Caching and accessing configuration data in ASP.NET MVC app - asp.net-mvc

I'm about to take a look at how to implement internationalisation for an ASP.NET MVC project. I'm looking at how to allow the user to change languages. My initial though is a dropdownlist containing each of the supported langauages. Whoever a few questions have come to mind:
How to store the list of supported languages? (e.g. just "en", "English"; "fr", "French" etc.) An xml file? .config files?
If I store this in a file I'll have to cache this (at startup I guess). So, what would be best, load the xml data into a list (somehow) and store this list in the System.Web.Cache? Application State?
How then to load this data into the view (for display in a dropdown)? Give the view direct access to the cache?
Just want to make sure I'm going in the right direction here...
Thank you.

First tell us please what you understand under "internationalization"?
a) You want to only have UI in several languages
b) You want to have the content in several languages
Have a look at this other question where I did some explanation on the topic: Advice on ASP.NET Multi-Lingual Strategy
To your questions:
I would advice storing a list of supported languages in DB where you can identify them by 2-character code. See ISO 3166-1 alpha-2 country codes.
If you're coming from the ASP.NET WebForms world, then the first thing to unlearn would be the statefulness and cache when you're dealing with ASP.NET MVC. You can store a simple single value in session or cookies that will indicate the currently active languages. The server side will generate a view in the required language based on this submitted values.
You instantiate a model in your controller to be passed to the view. In the model you can have a property - a list with available languages. The view can (and supposed to) consume the data to display from its model.
Regarding where you could store a cached version of languages. You can consider having somewhere a static list that will be initialized at the first request (like weak references). It will be shared then across all requests for a given application domain, meaning all the users of your web app will see the same values (until the domain gets recycled or the server restarted). If you were to use the server session, then the list would be created again for each new user session.

Related

architecture of Umbraco application

I m new to Umbraco, I have watched Umbraco.tv videos and want to use Umbraco in a project as a cms for managing and editing content. I am highly thankful for your guidance, time and for your thoughts on 3 questions:
How a Umbraco based data driven proejct should be architecutured ? For custom database tables do you use a separate database or same Umbraco database ?
How you work with custom data (non content) ? Do you make everything a document type, even if it is data which you are not going to create content of, for example a simple form submitted data ?
For DAL what technology or ORM you use ? Does Umbraco provide any API for saving simple data which is not a content or document type ?
Thank you so much once again.
1 The architecture question is important but it also has be considered against how complex the project needs to be.
I would usually recommend a separate database for non-Umbraco data since this keeps everything nicely independent and manageable especially as projects grow. It also means that CMS-specific data (i.e. content) can be kept separately from none-CMS data, e.g. user registrations.
However, if the project is small and isn't likely to grow, keep it simple. Use the same database and piggy back off Umbraco's implementation of the Petapoco ORM. For example:
ApplicationContext.DatabaseContext.Database.Save(new Thing());
Or
var item = ApplicationContext.DatabaseContext.Database.Single(thingId);
2 For custom data, again it's a matter of need, maintainability and simplicity. Only use document types for what needs to be and can be stored in the CMS. My personal rule is that if it isn't content or organises content then it doesn't belong in the CMS. For example news and news categories obviously belong in the CMS. However, the comments made on an article have no reason to in the CMS.
3 With regards to DAL, as I have said, Umbraco has an implementation of Petapoco that can be used out of the box. If the project is basic enough, just use that. There is little point in using anything else unless you need some separation and/or some additional grunt in which case I would recommend using NHibernate or EF.
In addition to the points above,
Use NuGet;
Use the MVC mode of Umbraco, as it will provide you with substantially more flexibility. Check out the Hybrid Framework as it provides a very good start point for a robust and flexible project architecture;
http://www.youtube.com/watch?v=0PtzyrEFG7I.
You always need a doctype in Umbraco, even if a page doesn't offer any WYSIWYG type editability
I would recommend using a Service Oriented Architecture, and the .dlls you drop into Umbraco can call the service. You can then deploy this service and have full control over how you do data access. Choose whatever method you want. Most modern sites use an ORM and it doesn't matter which, although nHIbernate and Entity Framework are the favourites. Don't be frightened to mix and match a more direct form of data access though as it can give you more control, especially in situations where performance optimisation of large queries is important.
If you're not familiar with adding custom .NET functionality into Umbraco, Trying out adding .NET user controls into Umbraco will give you a good start, and to help you to understand how you can utilise your own .dlls in Umbraco:
http://umbraco.com/help-and-support/video-tutorials/introduction-to-umbraco/developer-introduction/using-net-user-controls.aspx
Anything custom I put in the same database as my Umbraco installation, but everything in custom tables. I don't touch the Umbraco tabes, I would not want to affect my future version updates.
Form submitted data I store in my own custom tables, I avoid creating content nodes with those, it's often tricky and doesn't give me the flexebility I often need. What I do instead is create an "Admin" document type, that is behind login (hard coded access, but easy to hook up to Umbraco users / members if wanted) and use my own custom UI to display my stored custom data.
I use PetaPoto (http://www.toptensoftware.com/petapoco/), it's a micro ORM that is added through a single file (installation is so easy then), using the same db connection string. Then I create custom models as I need and with with parts of the MVC. I normally stay away from route hijacking and rather use Surface Controllers and ajax calls for almost everything.
Hope this helps!
You can use the database containing the Umbraco tables for tables not used in Umbraco. If there are no hosting problems for you using multiple databases then you can simply link to a second database in the web.config - this would be safer than using the default Umbraco database as Umbraco packages often add database tables & there could be naming conflicts.
Viewing non-Umbraco data (eg from a database) is best done by adding macros that access the data using standard .Net patterns (eg razor scripts, .Net User Controls) & then in Umbraco you add in a reference to the macro in the template (view). You can use multiple templates (views) for any document type; so if you have a document type called 'forms' that contains no data you can use the 'allowed templates' checkboxes to say which view(s) are valid for this document type. When you add a content item you must specify a doc type at the start, but the template (view) can be changed at any time.
If you are storing data any .Net ORM will work with Umbraco (see http://en.wikipedia.org/wiki/List_of_object-relational_mapping_software#.NET) I've used Linq to Sql, Subsonic & Dapper before now - but there are lots of options.
Take a look at my example using umbraco within Onion Architecture
https://github.com/afroukh/OnionCMS

ASP.NET MVC - Best way to retain search criteria across all pages

I am writing a real-estate related ASP.NET MVC application. I would like to auto populate the the user's (detected) location (city, state) when the site loads and allow them to search for using a different location. Location search is available on all pages on the site and I would like to preserve the most recent location value across all requests. Right now I'm planning to use cookies to preserve this data across all requests so it can be populated in all my views (via a layout file) but wonder if there is a better way of doing so.
I'd like to hear any suggestions.
Thank you,
JP
you could either use static variables or session variables. To use session variables in MVC you use:
HttpContext.Current.Session["Location"] = "East St. Louis";
To recall in your view with Razor you use:
#Session["Location"].ToString()

Using a client state model object for a hierarchical MVC web application

I'm building a large hierarchical web application and I need some help deciding on some best practices with leveraging MVC.
The application will have tabs at the top which control a sub page, and a query pane (off to the side).
There will be two templates for query panes, each used by different sub-pages. The sub-pages will be based on the selected tab with settings derived from the query panes.
Clicking on tabs or updating the query pane will update the sub-page section without refreshing the page.
I'm a bit new to MVC and what I don't quite understand is how I can leverage MVC methodologies to help me manage the web application's state (which consists of the selected tab, query options, and other page-specific options).
Currently I'm planning on initially setting up a model which stores the client state parameters (default values, or values obtained from a DB), and using it to load the page, consisting of several partial views. When anything is changed (tab/query/etc), the view will call a corresponding controller, passing back model parameters via post (I'm assuming there's no way to store session-specific client state models on the server-side?).
My question is:
Am I doing it right?
If not, what am I missing; and specifically, is there a way to store these session-specific state models server side so they don't have to be passed back to the server during every single page transaction?
If I understood everything you need Its a SPA (Single Page Application). This will provide a magic user experience, without full page reload, and low data traffic. But, requireds some MVVM framework (AngularJS, KnockoutJs, etc) and a lot of JavaScript coding. But the result is amazing. The guy behind this in MVC is John Papa, take a look in everything on his blog and you will win.
John Papa Blog
Hopes Its Help you

Right way to design a mobile ASP.NET MVC 4 data entry application with HTML5 offline capabilities

I'm creating a PoC of an ASP.NET MVC 4 mobile application (tablets) for field operators that needs to support offline capabilities. It's a relatively simple data entry application with a WebAPI backend that will use a SQL database as persistent storage.
This is the first time I'll be implementing offline support, so I'm a bit unsure of which is the best way to design this. Right now I can think of two ways, and not sure which one is better or more correct.
What I want is that whenever there is internet connectivity, the information is retrieved from the server. If the connection is lost, I want the latest information to be cached/served and available to display to the user.
Based on a few tests that I've made, the way the HTML5 offline support appears to work is that as soon as any of the Views is loaded, the browser will check the manifest and cache the version of each page at that moment of time. If my understanding is correct, that means that if the user adds new entries after the browser downloaded the manifest files, then those records would not show in the offline version.
One way I can think to work around that is to change the manifest version (since I'm generating it dynamically) every time it is served, but i feel that would defeat the purpose..or at least be incredibly inefficient bandwidth-wise.
The only other alternative that I can think of is not using regular MVC (not passing the records/db information as the model property or a ViewBag property) to display the records, but just retrieve the json for the page information needed and save in local storage. Then, generate the view using a template or javascript based on the information in localstorage. That would work either online or offline, wouldn't it?
I'm not sure if I'm approaching this the wrong way. Would building an SPA or a simple HTML/javascript application be more appropiate? Maybe using a JS framework? At this point I have some ability to determine what technology or framework we'd use.
Any guidance is greatly appreciated.
Based on a few tests that I've made, the way the HTML5 offline support appears to work is that as soon as any of the Views is loaded, the browser will check the manifest and cache the version of each page at that moment of time. If my understanding is correct, that means that if the user adds new entries after the browser downloaded the manifest files, then those records would not show in the offline version.
This is why you shouldn't use server side rendering. When the server mixes data with formatting before sending to the browser it means that the browser can't tell the difference between the two. HTML5 and Javascript are perfectly capable of handling templating on their own so why not let them?
Write your MVC application as nothing but a series of controller methods. Do not use "views" per-say. Instead use straight HTML files that use javascript to make AJAX requests to the server for data. Populate the HTML view with the data and you're golden. Then you can cache the HTML files because they're static and you can save the data separately for offline mode.
Adding to the accepted answer, there's a thorough post in this matter titled:
Build an HTML5 Offline Application with Application Cache, Web Storage and ASP.NET MVC.
The guy explains everything using a sample/complete app and do a great job explaining each and every point.
Besides that, if you really want to get the crème de la crème from modern web dev with offline capabilities, I'd suggest you take the 2 parts course by extraordinaire/star developer John Papa. It uses the SPA = Single Page Application UI paradigm.
The course is available at Pluralsight:
Building Apps with Angular and Breeze - Part 1
Building Apps with Angular and Breeze - Part 2
Enjoy as much as I'm enjoying it. Make sure you check this keyword/feature in the course: WIP = Work In Progress.
If you're interested, here's is a repo where I keep the updated course's code:
https://github.com/leniel/SPACodeCamper

Editing heirarchical data in ASP.NET MVC 2

Can anyone point me to some good resources that can help me understand the best way to work with hierarchical data in ASP.NET MVC 2?
I have an application under development that requires an interface allowing users to add, remove and modify children and grand-children of my root object. The user can make multiple changes without persistance. Only when they click "Save" will the entire object graph be saved.
I've seen one article that serialized the object and stored the data in a hidden field on the form but that seems really cludgy and I am dealing with a lot of data.
If I was doing this in standard ASP.NET, I'd be looking at using child windows and the like to display the edit pages and maintain an instance of the object being edited in Session - which is bad in and of itself. But I've been told we are using MVC as we are standardizing our platforms (but not moving up to MVC 3 yet).
Essentially I need that app to display the properties of my root which includes a child collection of objects. The UI should allow the user to add new items to the collection, remove existing items and 'open' an item for editing. These child items also contain their own list of grandchildren that is editable as well. All of this needs to go on without round-trips across the wire to persist data (its a distributed architecture with all data access behind a WCF service interface).
The examples on www.asp.net all persist the data each time a single change is made, i.e. each postback. But, that would require major schema changes and extra code to deal with temporary objects versus committed objects plus the overhead of the service calls each time. I'm looking for a better solution.
Have you considered looking at any client side libraries like Knockout.JS? I've found that it is excellent at manipulating collections and posting the final version as JSON. Here is an example of what you can do with it. Here is an article about how to integrate it with MVC 2. This is my absolute favorite JS library.

Resources