Monday, 7 December 2009

Building an ASP.NET MVC E-Commerce app with FLUX.net as the management tool

A friend of mine was recently contemplating using FLUX to build out an e-commerce application for a store. Whilst this is all well and good, and certainly very plausible, I started to think about how e-commerce sites are generally structured – they aren’t so much about open ended content where you add pages and articles together to form a site, they are far more rigid – more like structured catalogues with prescribed functionality.

As I’m also a fan of the MVC framework (I use it lots in my day job) I’ve been looking for an excuse to build something that uses MVC to render content fed from Flux.NET and I figure this is a good enough opportunity to explore this through a series of blog posts.

The site concept

We’ll manage products and categories in one website, which is standard ASP.NET and using the Flux content management system, then render these to our MVC store front application that has a prescribed structure by querying the content database for data in the appropriate places. Whereas we might normally use flux to dictate the form and structure, in this instance we’re going to use it purely as a management tool for data in our application.

Getting started

I began by downloading the flux starter package and creating a new empty solution. I added an ASP.NET website to the solution and then extracted the starter kit to it, trimming it right down to almost nothing (see image below) and reconfiguring /settings/flux.config to use a new database I created – FluxMVCTest. I ran the flux database script against this db to create my tables etc and then cleared down all the content through the admin tool.

image

With this running and responding on the development web server, I was reasonably happy to proceed and get the MVC site up in the solution too. I added this using the usual new project wizard and then stripped it back to it’s bare essentials also, which at this point gives me two sites in my solution as below;

Follow along source on codeplex

As usual, I’ve started a new project in codeplex for this, so hop along to http://fluxmvcstore.codeplex.com to get the source code so far.

Sunday, 15 November 2009

Overcoming cross domain issues between frames

In an application that I am building at the moment, I had need to “mash-up” content from different sources into a SharePoint site. This SharePoint site would;

  1. Use an iframe to pull in a list of content from another server
  2. Launch a modal dialog (AJAX style) overlaid on top of the SharePoint portal with an iframed form to action against the data
  3. Display validation errors from that form as another modal dialog (owned by the parent SharePoint page so as not to be size limited to the owning dialog).

A picture says a thousand words;

 image

In this case, we’d want the SharePoint host page to offer two methods

  • launchTool(url) – which will launch the given URL in an iframe within a modal jquery overlay (using boxy in my instance).
  • displayValidation(message) – which will be called by the form iframe to display the validation messages if the form can’t be submitted

And this is where we have a problem. If everything was on the same server, it would all be quite happy with this scenario and all the components would play together nicely but in this instance the content is spread across different servers. In this case, we’re not allowed to build this level of interaction using JavaScript as the content doesn’t fulfil the “same origin policy” – which means content from different domains can’t interrogate each others DOM nor invoke methods etc against each other.

image

Working around this isn’t just as simple as putting the sites into a different zone in the browser etc, we have to do something a little more innovative. If you are in total control of your infrastructure you can set all of the content servers to have a fully qualified domain name – eg: sp.mysite.com and forms.mysite.com and then use the document.domain property in JavaScript to set the domain to the higher level domain ie: mysite.com (document.domain = “mysite.com”) – but this does cause problems in my scenario where the SSRS report viewer web part in sharepoint doesn’t work (it does some funky stuff under the covers with iframes which fails miserably when we set document.domain in the host pages).

So, what can we do?

Well, the content from formserver can’t call anything within the sharepoint page and vice versa, but the formserver content could somehow create an iframe of it’s own with a special url on the sharepoint site that then calls the javascript we’re looking for in the sharepoint page. Clear as mud? No? here’s another diagram.

image

So, the list iframe doesn’t go back up the window stack to ask it’s parent to do something (which it has no rights to), it instead goes down – creating it’s own iframe and navigating it to a known page on the host and passes information to the page using a # bookmark. This known page then uses javascript on it’s page load event to determine what to do with the data passed in. If this indicates it should launch a modal dialog for a given URL, it passes the request to the overall parent, which, because it’s on the same domain, can be accessed.

It’s important to note the # part of the URL – by doing this, the cross domain receiver page can be fully cached in the browser, meaning there should be no lag between making the request down the stack to the parent actually processing that request.

I’ve wrapped the above into an API and built a sample that does the above, but also supports bi-directional message passing – so our host page has a list from another domain, it asks the host to display a form from yet another domain and this form requires confirmation, so it asks the host to display a message with options – the result of this message is then passed back down to the form.

Get the source code here: http://xdsframes.codeplex.com

(This isn’t a finished API in any way, just a starting off point for solving your own cross domain issues.)

Saturday, 3 October 2009

Flux.NET quick start package released

As promised, I've just wrapped up the content of my previous post, with some extras to produce a FLUX.NET quick start package. You can get it now on codeplex : http://flux.codeplex.com/Release/ProjectReleases.aspx?ReleaseId=26826#DownloadId=86138

To use it:

  1. Unzip the distribution into a folder on disk
  2. Create an application in IIS that maps to this folder (To work out of the box, the application should be hosted on localhost. If it isn’t, you will need to update the domain mappings in the website mapping tool in flux.)
  3. Create an SQL database called FluxStarterKit (You can actually call it whatever you like but you must change the connection strings in /settings/flux.config)
  4. Execute /documentation/flux.startup.sql against the SQL database

That’s it.

To view the website:

  1. Navigate to the IIS website you created above

To open the admin tool:

  1. Navigate to the IIS website you created above + /admin
  2. Login as admin@fluxcms.co.uk with a password of password

10 Step guide to getting a basic FLUX.NET CMS site up.

Today I decided to start a new site for a friend of mine, and being the owner of the FLUX.NET open source CMS, I obviously decided to use that. I figured I'd use this opportunity though to document a step by step guide to how to get flux.net setup from download to a single basic page being up and running, so here goes.

Step 1 - get the distro!

Go to http://flux.codeplex.com and get the latest release package (v2.0 release here)

Step 2 - Copy the files into your site

1. From the distribution, copy the /Src/flux_client folder into the folder that will be the root of your site

image

2. Next go into /Src/Deepcode.Flux.Admin/and copy the entire contents into a /admin directory in your site

image

3. Move the bin directory from admin, 1 level down, outside of the admin directory.

4. Finally, copy one of the web.config files from the distribution into your site. These are in the root of the distribution and are named FormsWeb.Config and WindowsWeb.Config. In this instance, I want to use forms authentication so I'll take FormsWeb.Config, copy it into my site and rename it to web.config.

5. At this point, you should have a site directory structure as follows;

image

Step 3 - configuring IIS and SQL for local development

1. Open IIS manager, right click on the default website and select to create a new application (I'm using IIS7 in this exercise):

image

2. Give the virtual directory an alias to use, set it's application pool to classic .NET app pool and point the physical path to the directory you've created in step 2.4 above.

image

3. Open SQL Server and create a database for your content. (You can get the installer tool to create it for you, but I prefer to create it by hand)...

Step 4 - install the application

1. Navigate to http://localhost/youralias/admin/setup/install to start the installation process and click next to open the setup database form:

image

2. Enter the parameters necessary to connect to the database you created earlier. For my local development environment, I setup my server as ".", fluxtest as the database name, ticked the database has already been created checkbox,  and specified windows authentication as the SQL authentication mechanism. Click next to go to authentication setup.

 image

3. As we're aiming for forms authentication here, I selected "Forms authentication"  and setup the details of my super user to correspond to myself. Click next to go to general setup;

image

4. Enter a name for the installation (which will appear in the administration tools), along with a default mail server and an address which flux can send email as. Click next to review your installation settings.

image

5. Click finish to begin installation. At this point the system will install the database scripts and generate configuration files. As such, the app pool users of your web application must have write access to the location of your site and the SQL user/windows user also must have permission to create tables and stored procedures etc in your database. Once complete, you should see a screen like that below and a flux.config file will appear in a new folder called settings in your application.

image

6. If you "click here to continue" flux is now completely setup and you should be able to access the admin tool and it's modules etc as below;

image

However, the application itself doesn't know anything about content at this point.... we need to configure it to make it do something next.

Step 5 - opening your site in visual studio.

In order to do anything useful we will need to open visual studio 2008 and open the web site. Open it directly from the local IIS installation. If it offers to update it to .NET 3.5, say yes so that we can utilise LINQ and other cool features in our rendering templates. With this done, you should now have an environment ready to roll;

image

Step 6 - the worlds most basic content type

Back in flux, If you open "content management" and select to add a new item of content, you will notice that it doesn't give you any options of what to add. That is because flux is completely agnostic of what data makes up content. It understands content concepts and how content should relate etc, but not necessarily what fields make up what types of content, for this we must give it some more information.

For the purposes of this demonstration we're going to build a CMS that can host multiple web sites, but only has basic HTML pages available to it. This keeps the scope simple so that I can actually complete this blog post in one sitting!!!!

1. In VS.NET Create a new folder under /settings called "types" and a folder within this called "Site".

2. Add a new XML file to this folder and name it "Site.type.config". Then paste the content below into it;

<NodeType ID="Site" Name="Website" LimitVersions="1">

    <AddForm>~/settings/types/Site/form.aspx</AddForm>
    <EditForm>~/settings/types/Site/form.aspx</EditForm>

    <Relationships AddRoot="true">
        <Add>Page</Add>
    </Relationships>

    <StaticData>
        <Template>~/settings/types/Site/render.aspx</Template>
    </StaticData>

</NodeType>

This defines a content type that we can add to our system. LimitVersions="1" tells flux only to keep a maximum of 1 version history of anything that is of this type. The Add/Edit form specifies the ASPX templates that will be used to add or edit content items of this type, which we'll build shortly. The relationships section determines how this content type relates to other items. In this case, we tell flux that sites can be added to the root node and within them, a content type called page can be added. Finally the static data template item specifies the rendering template when content of this type is rendered in html.

3. Now we need to create the ASPX file that will let us add and edit content in flux. Create a new web form called form.aspx in the /settings/types/Site directory and set it's content as follows;

<%@ Page Language="C#" Inherits="Deepcode.Flux.Core.UI.CMS.CMSContentForm" ValidateRequest="false"%>
<%@ Register TagPrefix="flux" Namespace="Deepcode.Flux.Core.UI.Controls" Assembly="Deepcode.Flux.Core"%>

<script runat="server">
    // Type ID being managed by this form
    protected override string FormTypeCode { get { return "Site"; } }

    // Do form setup
    protected override void SetupForm(int ContentID, int ParentID)
    {
        if (!Page.IsPostBack)
        {
            fChannelPath.NodeIDParent = ParentID;
            fChannelPath.NodeIDPathed = ContentID;
        }
    }

    // Save fields to content object
    protected override void SaveContent(ref Deepcode.Flux.Core.Systems.CMS.ContentObject save)
    {
        save.NodeName = this.fChannelPath.Text;
        save.NodeTitle = this.fChannelName.Text;
    }

    // Load fields from content object
    protected override void LoadContent(Deepcode.Flux.Core.Systems.CMS.ContentObject load)
    {
        this.fChannelName.Text = load.NodeTitle;
        this.fChannelPath.Text = load.NodeName;
    }
</script>

<html>
<head id="Head1" runat="server">
    <link href="../../../admin/Asset/Style/GeneralStyle.css" rel="Stylesheet" type="text/css" />
</head>
<body class="nopadshaded">
<form id="form1" runat="server">
<flux:HostTable ID="HostTable1" runat="server">

<%-- Summary --%>
<flux:ValidationSummarySection ID="ValidationSummarySection1" runat="server" HeaderText="Please correct the following errors"/>

<%-- Form area --%>
    <flux:Section ID="Section1" runat="server" Title="Add/Edit Website channel">
    <flux:ShadePadBox ID="ShadePadBox1" runat="server">

        <table cellspacing="0" cellpadding="3" border="0">
        <tr><td>Channel Name:</td>
            <td><asp:TextBox runat="server" ID="fChannelName" Width="300px" MaxLength="500"/></td>
            </tr>
        <tr><td>Channel Path:</td>
            <td><flux:NodePath runat="server" ID="fChannelPath" Width="200px" MaxLength="100"/></td>
            </tr>
        </table>

    </flux:ShadePadBox>
    </flux:Section>

<%-- Buttons --%>
<flux:Section ID="Section2" runat="server">
<flux:PadBox ID="PadBox1" CssClass="Pad5Button" runat="server">
    <asp:Button runat="server" ID="btnSave" Text="Save" OnClick="btnSave_Click" CssClass="button"/>&nbsp;&nbsp;&nbsp;
    <asp:Button runat="server" ID="btnCancel" Text="Cancel" OnClick="btnCancel_Click" CausesValidation="False"/>
</flux:PadBox>
</flux:Section>

<%-- Validators --%>
<asp:RequiredFieldValidator ID="RequiredFieldValidator1" Runat="server" ControlToValidate="fChannelName" 
    ErrorMessage="You must specify the title of this channel" Display="None"/>

<asp:RequiredFieldValidator ID="RequiredFieldValidator2" runat="server" ControlToValidate="fChannelPath"
    ErrorMessage="You must specify the path of this channel" Display="None"/>

</flux:HostTable>

</form>
</body>
</html>

There are a large number of flux specific controls in there, but it should be reasonably self explanatory. The base class CMSContentForm within flux provides everything necessary to wire our form up. It invokes Save/Load content methods and utilises the form type code to determine what it's editing. All we do is override these methods and add the fields in/out of the persistence mechanism.

4. We need to go ahead and create a Page content type now also, so we can publish something so add another folder within types called Page along with the configuration file "Page.type.config" and it's editor form "form.aspx", using the following content:

/Settings/types/page/Page.type.config

<NodeType ID="Page" Name="Web Page" LimitVersions="1">

    <AddForm>~/settings/types/Page/form.aspx</AddForm>
    <EditForm>~/settings/types/Page/form.aspx</EditForm>

    <Relationships AddRoot="false">
        <Add>Page</Add>
    </Relationships>

    <StaticData>
        <Template>~/settings/types/Page/render.aspx</Template>
    </StaticData>

</NodeType>

/Settings/types/page/Form.aspx

<%@ Page Language="C#" Inherits="Deepcode.Flux.Core.UI.CMS.CMSContentForm" ValidateRequest="false"%>
<%@ Register TagPrefix="flux" Namespace="Deepcode.Flux.Core.UI.Controls" Assembly="Deepcode.Flux.Core"%>

<script runat="server">
    // Type ID being managed by this form
    protected override string FormTypeCode { get { return "Page"; } }

    // Do form setup
    protected override void SetupForm(int ContentID, int ParentID)
    {
        if (!Page.IsPostBack)
        {
            fPath.NodeIDParent = ParentID;
            fPath.NodeIDPathed = ContentID;
        }
    }

    // Save fields to content object
    protected override void SaveContent(ref Deepcode.Flux.Core.Systems.CMS.ContentObject save)
    {
        save.NodeName = this.fPath.Text;
        save.NodeTitle = this.fPageTitle.Text;
        save.Fields["MenuTitle"] = this.fMenuTitle.Text;
        save.Fields["BodyContent"] = this.fHtml.Text;
    }

    // Load fields from content object
    protected override void LoadContent(Deepcode.Flux.Core.Systems.CMS.ContentObject load)
    {
        this.fPath.Text = load.NodeName;
        this.fPageTitle.Text = load.NodeTitle;
        this.fMenuTitle.Text = load.Fields["MenuTitle"];
        this.fHtml.Text = load.Fields["BodyContent"];
    }
</script>

<html>
<head id="Head1" runat="server">
    <link href="../../../admin/Asset/Style/GeneralStyle.css" rel="Stylesheet" type="text/css" />
</head>
<body class="nopadshaded">
<form id="form1" runat="server">
<flux:HostTable ID="HostTable1" runat="server">

<%-- Summary --%>
<flux:ValidationSummarySection ID="ValidationSummarySection1" runat="server" HeaderText="Please correct the following errors"/>

<%-- Form area --%>
    <flux:Section ID="Section1" runat="server" Title="Add/Edit Page">
    <flux:ShadePadBox ID="ShadePadBox1" runat="server">

        <table cellspacing="0" cellpadding="3" border="0">
        <tr><td>Path:</td>
            <td><flux:NodePath runat="server" ID="fPath" Width="200px" MaxLength="100"/></td>
            </tr>

        <tr><td>Page Title:</td>
            <td><asp:TextBox runat="server" ID="fPageTitle" Width="300px" MaxLength="500"/></td>
            </tr>
        
        <tr><td>Title for menus:</td>
            <td><asp:TextBox runat="server" ID="fMenuTitle" Width="300px" MaxLength="50"/></td>
            </tr>
        </table>

    </flux:ShadePadBox>
    </flux:Section>
    
    <flux:Section ID="Section2" runat="server" Title="Page Content">
        <asp:TextBox runat="server" ID="fHtml" Width="100%" Height="150px" />
    </flux:Section>

<%-- Buttons --%>
<flux:Section ID="Section3" runat="server">
<flux:PadBox ID="PadBox1" CssClass="Pad5Button" runat="server">
    <asp:Button runat="server" ID="btnSave" Text="Save" OnClick="btnSave_Click" CssClass="button"/>&nbsp;&nbsp;&nbsp;
    <asp:Button runat="server" ID="btnCancel" Text="Cancel" OnClick="btnCancel_Click" CausesValidation="False"/>
</flux:PadBox>
</flux:Section>

<%-- Validators --%>
<asp:RequiredFieldValidator ID="RequiredFieldValidator1" runat="server" ControlToValidate="fPath"
    ErrorMessage="You must specify the path for this page" Display="None"/>

<asp:RequiredFieldValidator ID="RequiredFieldValidator2" Runat="server" ControlToValidate="fPageTitle" 
    ErrorMessage="You must specify the title of this page" Display="None"/>


</flux:HostTable>

</form>
</body>
</html>

Step 7 - creating some content

image

1. Open the administration console (http://localhost/[youralias]/admin/) and select "Content Management" from the right hand menu.  You should be presented with an empty cms as follows;

image

2. Click "Add new" in the actions panel and you will be presented with the new "website channel" form which we defined earlier. This is presented immediately without asking what sort of content you wish to create as it's the only type that can be added to the root like this.

3. Enter test for the channel name and path as below and click save.

image

4. Now click the "test" node in the tree view to the left and then click "Add new". This will present the form for adding a page, again as defined above. Define a home page as follows;

image

5. Repeat the process to create another new page called "About us".

image

6. You should now have a site defined as follows;

image

7. And you will notice that all of your pages presently are shown with a green icon indicating they are checked out for editing by you.

image

8. In order for this content to appear on the site, it must be checked in, and optionally validated through the workflow processes. An easy way to check everything in quickly is to go back to the admin home page and select the items that are checked out to you and check them all in in one go. Alternatively, select each in turn and click check-in.

9. Once they are checked in they should present in blue;

image

Step 7a - using a WYSIWYG editor

A downside to what we've seen so far is our editor is free text. It doesn't allow us to write bold, italics, insert images etc. As such, we would normally obtain a WYSIWYG editor and embed that instead. In this instance, you could grab the latest version of FCKEditor, which is my favourite editor that works well across all browsers. Normally we'd download FCK editor's latest source and then apply flux specific changes to it, to allow the editor to use the standard flux asset and link browsers. Feel free to use this process, or to save time, you can download the basic starter kit which I am producing as a result of this article from our codeplex site. For those interested however, to use a WYSIWYG editor like this we would;

1. Download the latest FCKEditor and apply flux specific changes to it (replace the asset and link browser).

2. Extract it's contents to a folder in your site called /FCKEditor

3.  Copy the FredCK.FCKEditorV2.dll into your site's Bin directory.

4. Create a folder in settings called common and create a custom user control in there with the name fckEditor.ascx. The content of the ASCX should be as follows;

<%@ Control Language="C#" AutoEventWireup="true" %>
<%@ Register TagPrefix="fck" Namespace="FredCK.FCKeditorV2" Assembly="FredCK.FCKEditorV2" %>

<script runat="server">
    public string Value
    {
        get
        {
            return fContent.Value;
        }

        set
        {
            fContent.Value = value;
        }
    }
    
    protected void Page_Load(object sender, EventArgs e)
    {
        if (!Page.IsPostBack)
        {
            fContent.BasePath = "~/FCKEditor/";
            fContent.ToolbarSet = "Default";
        }
    }
</script>

<fck:FCKeditor runat="server" ID="fContent" Height="200px"/>

5. We then replace the textbox in form.aspx for the Page content type;

6. From here, the editor should allow us to display html etc.

image

Step 8 - setting up mappings and wiring it up

Flux is capable of doing URL re-writing for you, and is capable of being multi-homed against different domains etc as a result. In order to tell flux which folder to get it's content from for different domains and which page is the home page, we need to use the mapping tool.

image

1. Open website mappings and select to add a new mapping

2. The domain we're adding a mapping for during development here is localhost, so enter that into the domain field.

3. Select the "test" root node as the site's content root node - this determines which node contains all the content for this domain.

4. Select the "home" node as the site's home page node.

image

5. Click save. You have successfully setup the mapping between the domain and the content.

image

6. As flux doesn't force you to use the URL re-writer engine, we must now hook this up in code. Create a default.aspx web form in the root of your site and set it's content as follows; This simply redirects to the correct URI whenever someone accesses the site.

<%@ Page Language="C#" %>
<%@ Import Namespace="Deepcode.Flux.Core.Systems.URLRewriter"%>

<script runat="server">
protected void Page_Load()
{
    Rewriter rw = new Rewriter();
    rw.RewriteFromHome( Context );
}
</script>

Step 9 - Writing the rendering template for the worlds most basic content type

With the domain maps etc in place, we're ready now to show the content we've created to the world. If you navigate now to your site, you will get an error saying it can't find the render template for the page. That's because we still need to create it.

1. Create a new web form - /settings/types/page/render.aspx and set it's content as follows;

<%@ Page Language="C#"%>
<%@ Import Namespace="Deepcode.Flux.Core.Systems.CMS" %>
<%@ Import Namespace="System.Linq" %>
<script runat="server">
    protected ContentObject Content { get; private set; }

    protected override void OnLoad(EventArgs e)
    {
        Content = ContentQuery.GetByID(Int32.Parse(Request["nodeid"]));
        if (Content == null) Response.Redirect("~/", true);
    
        Page.Title = Content.NodeTitle; 
        html.Text = Content.Fields["BodyContent"];
        
        // Get the menu together by querying everything within my parent....
        ContentQuery query = new ContentQuery();
        query.BaseMatch.AND(new StaticFieldMatch(
            ContentStaticField.ParentID, 
            ContentQueryOperator.EQ, 
            Content.FK_ParentID));
    
        menu.DataSource = from c in query.GetMatching() 
                          select new 
                          { 
                              Url = Page.ResolveUrl(String.Format("~/{0}/{1}.aspx", 
                                      c.NodePath, c.NodeName)), 
                              Title = c.NodeTitle };
        menu.DataBind();
        
        
        base.OnLoad(e);
    }
</script>

<html>
<head runat="server"/>
<body>

    <div>
    <asp:Repeater ID="menu" runat="server">
        <ItemTemplate>
            <a href="<%#Eval("Url")%>">[<%#Eval("Title")%>]</a> &middot;
        </ItemTemplate>
    </asp:Repeater>
    </div>

    <asp:Literal ID="html" runat="server"/>

</body>
</html>

You should now be able to navigate to your site and view a basic page with a simple menu:

image

image

Step 10 - What next?

Well, that was the overview for how to get a basic CMS up and running. It's quite complex, but this means that flux isn't prescriptive about anything - if you don't want to use it to define the actual structure of your site, you don't have to, you can use it as a content repository, and ask it for content as necessary. In fact this approach works quite well when building complex sites. You can even setup flux to work with the excellent ASP.NET MVC framework. Simply have one site for the admin and another site (the MVC app) for the front end, which in turn uses the flux API's to get at it's data.

Whilst this was an overview, I am intent now on packaging this as a sample onto codeplex (http://flux.codeplex.com) so that it can be used as a starting point very quickly without repeating all the steps above.

Notes on live deployment

When installing your finished product onto your deployment site, chances are you are installing to a shared environment where IIS is already configured for you and the database is already created, but empty. In this case, all you need to do to deploy the first time is;

  • FTP all of your files up to the server
  • On the first instance, delete settings/flux.config so that the database scripts are executed
  • Login to http://[yoursite.com]/admin/setup/install/ to walk through the installation process
  • Create your content and domain mappings.....

From then on, if you change your template code etc, you just need to re-deploy your files.

Tuesday, 22 September 2009

Visual Studio 2008 crapping out when debugging?

Had a weird issue today where visual studio would blow up when I finished debugging a unit test. Turns out, the bug wasn’t anything to do with debugging or unit testing etc, but a bug in visual studio 2008 that may manifest if you changed your panel layouts. Each time I debugged the test, I was moving the immediate window from it’s default position to be docked with the other debug windows at the bottom of the screen and it was this that was causing the crash. There’s a hotfix available from the Microsoft bods:

http://code.msdn.microsoft.com/KB960075

Wednesday, 29 July 2009

NHibernate N:N relationships

NHibernate provides the facility to map out many-to-many relationships without the need for the intersection table to be represented in the object model. Take for example, the following schema of Users, Roles and the intersection table that links users to many roles and roles to many users;

image

We then have the following object model to represent this;

image

or, in code if you prefer;

public class User
{
    // Fields
    public virtual Guid Id { get; set; }
    public virtual string UserName { get; set; }
    public virtual string Email { get; set; }

    // Storage for the roles that this user belongs to
    private readonly IList<Role> _roles = new List<Role>();

    // Accessor to list of roles - note we don't return list
    // that we can add/remove from - we use methods to do that
    // so we can implement any business logic.
    public virtual ReadOnlyCollection<Role> Roles
    {
        get 
        { 
            return new ReadOnlyCollection<Role>(_roles);
        }
    }

    // Method to associate a role with this user
    public virtual void AddRole( Role role )
    {
        _roles.Add( role );
    }

    // Method to remove a role from this user
    public virtual void RemoveRole( Role role )
    {
        _roles.Remove(role);
    }
}
public class Role
{
    // Fields
    public virtual Guid Id { get; set; }
    public virtual string Name { get; set; }

    // Storage for the users that have this role
    private readonly IList<User> _users = new List<User>();

    // Returns list of users belonging to this role
    // Notice this can't be added or removed from.
    // Nor is there any logic to add users into roles
    // as the link is maintained from the User object.
    public virtual ReadOnlyCollection<User> Users
    {
        get
        {
            return new ReadOnlyCollection<User>(_users);
        }
    }
}

Notice that the relationship between users and roles is maintained and owned in the user object, not role. The role object just gives us information about which users belong to a role, but we can’t add or remove users into it without loading the user and accessing it’s Add/Remove Role methods. We can map this out as follows (notice, I’m using Fluent mappings for NHibernate here);

public class UserMap : ClassMap<User>
{
    public UserMap()
    {
        base.Id(x => x.Id).GeneratedBy.GuidComb();
        base.Map(x => x.UserName);
        base.Map(x => x.Email);

        HasManyToMany<Role>(Reveal.Property<User>("_roles"))
            .LazyLoad()
            .AsBag()
            .WithTableName("UserRoles")
            .WithParentKeyColumn("UserId")
            .WithChildKeyColumn("RoleId")
            .Access.AsReadOnlyPropertyThroughCamelCaseField(Prefix.Underscore)
            .Cascade.All();
    }
}

public class RoleMap : ClassMap<Role>
{
    public RoleMap()
    {
        base.Id(x => x.Id).GeneratedBy.GuidComb();
        base.Map(x => x.Name);

        HasManyToMany<User>(Reveal.Property<Role>("_users"))
            .LazyLoad()
            .AsBag().Inverse()
            .WithTableName("UserRoles")
            .WithParentKeyColumn("RoleId")
            .WithChildKeyColumn("UserId")
            .Cascade.None();
    }
}

Notice that the role map has .Inverse() set, which indicates that it is not the owner of the relationship.

That is pretty much all there is to it, NHibernate will take care of maintaining the relationship table for you.

Tuesday, 28 July 2009

Query NHibernate – Simple query patterns

There are a variety of ways to query NHibernate, including using HQL, the criteria API, Linq2NH, and the criteria API tied up with the Lambda extensions. My preference is the criteria API as it’s a powerful way of expressing dynamic queries. I’m going to explore how we go about implementing some of the more basic query patterns using this API;

Given the following database as an example, where a location can be split into multiple sub-locations, each sub-location belonging to a cost centre, we’ll drive out some simple queries;

 image

First off, a simple location query by name;

ICriteria c =
    session.CreateCriteria(typeof(Location))
    .Add(Restrictions.Eq("Name", "Manchester"));

Next, a query that will check the name contains particular text and the postcode starts with text provided.

ICriteria c =
    session.CreateCriteria(typeof(Location))
    .Add( Restrictions.Conjunction()
        .Add(Restrictions.Like("Name", "%manchester%"))
        .Add(Restrictions.Like("Postcode", "M60%")));

How about looking for all sub locations where it’s parent location has a postcode of M60*?

ICriteria c = session.CreateCriteria(typeof(SubLocation))
    .CreateAlias("Location", "l")
    .Add(Restrictions.Like("l.Postcode", "M60%"));

Going further, how about finding all sub locations where some text can be found either in the sub location name, the location name, location address, location postcode or cost centre name? Simples….

ICriteria criteria = session.CreateCriteria(typeof(SubLocation))
    .CreateAlias("Location", "l")
    .CreateAlias("CostCentre", "cc")
    .Add(Restrictions.Disjunction()
        .Add(Restrictions.Like("Name", freeTextToMatch))
        .Add(Restrictions.Like("l.Name", freeTextToMatch))
        .Add(Restrictions.Like("l.Address", freeTextToMatch))
        .Add(Restrictions.Like("l.Postcode", freeTextToMatch))
        .Add(Restrictions.Like("cc.Name", freeTextToMatch)));

What about finding all locations that have more than one sub location?

DetachedCriteria subquery = DetachedCriteria.For(typeof(SubLocation))
   .Add(Restrictions.EqProperty("l.Id", "Location.Id"))
    .SetProjection(Projections.Count("Id"));

ICriteria criteria = session.CreateCriteria(typeof(Location), "l")
    .Add(Subqueries.Lt(1, subquery));

It’s a bit more complicated, but the first bit creates a reusable query that will look for sub locations where the sub location’s parent location Id is of the same value as the Id property of the aliased object “l” passed into the query. It then projects the results into a count against the Id property.

The second part then creates a criteria that will find all locations, aliased as “l” (for the sub query above), and execute the given subquery, ensuring that the result is > 1.

Finally, getting a bit more complex, find all locations having more than one sub location in cost centre named “ENGINEERING”?

DetachedCriteria subquery = DetachedCriteria.For(typeof(SubLocation))
        .CreateAlias("CostCentre", "cc")
        .Add(Restrictions.EqProperty("l.Id", "Location.Id"))
        .Add(Restrictions.Eq("cc.Name", "ENGINEERING"))
        .SetProjection(Projections.Count("Id"));

ICriteria criteria = session.CreateCriteria(typeof(Location), "l")
        .Add(Subqueries.Lt(1, subquery));

Rhino – Mock versus Stub

Questions always pop up on when we should use mocks and when we should use stubs.

To summarise the differences;

  • Stubs
    • Provide canned answers to calls made during tests
    • Don’t respond to anything else
  • Mocks
    • Setup expectations - behaviour that will be verified as part of the test

Links to articles that discuss same:

Saturday, 9 May 2009

Why-o-why is this so difficult!

This is really getting my goat today – I’ve been looking around for a week now for some way of sucking out the content of my old blog (community server 2.1, self hosted) and spitting it out into my new host (blogger.com). I want to take across all of the content, including attachments, images, posting dates and tags etc, but there is nothing out there that seems to do a particularly adequate job!

Blogging has been around for a while now, and regardless of the type of host we use, blog articles are largely all the same – they have a title, some rich text, inline images, tags, attachments and a publishing date time. How hard can it be to come up with a canonical format to easily move this content into and out of different hosts and have all the hosts support that standard.

Hard is all I can assume as it’s not been done in the years that blogging has been around, with the exception of half hearted attempts at standardisation like with BlogML, which doesn’t deliver on the objectives I’ve mentioned.

Wednesday, 4 March 2009

Using Regular Expressions to identify HTML tags

In many situations we might find ourselves looking to find specific HTML tags within a fragment in order to process them in some way. A good example would be a HTML “white list” of tags that you want to pre-process to allow through anti cross site scripting encoding.

The following regular expression will find all tags specified (highlighted red) in the input, regardless of them being start tags, end tags, self terminating tags and irrespective of the number of attributes etc.

<[/]?(P|B|I)(/>|\s/>|\s[\S|\s]*?>|>)

This will also return correct results when looking for <I> tags in that it won’t incorrectly allow <IMG> to be a positive match.