We Keep Coding

How to use implicit operators in class for passing parameter in web api c#

I have Web API which throws an error if one of the parameters passed in a URL query is null. Let's say I cant overwrite anything to the above validation.
So that, I want to have a Web API as below, the parameter c is string and the API call may pass an empty string which instantiates parameter c as null and not "".
[Route("{id}")]
[HttpGet]
public IdDTO GetIdDetails(
int Id,
[FromUri] int a,
[FromUri] string b,
[FromUri] string c,
{
//doing something here...
}
A typical API call is localhost:3000/123?a=1&b=abc&c=
I also changed the data type of c to a custom datatype ReqString ([FromUri] ReqString <string> c) as below
public struct ReqString<T>
{
private T _value;
public ReqString(T s)
{
_value = s;
}
public static implicit operator ReqString<T>(T s) => new ReqString<T>(s);
public static implicit operator string(ReqString<T> s)
{
if(typeof(T).Equals(typeof(string)))
{
return s._value as string == null ? "" : s._value as string;
}
else
{
return s._value.ToString();
}
}
}
The problem now is, the value for c is "". But if I pass values for parameter c in API URL query, it is still "" and not the value passed
for eg: localhost:3000/123?a=1&b=abc&c=def
ie, the _value is still null
So my question is How can I use implicit operators to instantiate _value with query value?
UPDATE
The custom datatype works when instantiated in a statement as below, but I want to get similar result in a function call parameter
ReqString rStr = "testing";

As per Michael comment, I took a look into ModalBinder. I was able to resolve my issue. Below is the sample solution I used
public class ReqStringModelBinder : IModelBinder
{
public bool BindModel(HttpActionContext actionContext, ModelBindingContext bindingContext)
{
if (bindingContext.ModelType != typeof(ReqString))
{
return false;
}
ValueProviderResult val = bindingContext.ValueProvider.GetValue(
bindingContext.ModelName);
if (val == null)
{
return false;
}
string key = val.RawValue as string;
if (key == null)
{
bindingContext.ModelState.AddModelError(
bindingContext.ModelName, "Wrong value type");
return false;
}
ReqString result;
if (ReqString.TryParse(key, out result))
{
bindingContext.Model = result;
return true;
}
bindingContext.ModelState.AddModelError(
bindingContext.ModelName, "Cannot convert value to ReqString");
return false;
}
}
[ModelBinder(typeof(ReqStringModelBinder))]
public class ReqString
{
public string value { get; set; }
public static bool TryParse(string s, out ReqString result)
{
result = null;
result = new ReqString() { value = String.IsNullOrEmpty(s) ? "" : s };
return true;
}
public static implicit operator string(ReqString reqString) => reqString.value;
}

Httpcontext.current.session becoming null

Httpcontext.current.session becoming null in local code, session not expired still in active and I didn't clear session manually when click on button controller will receive post request and after authenticating user setting user details in httpcontext.current.session, after setting if I send another post request all values are becoming null
internal class AspDotNet : Base
{
internal AspDotNet(String key)
: base(key)
{
}
internal override void Set(Object value)
{
this.Value = value;
HttpContext.Current.Session[base.Key] = value;
}
internal override Object Get()
{
this.Value = HttpContext.Current.Session[base.Key];
return this.Value;
}
internal override void Remove()
{
HttpContext.Current.Session.Remove(base.Key);
}
}
}

can you try like this. I believe static class will work for you.
public static class SessionPersiter
{
public static int? UserID
{
get
{
if (HttpContext.Current == null)
return (Int32?)null;
var sessionData = HttpContext.Current.Session[CommonConstants.UserIDSessionKey];
if (sessionData == null)
return (Int32?)null;
return (int)sessionData;
}
set
{
HttpContext.Current.Session[CommonConstants.UserIDSessionKey] = value;
}
}
public static UserDTO AuthenticatedUser
{
get
{
if (HttpContext.Current == null)
return null;
var sessionData = HttpContext.Current.Session[CommonConstants.AuthenticatedUserKey];
if (sessionData == null)
return null;
return (UserDTO)sessionData;
}
set
{
HttpContext.Current.Session[CommonConstants.AuthenticatedUserKey] = value;
}
}
public static long? LastLoginID
{
get
{
if (HttpContext.Current == null)
return (long?)null;
var sessionData = HttpContext.Current.Session[CommonConstants.LastLoginIDSessionKey];
if (sessionData == null)
return (long?)null;
return (long)sessionData;
}
set
{
HttpContext.Current.Session[CommonConstants.LastLoginIDSessionKey] = value;
}
}
}
and call like this
var userToken = SessionPersiter.AuthenticatedUser.Token;

Call WebApi action with a list of parameters [duplicate]

I have an ASP.NET Web API (version 4) REST service where I need to pass an array of integers.
Here is my action method:
public IEnumerable<Category> GetCategories(int[] categoryIds){
// code to retrieve categories from database
}
And this is the URL that I have tried:
/Categories?categoryids=1,2,3,4

You just need to add [FromUri] before parameter, looks like:
GetCategories([FromUri] int[] categoryIds)
And send request:
/Categories?categoryids=1&categoryids=2&categoryids=3

As Filip W points out, you might have to resort to a custom model binder like this (modified to bind to actual type of param):
public IEnumerable<Category> GetCategories([ModelBinder(typeof(CommaDelimitedArrayModelBinder))]long[] categoryIds)
{
// do your thing
}
public class CommaDelimitedArrayModelBinder : IModelBinder
{
public bool BindModel(HttpActionContext actionContext, ModelBindingContext bindingContext)
{
var key = bindingContext.ModelName;
var val = bindingContext.ValueProvider.GetValue(key);
if (val != null)
{
var s = val.AttemptedValue;
if (s != null)
{
var elementType = bindingContext.ModelType.GetElementType();
var converter = TypeDescriptor.GetConverter(elementType);
var values = Array.ConvertAll(s.Split(new[] { ","},StringSplitOptions.RemoveEmptyEntries),
x => { return converter.ConvertFromString(x != null ? x.Trim() : x); });
var typedValues = Array.CreateInstance(elementType, values.Length);
values.CopyTo(typedValues, 0);
bindingContext.Model = typedValues;
}
else
{
// change this line to null if you prefer nulls to empty arrays
bindingContext.Model = Array.CreateInstance(bindingContext.ModelType.GetElementType(), 0);
}
return true;
}
return false;
}
}
And then you can say:
/Categories?categoryids=1,2,3,4 and ASP.NET Web API will correctly bind your categoryIds array.

I recently came across this requirement myself, and I decided to implement an ActionFilter to handle this.
public class ArrayInputAttribute : ActionFilterAttribute
{
private readonly string _parameterName;
public ArrayInputAttribute(string parameterName)
{
_parameterName = parameterName;
Separator = ',';
}
public override void OnActionExecuting(HttpActionContext actionContext)
{
if (actionContext.ActionArguments.ContainsKey(_parameterName))
{
string parameters = string.Empty;
if (actionContext.ControllerContext.RouteData.Values.ContainsKey(_parameterName))
parameters = (string) actionContext.ControllerContext.RouteData.Values[_parameterName];
else if (actionContext.ControllerContext.Request.RequestUri.ParseQueryString()[_parameterName] != null)
parameters = actionContext.ControllerContext.Request.RequestUri.ParseQueryString()[_parameterName];
actionContext.ActionArguments[_parameterName] = parameters.Split(Separator).Select(int.Parse).ToArray();
}
}
public char Separator { get; set; }
}
I am applying it like so (note that I used 'id', not 'ids', as that is how it is specified in my route):
[ArrayInput("id", Separator = ';')]
public IEnumerable<Measure> Get(int[] id)
{
return id.Select(i => GetData(i));
}
And the public url would be:
/api/Data/1;2;3;4
You may have to refactor this to meet your specific needs.

In case someone would need - to achieve same or similar thing(like delete) via POST instead of FromUri, use FromBody and on client side(JS/jQuery) format param as $.param({ '': categoryids }, true)
c#:
public IHttpActionResult Remove([FromBody] int[] categoryIds)
jQuery:
$.ajax({
type: 'POST',
data: $.param({ '': categoryids }, true),
url: url,
//...
});
The thing with $.param({ '': categoryids }, true) is that it .net will expect post body to contain urlencoded value like =1&=2&=3 without parameter name, and without brackets.

Easy way to send array params to web api
API
public IEnumerable<Category> GetCategories([FromUri]int[] categoryIds){
// code to retrieve categories from database
}
Jquery : send JSON object as request params
$.get('api/categories/GetCategories',{categoryIds:[1,2,3,4]}).done(function(response){
console.log(response);
//success response
});
It will generate your request URL like
../api/categories/GetCategories?categoryIds=1&categoryIds=2&categoryIds=3&categoryIds=4

You may try this code for you to take comma separated values / an array of values to get back a JSON from webAPI
public class CategoryController : ApiController
{
public List<Category> Get(String categoryIDs)
{
List<Category> categoryRepo = new List<Category>();
String[] idRepo = categoryIDs.Split(',');
foreach (var id in idRepo)
{
categoryRepo.Add(new Category()
{
CategoryID = id,
CategoryName = String.Format("Category_{0}", id)
});
}
return categoryRepo;
}
}
public class Category
{
public String CategoryID { get; set; }
public String CategoryName { get; set; }
}
Output :
[
{"CategoryID":"4","CategoryName":"Category_4"},
{"CategoryID":"5","CategoryName":"Category_5"},
{"CategoryID":"3","CategoryName":"Category_3"}
]

ASP.NET Core 2.0 Solution (Swagger Ready)
Input
DELETE /api/items/1,2
DELETE /api/items/1
Code
Write the provider (how MVC knows what binder to use)
public class CustomBinderProvider : IModelBinderProvider
{
public IModelBinder GetBinder(ModelBinderProviderContext context)
{
if (context == null)
{
throw new ArgumentNullException(nameof(context));
}
if (context.Metadata.ModelType == typeof(int[]) || context.Metadata.ModelType == typeof(List<int>))
{
return new BinderTypeModelBinder(typeof(CommaDelimitedArrayParameterBinder));
}
return null;
}
}
Write the actual binder (access all sorts of info about the request, action, models, types, whatever)
public class CommaDelimitedArrayParameterBinder : IModelBinder
{
public Task BindModelAsync(ModelBindingContext bindingContext)
{
var value = bindingContext.ActionContext.RouteData.Values[bindingContext.FieldName] as string;
// Check if the argument value is null or empty
if (string.IsNullOrEmpty(value))
{
return Task.CompletedTask;
}
var ints = value?.Split(',').Select(int.Parse).ToArray();
bindingContext.Result = ModelBindingResult.Success(ints);
if(bindingContext.ModelType == typeof(List<int>))
{
bindingContext.Result = ModelBindingResult.Success(ints.ToList());
}
return Task.CompletedTask;
}
}
Register it with MVC
services.AddMvc(options =>
{
// add custom binder to beginning of collection
options.ModelBinderProviders.Insert(0, new CustomBinderProvider());
});
Sample usage with a well documented controller for Swagger
/// <summary>
/// Deletes a list of items.
/// </summary>
/// <param name="itemIds">The list of unique identifiers for the items.</param>
/// <returns>The deleted item.</returns>
/// <response code="201">The item was successfully deleted.</response>
/// <response code="400">The item is invalid.</response>
[HttpDelete("{itemIds}", Name = ItemControllerRoute.DeleteItems)]
[ProducesResponseType(typeof(void), StatusCodes.Status204NoContent)]
[ProducesResponseType(typeof(void), StatusCodes.Status404NotFound)]
public async Task Delete(List<int> itemIds)
=> await _itemAppService.RemoveRangeAsync(itemIds);
EDIT: Microsoft recommends using a TypeConverter for these kids of operations over this approach. So follow the below posters advice and document your custom type with a SchemaFilter.

Instead of using a custom ModelBinder, you can also use a custom type with a TypeConverter.
[TypeConverter(typeof(StrListConverter))]
public class StrList : List<string>
{
public StrList(IEnumerable<string> collection) : base(collection) {}
}
public class StrListConverter : TypeConverter
{
public override bool CanConvertFrom(ITypeDescriptorContext context, Type sourceType)
{
return sourceType == typeof(string) || base.CanConvertFrom(context, sourceType);
}
public override object ConvertFrom(ITypeDescriptorContext context, CultureInfo culture, object value)
{
if (value == null)
return null;
if (value is string s)
{
if (string.IsNullOrEmpty(s))
return null;
return new StrList(s.Split(','));
}
return base.ConvertFrom(context, culture, value);
}
}
The advantage is that it makes the Web API method's parameters very simple. You dont't even need to specify [FromUri].
public IEnumerable<Category> GetCategories(StrList categoryIds) {
// code to retrieve categories from database
}
This example is for a List of strings, but you could do categoryIds.Select(int.Parse) or simply write an IntList instead.

I originally used the solution that #Mrchief for years (it works great). But when when I added Swagger to my project for API documentation my end point was NOT showing up.
It took me a while, but this is what I came up with. It works with Swagger, and your API method signatures look cleaner:
In the end you can do:
// GET: /api/values/1,2,3,4
[Route("api/values/{ids}")]
public IHttpActionResult GetIds(int[] ids)
{
return Ok(ids);
}
WebApiConfig.cs
public static class WebApiConfig
{
public static void Register(HttpConfiguration config)
{
// Allow WebApi to Use a Custom Parameter Binding
config.ParameterBindingRules.Add(descriptor => descriptor.ParameterType == typeof(int[]) && descriptor.ActionDescriptor.SupportedHttpMethods.Contains(HttpMethod.Get)
? new CommaDelimitedArrayParameterBinder(descriptor)
: null);
// Allow ApiExplorer to understand this type (Swagger uses ApiExplorer under the hood)
TypeDescriptor.AddAttributes(typeof(int[]), new TypeConverterAttribute(typeof(StringToIntArrayConverter)));
// Any existing Code ..
}
}
Create a new class: CommaDelimitedArrayParameterBinder.cs
public class CommaDelimitedArrayParameterBinder : HttpParameterBinding, IValueProviderParameterBinding
{
public CommaDelimitedArrayParameterBinder(HttpParameterDescriptor desc)
: base(desc)
{
}
/// <summary>
/// Handles Binding (Converts a comma delimited string into an array of integers)
/// </summary>
public override Task ExecuteBindingAsync(ModelMetadataProvider metadataProvider,
HttpActionContext actionContext,
CancellationToken cancellationToken)
{
var queryString = actionContext.ControllerContext.RouteData.Values[Descriptor.ParameterName] as string;
var ints = queryString?.Split(',').Select(int.Parse).ToArray();
SetValue(actionContext, ints);
return Task.CompletedTask;
}
public IEnumerable<ValueProviderFactory> ValueProviderFactories { get; } = new[] { new QueryStringValueProviderFactory() };
}
Create a new class: StringToIntArrayConverter.cs
public class StringToIntArrayConverter : TypeConverter
{
public override bool CanConvertFrom(ITypeDescriptorContext context, Type sourceType)
{
return sourceType == typeof(string) || base.CanConvertFrom(context, sourceType);
}
}
Notes:
https://stackoverflow.com/a/47123965/862011 pointed me in the right direction
Swagger was only failing to pick my comma delimited end points when using the [Route] attribute

public class ArrayInputAttribute : ActionFilterAttribute
{
private readonly string[] _ParameterNames;
/// <summary>
///
/// </summary>
public string Separator { get; set; }
/// <summary>
/// cons
/// </summary>
/// <param name="parameterName"></param>
public ArrayInputAttribute(params string[] parameterName)
{
_ParameterNames = parameterName;
Separator = ",";
}
/// <summary>
///
/// </summary>
public void ProcessArrayInput(HttpActionContext actionContext, string parameterName)
{
if (actionContext.ActionArguments.ContainsKey(parameterName))
{
var parameterDescriptor = actionContext.ActionDescriptor.GetParameters().FirstOrDefault(p => p.ParameterName == parameterName);
if (parameterDescriptor != null && parameterDescriptor.ParameterType.IsArray)
{
var type = parameterDescriptor.ParameterType.GetElementType();
var parameters = String.Empty;
if (actionContext.ControllerContext.RouteData.Values.ContainsKey(parameterName))
{
parameters = (string)actionContext.ControllerContext.RouteData.Values[parameterName];
}
else
{
var queryString = actionContext.ControllerContext.Request.RequestUri.ParseQueryString();
if (queryString[parameterName] != null)
{
parameters = queryString[parameterName];
}
}
var values = parameters.Split(new[] { Separator }, StringSplitOptions.RemoveEmptyEntries)
.Select(TypeDescriptor.GetConverter(type).ConvertFromString).ToArray();
var typedValues = Array.CreateInstance(type, values.Length);
values.CopyTo(typedValues, 0);
actionContext.ActionArguments[parameterName] = typedValues;
}
}
}
public override void OnActionExecuting(HttpActionContext actionContext)
{
_ParameterNames.ForEach(parameterName => ProcessArrayInput(actionContext, parameterName));
}
}
Usage:
[HttpDelete]
[ArrayInput("tagIDs")]
[Route("api/v1/files/{fileID}/tags/{tagIDs}")]
public HttpResponseMessage RemoveFileTags(Guid fileID, Guid[] tagIDs)
{
_FileRepository.RemoveFileTags(fileID, tagIDs);
return Request.CreateResponse(HttpStatusCode.OK);
}
Request uri
http://localhost/api/v1/files/2a9937c7-8201-59b7-bc8d-11a9178895d0/tags/BBA5CD5D-F07D-47A9-8DEE-D19F5FA65F63,BBA5CD5D-F07D-47A9-8DEE-D19F5FA65F63

If you want to list/ array of integers easiest way to do this is accept the comma(,) separated list of string and convert it to list of integers.Do not forgot to mention [FromUri] attriubte.your url look like:
...?ID=71&accountID=1,2,3,289,56
public HttpResponseMessage test([FromUri]int ID, [FromUri]string accountID)
{
List<int> accountIdList = new List<int>();
string[] arrAccountId = accountId.Split(new char[] { ',' });
for (var i = 0; i < arrAccountId.Length; i++)
{
try
{
accountIdList.Add(Int32.Parse(arrAccountId[i]));
}
catch (Exception)
{
}
}
}

I have created a custom model binder which converts any comma separated values (only primitive, decimal, float, string) to their corresponding arrays.
public class CommaSeparatedToArrayBinder<T> : IModelBinder
{
public bool BindModel(HttpActionContext actionContext, ModelBindingContext bindingContext)
{
Type type = typeof(T);
if (type.IsPrimitive || type == typeof(Decimal) || type == typeof(String) || type == typeof(float))
{
ValueProviderResult val = bindingContext.ValueProvider.GetValue(bindingContext.ModelName);
if (val == null) return false;
string key = val.RawValue as string;
if (key == null) { bindingContext.ModelState.AddModelError(bindingContext.ModelName, "Wrong value type"); return false; }
string[] values = key.Split(',');
IEnumerable<T> result = this.ConvertToDesiredList(values).ToArray();
bindingContext.Model = result;
return true;
}
bindingContext.ModelState.AddModelError(bindingContext.ModelName, "Only primitive, decimal, string and float data types are allowed...");
return false;
}
private IEnumerable<T> ConvertToDesiredArray(string[] values)
{
foreach (string value in values)
{
var val = (T)Convert.ChangeType(value, typeof(T));
yield return val;
}
}
}
And how to use in Controller:
public IHttpActionResult Get([ModelBinder(BinderType = typeof(CommaSeparatedToArrayBinder<int>))] int[] ids)
{
return Ok(ids);
}

Make the method type [HttpPost], create a model that has one int[] parameter, and post with json:
/* Model */
public class CategoryRequestModel
{
public int[] Categories { get; set; }
}
/* WebApi */
[HttpPost]
public HttpResponseMessage GetCategories(CategoryRequestModel model)
{
HttpResponseMessage resp = null;
try
{
var categories = //your code to get categories
resp = Request.CreateResponse(HttpStatusCode.OK, categories);
}
catch(Exception ex)
{
resp = Request.CreateErrorResponse(HttpStatusCode.InternalServerError, ex);
}
return resp;
}
/* jQuery */
var ajaxSettings = {
type: 'POST',
url: '/Categories',
data: JSON.serialize({Categories: [1,2,3,4]}),
contentType: 'application/json',
success: function(data, textStatus, jqXHR)
{
//get categories from data
}
};
$.ajax(ajaxSettings);

Or you could just pass a string of delimited items and put it into an array or list on the receiving end.

I addressed this issue this way.
I used a post message to the api to send the list of integers as data.
Then I returned the data as an ienumerable.
The sending code is as follows:
public override IEnumerable<Contact> Fill(IEnumerable<int> ids)
{
IEnumerable<Contact> result = null;
if (ids!=null&&ids.Count()>0)
{
try
{
using (var client = new HttpClient())
{
client.BaseAddress = new Uri("http://localhost:49520/");
client.DefaultRequestHeaders.Accept.Clear();
client.DefaultRequestHeaders.Accept.Add(new MediaTypeWithQualityHeaderValue("application/json"));
String _endPoint = "api/" + typeof(Contact).Name + "/ListArray";
HttpResponseMessage response = client.PostAsJsonAsync<IEnumerable<int>>(_endPoint, ids).Result;
response.EnsureSuccessStatusCode();
if (response.IsSuccessStatusCode)
{
result = JsonConvert.DeserializeObject<IEnumerable<Contact>>(response.Content.ReadAsStringAsync().Result);
}
}
}
catch (Exception)
{
}
}
return result;
}
The receiving code is as follows:
// POST api/<controller>
[HttpPost]
[ActionName("ListArray")]
public IEnumerable<Contact> Post([FromBody]IEnumerable<int> ids)
{
IEnumerable<Contact> result = null;
if (ids != null && ids.Count() > 0)
{
return contactRepository.Fill(ids);
}
return result;
}
It works just fine for one record or many records. The fill is an overloaded method using DapperExtensions:
public override IEnumerable<Contact> Fill(IEnumerable<int> ids)
{
IEnumerable<Contact> result = null;
if (ids != null && ids.Count() > 0)
{
using (IDbConnection dbConnection = ConnectionProvider.OpenConnection())
{
dbConnection.Open();
var predicate = Predicates.Field<Contact>(f => f.id, Operator.Eq, ids);
result = dbConnection.GetList<Contact>(predicate);
dbConnection.Close();
}
}
return result;
}
This allows you to fetch data from a composite table (the id list), and then return the records you are really interested in from the target table.
You could do the same with a view, but this gives you a little more control and flexibility.
In addition, the details of what you are seeking from the database are not shown in the query string. You also do not have to convert from a csv file.
You have to keep in mind when using any tool like the web api 2.x interface is that the get, put, post, delete, head, etc., functions have a general use, but are not restricted to that use.
So, while post is generally used in a create context in the web api interface, it is not restricted to that use. It is a regular html call that can be used for any purpose permitted by html practice.
In addition, the details of what is going on are hidden from those "prying eyes" we hear so much about these days.
The flexibility in naming conventions in the web api 2.x interface and use of regular web calling means you send a call to the web api that misleads snoopers into thinking you are really doing something else. You can use "POST" to really retrieve data, for example.

My solution was to create an attribute to validate strings, it does a bunch of extra common features, including regex validation that you can use to check for numbers only and then later I convert to integers as needed...
This is how you use:
public class MustBeListAndContainAttribute : ValidationAttribute
{
private Regex regex = null;
public bool RemoveDuplicates { get; }
public string Separator { get; }
public int MinimumItems { get; }
public int MaximumItems { get; }
public MustBeListAndContainAttribute(string regexEachItem,
int minimumItems = 1,
int maximumItems = 0,
string separator = ",",
bool removeDuplicates = false) : base()
{
this.MinimumItems = minimumItems;
this.MaximumItems = maximumItems;
this.Separator = separator;
this.RemoveDuplicates = removeDuplicates;
if (!string.IsNullOrEmpty(regexEachItem))
regex = new Regex(regexEachItem, RegexOptions.Compiled | RegexOptions.Singleline | RegexOptions.IgnoreCase);
}
protected override ValidationResult IsValid(object value, ValidationContext validationContext)
{
var listOfdValues = (value as List<string>)?[0];
if (string.IsNullOrWhiteSpace(listOfdValues))
{
if (MinimumItems > 0)
return new ValidationResult(this.ErrorMessage);
else
return null;
};
var list = new List<string>();
list.AddRange(listOfdValues.Split(new[] { Separator }, System.StringSplitOptions.RemoveEmptyEntries));
if (RemoveDuplicates) list = list.Distinct().ToList();
var prop = validationContext.ObjectType.GetProperty(validationContext.MemberName);
prop.SetValue(validationContext.ObjectInstance, list);
value = list;
if (regex != null)
if (list.Any(c => string.IsNullOrWhiteSpace(c) || !regex.IsMatch(c)))
return new ValidationResult(this.ErrorMessage);
return null;
}
}

I just added the Query key (Refit lib) in the property for the request.
[Query(CollectionFormat.Multi)]
public class ExampleRequest
{
[FromQuery(Name = "name")]
public string Name { get; set; }
[AliasAs("category")]
[Query(CollectionFormat.Multi)]
public List<string> Categories { get; set; }
}

All other solutions need too much work. I was trying to use IEnumerable<long> or long[] in a HttpGet method parameter, but I see no point of doing all the work just to make the signature of the handler method parameter long[]. I ended up just making it string, and then separated it within the handler. Took me one line.
public async Task<IActionResult> SomeHandler(string idsString)
{
var ids = idsString.Split(',').Select(x => long.Parse(x));
Now you can just pass the numbers like
.../SomeHandler?idsString=123,456,789,012

Async Foolproof Validation

When using MVC Foolproof Validation, is it possible to have a custom validation asynchronous? I am wanting to retrieve an object from a DbSet in a custom validation, to check if the object exists.
Is this possible? If not, is there a work around that achieves the same result?
Here is my code:
public class RequiredThumbnailAssetId : ModelAwareValidationAttribute
{
static RequiredThumbnailAssetId() { Register.Attribute(typeof(RequiredThumbnailAssetId)); }
public async override Task<bool> IsValid(object value, object container)
{
var model = (IGridViewItemWithAssetId)container;
if (model.thumbnailAssetId <= 0)
{
return false;
}
CanFindLocationDatabaseContext db = new CanFindLocationDatabaseContext();
Asset asset = await db.assets.FindAsync(model.thumbnailAssetId);
if (asset == null)
{
return false;
}
return true;
}
}
Here is the error that I am getting:
'CanFindLocation.CustomValidations.RequiredThumbnailAssetId.IsValid(object,
object)': return type must be 'bool' to match overridden member
'Foolproof.ModelAwareValidationAttribute.IsValid(object, object)'
Thanks in advance.

You could use the classes defined in the post How would I run an async Task<T> method synchronously? and then separate the override you need and the async call out into separate methods. This would allow you to call the synchronous override from the async request.
public class RequiredThumbnailAssetId : ModelAwareValidationAttribute
{
static RequiredThumbnailAssetId() { Register.Attribute(typeof(RequiredThumbnailAssetId)); }
public async override Task<bool> CheckValid(object value, object container)
{
var valid = AsyncHelpers.RunSync<bool>(() => isValid(value,container));
return valid;
}
public override bool IsValid(object value, object container)
{
var model = (IGridViewItemWithAssetId)container;
if (model.thumbnailAssetId <= 0)
{
return false;
}
CanFindLocationDatabaseContext db = new CanFindLocationDatabaseContext();
Asset asset = await db.assets.FindAsync(model.thumbnailAssetId);
if (asset == null)
{
return false;
}
return true;
}

How optimize by lambda expression

I have a very similar function is only one previous report and the other future, how can I optimize and write beautiful?
public bool AnyPreviousReportByGroup(int groupID)
{
if(this.GroupID == groupID)
{
return true;
}
else
{
return PreviousReport.AnyPreviousReportByGroup(groupID);
}
}
public bool AnyNextReportByGroup(int groupID)
{
if (this.GroupID == groupID)
{
return true;
}
else
{
return NextReport.AnyNextReportByGroup(groupID);
}
}

The following code is a conciser way of achieving the same thing:
public bool AnyPreviousReportByGroup(int groupID)
{
return this.GroupID == groupID ||
this.PreviousReport != null &&
this.PreviousReport.AnyPreviousReportByGroup(groupID);
}
If you really want to use lambda expressions, here's a possible way:
public bool AnyReportByGroup(int groupID, Func<Report, Report> getOtherReport)
{
if (this.GroupID == groupID)
return true;
Report other = getOtherReport(this);
return other != null &&
other.AnyReportByGroup(groupID, getOtherReport);
}
You could then call this helper method using lambda expressions:
bool anyPrevious = this.AnyReportByGroup(groupID, report => report.PreviousReport);
bool anyNext = this.AnyReportByGroup(groupID, report => report.NextReport);

private bool AnyReportByGroup(int groupID, Func<int, bool> callback)
{
if (this.GroupID == groupID)
{
return true;
}
else
{
return callback(groupID);
}
}
public bool AnyPreviousReportByGroup(int groupID)
{
return AnyReportByGroup(groupID, gid => PreviousReport.AnyPreviousReportByGroup(gid));
}
public bool AnyNextReportByGroup(int groupID)
{
return AnyReportByGroup(groupID, gid => NextReport.AnyNextReportByGroup(gid));
}
But, I hope, that these methods are just a sample, and in your real code they're more complex.
Otherwise, I can't understand, what do you try to optimize.

May be like this
public enum ReportType{ Next, Previous }
public bool AnyReportByGroup(int groupID, ReportType type)
{
if(this.GroupID == groupID)
return true;
else
{
switch(type)
{
case ReportType.Next:
return NextReport.AnyNextReportByGroup(groupID);
case ReportType.Previous:
return NextReport.AnyPreviousReportByGroup(groupID);
}
return false;
}
}

Here's a non-recursive solution, assuming that you want to return false when you run out of reports:
public bool AnyPreviousReportByGroup(int groupID)
{
return GetEventualValue(groupID, r => r.PreviousReport);
}
public bool AnyNextReportByGroup(int groupID)
{
return GetEventualValue(groupID, r => r.NextReport);
}
public bool GetEventualValue(int groupID, Func<Report, Report> nextReport)
{
Report report = this;
while (report != null && report.GroupID != groupID)
{
report = nextReport(report);
}
return report != null;
}

Standard way to traverse linked lists is like so:
public bool AnyPreviousReportByGroup(int groupID)
{
var item = this;
do
{
if (item.GroupId == groupID)
{
return true;
}
item = item.PreviousReport;
} while (item != null);
return false;
}
public bool AnyNextReportByGroup(int groupID)
{
var item = this;
do
{
if (item.GroupId == groupID)
{
return true;
}
item = item.NextReport;
} while (item != null);
return false;
}
This has a benefit of not creating potentially massive call stacks the way a recusive approach would.
This also fixes your code, where it never returned false, it would just NPE.
Now we can refactor as you requested:
private bool AnyReportByGroup(int groupID, bool searchForward)
{
var item = this;
do
{
if (item.GroupId == groupID)
{
return true;
}
item = searchForward ? item.NextReport : item.PreviousReport;
} while (item != null);
return false;
}
public bool AnyPreviousReportByGroup(int groupID)
{
return AnyReportByGroup(groupID, false);
}
public bool AnyNextReportByGroup(int groupID)
{
return AnyReportByGroup(groupID, true);
}