How to select multiple class properties in LINQ Expression? - linq

If I have a class like this
`
class Person
{
public string First;
public string Last;
public bool IsMarried;
public int Age;
}`
Then how can I write a LINQ Expression where I could select properties of a Person. I want to do something like this (user can enter 1..n properties)
SelectData<Person>(x=>x.First, x.Last,x.Age);
What would be the input expression of my SelectData function ?
SelectData(Expression<Func<TEntity, List<string>>> selector); ?
EDIT
In my SelectData function I want to extract property names and then generate SELECT clause of my SQL Query dynamically.
SOLUTION
Ok, so what I have done is to have my SelectData as
public IEnumerable<TEntity> SelectData(Expression<Func<TEntity, object>> expression)
{
NewExpression body = (NewExpression)expression.Body;
List<string> columns = new List<string>();
foreach(var arg in body.Arguments)
{
var exp = (MemberExpression)arg;
columns.Add(exp.Member.Name);
}
//build query
And to use it I call it like this
ccc<Person>().SelectData(x => new { x.First, x.Last, x.Age });
Hopefully it would help someone who is looking :)
Thanks,
IY

I think it would be better to use delegates instead of Reflection. Apart from the fact that delegates will be faster, the compiler will complain if you try to fetch property values that do not exist. With reflection you won't find errors until run time.
Luckily there is already something like that. it is implemented as an extension function of IEnumerable, and it is called Select (irony intended)
I think you want something like this:
I have a sequence of Persons, and I want you to create a Linq
statement that returns per Person a new object that contains the
properties First and Last.
Or:
I have a sequence of Persns and I want you to create a Linq statement
that returns per Person a new object that contains Age, IsMarried,
whether it is an adult and to make it difficult: one Property called
Name which is a combination of First and Last
The function SelectData would be something like this:
IEnumerable<TResult> SelectData<TSource, TResult>(this IEnumerable<TSource> source,
Func<TSource, TResult> selector)
{
return source.Select(selector);
}
Usage:
problem 1: return per Person a new object that contains the
properties First and Last.
var result = Persons.SelectData(person => new
{
First = person.First,
Last = person.Last,
});
problem 2: return per Person a new object that contains Age, IsMarried, whether he is an adult and one Property called Name which is a combination
of First and Last
var result = Persons.SelectData(person => new
{
Age = person.Name,
IsMarried = person.IsMarried,
IsAdult = person.Age > 21,
Name = new
{
First = person.First,
Last = person.Last,
},
});
Well let's face it, your SelectData is nothing more than Enumerable.Select
You could of course create a function where you'd let the caller provide a list of properties he wants, but (1) that would limit his possibilities to design the end result and (2) it would be way more typing for him to call the function.
Instead of:
.Select(p => new
{
P1 = p.Property1,
P2 = p.Property2,
}
he would have to type something like
.SelectData(new List<Func<TSource, TResult>()
{
p => p.Property1, // first element of the property list
p -> p.Property2, // second element of the property list
}
You won't be able to name the returned properties, you won't be able to combine several properties into one:
.Select(p => p.First + p.Last)
And what would you gain by it?
Highly discouraged requirement!

You could achive similar result using Reflection and Extension Method
Model:
namespace ConsoleApplication2
{
class Person
{
public string First { get; set; }
public string Last { get; set; }
public bool IsMarried { get; set; }
public int Age { get; set; }
}
}
Service:
using System.Collections.Generic;
using System.Linq;
namespace Test
{
public static class Service
{
public static IQueryable<IQueryable<KeyValuePair<string, object>>> SelectData<T>(this IQueryable<T> queryable, string[] properties)
{
var queryResult = new List<IQueryable<KeyValuePair<string, object>>>();
foreach (T entity in queryable)
{
var entityProperties = new List<KeyValuePair<string, object>>();
foreach (string property in properties)
{
var value = typeof(T).GetProperty(property).GetValue(entity);
var entityProperty = new KeyValuePair<string, object>(property, value);
entityProperties.Add(entityProperty);
}
queryResult.Add(entityProperties.AsQueryable());
}
return queryResult.AsQueryable();
}
}
}
Usage:
using System;
using System.Collections.Generic;
using System.Linq;
namespace Test
{
class Program
{
static void Main(string[] args)
{
var list = new List<Person>()
{
new Person()
{
Age = 18,
First = "test1",
IsMarried = false,
Last = "test2"
},
new Person()
{
Age = 40,
First = "test3",
IsMarried = true,
Last = "test4"
}
};
var queryableList = list.AsQueryable();
string[] properties = { "Age", "Last" };
var result = queryableList.SelectData(properties);
foreach (var element in result)
{
foreach (var property in element)
{
Console.WriteLine($"{property.Key}: {property.Value}");
}
}
Console.ReadKey();
}
}
}
Result:
Age: 18
Last: test2
Age: 40
Last: test4

Related

Intersection of arrays in LINQ to CosmosDB

I'm trying find all items in my database that have at least one value in an array that matches any value in an array that I have in my code (the intersection of the two arrays should not be empty).
Basically, I'm trying to achieve this :
public List<Book> ListBooks(string partitionKey, List<string> categories)
{
return _client.CreateDocumentQuery<Book>(GetCollectionUri(), new FeedOptions
{
PartitionKey = new PartitionKey(partitionKey)
})
.Where(b => b.Categories.Any(c => categories.Contains(c))
.ToList();
}
With the Book class looking like this :
public class Book
{
public string id {get;set;}
public string Title {get;set;}
public string AuthorName {get;set;}
public List<string> Categories {get;set;}
}
However the SDK throws an exception saying that Method 'Any' is not supported when executing this code.
This doesn't work either :
return _client.CreateDocumentQuery<Book>(GetCollectionUri(), new FeedOptions
{
PartitionKey = new PartitionKey(partitionKey)
})
.Where(b => categories.Any(c => b.Categories.Contains(c))
.ToList();
The following code works because there's only one category to find :
public List<Book> ListBooksAsync(string category)
{
return _client.CreateDocumentQuery<Book>(GetCollectionUri())
.Where(b => b.Categories.Contains(category))
.ToList();
}
In plain SQL, I can queue multiple ARRAY_CONTAINS with several OR the query executes correctly.
SELECT * FROM root
WHERE ARRAY_CONTAINS(root["Categories"], 'Humor')
OR ARRAY_CONTAINS(root["Categories"], 'Fantasy')
OR ARRAY_CONTAINS(root["Categories"], 'Legend')
I'm trying to find the best way to achieve this with LINQ, but I'm not even sure it's possible.
In this situation I've used a helper method to combine expressions in a way that evaluates to SQL like in your final example. The helper method 'MakeOrExpression' below lets you pass a number of predicates (in your case the individual checks for b.Categories.Contains(category)) and produces a single expression you can put in the argument to .Where(expression) on your document query.
class Program
{
private class Book
{
public string id { get; set; }
public string Title { get; set; }
public string AuthorName { get; set; }
public List<string> Categories { get; set; }
}
static void Main(string[] args)
{
var comparison = new[] { "a", "b", "c" };
var target = new Book[] {
new Book { id = "book1", Categories = new List<string> { "b", "z" } },
new Book { id = "book2", Categories = new List<string> { "s", "t" } },
new Book { id = "book3", Categories = new List<string> { "z", "a" } } };
var results = target.AsQueryable()
.Where(MakeOrExpression(comparison.Select(x => (Expression<Func<Book, bool>>)(y => y.Categories.Contains(x))).ToArray()));
foreach (var result in results)
{
// Should be book1 and book3
Console.WriteLine(result.id);
}
Console.ReadLine();
}
private static Expression<Func<T,bool>> MakeOrExpression<T>(params Expression<Func<T,bool>>[] inputExpressions)
{
var combinedExpression = inputExpressions.Skip(1).Aggregate(
inputExpressions[0].Body,
(agg, x) => Expression.OrElse(agg, x.Body));
var parameterExpression = Expression.Parameter(typeof(T));
var replaceParameterVisitor = new ReplaceParameterVisitor(parameterExpression,
Enumerable.SelectMany(inputExpressions, ((Expression<Func<T, bool>> x) => x.Parameters)));
var mergedExpression = replaceParameterVisitor.Visit(combinedExpression);
var result = Expression.Lambda<Func<T, bool>>(mergedExpression, parameterExpression);
return result;
}
private class ReplaceParameterVisitor : ExpressionVisitor
{
private readonly IEnumerable<ParameterExpression> targetParameterExpressions;
private readonly ParameterExpression parameterExpression;
public ReplaceParameterVisitor(ParameterExpression parameterExpressionParam, IEnumerable<ParameterExpression> targetParameterExpressionsParam)
{
this.parameterExpression = parameterExpressionParam;
this.targetParameterExpressions = targetParameterExpressionsParam;
}
public override Expression Visit(Expression node)
=> targetParameterExpressions.Contains(node) ? this.parameterExpression : base.Visit(node);
}
}

Unable to cast object of type WhereSelectListIterator

I have been attempting to figure out why a Linq query that returns a list of U.S. States formatted for a drop down list will not cast to a List when the code returns to the calling method. The error that I get is:
Unable to cast object of type 'WhereSelectListIterator'2[StateListing.States,<>f__AnonymousTypea'2[System.String,System.String]]' to type 'System.Collections.Generic.List`1[StateListing.States]'
The namespace StateListing from the error, is a dll library that has a class called States returning an IEnumerable List of states shown below.
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
namespace StateListing
{
public class States
{
public string StateAbbriviation { get; set; }
public int StateID { get; set; }
public string StateName { get; set; }
static int cnt = 0;
public static IEnumerable<States> GetStates()
{
return new List<States>
{
new States
{
StateAbbriviation = "AL",
StateID=cnt++,
StateName = "Alabama"
},
new States
{
StateAbbriviation = "AL",
StateID=cnt++,
StateName = "Alaska"
}
//Continued on with the rest of states
}.AsQueryable();
}
}
}
In my control I make a call to GetStates that returns a List of States from the class library above.
[HttpPost]
public JsonResult GetStateOptions()
{
try
{
//Return a list of options for dropdown list
var states = propertyRepository.GetStates();
return Json(new { Result = "OK", options = states });
}
In the property repository class I have two methods one to get the StateList from the library, and another to format the listing of states for a drop down list in my view.
public List<States> GetStateList()
{
var items = (from s in States.GetStates()
select s).ToList();
return items;
}
List<States> IPropertyRepository.GetStates()
{
try
{
List<States> RawStates = GetStateList();
var stateList = RawStates.Select(c => new { DisplayText = c.StateName, Value = c.StateID.ToString() });
return (List<States>)stateList; //<=== Error
}
The error occurs when the code reaches the return within the GetStates method.
Any help with this casting problem explaining what I'm doing wrong would be appreciated.
This is the problem:
var stateList = RawStates.Select(c => new { DisplayText = c.StateName,
Value = c.StateID.ToString() });
return (List<States>)stateList;
Two issues:
Select doesn't return a List<T>
You're not= selecting States objects; you're selecting an anonymous type
The first is fixable using ToList(); the second is fixable either by changing your Select call or by changing your method's return type. It's not really clear what you really want to return, given that States doesn't have a DisplayText or Value property.
I would expect a method of GetStates to return the states - in which case you've already got GetStatesList() which presumably does what you want already.
Basically, you need to think about the type you really want to return, and make both your method return type and the method body match that.
You are projecting your LINQ query to an anonymmous object and not to a State list which obviously cannot work. The 2 types are incompatible. So start by modifying your repository layer and get rid of the GetStateList method:
public class PropertyRepository: IPropertyRepository
{
public List<States> GetStates()
{
return States.GetStates().ToList();
}
}
and then project to the desired structure in your controller:
[HttpPost]
public JsonResult GetStateOptions()
{
var states = propertyRepository.GetStateList();
var options = states.Select(x => new
{
DisplayText = c.StateName,
Value = c.StateID.ToString()
}).ToList();
return Json(new { Result = "OK", options = states });
}

Attempting to perform complicated anonymous methods is failing

Update (Stupidity Fail)
So then, in all of my convoluted formula code, I neglected the fundamental principles of C#.
Methods may return a value.
static dynamic Construct<T>(T expression){
return expression;
}
Then just use that, instead of a variable ...
Method = Construct<Action<Context, string, int>>(
(context, key, change) =>
{
context.Saved[key] += change;
Console.WriteLine("{0}'s saved value of {1} was changed by {2}, resulting in {3}",
context.Name, key, change, context.Saved[key]);
}
)
I have a situation where I need to call upon methods that don't exist as compiled methods, but rather need to be able to accept an array of parameters and execute as an anonymous function. I thought I had it worked out, but I am running into an issue with the following..
public static IDictionary<string, Function> Expressions =
new Dictionary<string, Function> {
{
"Increase [X] by value of [Y]",
new Function {
Name = "Increase [X] by [Y]",
Parameters = 2,
Types = new List<Type>{
typeof(Param),
typeof(Param)
},
Method = (Expression<Func<Context, Param, Param, bool>>)
((context, x, y) => {
Console.WriteLine("test"); // this is where I need to do stuff...
})
}
}
};
I am being told that a Method name is expected on this. The problem is that Context will be passed in by the object that takes the function and runs its method, because the Context object cannot be pre-bound (it has to be late bound). So basically I package up the trailing 2 parameters (Param) and (Param) in this case and create a function to execute against them.
The database stores those parameters, and then invokes the method passing in the appropriate Context as the first parameter by using Compile().DynamicInvoke(object[] params).
Can anyone give me a hand here as to why I cannot put any kind of logic in between my { }?
UPDATE
Okay, since I've been told this example is unclear, here is an entire program running start to finish that illustrates what I am trying to accomplish.
public class Program {
static void Main(string[] args) {
// simple object stored in database.
var ctx = new Context {
Name = "Ciel",
Saved = new Dictionary<string, int> {
{ "First", 10 },
{ "Second", 20 }
}
};
// simple object stored in database.
var rule = new Rule {
Equations = new List<Equation> {
new Equation {
Parameters = new List<object>{
"First",
5
},
Name = "Increase [X] by value of [Y]"
}
}
};
// =======================================
// runtime environment!!!
// =======================================
var method = Evaluations.Expressions[rule.Equations[0].Name].Method;
var parameters = rule.Equations[0].Parameters;
// insert the specific context as the first parameter.
parameters.Insert(0, ctx);
method.DynamicInvoke(parameters.ToArray());
Console.ReadLine();
}
}
public class Function {
public string Name { get; set; }
public dynamic Method { get; set; }
}
public class Equation {
public string Name { get; set; }
// these objects will be simple enough to serialize.
public IList<object> Parameters { get; set; }
public Function Function { get; set; }
}
public class Context {
public string Name { get; set; }
// this is a crude example, but it serves the demonstration purposes.
public IDictionary<string, int> Saved { get; set; }
}
public class Rule {
// again, a crude example.
public IList<Equation> Equations { get; set; }
}
public static class Evaluations {
static Action<Context, string, int> expr = (context, key, change) =>
{
context.Saved[key] += change;
Console.WriteLine("{0}'s saved value of {1} was changed by {2}, resulting in {3}",
context.Name, key, change, context.Saved[key]);
};
public static IDictionary<string, Function> Expressions =
new Dictionary<string, Function> {
{
"Increase [X] by value of [Y]",
new Function {
Name = "Increase [X] by [Y]",
Method = expr
}
}
};
}
Four problems:
You're trying to create an expression tree from a lambda expression with a statement body (i.e. braces). C# doesn't allow this - you can only convert a statement lambda into a delegate, not an expression tree
Your lambda body doesn't return a Boolean value
You're trying to call an Expression<Func<Context, Param, Param, bool>> as if it were a method with a bool parameter. It's not at all clear what you're trying to do there.
Even if the third point were valid, I suspect you'd need more brackets.
If you refactor your code to make it a little more readable and manageable, you'll probably be well on your way to solving your problem. Rather than having one mammoth C# statement with a single semicolon, split it up into several lines. Something like this:
public static Dictionary<string, Function> Expressions = getExpressions();
private static Dictionary<string, Function> getExpressions()
{
var method = (Expression<Func<Context, Param, Param, bool>>)
((context, x, y) => {
Console.WriteLine("test"); // this is where I need to do stuff...
})(true);
var func = new Function()
{
Name = "Increase [X] by [Y]",
Parameters = 2,
Types = new List<Type>
{
typeof(Param),
typeof(Param)
},
Method = method
};
var dict = new Dictionary<string, Function>();
dict["Increase [X] by value of [Y]"] = func;
return dict;
}
Note: my syntax could be incorrect, but you get the general idea.

Sorting a List of Class with LINQ

I have a List<MyClass> and I want to sort it by DateTime CreateDate attribute of MyClass.
Is that possible with LINQ ?
Thanks
To sort the existing list:
list.Sort((x,y) => x.CreateDate.CompareTo(y.CreateDate));
It is also possible to write a Sort extension method, allowing:
list.Sort(x => x.CreateDate);
for example:
public static class ListExt {
public static void Sort<TSource, TValue>(
this List<TSource> list,
Func<TSource, TValue> selector) {
if (list == null) throw new ArgumentNullException("list");
if (selector == null) throw new ArgumentNullException("selector");
var comparer = Comparer<TValue>.Default;
list.Sort((x,y) => comparer.Compare(selector(x), selector(y)));
}
}
You can enumerate it in sorted order:
IEnumerable<MyClass> result = list.OrderBy(element => element.CreateDate);
You can also use ToList() to convert to a new list and reassign to the original variable:
list = list.OrderBy(element => element.CreateDate).ToList();
This isn't quite the same as sorting the original list because if anyone still has a reference to the old list they won't see the new ordering. If you actually want to sort the original list then you need to use the List<T>.Sort method.
Here is a sample:
using System.Collections.Generic;
using System.Linq;
namespace Demo
{
public class Test
{
public void SortTest()
{
var myList = new List<Item> { new Item { Name = "Test", Id = 1, CreateDate = DateTime.Now.AddYears(-1) }, new Item { Name = "Other", Id = 1, CreateDate = DateTime.Now.AddYears(-2) } };
var result = myList.OrderBy(x => x.CreateDate);
}
}
public class Item
{
public string Name { get; set; }
public int Id { get; set; }
public DateTime CreateDate { get; set; }
}
}
Sure the other answers with .OrderBy() work, but wouldn't you rather make your source item inherit from IComparable and just call .Sort()?
class T {
public DateTime CreatedDate { get; set; }
}
to use:
List<T> ts = new List<T>();
ts.Add(new T { CreatedDate = DateTime.Now });
ts.Add(new T { CreatedDate = DateTime.Now });
ts.Sort((x,y) => DateTime.Compare(x.CreatedDate, y.CreatedDate));

How to do a "where in values" in LINQ-to-Entities 3.5

Does anybody know how to apply a "where in values" type condition using LINQ-to-Entities? I've tried the following but it doesn't work:
var values = new[] { "String1", "String2" }; // some string values
var foo = model.entitySet.Where(e => values.Contains(e.Name));
I believe this works in LINQ-to-SQL though? Any thoughts?
Update: found out how to do this. And EF will generate the appropriate SQL on the database. I'm not sure if this is for EF4 only but I got the tip from Entity Framework 4.0 Recipes
var listOfIds=GetAListOfIds();
var context=CreateEntityFrameworkObjectContext();
var results = from item in context.Items
where listOfIds.Contains(item.Category.Id)
select item;
//results contains the items with matching category Ids
This query generates the correct in clause on the server side. I haven't tested it with EF 3.5 but it does work with EF4.
NB: The values passed into the in clause are NOT parameters so make sure you validate your inputs.
It is somewhat of a shame that Contains is not supported in Linq to Entities.
IN and JOIN are not the same operator (Filtering by IN never changes the cardinality of the query).
Contains is not supported in EF at this time.
FYI:
If you are using ESql you are able to use in operation.
I don't have VS 2008 With me but code should be something like following:
var ids = "12, 34, 35";
using (context = new Entites())
{
var selectedProducts = context.CreateQuery<Products>(
String.Format("select value p from [Entities].Products as p
where p.productId in {{{0}}}", ids)).ToList();
...
}
For the cases when you want to use expressions when querying your data, you can use the following extension method (adapted after http://social.msdn.microsoft.com/forums/en-US/adodotnetentityframework/thread/095745fe-dcf0-4142-b684-b7e4a1ab59f0/):
using System;
using System.Collections.Generic;
using System.Linq;
using System.Linq.Expressions;
using System.Data.Objects;
namespace Sample {
public static class Extensions {
public static IQueryable<T> ExtWhereIn<T, TValue>(this ObjectQuery<T> query,
Expression<Func<T, TValue>> valueSelector,
IEnumerable<TValue> values) {
return query.Where(BuildContainsExpression<T, TValue>(valueSelector, values));
}
public static IQueryable<T> ExtWhereIn<T, TValue>(this IQueryable<T> query,
Expression<Func<T, TValue>> valueSelector,
IEnumerable<TValue> values) {
return query.Where(BuildContainsExpression<T, TValue>(valueSelector, values));
}
private static Expression<Func<TElement, bool>> BuildContainsExpression<TElement, TValue>(
Expression<Func<TElement, TValue>> valueSelector, IEnumerable<TValue> values) {
if (null == valueSelector) { throw new ArgumentNullException("valueSelector"); }
if (null == values) { throw new ArgumentNullException("values"); }
ParameterExpression p = valueSelector.Parameters.Single();
// p => valueSelector(p) == values[0] || valueSelector(p) == ...
if (!values.Any()) {
return e => false;
}
var equals = values.Select(value => (Expression)Expression.Equal(valueSelector.Body, Expression.Constant(value, typeof(TValue))));
var body = equals.Aggregate<Expression>((accumulate, equal) => Expression.Or(accumulate, equal));
return Expression.Lambda<Func<TElement, bool>>(body, p);
}
}
class Program {
static void Main(string[] args) {
List<int> fullList = new List<int>();
for (int i = 0; i < 20; i++) {
fullList.Add(i);
}
List<int> filter = new List<int>();
filter.Add(2);
filter.Add(5);
filter.Add(10);
List<int> results = fullList.AsQueryable().ExtWhereIn<int, int>(item => item, filter).ToList();
foreach (int result in results) {
Console.WriteLine(result);
}
}
}
}
Using the extensions is really easy (as you can see in the sample). To use it on a database object, assuming you are filtering a table called "Product" by more than one id, you could do something like that:
class Product {
public int Id { get; set; }
/// ... other properties
}
List<Product> GetProducts(List<int> productIds) {
using (MyEntities context = new MyEntities()) {
return context.Products.ExtWhereIn<Product, int>(product => product.Id, productIds).ToList();
}
}
Using the where method doesn't alway work
var results = from p in db.Products
where p.Name == nameTextBox.Text
select p;
Yes it does translate to SQL, it generates a standard IN statement like this:
SELECT [t0].[col1]
FROM [table] [t0]
WHERE [col1] IN ( 'Value 1', 'Value 2')

Resources