I'm trying to set some caching HTTP headers for all requests that return an "HTML page".
I could do that in some global place such as BeginRequest or in a global MVC filter (as suggested in that question).
For that to work I must differentiate HTML pages from other requests that are supposed to be cached (mostly resources I think but I'm not sure).
I am unsure how to define "HTML page" in a rigorous way so that I could put it into an algorithm. How could I detect such requests?
I really do not want to mark up all MVC actions that I write individually. That's tedious and I could forget something.
I think the best way is with filters, then I'd use a regex to check if the result has HTML:
Create a filter:
public class NoCacheForHTMLResult : IActionFilter
{
public void OnActionExecuted(ActionExecutedContext filterContext)
{
Regex tagRegex = new Regex(#"<[^>]+>");
string response = filterContext.RequestContext.HttpContext.Response.Output.ToString();
bool hasHtml = tagRegex.IsMatch(response);
if (hasHtml)
{
filterContext.HttpContext.Response.Cache.SetExpires(DateTime.UtcNow.AddDays(-1));
filterContext.HttpContext.Response.Cache.SetValidUntilExpires(false);
filterContext.HttpContext.Response.Cache.SetRevalidation(HttpCacheRevalidation.AllCaches);
filterContext.HttpContext.Response.Cache.SetCacheability(HttpCacheability.NoCache);
filterContext.HttpContext.Response.Cache.SetNoStore();
}
}
public void OnActionExecuting(ActionExecutingContext filterContext)
{
}
}
...and add in the Global Filters (inside your App_Start):
public class FilterConfig
{
public static void RegisterGlobalFilters(GlobalFilterCollection filters)
{
filters.Add(new NoCacheForHTMLResult());
}
}
I'm not a regex expert, I got this regex from this question, but it looks we need to improve this, by checking if there is at least an HTML tag, otherwise if the result was XML, it looks the regex would match (though you might want to set no cache for XML as well, it's up to you).
The code for disabling cache I got in this answer, you might have a different code.
Hope this helps you.
Related
New to WebFlux, reactive, and handlers.
I am able to get a Mono<> from a ServerRequest and process the contained POJO to add a new tuple to a database. But, it seems like there should be a "better" or "more accepted" way to write this code.
Any help/input with the code in AccountRequestHandler would be appreciated, especially with explanations of the rationale behind the recommend change(s).
Router implementation (stripped down to only "POST")...
#Configuration
public class AccountRequestRouter {
#Bean
public RouterFunction<ServerResponse> route(AccountRequestHandler requestHandler) {
return nest(path("/v2"),
nest(accept(APPLICATION_JSON),
.andRoute(RequestPredicates.POST("/accounts"), requestHandler::addAccount)
));
}
}
Handler implementation...
The code where I'm actually doing the add, and then separately creating a ServerResponse, is what I'm focused on. It seems "clunky", especially since AccountService.addAccount() returns a Mono on completion.
#Component
public class AccountRequestHandler {
#Autowired
private mil.navy.ccop.service.accounts.account.AccountService accountService;
public Mono<ServerResponse> addAccount(ServerRequest request) {
return request.bodyToMono(Account.class).flatMap(account -> {
accountService.addAccount(account);
return ServerResponse.ok().build();
})
.switchIfEmpty(ServerResponse.badRequest()
.contentType(APPLICATION_JSON)
.build(Mono.empty()));
}
}
AccountService implementation (again, stripped down)...
#Service
class AccountService {
#Autowired
private AccountRepository accounts;
public AccountService() {
}
public Mono<Void> addAccount(Account account) {
Account proxy;
// make sure that accountId is set to support auto-generation of synthetic key value
proxy = new Account(-1, account.getShortName(), account.getLongName(), account.getDescription());
accounts.save(proxy);
return Mono.empty();
}
}
Appreciating all the help in ramping up on this style of programming....
well first of all, you have 2 addAccount, that can be a bit confusing.
Second of all, what kind of "repository" are you writing too? if its an sql repo you need to properly wrap it in a Mono.fromCallable() otherwise it will block the Reactive thread pool and you can have really bad performance.
Yes there are other ways of doing things. A lot of people tend to do things in flatmap or map and sure it is completely possible to do things here, but for the semantics i'd say it is less good.
map and flatmap are usually used to perform some sort of computation on the inner value of the mono and then return the same or a new value and or type inside the mono.
i would rewrite this like such.
return void here:
public void addAccount(Account account) {
Account proxy;
// make sure that accountId is set to support auto-generation of synthetic key value
proxy = new Account(-1, account.getShortName(), account.getLongName(), account.getDescription());
accounts.save(proxy);
}
And here:
public Mono<ServerResponse> addAccount(ServerRequest request) {
return request.bodyToMono(Account.class)
.doOnSuccess(account -> {
accountService.addAccount(account);
}).then(ServerResponse.ok().build())
.switchIfEmpty(ServerResponse.badRequest()
.contentType(APPLICATION_JSON)
.build());
}
there are a number of different doOn methods that are ment to be used to consume and do "side effects" on things. Like doOnSuccess, doOnError, doOnCancel etc. etc.
you also have then and thenReturn which will just return whatever you put in them. Then returns whatever Mono you put in it. thenReturn wraps whatever value you put into it into a Mono and returns it.
I'm trying to add custom http headers to a webview client (for authorization).
It seems to work in some cases, I'am able to login to a webpage without entering username and password, and I get redirected to another page. But when the page is calling other resources to get elements populated with data an error is thrown and OnReceivedHttpError is invoked. The error I'm getting is 401 unauthorized and when i look through the headers on the IWebResourceRequest i can't see the authorization headers at all.
Am I missing something or have anyone had same problems ?
Using Xamarin Forms 2.3.3.180 and targeting API 21 (Android 5.0 Lollipop), compile with Android 7.1 Nougat.
I've tried in postman to add headers to request and it works perfectly.
Renderer:
public class MyWebViewRenderer : WebViewRenderer
{
private MyWebViewClient _webViewClient;
protected override void OnElementChanged(ElementChangedEventArgs<Xamarin.Forms.WebView> e)
{
base.OnElementChanged(e);
if(_webViewClient == null)
_webViewClient = new MyWebViewClient();
Control.SetWebViewClient(_webViewClient);
Control.LongClickable = false;
Control.HapticFeedbackEnabled = false;
Control.Settings.JavaScriptEnabled = true;
var data = Encoding.GetEncoding("ISO-8859-1").GetBytes("username:password");
var base64string = Base64.EncodeToString(data, Base64Flags.NoWrap);
var headers = new Dictionary<string, string>();
headers.Add("Authorization", $"Basic {base64string}")
Control.LoadUrl(Control.Url, headers);
}
}
WebViewClient:
public override bool ShouldOverrideUrlLoading(WebView view, string url)
{
WebView.SetWebContentsDebuggingEnabled(true);
var data = Encoding.GetEncoding("ISO-8859-1").GetBytes("username:password");
var base64string = Base64.EncodeToString(data, Base64Flags.NoWrap);
var headers = new Dictionary<string, string>();
headers.Add("Authorization", $"Basic {base64string}")
view.LoadUrl(url, headers);
return true;
}
public override WebResourceResponse ShouldInterceptRequest(WebView view, IWebResourceRequest urlResource)
{
//headers does not always contains authorization header, so let's add it.
if (!urlResource.RequestHeaders.ContainsKey("authorization") && !urlResource.RequestHeaders.ContainsKey("Authorization"))
{
var data = Encoding.GetEncoding("ISO-8859-1").GetBytes("username:password");
var base64string = Base64.EncodeToString(data, Base64Flags.NoWrap);
urlResource.RequestHeaders.Add("Authorization", $"{base64string}");
}
return base.ShouldInterceptRequest(view, urlResource);
}
public override void OnReceivedHttpError(WebView view, IWebResourceRequest request, WebResourceResponse errorResponse)
{
base.OnReceivedHttpError(view, request, errorResponse);
}
If you only need the headers on the get requests, the code below will work. However POST requests are a different issue. I needed to do a similar thing (with all requests, not just GET), and all I can say is that there's not straightforward solution, at least not one that I've found (and I've tried everything short of writing my own network driver). I've tried so many methods (ShouldOverrideUrlLoading, ShouldInterceptRequest, custom LoadUrl and PostUrl etc.) and none of them give a 100% solution. There is a lot of misinformation about this so I think some clarification is needed since I've spent two days on this without success.
So here's what I've learned:
If you only need the headers in the GET requests, that's trivial. Simply create an implementation of WebViewClient and override ShouldOverrideUrlLoading like this:
[assembly: ExportRenderer(typeof(Xamarin.Forms.WebView), typeof(App.Android.HybridWebViewRenderer))]
namespace App.Android
{
public class HybridWebViewRenderer : WebViewRenderer
{
public HybridWebViewRenderer(Context context) : base(context)
{
}
protected override void OnElementChanged(ElementChangedEventArgs<Xamarin.Forms.WebView> e)
{
base.OnElementChanged(e);
Control.SetWebViewClient(new CustomWebViewClient());
}
}
public class CustomWebViewClient : WebViewClient
{
public override bool ShouldOverrideUrlLoading(Android.Webkit.WebView view, string url)
{
Dictionary<string, string> headers = new Dictionary<string, string>
{
["Name"] = "value"
};
view.LoadUrl(url, headers);
return true;
}
}
}
If, however, you need the headers in other requests (specifically POST requests) there really isn't a perfect solution. Many answers tell you to override ShouldInterceptRequest but this is unlikely to help. ShouldInterceptRequest provides an IWebResourceRequest which contains the URL of the request, the method (i.e. POST) and the headers. There are answers out there which state that adding the headers by doing request.Headers.Add("Name", "Value") is a viable solution but this is wrong. The IWebResourceRequest is not used by the WebView's internal logic so modifying it is useless!
You can write your own HTTP client in ShouldInterceptRequest which includes your own headers to perform the requests and return a WebResourceResponse object. Again, this works for GET requests, but the problem with this is that even though we can intercept a POST request, we cannot determine the content in the request as the request content is not included in the IWebResourceRequest object. As a result, we cannot accurately perform the request manually. So, unless the content of the POST request is unimportant or can somehow be fetched, this method is not viable.
An additional note on this method: returning null tells the WebView to handle the request for us. In other words 'I don't want to intercept the request'. If the return is not null however, the WebView will display whatever is in the WebResourceResponse object.
I also tried overriding the PostUrl and LoadUrl methods in the WebView itself. These methods are not called by the internal logic, so unless you are calling them yourself, this does not work.
So what can be done? There are a few hacky solutions (see github.com/KeejOow/android-post-webview) to get around this problem, but they rely on javascript and are not suitable in all cases (I have read that they don't work with forms). If you want to use them in Xamarin, you're going to need to adapt the code for C# anyway, and there is no guarantee that it will solve your problem.
I'm writing this so no one else has to waste countless hours finding a solution that doesn't really exist.
If only the Android devs had decided to include the POST content in the IWebResourceRequest object...
And apologies for the length, if you've read to this point, you're probably as desperate as I was.
I'm using Elmah for logging exceptions on my MVC application using Alex Beletsky's elmah-mvc NuGet package.
The application registers some global filters, applied on each action called.
Is there a way to prevent some of those filters from being applied when calling the Elmah.Mvc.ElmahController error log page (foo.com/elmah) ?
A test like below works, of course, but I'm looking for a more elegant way that would not involve modifying the filter (nor the source code from Elmah / Elmah MVC). Is it even possible ?
public class FooAttribute : FilterAttribute, IActionFilter
{
// ...
public void OnActionExecuting(ActionExecutingContext filterContext)
{
if (filterContext.Controller is ElmahController)
{
return;
}
// do stuff
}
}
I know that attributes can't be added or removed at runtime.
I thought of wrapping the ElmahController in a new one where I could add an exclusion filter, but I'm not sure how (if possible) to change the web.config to reference this wrapper instead of the original controller.
You could register your global filters through a custom IFilterProvider:
public class MyFilterProvider : IFilterProvider
{
public IEnumerable<Filter> GetFilters(ControllerContext controllerContext, ActionDescriptor actionDescriptor)
{
if (controllerContext.Controller is ElmahController)
{
return Enumerable.Empty<Filter>();
}
return ... the collection of your global filters
}
}
and in your Application_Start instead of calling:
FilterConfig.RegisterGlobalFilters(GlobalFilters.Filters);
you would call:
FilterProviders.Providers.Add(new MyFilterProvider());
I have an ASP.NET MVC 4 app that i'm incorporating an OData API into. This is running the 2012.2 stuff with the larger OData support.
I did not use a separate area for this...that might have been a mistake but my app is small and area seemed overkill.
I've got my controllers setup correctly and an example path to my Segments collection (segments is a type in my domain) is "/odata/Segments". This loads as expected and is working.
On my homepage i'm trying to add a link to this resource using Razor's Html.ActionLink (or RouteLink) but it seems the OData controllers layout doesn't quite work with those methods because the controllers are prefixed with "odata" when registered in WebAPIConfig:
config.Routes.MapODataRoute("OData Route", "odata", model );
I can trick the method to construct the correct url by pretending there's an odata controller when there certainly isn't one (as far as i know) with something like this:
#Html.RouteLink("Segments", "Segments", "odata")
but that seems like a hack.
I don't quite understand the ASP.NET routing plumbing well enough to understand how that prefix passed to MapODataRoute is being incorporated into the MVC chain so that i can use the "right" razor method the "right" way.
just for kicks, here's my SegmentsController:
public class SegmentsController : EntitySetController<Segment, long>
{
private MarketerDB db = new MarketerDB();
// GET api/segments
override public IQueryable<Segment> Get()
{
return db.Segments.AsQueryable();
}
protected override Segment GetEntityByKey(long key)
{
return db.Segments.Find(key);
}
public IQueryable<Affiliate> GetAffiliates([FromODataUri] long key)
{
return this.GetEntityByKey(key).Affiliates.AsQueryable();
}
protected override void Dispose(bool disposing)
{
db.Dispose();
base.Dispose(disposing);
}
}
We have an ODataLink method on System.Web.Http.UrlHelper but we forgot to add one to the MVC System.Web.Mvc.UrlHelper. Till we add it, you can use this extension method,
namespace System.Web.Mvc
{
public static class UrlHelperExtensions
{
private static IODataPathHandler _pathHandler = new DefaultODataPathHandler();
public static string ODataUrl(this UrlHelper urlHelper, string routeName, params ODataPathSegment[] segments)
{
string odataPath = _pathHandler.Link(new ODataPath(segments));
return urlHelper.HttpRouteUrl(
routeName,
new RouteValueDictionary() { { ODataRouteConstants.ODataPath, odataPath } });
}
}
}
and call it from your razor views by doing something like (assuming there is an entityset customers and you want to put the navigation link to orders on customers(42)),
#Url.ODataUrl("odata", new EntitySetPathSegment("customers"), new KeyValuePathSegment("42"), new NavigationPathSegment("orders"))
Make sure you have an #using System.Web.Http.OData.Routing directive in your razor view.
In my applications, I often have to use relative paths. For example, when I reference JQuery, I usually do so like this:
<script type="text/javascript" src="../Scripts/jquery-1.2.6.js"></script>
Now that I'm making the transition to MVC, I need to account for the different paths a page might have, relative to the root. This was of course an issue with URL rewriting in the past, but I managed to work around it by using consistent paths.
I'm aware that the standard solution is to use absolute paths such as:
<script type="text/javascript" src="/Scripts/jquery-1.2.6.js"></script>
but this will not work for me as during the development cycle, I have to deploy to a test machine on which the app will run in a virtual directory. Root relative paths don't work when the root changes. Also, for maintenance reasons, I cannot simply change out all the paths for the duration of deploying the test - that would be a nightmare in itself.
So what's the best solution?
Edit:
Since this question is still receiving views and answers, I thought it might be prudent to update it to note that as of Razor V2, support for root-relative urls is baked in, so you can use
<img src="~/Content/MyImage.jpg">
without any server-side syntax, and the view engine automatically replaces ~/ with whatever the current site root is.
Try this:
<script type="text/javascript" src="<%=Url.Content("~/Scripts/jquery-1.2.6.js")%>"></script>
Or use MvcContrib and do this:
<%=Html.ScriptInclude("~/Content/Script/jquery.1.2.6.js")%>
While an old post, new readers should know that Razor 2 and later (default in MVC4+) completely resolves this problem.
Old MVC3 with Razor 1:
Application home page
New MVC4 with Razor 2 and later:
Application home page
No awkward Razor function-like syntax.
No non-standard markup tags.
Prefixing a path in any HTML attributes with a tilde ('~') tells Razor 2 to "just make it work" by substituting the correct path. It's great.
Breaking change - MVC 5
Watch out for a breaking change change in MVC 5 (from the MVC 5 release notes)
Url Rewrite and Tilde(~)
After upgrading to ASP.NET Razor 3 or ASP.NET MVC 5, the tilde(~)
notation may no longer work correctly if you are using URL rewrites.
The URL rewrite affects the tilde(~) notation in HTML elements such as
<A/>, <SCRIPT/>, <LINK/>, and as a result the tilde no longer maps to
the root directory.
For example, if you rewrite requests for asp.net/content to asp.net,
the href attribute in
<A href="~/content/"/> resolves to
/content/content/ instead of /. To suppress this change, you can set
the IIS_WasUrlRewritten context to false in each Web Page or in
Application_BeginRequest in Global.asax.
They don't actually explain how to do it, but then I found this answer:
If you are running in IIS 7 Integrated Pipeline mode try putting the
following in your Global.asax:
protected void Application_BeginRequest(object sender, EventArgs e)
{
Request.ServerVariables.Remove("IIS_WasUrlRewritten");
}
Note: You may want to check Request.ServerVariables actually contains IIS_WasUrlRewritten first to be sure this is what your problem is.
PS. I thought I had a situation where this was happening to me and I was getting src="~/content/..." URLS generated into my HTML - but it turned out something just wasn't refreshing when my code was being compiled. Editing and resaving the Layout and page cshtml files somehow triggered something to work.
In ASP.NET I usually use <img src='<%= VirtualPathUtility.ToAbsolute("~/images/logo.gif") %>' alt="Our Company Logo"/>.
I don't see why a similar solution shouldn't work in ASP.NET MVC.
<script src="<%=ResolveUrl("~/Scripts/jquery-1.2.6.min.js") %>" type="text/javascript"></script>
Is what I used. Change path to match your example.
For what it's worth, I really hate the idea of littering my app with server tags just to resolve paths, so I did a bit more research and opted to use something I'd tried before for rewriting links - a response filter. In this way, I can prefix all absolute paths with a known prefix and replace it at runtime using the Response.Filter object and not have to worry about unnecessary server tags. The code is posted below in case it will help anyone else.
using System;
using System.IO;
using System.Text;
using System.Text.RegularExpressions;
using System.Web;
namespace Demo
{
public class PathRewriter : Stream
{
Stream filter;
HttpContext context;
object writeLock = new object();
StringBuilder sb = new StringBuilder();
Regex eofTag = new Regex("</html>", RegexOptions.IgnoreCase | RegexOptions.Compiled);
Regex rootTag = new Regex("/_AppRoot_", RegexOptions.IgnoreCase | RegexOptions.Compiled);
public PathRewriter(Stream filter, HttpContext context)
{
this.filter = filter;
this.context = context;
}
public override void Write(byte[] buffer, int offset, int count)
{
string temp;
lock (writeLock)
{
temp = Encoding.UTF8.GetString(buffer, offset, count);
sb.Append(temp);
if (eofTag.IsMatch(temp))
RewritePaths();
}
}
public void RewritePaths()
{
byte[] buffer;
string temp;
string root;
temp = sb.ToString();
root = context.Request.ApplicationPath;
if (root == "/") root = "";
temp = rootTag.Replace(temp, root);
buffer = Encoding.UTF8.GetBytes(temp);
filter.Write(buffer, 0, buffer.Length);
}
public override bool CanRead
{
get { return true; }
}
public override bool CanSeek
{
get { return filter.CanSeek; }
}
public override bool CanWrite
{
get { return true; }
}
public override void Flush()
{
return;
}
public override long Length
{
get { return Encoding.UTF8.GetBytes(sb.ToString()).Length; }
}
public override long Position
{
get { return filter.Position; }
set { filter.Position = value; }
}
public override int Read(byte[] buffer, int offset, int count)
{
return filter.Read(buffer, offset, count);
}
public override long Seek(long offset, SeekOrigin origin)
{
return filter.Seek(offset, origin);
}
public override void SetLength(long value)
{
throw new NotImplementedException();
}
}
public class PathFilterModule : IHttpModule
{
public void Dispose()
{
return;
}
public void Init(HttpApplication context)
{
context.ReleaseRequestState += new EventHandler(context_ReleaseRequestState);
}
void context_ReleaseRequestState(object sender, EventArgs e)
{
HttpApplication app = sender as HttpApplication;
if (app.Response.ContentType == "text/html")
app.Response.Filter = new PathRewriter(app.Response.Filter, app.Context);
}
}
}
The Razor view engine for MVC 3 makes it even easier and cleaner to use virtual-root relative paths that are properly resolved at run-time. Just drop the Url.Content() method into the href attribute value and it will resolve properly.
Application home page
Like Chris, I really can't stand having to put bloated server-side tags inside my clean markup just purely to tell the stupid thing to look from the root upwards. That should be a very simple, reasonable thing to ask for. But I also hate the idea of having to go to the effort of writing any custom C# classes to do such a simple thing, why should I have to? What a waste of time.
For me, I simply compromised on "perfection" and hardcoded the virtual directory's root path name inside my path references. So like this:
<script type="text/javascript" src="/MyProject/Scripts/jquery-1.2.6.js"></script>
No server-side processing or C# code required to resolve the URL, which is best for performance although I know it would be negligible regardless. And no bloated ugly server-side chaos in my nice clean markup.
I'll just have to live with knowing that this is hardcoded and will need to be removed when the thing migrates to a proper domain instead of http://MyDevServer/MyProject/
Cheers
I use a simple helper method. You can easily use it in the Views and Controllers.
Markup:
About Us
Helper method:
public static string Root()
{
if (HttpContext.Current.Request.Url.Host == "localhost")
{
return "";
}
else
{
return "/productionroot";
}
}