I am using Linq to filter some things and some times I need to use reflection to get the value. Here is the example:
//...
PropertyType[] properties = myType.GetProperties();
var filtered = properties.Where(p=>p.PropertyType==typeof(MyMetaData)
&& ((MyType)p.GetValue(obj)).Name=="name"
&& ((MyType)p.GetValue(obj)).Length==10
).ToList();
//...
In my example I am using GetValue() method more than one time. Is there way if I can use variable to store it? I think that will help with performance.
It looks like that to include some variable in a LINQ we have to use the expression query, not method query, like this:
var filtered = (from x in properties
let a = (x.PropertyType is MyType) ? (MyType) x.GetValue(obj) : null
where a != null && a.Name == "name" && a.Length == 10).ToList();
I think this also works for method query with some Select:
var filtered = properties.Select(p=> new {p, a = (p.PropertyType is MyType) ? (MyType) p.GetValue(obj) : null})
.Where(x=>x.a != null && x.a.Name == "name" && x.a.Length == 10)
.Select(x=>x.p).ToList();
Something like following should work with method expression(lambda syntax)
var filtered = properties.Where(p =>
{
if(!p.PropertyType is MyMetaData)
{
return false;
}
var item = (MyType) p.GetValue(obj);
return item.Name == "name"
&& item.Length == 10
}
).ToList();
Related
I have a query that in T-SQL it is
SELECT *
FROM rpm_scrty_rpm_usr ru
WHERE ru.inact_ind = 'N'
AND email_id IS NOT NULL
AND wwid IS NULL
AND LTRIM(RTRIM (email_id)) <> ''
AND dflt_ste_id NOT IN (25,346,350,352,353,354,355,357,358,366,372,411)
When I have been converting it to LINQ, I have everything except the "NOT IN"
var querynonSystem = (from ru in Rpm_scrty_rpm_usrs
where ru.Inact_ind == "N" && ru.Email_id != null && ru.Wwid == null && ru.Email_id.Trim() != ""
&& ru.Dflt_ste_id != 25
select ru).Count();
I did temporarily put in this line && ru.Dflt_ste_id != 25
However I need to have AND dflt_ste_id NOT IN (25,346,350,352,353,354,355,357,358,366,372,411)
I am seeing a lot of different code like
this lambda where !(list2.Any(item2 => item2.Email == item1.Email))
Then var otherObjects = context.ItemList.Where(x => !itemIds.Contains(x.Id));
For my linq query, how can I do this Not In in simple manner?
You can use Contains with !. In addition, if you just want to count rows, you can use Count.
var ids = new List<int> {25, 346, 350, 352, 353, 354, 355, 357, 358, 366, 372, 411};
var querynonSystem = XXXcontext.Rpm_scrty_rpm_usrs.Count(x =>
x.Inact_ind == "N" &&
x.Email_id != null &&
x.Wwid == null &&
x.Email_id.Trim() != "" &&
!ids.Contains(x.Dflt_ste_id));
From comment: if you want to retrieve all, you can still use Where and Select.
var querynonSystem = XXXcontext.Rpm_scrty_rpm_usrs.Where(x =>
x.Inact_ind == "N" &&
x.Email_id != null &&
x.Wwid == null &&
x.Email_id.Trim() != "" &&
!ids.Contains(x.Dflt_ste_id)).Select(x => x).ToList();
FYI: you cannot call Rpm_scrty_rpm_usrs table class to query. Instead, you need DbContext or some other repository.
There is no "not in" operator unless the type of the query is the same as the type you want to filter against (in which case you could use except). Here it is not. You're working on an IEnumerable and you want to filter on it's ID so a list of int. The where with lambda and contains is your best bet and WILL be translated to an in on the SQL side by most providers.
var FilterIds = new List<int>{1,2,3,4,344,3423525};
var querynonSystem = (from ru in Rpm_scrty_rpm_usrs
where ru.Inact_ind == "N" && ru.Email_id != null && ru.Wwid == null && ru.Email_id.Trim() != ""
&& ru.Dflt_ste_id != 25
select ru)
// Use this
.Where(ru=>!FilterIds.Any(id=>ru.dflt_ste_id ==id))
// Or this
.Where(ru=>!FilterIds.Contains(ru.dflt_ste_id))
.Count();
I was originally using a foreach loop and then for each element in the loop, I perform a LINQ query like so:
foreach (MyObject identifier in identifiers.Where(i => i.IsMarkedForDeletion == false))
{
if (this.MyEntities.Identifiers.Where(pi => identifier.Field1 == pi.Field1 && identifier.Field2 == pi.Field2 && identifier.Field3 == pi.Field3).Any())
{
return false;
}
}
return true;
Then I modified it like so:
if (identifiers.Any(i => !i.IsMarkedForDeletion && this.MyEntities.Identifiers.Where(pi => i.Field1 == pi.Field1 && i.Field2 == pi.Field2 && i.Field3 == pi.Field3).Any()))
{
return false;
}
return true;
My question is this still the wrong way to use LINQ? Basically, I want to eliminate the need for the foreach loop (which seems like I should be able to get rid of it) and also make the DB query faster by not performing separate DB queries for each element of a list. Instead, I want to perform one query for all elements. Thanks!
You can change your code in this way, and it will be converted to SQL statement as expected.
To prevent runtime errors during transformation, it will be better to save DBSet to the IQueryable variable; identifiers should be IQueryable too, so you should change your code into something like this (to be honest, Resharper converted your foreach in this short labda):
IQueryable<MyObject2> identifiers = MyEntities.Identifiers.Where(i => i.IsMarkedForDeletion == false);
IQueryable<MyObject2> ids = MyEntities.Identifiers.AsQueryable();
return identifiers.All(identifier => !ids.Any(pi => identifier.Field1 == pi.Field1 && identifier.Field2 == pi.Field2 && identifier.Field3 == pi.Field3));
If identifiers is in memory collection you can change code in this way (hope that fields are string):
IQueryable<MyObject2> ids = MyEntities.Identifiers.AsQueryable();
string[] values = identifiers.Where(i => i.IsMarkedForDeletion == false).Select(i => String.Concat(i.Field1, i.Field2, i.Field3)).ToArray();
return !ids.Any(i => values.Contains(i.Field1 + i.Field2 + i.Field3));
Unfortunately your modified version will be executed exactly the same way (i.e. multiple database queries) as in the original foreach approach because EF does not support database query with joins to in memory collection (except for primitive and enumeration type collections), so if you try the most logical way
bool result = this.MyEntities.Identifiers.Any(pi => identifiers.Any(i =>
!i.IsMarkedForDeletion &&
i.Field1 == pi.Field1 && i.Field2 == pi.Field2 && i.Field3 == pi.Field3));
you'll get
NotSupportedException: Unable to create a constant value of type 'YourType'. Only primitive types or enumeration types are supported in this context.
The only way to let EF execute a single database query is to manually build a LINQ query with Concat per each item from in memory collection, like this
IQueryable<Identifier> query = null;
foreach (var item in identifiers.Where(i => !i.IsMarkedForDeletion))
{
var i = item;
var subquery = this.MyEntities.Identifiers.Where(pi =>
pi.Field1 == i.Field1 && pi.Field2 == i.Field2 && pi.Field3 == i.Field3);
query = query != null ? query.Concat(subquery) : subquery;
}
bool result = query != null && query.Any();
See Logging and Intercepting Database Operations of how to monitor the EF actions.
I would use it as follows:
if (identifiers.Where(i => !i.IsMarkedForDeletion &&
this.MyEntities.Identifiers.Field1 == i.Field1 &&
this.MyEntities.Identifiers.Field2 == i.Field2 &&
this.MyEntities.Identifiers.Field3 == i.Field3).Any()))
{
return false;
}
return true;
I hope this helps. Even though it is more to type out, it is more understandable and readable then using multiple 'where' statements.
I've got a method which can accept an optional int? value as a number of items to Take from a collection. I want to return all items if a null value is passed. Right now I have to duplicate my query to accomplish this
if(take == null)
{
x = db.WalkingDeadEps.Where(x => x.BicyclesCouldHaveSavedLives == true).ToList()
}
else
{
x = db.WalkingDeadEps.Where(x => x.BicyclesCouldHaveSavedLives == true).Take(take).ToList()
}
Is there a simpler way? Something like this?
.Take(take != null ? take : "all")
with Linq you have the option to store your query in variables. it will not be executed until you call ToList or equivalent methods on it.
var query = db.WalkingDeadEps.Where(x => x.BicyclesCouldHaveSavedLives == true);
x = take.HasValue ? query.Take(take.Value).ToList() : query.ToList();
I am trying to extend my linq query with additional search criteria to filter the data by sending also a List<Listitem> to the function for processing. The List can contain 1 or more items and the objective is to retreive all items which match any criteria.
Since i am sending several search criteria to the function the goal is to make a more accurate filter result the more information i am sending to the filter. If one or several criterias are empty then the filter will get less accurate results.
Exception is raised every time i execute following code, and I cant figure out how to solve the using statement to include the List<ListItem>. Appreciate all the help in advance!
Exception: Unable to create a constant value of type 'System.Web.UI.WebControls.ListItem'. Only primitive types or enumeration types are supported in this context.
using (var db = new DL.ENTS())
{
List<DL.PRODUCTS> products =
(from a in db.PRODUCTS
where (description == null || description == "" ||
a.DESCRIPTION.Contains(description)) &&
(active == null || active == "" || a.ACTIVE.Equals(active, StringComparison.CurrentCultureIgnoreCase)) &&
(mID == null || mID == "" || a.MEDIA_ID == mID) &&
(mID == null || objTypes.Any(s => s.Value == a.OBJECTS)) //Exception here!
select a).ToList<DL.PRODUCTS>();
return products;
}
Pass collection of primitive values to expression:
using (var db = new DL.ENTS())
{
var values = objTypes.Select(s => s.Value).ToArray();
List<DL.PRODUCTS> products =
(from a in db.PRODUCTS
where (description == null || description == "" || a.DESCRIPTION.Contains(description)) &&
(active == null || active == "" || a.ACTIVE.Equals(active, StringComparison.CurrentCultureIgnoreCase)) &&
(mID == null || mID == "" || a.MEDIA_ID == mID) &&
(mID == null || values.Contains(a.OBJECTS))
select a).ToList<DL.PRODUCTS>();
return products;
}
That will generate SQL IN clause.
Note - you can use lambda syntax to compose query by adding filters based on some conditions:
var products = db.PRODUCTS;
if (!String.IsNullOrEmpty(description))
products = products.Where(p => p.DESCRIPTION.Contains(description));
if (!String.IsNullOrEmpty(active))
products = products.Where(p => p.ACTIVE.Equals(active, StringComparison.CurrentCultureIgnoreCase)));
if (!String.IsNullOrEmpty(mID))
products = products.Where(p => p.MEDIA_ID == mID);
if (mID != null)
products = products.Where(p => values.Contains(p.OBJECTS));
return products.ToList();
Linq isn't able to convert the predicate on ListItem to something useful to Sql.
I would suggest that you pre-project the values of the ListItems into a simple List<string> before using this with Contains (which is converted to IN)
var listValues = objTypes.Select(_ => _.Value).ToList();
List<DL.PRODUCTS> products = ...
listValues.Contains(a.OBJECTS))
I use HtmlAgilityPack to get data from a website for a hobby project. I want to get the shoe article number from a site that sells shoes.
But my linq query doesn't return a string. Instead it returns the type:
System.Linq.Enumerable.WhereSelectEnumerableIterator<HtmlAgilityPack.HtmlNode,string>
How can I get the query to simply return a string?
foreach (var node in query)
{
Shoe shoe = new Shoe();
var num = from x in node.Descendants()
where x.Name == "img" && x.Attributes.Contains("class") && x.Attributes["class"].Value == "thumb lazy"
select x.Attributes["title"].Value.Substring(11);
shoe.articleNumber = Convert.ToInt32(num); //error
shoes.Add(shoe);
}
The error: InvalidCastException was unhandled.
Unable to cast object of type
'WhereSelectEnumerableIterator`2[HtmlAgilityPack.HtmlNode,System.String]'
to type 'System.IConvertible'.
Your LINQ query returns a collection. Use First/FirstOrDefault/Single/SingleOrDefault to get only one element from the collection:
var num = (from x in node.Descendants()
where x.Name == "img" && x.Attributes.Contains("class") && x.Attributes["class"].Value == "thumb lazy"
select x.Attributes["title"]).First().Value.Substring(11);
Assuming there's only going to be one entry that matches your query, you just need to modify it to use FirstOrDefault() or First():
foreach (var node in query)
{
Shoe shoe = new Shoe();
var num = (from x in node.Descendants()
where x.Name == "img" && x.Attributes.Contains("class") && x.Attributes["class"].Value == "thumb lazy"
select x.Attributes["title"].Value.Substring(11)).First();
shoe.articleNumber = Convert.ToInt32(num);
shoes.Add(shoe);
}
Just bear in mind that the above will fail with an exception if the item is not present.
your query does not return a single result but a list of results. In your case this will be a list containing a single id. You should modify it to this:
var num = (from x in node.Descendants()
where x.Name == "img" && x.Attributes.Contains("class") && x.Attributes["class"].Value == "thumb lazy"
select x.Attributes["title"].Value.Substring(11)).FirstOrDefault();
Difference between single, singleorDefault ,first ,FirstorDefault
Here