/using-flex-with-rubyonrails/ This is the second blog post in a series of three posts comparing Flex with Silverlight. In my previous post I demonstrated how to integrate a scaffolded Rails application with Silverlight. Here I will be showing the same with Flex. The steps needed to create the Rails application are also mentioned here.

 

Creating a new Flex project:

Open Flex builder and create a new Flex project.

image

Leave all the options to default and click Finish. A blank project is now created for you.

Add a List control to the default mxml file. It should look like:

<?xml version="1.0" encoding="utf-8"?>
<mx:Application xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical">
    <mx:List id="userList" width="200" height="200"/>
</mx:Application>

 

Now create a script tag inside the source and add a Bindable element to it:

[Bindable]
private var users:XMLListCollection = new XMLListCollection();

I am using an XMLListCollection because I can easily add/remove elements from it.

 

Now send request to the server and bind the result to the list box:

First create a new HTTPService element:

<mx:HTTPService id="userIndexService" resultFormat="e4x" result="userIndexServiceResult(event)" url="http://localhost:3000/users.xml"/>

 

and handle its result:

private function userIndexServiceResult(event:ResultEvent):void{
    for each(var xml:XML in event.result.user as XMLList){
        users.addItem(xml);
    }
}

 

To invoke this service we need to call its “send()” method with appropriate paramater, which in this case are none.

So the code now looks like:

<?xml version="1.0" encoding="utf-8"?>
<mx:Application xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical" initialize="sendServiceRequest(event)">
    <mx:Script>
        <![CDATA[
            import mx.rpc.events.ResultEvent;
            import mx.collections.XMLListCollection;
            [Bindable]
            private var users:XMLListCollection = new XMLListCollection();

            private function sendServiceRequest(event:Event):void{
                userIndexService.send();
            }
            private function userIndexServiceResult(event:ResultEvent):void{
                for each(var xml:XML in event.result.user as XMLList){
                    users.addItem(xml);
                }
            }
        ]]>
    </mx:Script>
    <mx:HTTPService id="userIndexService" resultFormat="e4x" result="userIndexServiceResult(event)" url="http://localhost:3000/users.xml"/>
    <mx:List dataProvider="{users}" id="userList" width="200" height="200" labelField="name"/>
</mx:Application>

 

And the output looks like:

image

Notice that we are sending the index request in the initialize method of the application.

Now to add controls which show the currently selected element. Its really easy. Just add the folowwing to the code and DataBinding will do the rest:

<mx:HBox>
    <mx:Label text="Name" width="50"/>
    <mx:TextInput text="{userList.selectedItem.name}" width="150" id="userName"/>
</mx:HBox>
<mx:HBox>
    <mx:Label text="Address" width="50"/>
    <mx:TextInput text="{userList.selectedItem.address}" width="150" id="userAddress"/>
</mx:HBox>
 

Now it loooks like:

image

 
Implementing the full CRUD:

Now we need to add the other three HTTPServices.

<mx:HTTPService id="userIndexService" resultFormat="e4x" result="userIndexServiceResult(event)" url="http://localhost:3000/users.xml"/>
<mx:HTTPService id="userCreateService" resultFormat="e4x" result="userCreateServiceResult(event)" method="POST" url="http://localhost:3000/users.xml"/>
<mx:HTTPService id="userUpdateService" resultFormat="e4x" method="POST"/>
<mx:HTTPService id="userDeleteService" resultFormat="e4x" method="POST"/>

 

And their corresponding callbacks.

private function userCreateServiceResult(event:ResultEvent):void{
    users.addItem(event.result as XML);
}

 

Some controls to invoke these services:

<mx:HBox>
    <mx:Button label="Create" id="createUser" click="sendServiceRequest(event)"/>
    <mx:Button label="Update" id="updateUser" click="sendServiceRequest(event)"/>
    <mx:Button label="Delete" id="deleteUser" click="sendServiceRequest(event)"/>
</mx:HBox>

 

And their event handlers (just one in this case):

private function sendServiceRequest(event:Event):void{
    if(event.target.id == 'createUser'){
        userCreateService.send({'user[name]': userName.text, 'user[address]': userAddress.text});
    }else if(event.target.id == 'updateUser'){
        userUpdateService.url = 'http://localhost:3000/users/'+ userList.selectedItem.id +'.xml';
        userUpdateService.send({'user[name]': userName.text, 'user[address]': userAddress.text, _method: 'put'});
        users[userList.selectedIndex].name = userName.text;
        users[userList.selectedIndex].address = userAddress.text;
    }else if(event.target.id == 'deleteUser'){
        userDeleteService.url = 'http://localhost:3000/users/'+ userList.selectedItem.id +'.xml';
        userDeleteService.send({_method: 'delete'});
        users.removeItemAt(userList.selectedIndex);
    }else{
        userIndexService.send();
    }
}

 

Little bit of  knowledge of how REST is implemented in Rails is required to understand the parameters to the     “send()” method of HTTPService class is required, which is out of scope of this blog post.

The TestFrontEnd.mxml should now have the folowwing code:

<?xml version="1.0" encoding="utf-8"?>
<mx:Application xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical" initialize="sendServiceRequest(event)">
    <mx:Script>
        <![CDATA[
            import mx.collections.XMLListCollection;
            import mx.controls.Alert;
            import mx.rpc.events.ResultEvent;
            import mx.collections.ArrayCollection;
            [Bindable]
            private var users:XMLListCollection = new XMLListCollection();

            private function sendServiceRequest(event:Event):void{
                if(event.target.id == 'createUser'){
                    userCreateService.send({'user[name]': userName.text, 'user[address]': userAddress.text});
                }else if(event.target.id == 'updateUser'){
                    userUpdateService.url = 'http://localhost:3000/users/'+ userList.selectedItem.id +'.xml';
                    userUpdateService.send({'user[name]': userName.text, 'user[address]': userAddress.text, _method: 'put'});
                    users[userList.selectedIndex].name = userName.text;
                    users[userList.selectedIndex].address = userAddress.text;
                }else if(event.target.id == 'deleteUser'){
                    userDeleteService.url = 'http://localhost:3000/users/'+ userList.selectedItem.id +'.xml';
                    userDeleteService.send({_method: 'delete'});
                    users.removeItemAt(userList.selectedIndex);
                }else{
                    userIndexService.send();
                }
            }

            private function userIndexServiceResult(event:ResultEvent):void{
                for each(var xml:XML in event.result.user as XMLList){
                    users.addItem(xml);
                }
            }
            private function userCreateServiceResult(event:ResultEvent):void{
                users.addItem(event.result as XML);
            }
        ]]>
    </mx:Script>

    <mx:HTTPService id="userIndexService" resultFormat="e4x" result="userIndexServiceResult(event)" url="http://localhost:3000/users.xml"/>
    <mx:HTTPService id="userCreateService" resultFormat="e4x" result="userCreateServiceResult(event)" method="POST" url="http://localhost:3000/users.xml"/>
    <mx:HTTPService id="userUpdateService" resultFormat="e4x" method="POST"/>
    <mx:HTTPService id="userDeleteService" resultFormat="e4x" method="POST"/>

    <mx:List dataProvider="{users}" id="userList" width="200" height="200" labelField="name"/>

    <mx:HBox>
        <mx:Label text="Name" width="50"/>
        <mx:TextInput text="{userList.selectedItem.name}" width="150" id="userName"/>
    </mx:HBox>
    <mx:HBox>
        <mx:Label text="Address" width="50"/>
        <mx:TextInput text="{userList.selectedItem.address}" width="150" id="userAddress"/>
    </mx:HBox>
    <mx:HBox>
        <mx:Button label="Create" id="createUser" click="sendServiceRequest(event)"/>
        <mx:Button label="Update" id="updateUser" click="sendServiceRequest(event)"/>
        <mx:Button label="Delete" id="deleteUser" click="sendServiceRequest(event)"/>
    </mx:HBox>

</mx:Application>

 

This is the end result:

image

As I mentioned in my last blog post it is not the best way to make this application as there is no error handling. But I think this is the simplest way to make this application.

You also might have noticed Flex requires a lot less lines of code as compared to Silverlight. But that is also dependent on my coding approach.

Full code for this application can be downloaded from:

]]> http://allyourcodearebelongto.me/blog/2008/03/26/using-flex-with-rubyonrails/feed/ 3