My code:
string name;
db.Entities.Where(m=>m.name==(name??m.name))
throw:Should be CHAR, but get NCLOB.
//----------------
decimal? id;
db.Entities.Where(m=>m.id==(id??m.id))
The code is working.
my database is oracle.
What should I do to make the code working.?
Thanks.
Beware that in Oracle an empty string is NULL.
I don't know about Oracle + EF and type mappings. But this is not the way to make predicates optional anyway, because now the part x ?? x.y will always be part of the expression that is translated into SQL. This may hit query optimization and performance.
The way to do this is:
IQueryable<Entity> query = db.Entities;
if (!string.IsNullOrWhiteSpace(name))
{
query = query.Where(m => m.Name == name);
}
Related
I am using jqgrid in MVC 4. I have written a method for getting a list in linq.
In my function I am getting all values of jqgrid with search criteria i.e. Operator AND/OR
, operations Equals to, not equals to etc. Here I am also getting the column name like Name, City, State etc.
My problem is I can't pass the column name directly in linq query i.e. I have to use the column name as x => x.Name
switch (rule.field)
{
case "Name":
query = query.Where(x => x.Name.StartsWith(rule.data, StringComparison.OrdinalIgnoreCase));
break;
case "Course":
query = query.Where(x => x.Course.StartsWith(rule.data, StringComparison.OrdinalIgnoreCase));
break;
}
In rule.field I am getting column Name i.e. Name, city, state etc. I want to pass the column name which I am getting in rule.filed in LINQ query instead of x =>x.Name.
Is there any way to do it so I can avoid writing switch cases?
You can use System.Linq.Dynamic, which can be installed as a Nuget package along with string.Format. The syntax would then look something like..
var newquery = query.AsQueryable()
.Where(
string.Format("{0}.ToUpper().StartsWith(#0)", rule.field)
,rule.data.ToUpper());
You could always use reflection:
query = query.ToList().Where(p => {
var field = p.GetType().GetProperty(rule.field);
var value = (String) field.GetValue(p);
return value.StartsWith(rule.data, StringComparison.OrdinalIgnoreCase);
});
Warning: Reflection is slow. Only use this for short lists. Since you're dealing with UI rendering, I'm assuming this won't be a problem.
edit: My example is assuming all properties are indeed properties (and not fields), and that all properties are Strings. You may need to alter the code for your specific case.
My Linq to entities query is written as below.
The datatype of DATECOLUMN1 in my ORACLE database is of string.
Datetime FilterStartDate = DateTime.Now;
var query = from c in db.TABLE1
join l in db.TABLE2 on c.FK equals l.PK
where (FilterStartDate >= DateTime.ParseExact(l.DATECOLUMN1, "dd/MM/yyyy", CultureInfo.InvariantCulture) : false) == true
select c;
Writing above query gives me an error of not supported. How can I convert DATECOLUMN1 into a datetime to compare it.
P.S. I do not have control over database schema, so changing datatype of column in Oracle database is not a feasible solution for me.
In you Model, add the following property to your partial class TABLE2:
public DateTime DATECOLUMN1_NEW
{
get
{
return DateTime.ParseExact(DATECOLUMN1, "dd/MM/yyyy", CultureInfo.InvariantCulture);
}
set { }
}
Then, in you LINQ query, use DATECOLUMN1_NEW (it's already in DateTime format) in place of DATECOLUMN1.
Erm.. I think the problem you are having is that you are putting ": false" in there.
It looks like you are trying to use a condtional operator (?:) but you forgot the "?".
I don't think you actually need this as you are just trying to determine if the date is greater or not. Also if ParseExact fails it will throw an exception (not what you want) so you should use TryParse instead and handle the true/false returned and the out value to determine whether or not the date is (a) Actually a date (b) less then FilterStartDate.
You can use two alternatives:
Use the function described in the answer here: How to I use TryParse in a linq query of xml data?
Use the following fluent syntax version which I think is more readable.
var query = db.Table1.Join(db.Table2, x => x.FK, y => y.PK, (x, y) => x).Where(x =>
{
DateTime Result;
DateTime.TryParse(x.Date, out Result);
return DateTime.TryParse(x.Date, out Result) && FilterStartDate >= Result;
});
I have a dynamic select statement thus:
"new(PurchaseOrderID as ID_PK, PContractNo + GoodsSupplier.AssociatedTo.DisplayName as Search_Results)"
As can be seen I wish to concatenate the 'PContractNo' and 'GoodsSupplier.AssociatedTo.DisplayName' fields into one returned field named 'Search_Results'. It is important that these two fields are combined.
However the Linq library complains regarding the '+', which the expression parser brings back as a 'Concat(etc...), which of course cannot be translated into a store expression.
Obviously therefore I would like some help regarding just how i should format the select string in order to do what I want. I've tried many things!
Any help would be greatly appreciated!
Thank You, Ian Mac
Create a new Class like
public class A
{
public String k;
public String v;
}
and use linq to join
res = from a in list
select new A
{
k = a.Key,
v = String.Concat(a.Key,a.Value)
};
I have a linq Entity called Enquiry, which has a property: string DateSubmitted.
I'm writing an app where I need to return IQueryable for Enquiry that have a DateSubmitted within a particular date range.
Ideally I'd like to write something like
IQueryable<Enquiry> query = Context.EnquirySet.AsQueryable<Enquiry>();
int dateStart = int.Parse("20090729");
int dateEnd = int.Parse("20090930");
query = (from e in query
where(enq => int.Parse(enq.DateSubmitted) < dateEnd)
where(enq => int.Parse(enq.DateSubmitted) > dateStart)
select e);
Obviously Linq to EF doesn't recognise int.Parse, so I think I can achieve what I want with an Expression method that returns a predicate???
I've been playing around with PredicateBuilder and looking all over but I've successfully fried my brains trying to work this out. Sure I could add another property to my Entity and convert it there but I'd really like to understand this. Can anyone explain or give an example/link that doesn't fry my brains?
Thanks in advance
Mark
If you know your date strings are valid, and they're really in that order (which is a natural sort order) you might be able to get away with string comparisons:
IQueryable<Enquiry> query = Context.EnquirySet.AsQueryable<Enquiry>();
string dateStart ="20090729";
string dateEnd = "20090930";
query = (from e in query
where(enq => enq.DateSubmitted.CompareTo(dateEnd)) < 0)
where(enq => enq.DateSubmitted.CompareTo(dateStart)) > 0)
select e);
I have a need to replace multiple strings with others in a query
from p in dx.Table
where p.Field.Replace("A", "a").Replace("B", "b").ToLower() = SomeVar
select p
Which provides a nice single SQL statement with the relevant REPLACE() sql commands.
All good :)
I need to do this in a few queries around the application... So i'm looking for some help in this regard; that will work as above as a single SQL hit/command on the server
It seems from looking around i can't use RegEx as there is no SQL eq
Being a LINQ newbie is there a nice way for me to do this?
eg is it possible to get it as a IQueryable "var result" say and pass that to a function to add needed .Replace()'s and pass back? Can i get a quick example of how if so?
EDIT: This seems to work! does it look like it would be a problem?
var data = from p in dx.Videos select p;
data = AddReplacements(data, checkMediaItem);
theitem = data.FirstOrDefault();
...
public IQueryable<Video> AddReplacements(IQueryable<Video> DataSet, string checkMediaItem)
{
return DataSet.Where(p =>
p.Title.Replace(" ", "-").Replace("&", "-").Replace("?", "-") == checkMediaItem);
}
Wouldn't it be more performant to reverse what you are trying to do here, ie reformat the string you are checking against rather than reformatting the data in the database?
public IQueryable<Video> AddReplacements(IQueryable<Video> DataSet, string checkMediaItem)
{
var str = checkMediaItem.Replace("-", "?").Replace("&", "-").Replace("-", " "));
return DataSet.Where(p => p.Title == str);
}
Thus you are now comparing the field in the database with a set value, rather than scanning the table and transforming the data in each row and comparing that.