I have a List<object> that I want to access in my class.
The list is a property, and has a get and set.
When I want a single item from the list, I want it to lazy load in the single item I am requesting from a database.
To do this - I want to check if the item being requested exists in the list, if not, I want to add it to the list from the database (if it exists in the database) - then I return the list.
The problem is, I don't think the getter knows what items are being 'gotten', it just knows that the list itself is being requested. So while I can easily lazy load in the entire list, I can't "download and cache as you go".
It was looking very neat too - I could expose the list as a property, and let the property handle all of the loading logic.
In less trivial examples, it's a bit more useless I suppose - like if I wanted to run a query - I would always have to do it against the whole data set otherwise I'd miss results.
Currently, I just download and cache every result from the database when the property is called.
Is there a better approach?
Was thinking mainly about pagination. I.e. want items 0-9, items 10-19, etc
Since we're talking about pagination, we must understand, that pagination is neither retrieving single item from the list, nor indexed list access.
Pagination usually allows user to:
fetch data from data source using small portions (pages);
set page size;
fetch next page (next N pages);
fetch previous page (previous N pages);
fetch particular page (go to page N).
Note, that pagination is about reading data, not about inserting new one.
Basically, pagination requires:
paged data source. This can be database, file, in-memory data, that allows to get data portion. E.g. SQL Server allows this via OFFSET... FETCH NEXT in TSQL;
paged data source API to access particular page. E.g. IQueryable<T> and its extensions Skip and Take;
pagination service. It implements pagination logic;
pagination UI controls.
There can be many specific things. Pagination service and UI depends on what framework do you use (ASP .NET, WPF, etc).
Here's the simple pagination service implementation:
/// <summary>
/// Represents paged data source.
/// </summary>
/// <typeparam name="T">
/// Data item type.
/// </typeparam>
public interface IPagedDataSource<T>
{
/// <summary>
/// Fetches particular page from data source.
/// </summary>
/// <param name="pageNumber">
/// Page number.
/// </param>
/// <param name="pageSize">
/// Page size.
/// </param>
/// <returns></returns>
IReadOnlyList<T> GetPage(int pageNumber, int pageSize);
}
/// <summary>
/// Represents paged data source, built on top of <see cref="IQueryable{T}"/> instance.
/// </summary>
/// <typeparam name="T">
/// Data item type.
/// </typeparam>
public sealed class QueryablePagedDataSource<T> : IPagedDataSource<T>
{
private readonly IQueryable<T> query;
/// <summary>
/// Initializes <see cref="QueryablePagedDataSource{T}"/> instance.
/// </summary>
/// <param name="query">
/// <see cref="IQueryable{T}"/> to use as paged data source.
/// </param>
public QueryablePagedDataSource(IQueryable<T> query)
{
this.query = query;
}
/// <summary>
/// Fetches particular page from data source.
/// </summary>
/// <param name="pageNumber">
/// Page number.
/// </param>
/// <param name="pageSize">
/// Page size.
/// </param>
/// <returns></returns>
public IReadOnlyList<T> GetPage(int pageNumber, int pageSize) => query
.Skip(pageNumber * pageSize)
.Take(pageSize)
.ToList();
}
/// <summary>
/// Implements basic pagination logic.
/// </summary>
/// <typeparam name="T">
/// Data item type.
/// </typeparam>
public sealed class Paginator<T>
{
private readonly int pageSize;
private readonly IPagedDataSource<T> dataSource;
private readonly IDictionary<int, IReadOnlyList<T>> pages;
/// <summary>
/// Initializes <see cref="Paginator{T}"/> instance.
/// </summary>
/// <param name="dataSource">
/// Paged data source.
/// </param>
/// <param name="pageSize">
/// Page size.
/// </param>
public Paginator(IPagedDataSource<T> dataSource, int pageSize)
{
this.dataSource = dataSource;
this.pageSize = pageSize;
pages = new Dictionary<int, IReadOnlyList<T>>();
}
/// <summary>
/// Gets current page number.
/// </summary>
public int CurrentPageNumber { get; private set; }
/// <summary>
/// Gets current page.
/// </summary>
public IReadOnlyList<T> CurrentPage
{
get
{
if (pages.Count == 0)
{
NavigateTo(0);
}
return pages[CurrentPageNumber];
}
}
/// <summary>
/// Navigates to particular page.
/// </summary>
/// <param name="pageNumber">
/// Page number to navigate to.
/// </param>
public void NavigateTo(int pageNumber)
{
CurrentPageNumber = pageNumber;
if (!pages.ContainsKey(CurrentPageNumber))
{
var page = dataSource.GetPage(CurrentPageNumber, pageSize);
pages.Add(CurrentPageNumber, page);
}
}
/// <summary>
/// Navigates to the next page.
/// </summary>
public void Next() => NavigateTo(CurrentPageNumber + 1);
/// <summary>
/// Navigates to the previous page.
/// </summary>
public void Previous()
{
if (CurrentPageNumber > 1)
{
NavigateTo(CurrentPageNumber - 1);
}
}
/// <summary>
/// Exports fetched data as a list.
/// </summary>
/// <returns>
/// A list, containing fetched data. Data is ordered by page number.
/// </returns>
public List<T> ToList() => pages
.OrderBy(_ => _.Key)
.SelectMany(_ => _.Value)
.ToList();
}
Try to play with it (e.g. create data source from collection using AsQueryable extension). Maybe it will be useful for you.
Note, that there are no null-checking and range-checking.
In Summary: I Want to replace IFormFile IFormFileCollection with my own classes not attached to Asp .Net because my view models are in different project with poco classes. My custom classes are ICommonFile, ICommonFileCollection, IFormFile (not Asp .net core class) and IFormFileCollection.
i will share it here:
ICommonFile.cs
/// <summary>
/// File with common Parameters including bytes
/// </summary>
public interface ICommonFile
{
/// <summary>
/// Stream File
/// </summary>
Stream File { get; }
/// <summary>
/// Name of the file
/// </summary>
string Name { get; }
/// <summary>
/// Gets the file name with extension.
/// </summary>
string FileName { get; }
/// <summary>
/// Gets the file length in bytes.
/// </summary>
long Length { get; }
/// <summary>
/// Copies the contents of the uploaded file to the <paramref name="target"/> stream.
/// </summary>
/// <param name="target">The stream to copy the file contents to.</param>
void CopyTo(Stream target);
/// <summary>
/// Asynchronously copies the contents of the uploaded file to the <paramref name="target"/> stream.
/// </summary>
/// <param name="target">The stream to copy the file contents to.</param>
/// <param name="cancellationToken">Enables cooperative cancellation between threads</param>
Task CopyToAsync(Stream target, CancellationToken cancellationToken = default(CancellationToken));
}
ICommonFileCollection.cs
/// <inheritdoc />
/// <summary>
/// Represents the collection of files.
/// </summary>
public interface ICommonFileCollection : IReadOnlyList<ICommonFile>
{
/// <summary>
/// File Indexer by name
/// </summary>
/// <param name="name">File name index</param>
/// <returns>File with related file name index</returns>
ICommonFile this[string name] { get; }
/// <summary>
/// Gets file by name
/// </summary>
/// <param name="name">file name</param>
/// <returns>File with related file name index</returns>
ICommonFile GetFile(string name);
/// <summary>
/// Gets Files by name
/// </summary>
/// <param name="name"></param>>
/// <returns>Files with related file name index</returns>
IReadOnlyList<ICommonFile> GetFiles(string name);
}
IFormFile.cs
/// <inheritdoc />
/// <summary>
/// File transferred by HttpProtocol, this is an independent
/// Asp.net core interface
/// </summary>
public interface IFormFile : ICommonFile
{
/// <summary>
/// Gets the raw Content-Type header of the uploaded file.
/// </summary>
string ContentType { get; }
/// <summary>
/// Gets the raw Content-Disposition header of the uploaded file.
/// </summary>
string ContentDisposition { get; }
}
IFormFileCollection.cs
/// <summary>
/// File Collection transferred by HttpProtocol, this is an independent
/// Asp.net core implementation
/// </summary>
public interface IFormFileCollection
{
//Use it when you need to implement new features to Form File collection over HttpProtocol
}
I finally created my model binders successfully, i will share it too:
FormFileModelBinderProvider.cs
/// <inheritdoc />
/// <summary>
/// Model Binder Provider, it inspects
/// any model when the request is triggered
/// </summary>
public class FormFileModelBinderProvider : IModelBinderProvider
{
/// <inheritdoc />
/// <summary>
/// Inspects a Model for any CommonFile class or Collection with
/// same class if exist the FormFileModelBinder initiates
/// </summary>
/// <param name="context">Model provider context</param>
/// <returns>a new Instance o FormFileModelBinder if type is found otherwise null</returns>
public IModelBinder GetBinder(ModelBinderProviderContext context)
{
if (context == null) throw new ArgumentNullException(nameof(context));
if (!context.Metadata.IsComplexType) return null;
var isSingleCommonFile = IsSingleCommonFile(context.Metadata.ModelType);
var isCommonFileCollection = IsCommonFileCollection(context.Metadata.ModelType);
if (!isSingleCommonFile && !isCommonFileCollection) return null;
return new FormFileModelBinder();
}
/// <summary>
/// Checks if object type is a CommonFile Collection
/// </summary>
/// <param name="modelType">Context Meta data ModelType</param>
/// <returns>If modelType is a collection of CommonFile returns true otherwise false</returns>
private static bool IsCommonFileCollection(Type modelType)
{
if (typeof(ICommonFileCollection).IsAssignableFrom(modelType))
{
return true;
}
var hasCommonFileArguments = modelType.GetGenericArguments()
.AsParallel().Any(t => typeof(ICommonFile).IsAssignableFrom(t));
if (typeof(IEnumerable).IsAssignableFrom(modelType) && hasCommonFileArguments)
{
return true;
}
if (typeof(IAsyncEnumerable<object>).IsAssignableFrom(modelType) && hasCommonFileArguments)
{
return true;
}
return false;
}
/// <summary>
/// Checks if object type is CommonFile or an implementation of ICommonFile
/// </summary>
/// <param name="modelType"></param>
/// <returns></returns>
private static bool IsSingleCommonFile(Type modelType)
{
if (modelType == typeof(ICommonFile) || modelType.GetInterfaces().Contains(typeof(ICommonFile)))
{
return true;
}
return false;
}
}
FormFileModelBinder.cs
/// <inheritdoc />
/// <summary>
/// Form File Model binder
/// Parses the Form file object type to a commonFile
/// </summary>
public class FormFileModelBinder : IModelBinder
{
/// <summary>
/// Expression to map IFormFile object type to CommonFile
/// </summary>
private readonly Func<Microsoft.AspNetCore.Http.IFormFile, ICommonFile> _expression;
/// <summary>
/// FormFile Model binder constructor
/// </summary>
public FormFileModelBinder()
{
_expression = x => new CommonFile(x.OpenReadStream(), x.Length, x.Name, x.FileName);
}
/// <inheritdoc />
/// <summary>
/// It Binds IFormFile to Common file, getting the file
/// from the binding context
/// </summary>
/// <param name="bindingContext">Http Context</param>
/// <returns>Completed Task</returns>
// TODO: Bind this context to ICommonFile or ICommonFileCollection object
public Task BindModelAsync(ModelBindingContext bindingContext)
{
dynamic model;
if (bindingContext == null) throw new ArgumentNullException(nameof(bindingContext));
var formFiles = bindingContext.ActionContext.HttpContext.Request.Form.Files;
if (!formFiles.Any()) return Task.CompletedTask;
if (formFiles.Count > 1)
{
model = formFiles.AsParallel().Select(_expression);
}
else
{
model = new FormFileCollection();
model.AddRange(filteredFiles.AsParallel().Select(_expression));
}
bindingContext.Result = ModelBindingResult.Success(model);
return Task.CompletedTask;
}
}
Actually Everything is working good except when i have Nested Models. I share an example of my models I'm using and I'll do some comments with working scenarios and which don't
Test.cs
public class Test
{
//It's Working
public ICommonFileCollection Files { get; set; }
//It's Working
public ICommonFileCollection Files2 { get; set; }
//This is a nested model
public TestExtra TestExtra { get; set; }
}
TestExtra.cs
public class TestExtra
{
//It's not working
public ICommonFileCollection Files { get; set; }
}
Actually when i make a request to my API I've got the following (Screenshot):
I'm sharing a screenshot of my postman request too for clarifying my request is good.
If there is any subjection to make this work with nested model it would be great.
Asp Net Core Model Binder won't bind model with one property only, if you have one property in a class, that property will be null but when you add two or more will bind it. my mistake i had one one property in a nested class. The entire code is correct.
I m hitting the following error on my new windows 10 universal app C#/XAML:
An exception of type 'System.InvalidCastException' occurred in GalaSoft.MvvmLight.Platform.dll but was not handled in user code
Additional information: Unable to cast object of type '' to type 'Windows.UI.Xaml.Controls.Frame'.
on the following navigating command in one of my page's view model:
_navigationService.NavigateTo(ViewModelLocator.MedicineBoxPageKey);
I am trying to have a hamburger menu style navigation (see this sample). app by Microsoft on an example of how to do this) to:
1- have a convenient solution shared across all my pages. The sample mentioned above uses an AppShell Page as the root of the app instead of a Frame, that encapsulates the navigation menu and some behavior of the back button. That would be ideal.
2- Use the MVVM-Light navigation service to handle all the navigation from my view model conveniently.
Here is how the App.xml.Cs initializes the shell page onLaunched:
AppShell shell = Window.Current.Content as AppShell;
// Do not repeat app initialization when the Window already has content,
// just ensure that the window is active
if (shell == null)
{
// Create a a AppShell to act as the navigation context and navigate to the first page
shell = new AppShell();
// Set the default language
shell.Language = Windows.Globalization.ApplicationLanguages.Languages[0];
shell.AppFrame.NavigationFailed += OnNavigationFailed;
if (e.PreviousExecutionState == ApplicationExecutionState.Terminated)
{
//TODO: Load state from previously suspended application
}
}
// Place our app shell in the current Window
Window.Current.Content = shell;
if (shell.AppFrame.Content == null)
{
// When the navigation stack isn't restored, navigate to the first page
// suppressing the initial entrance animation.
shell.AppFrame.Navigate(typeof(MedicinesStorePage), e.Arguments, new Windows.UI.Xaml.Media.Animation.SuppressNavigationTransitionInfo());
}
// Ensure the current window is active
Window.Current.Activate();
And here is the AppShell class definition:
public sealed partial class AppShell : Page
{
public static AppShell Current = null;
public AppShell()
{
this.InitializeComponent();
}
}
From what I have tried so far, the mvvm-light navigation service only works when a Frame is used a root of the app and note a Page (otherwise we get this casting bug).
However using a Frame does not seem to be a option either since as the sample app puts it:
Using a Page as the root for the app provides a design time experience as well as ensures that
when it runs on Mobile the app content won't appear under the system's StatusBar which is visible
by default with a transparent background. It will also take into account the presence of software
navigation buttons if they appear on a device. An app can opt-out by switching to UseCoreWindow.
I also tried to overide the navigationTo method from the mvvm-light navigation service but the bug seems to occur before I could catch it.
Does anyone has a solution to use the mvvm-light navigation service and a shell page as the app root (that manages the hamburger menu, etc.)?
Thanks a lot!
I talked to Laurent Bugnion and he recommended me to implemented my own navigation service who handles the navigation. For this I made a PageNavigationService who implements the INavigationService Interface of MVVM Light.
public class PageNavigationService : INavigationService
{
/// <summary>
/// The key that is returned by the <see cref="CurrentPageKey" /> property
/// when the current Page is the root page.
/// </summary>
public const string RootPageKey = "-- ROOT --";
/// <summary>
/// The key that is returned by the <see cref="CurrentPageKey" /> property
/// when the current Page is not found.
/// This can be the case when the navigation wasn't managed by this NavigationService,
/// for example when it is directly triggered in the code behind, and the
/// NavigationService was not configured for this page type.
/// </summary>
public const string UnknownPageKey = "-- UNKNOWN --";
private readonly Dictionary<string, Type> _pagesByKey = new Dictionary<string, Type>();
/// <summary>
/// The key corresponding to the currently displayed page.
/// </summary>
public string CurrentPageKey
{
get
{
lock (_pagesByKey)
{
var frame = ((AppShell) Window.Current.Content).AppFrame;
if (frame.BackStackDepth == 0)
{
return RootPageKey;
}
if (frame.Content == null)
{
return UnknownPageKey;
}
var currentType = frame.Content.GetType();
if (_pagesByKey.All(p => p.Value != currentType))
{
return UnknownPageKey;
}
var item = _pagesByKey.FirstOrDefault(
i => i.Value == currentType);
return item.Key;
}
}
}
/// <summary>
/// If possible, discards the current page and displays the previous page
/// on the navigation stack.
/// </summary>
public void GoBack()
{
var frame = ((Frame) Window.Current.Content);
if (frame.CanGoBack)
{
frame.GoBack();
}
}
/// <summary>
/// Displays a new page corresponding to the given key.
/// Make sure to call the <see cref="Configure" />
/// method first.
/// </summary>
/// <param name="pageKey">
/// The key corresponding to the page
/// that should be displayed.
/// </param>
/// <exception cref="ArgumentException">
/// When this method is called for
/// a key that has not been configured earlier.
/// </exception>
public void NavigateTo(string pageKey)
{
NavigateTo(pageKey, null);
}
/// <summary>
/// Displays a new page corresponding to the given key,
/// and passes a parameter to the new page.
/// Make sure to call the <see cref="Configure" />
/// method first.
/// </summary>
/// <param name="pageKey">
/// The key corresponding to the page
/// that should be displayed.
/// </param>
/// <param name="parameter">
/// The parameter that should be passed
/// to the new page.
/// </param>
/// <exception cref="ArgumentException">
/// When this method is called for
/// a key that has not been configured earlier.
/// </exception>
public void NavigateTo(string pageKey, object parameter)
{
lock (_pagesByKey)
{
if (!_pagesByKey.ContainsKey(pageKey))
{
throw new ArgumentException(
string.Format(
"No such page: {0}. Did you forget to call NavigationService.Configure?",
pageKey),
"pageKey");
}
var shell = ((AppShell) Window.Current.Content);
shell.AppFrame.Navigate(_pagesByKey[pageKey], parameter);
}
}
/// <summary>
/// Adds a key/page pair to the navigation service.
/// </summary>
/// <param name="key">
/// The key that will be used later
/// in the <see cref="NavigateTo(string)" /> or <see cref="NavigateTo(string, object)" /> methods.
/// </param>
/// <param name="pageType">The type of the page corresponding to the key.</param>
public void Configure(string key, Type pageType)
{
lock (_pagesByKey)
{
if (_pagesByKey.ContainsKey(key))
{
throw new ArgumentException("This key is already used: " + key);
}
if (_pagesByKey.Any(p => p.Value == pageType))
{
throw new ArgumentException(
"This type is already configured with key " + _pagesByKey.First(p => p.Value == pageType).Key);
}
_pagesByKey.Add(
key,
pageType);
}
}
}
Basicly it's a copy of his implementation. But instead of parsing to a Frame I parse to an AppShell and use the AppFrame Property to navigate.
I put this to my ViewModelLocator. Instead of:
var navigationService = new NavigationService();
I will just use:
var navigationService = new PageNavigationService();
EDIT: I Noticed that there is an excpetion in the NavMenuListView when you use the backkey after you navigated with the new navigationservice since the selected item is null. I fixed it with adjusting the SetSelectedItem Method and adding a nullcheck in the for loop after the cast:
public void SetSelectedItem(ListViewItem item)
{
var index = -1;
if (item != null)
{
index = IndexFromContainer(item);
}
for (var i = 0; i < Items.Count; i++)
{
var lvi = (ListViewItem) ContainerFromIndex(i);
if(lvi == null) continue;
if (i != index)
{
lvi.IsSelected = false;
}
else if (i == index)
{
lvi.IsSelected = true;
}
}
}
But there might be a more elegant solution than this.
I'm setting parameters for a report that I'm showing in ReportViewer control, and the parameters are setting properly and the report is running with the proper parameters, however the actual controls that provide the report criteria at the top of the ReportViewer are not selected. Why aren't the proper items selected in the criteria, even when the report is properly running with the criteria that I set?
ReportParameter month = new ReportParameter("month", "September 2011");
SsrsReportInfo reportInfo = new SsrsReportInfo("Summary", "http://server/ReportServer/", "/MyFolder/Summary", month);
this.reportViewer1.ServerReport.ReportPath = reportInfo.ReportPath;
this.reportViewer1.ServerReport.ReportServerUrl = new Uri(reportInfo.ReportServerUrl);
if (reportInfo.Parameters != null)
{
this.reportViewer1.ServerReport.SetParameters(reportInfo.Parameters);
}
this.reportViewer1.RefreshReport();
Here's the code for the reportInfo class:
/// <summary>
/// SSRS report information for report viewer.
/// </summary>
public class SsrsReportInfo
{
/// <summary>
/// Initializes a new instance of the <see cref="SsrsReportInfo"/> class.
/// </summary>
/// <param name="reportName">Name of the report.</param>
/// <param name="reportServerUrl">The report server URL.</param>
/// <param name="reportPath">The report path.</param>
public SsrsReportInfo(string reportName, string reportServerUrl, string reportPath)
: this(reportName, reportServerUrl, reportPath, null)
{
}
/// <summary>
/// Initializes a new instance of the <see cref="SsrsReportInfo"/> class.
/// </summary>
/// <param name="reportName">Name of the report.</param>
/// <param name="reportServerUrl">The report server URL.</param>
/// <param name="reportPath">The report path.</param>
/// <param name="reportParameters">The report parameters.</param>
public SsrsReportInfo(string reportName, string reportServerUrl, string reportPath, params ReportParameter[] reportParameters)
{
this.ReportName = reportName;
this.ReportServerUrl = reportServerUrl;
this.ReportPath = reportPath;
this.Parameters = reportParameters;
}
/// <summary>
/// Gets or sets the name of the report.
/// </summary>
/// <value>The name of the report.</value>
public string ReportName
{
get;
set;
}
/// <summary>
/// Gets or sets the report server URL.
/// </summary>
/// <value>The report server URL.</value>
public string ReportServerUrl
{
get;
set;
}
/// <summary>
/// Gets or sets the report path.
/// </summary>
/// <value>The report path.</value>
public string ReportPath
{
get;
set;
}
/// <summary>
/// Gets or sets the parameters.
/// </summary>
/// <value>The parameters.</value>
public ReportParameter[] Parameters
{
get;
set;
}
}
Thanks,
Mark
I've figured out the issue here. I had the code setting the path, url, parameters and refreshing the report within a Form constructor. I moved it into the Form.Load event and it works fine now. Reports still run properly, but now the parameters are properly set also within the criteria portion on the top of the ReportViewer.
I had the same usage shown here: http://technet.microsoft.com/es-es/library/aa337089(SQL.90).aspx, however noticed that they did it in the Form.Load event, and I tried that, and it worked. I could probably do it in the ReportViewer.Load event also, the reason for this is probably that the controls aren't drawn up on screen yet before setting the values.
What cool functionality and methods do you add to your ASP.net BasePage : System.Web.UI.Page classes?
Examples
Here's something I use for authentication, and I'd like to hear your opinions on this:
protected override void OnPreInit(EventArgs e)
{
base.OnPreInit(e);
// Authentication code omitted... Essentially same as below.
if (_RequiresAuthentication && !(IsAuthorized))
{
RespondForbidden("You do not have permissions to view this page.", UnauthorizedRedirect);
return;
}
}
// This function is overridden in each page subclass and fitted to each page's
// own authorization requirements.
// This also allows cascading authorization checks,
// e.g: User has permission to view page? No - base.IsAuthorized - Is user an admin?
protected virtual bool IsAuthorized
{
get { return true; }
}
My BasePage class contains an instance of this class:
public class StatusCodeResponse {
public StatusCodeResponse(HttpContext context) {
this._context = context;
}
/// <summary>
/// Responds with a specified status code, and if specified - transfers to a page.
/// </summary>
private void RespondStatusCode(HttpContext context, System.Net.HttpStatusCode status, string message, string transfer)
{
if (string.IsNullOrEmpty(transfer))
{
throw new HttpException((int)status, message);
}
context.Response.StatusCode = (int)status;
context.Response.StatusDescription = message;
context.Server.Transfer(transfer);
}
public void RespondForbidden(string message, string transfer)
{
RespondStatusCode(this._context, System.Net.HttpStatusCode.Forbidden, message, transfer);
}
// And a few more like these...
}
As a side note, this could be accomplished using extension methods for the HttpResponse object.
And another method I find quite handy for parsing querystring int arguments:
public bool ParseId(string field, out int result)
{
return (int.TryParse(Request.QueryString[field], out result) && result > 0);
}
Session related stuff, some complex object in the BasePage that maps to a session, and expose it as a property.
Doing stuff like filling a crumble pad object.
But most important: do not make your basepage into some helper class. Don't add stuff like ParseId(), that's just ridiculous.
Also, based on the first post: make stuff like IsAuthorized abstract. This way you don't create giant security holes if someone forgets that there is some virtual method.
using System;
using System.Collections.Generic;
using System.Text.RegularExpressions;
using System.Web;
using System.Web.UI;
using System.Web.UI.HtmlControls;
using System.Web.UI.WebControls;
namespace MySite
{
/// <summary>
/// Base class with properties for meta tags for content pages
/// http://www.codeproject.com/KB/aspnet/PageTags.aspx
/// http://weblogs.asp.net/scottgu/archive/2005/08/02/421405.aspx
/// </summary>
public partial class BasePage : System.Web.UI.Page
{
private string keywords;
private string description;
/// <SUMMARY>
/// Gets or sets the Meta Keywords tag for the page
/// </SUMMARY>
public string Meta_Keywords
{
get
{
return keywords;
}
set
{
// Strip out any excessive white-space, newlines and linefeeds
keywords = Regex.Replace(value, "\\s+", " ");
}
}
/// <SUMMARY>
/// Gets or sets the Meta Description tag for the page
/// </SUMMARY>
public string Meta_Description
{
get
{
return description;
}
set
{
// Strip out any excessive white-space, newlines and linefeeds
description = Regex.Replace(value, "\\s+", " ");
}
}
// Constructor
// Add an event handler to Init event for the control
// so we can execute code when a server control (page)
// that inherits from this base class is initialized.
public BasePage()
{
Init += new EventHandler(BasePage_Init);
}
// Whenever a page that uses this base class is initialized,
// add meta keywords and descriptions if available
void BasePage_Init(object sender, EventArgs e)
{
if (!String.IsNullOrEmpty(Meta_Keywords))
{
HtmlMeta tag = new HtmlMeta();
tag.Name = "keywords";
tag.Content = Meta_Keywords;
Header.Controls.Add(tag);
}
if (!String.IsNullOrEmpty(Meta_Description))
{
HtmlMeta tag = new HtmlMeta();
tag.Name = "description";
tag.Content = Meta_Description;
Header.Controls.Add(tag);
}
}
}
}
Along with the metadata already mentioned (mostly obsolete in ASP.NET 4.0 with the new Page.MetaDescription and Page.MetaKeywords properties), I've also had methods to add other header links to my page such as specific ones for adding page specific CSS, or things like cannonical links, RSS links, etc:
/// <overloads>
/// Adds a CSS link to the page. Useful when you don't have access to the
/// HeadContent ContentPlaceHolder. This method has 4 overloads.
/// </overloads>
/// <summary>
/// Adds a CSS link.
/// </summary>
/// <param name="pathToCss">The path to CSS file.</param>
public void AddCss(string pathToCss) {
AddCss(pathToCss, string.Empty);
}
/// <summary>
/// Adds a CSS link in a specific position.
/// </summary>
/// <param name="pathToCss">The path to CSS.</param>
/// <param name="position">The postion.</param>
public void AddCss(string pathToCss, int? position) {
AddCss(pathToCss, string.Empty, position);
}
/// <summary>
/// Adds a CSS link to the page with a specific media type.
/// </summary>
/// <param name="pathToCss">The path to CSS file.</param>
/// <param name="media">The media type this stylesheet relates to.</param>
public void AddCss(string pathToCss, string media) {
AddHeaderLink(pathToCss, "text/css", "Stylesheet", media, null);
}
/// <summary>
/// Adds a CSS link to the page with a specific media type in a specific
/// position.
/// </summary>
/// <param name="pathToCss">The path to CSS.</param>
/// <param name="media">The media.</param>
/// <param name="position">The postion.</param>
public void AddCss(string pathToCss, string media, int? position) {
AddHeaderLink(pathToCss, "text/css", "Stylesheet", media, position);
}
/// <overloads>
/// Adds a general header link. Useful when you don't have access to the
/// HeadContent ContentPlaceHolder. This method has 3 overloads.
/// </overloads>
/// <summary>
/// Adds a general header link.
/// </summary>
/// <param name="href">The path to the resource.</param>
/// <param name="type">The type of the resource.</param>
public void AddHeaderLink(string href, string type) {
AddHeaderLink(href, type, string.Empty, string.Empty, null);
}
/// <summary>
/// Adds a general header link.
/// </summary>
/// <param name="href">The path to the resource.</param>
/// <param name="type">The type of the resource.</param>
/// <param name="rel">The relation of the resource to the page.</param>
public void AddHeaderLink(string href, string type, string rel) {
AddHeaderLink(href, type, rel, string.Empty, null);
}
/// <summary>
/// Adds a general header link.
/// </summary>
/// <param name="href">The path to the resource.</param>
/// <param name="type">The type of the resource.</param>
/// <param name="rel">The relation of the resource to the page.</param>
/// <param name="media">The media target of the link.</param>
public void AddHeaderLink(string href, string type, string rel, string media)
{
AddHeaderLink(href, type, rel, media, null);
}
/// <summary>
/// Adds a general header link.
/// </summary>
/// <param name="href">The path to the resource.</param>
/// <param name="type">The type of the resource.</param>
/// <param name="rel">The relation of the resource to the page.</param>
/// <param name="media">The media target of the link.</param>
/// <param name="position">The postion in the control order - leave as null
/// to append to the end.</param>
public void AddHeaderLink(string href, string type, string rel, string media,
int? position) {
var link = new HtmlLink { Href = href };
if (0 != type.Length) {
link.Attributes.Add(HtmlTextWriterAttribute.Type.ToString().ToLower(),
type);
}
if (0 != rel.Length) {
link.Attributes.Add(HtmlTextWriterAttribute.Rel.ToString().ToLower(),
rel);
}
if (0 != media.Length) {
link.Attributes.Add("media", media);
}
if (null == position || -1 == position) {
Page.Header.Controls.Add(link);
}
else
{
Page.Header.Controls.AddAt((int)position, link);
}
}
Culture initialization by overriding InitializeCulture() method (set culture and ui culture from cookie or DB).
Some of my applications are brandable, then here I do some "branding" stuff too.
I use this methot and thanks for yours,
/// <summary>
/// Displays the alert.
/// </summary>
/// <param name="message">The message to display.</param>
protected virtual void DisplayAlert(string message)
{
ClientScript.RegisterStartupScript(
GetType(),
Guid.NewGuid().ToString(),
string.Format("alert('{0}');", message.Replace("'", #"\'")),
true
);
}
/// <summary>
/// Finds the control recursive.
/// </summary>
/// <param name="id">The id.</param>
/// <returns>control</returns>
protected virtual Control FindControlRecursive(string id)
{
return FindControlRecursive(id, this);
}
/// <summary>
/// Finds the control recursive.
/// </summary>
/// <param name="id">The id.</param>
/// <param name="parent">The parent.</param>
/// <returns>control</returns>
protected virtual Control FindControlRecursive(string id, Control parent)
{
if (string.Compare(parent.ID, id, true) == 0)
return parent;
foreach (Control child in parent.Controls)
{
Control match = FindControlRecursive(id, child);
if (match != null)
return match;
}
return null;
}
Putting authorization code in a base page is generally not a good idea. The problem is, what happens if you forget to derive a page that needs authorization from the base page? You will have a security hole.
It's much better to use an HttpModule, so that you can intercept requests for all pages, and make sure users are authorized even before the HttpHandler has a chance to run.
Also, as others have said, and in keeping with OO principles, it's better to only have methods in your base page that actually relate to the Page itself. If they don't reference "this," they should probably be in a helper class -- or perhaps be extension methods.
I inherit from System.Web.UI.Page when I need certain properties and every page. This is good for aweb application that implements a login. In the membership pages I use my own base class to get access to Properties like UserID, UserName etc. These properties wrap Session Variables
Here are some examples (sans code) that I use a custom base class for:
Adding a page filter (e.g. replace "{theme}" with "~/App_Theme/[currentTheme]"),
Adding a Property and handling for Auto Titling pages based upon Site Map,
Registering specialized logging (could probably be redone via different means),
Adding methods for generalized input(Form/Querystring) validation, with blanket redirector: AddRequiredInput("WidgetID", PageInputType.QueryString, typeof(string)),
Site Map Helpers, allowing for things like changing a static "Edit Class" into something context related like "Edit Fall '10 Science 101"
ViewState Helpers, allowing me to register variable on the page to a name and have it automatically populate that variable from the viewstate or a default, and save the value back out to the viewstate at the end of the request.
Custom 404 Redirector, where I can pass an exception or message (or both) and it will go to a page I have predefined to nicely display and log it.
I personally like #5 the most because a) updating the SiteMap is ugly and I prefer not to have clutter the page, making it more readable, b) It makes the SiteMap much more user friendly.
,
Please refer Getting page specific info in ASP.Net Base Page
public abstract string AppSettingsRolesName { get; }
List<string> authorizedRoles = new List<string>((ConfigurationManager.AppSettings[AppSettingsRolesName]).Split(','))
if (!authorizedRoles.Contains(userRole))
{
Response.Redirect("UnauthorizedPage.aspx");
}
In derived Page
public override string AppSettingsRolesName
{
get { return "LogsScreenRoles"; }
}