Following are the columns name and its data type:
TemplateName string
TaskName string
AvgDays string (but contains int values)
my LINQ query:
DataTable newTable = new DataTable();
newTable = (from r in table.AsEnumerable()
group r by r.Field<string>("TemplateName") into templates
let totalDays = templates.Sum(r => r.Field<int>("AvgDays"))
from t in templates
group t by new
{
TemplateName = templates.Key,
TaskName = t.Field<string>("TaskName"),
TotalDays = totalDays
} into tasks
select new
{
tasks.Key.TemplateName,
tasks.Key.TaskName,
AvgDays = tasks.Sum(r => r.Field<int>("AvgDays")),
tasks.Key.TotalDays
}).CopyToDataTable();
I am getting error after execution of query. Error is "Specified cast is not valid.".
Please let me know where am I doing wrong.
I guess Field<int> causes the problem, because the field is not really an int. As you said before, it's a string.
Try following
AvgDays = tasks.Sum(r => int.Parse(r.Field<string>("AvgDays"))),
Field<T> does not perform some magic transformation between different types. It's implemented like that:
return (T)((object)value);
In practice there is a little bit more code, but logic is the same. It just tried to cast your value to desired type, and as you probably know, casting string to int does not work.
Related
I have some errors using Linq on DataTable and I couldn't figure it out how to solve it. I have to admit that i am pretty new to Linq and I searched the forum and Internet and couldn't figure it out. hope you can help.
I have a DataTable called campaign with three columns: ID (int), Product (string), Channel (string). The DataTable is already filled with data. I am trying to select a subset of the campaign records which satisfied the conditions selected by the end user. For example, the user want to list only if the Product is either 'EWH' or 'HEC'. The selection criteria is dynaically determined by the end user.
I have the following C# code:
private void btnClick()
{
IEnumerable<DataRow> query =
from zz in campaign.AsEnumerable()
orderby zz.Field<string>("ID")
select zz;
string whereClause = "zz.Field<string>(\"Product\") in ('EWH','HEC')";
query = query.Where(whereClause);
DataTable sublist = query.CopyToDataTable<DataRow>();
}
But it gives me an error on line: query = query.Where(whereClause), saying
No property or field 'zz' exists in type 'DataRow'".
If I changed to:
string whereClause = "Product in ('EWH','HEC')"; it will say:
No property or field 'Product' exists in type 'DataRow'
Can anyone help me on how to solve this problem? I feel it could be a pretty simple syntax change, but I just don't know at this time.
First, this line has an error
orderby zz.Field<string>("ID")
because as you said, your ID column is of type int.
Second, you need to learn LINQ query syntax. Forget about strings, the same way you used from, orderby, select in the query, you can also use where and many other operators. Also you'll need to learn the equivalent LINQ constructs for SQL-ish things, like for instance IN (...) is mapped to Enumerable.Contains etc.
With all that being said, here is your query
var productFilter = new[] { "EWH", "HEC" };
var query =
from zz in campaign.AsEnumerable()
where productFilter.Contains(zz.Field<string>("Product"))
orderby zz.Field<int>("ID")
select zz;
Update As per your comment, if you want to make this dynamic, then you need to switch to lambda syntax. Multiple and criteria can be composed by chaining multiple Where clauses like this
List<string> productFilter = ...; // coming from outside
List<string> channelFilter = ...; // coming from outside
var query = campaign.AsEnumerable();
// Apply filters if needed
if (productFilter != null && productFilter.Count > 0)
query = query.Where(zz => productFilter.Contains(zz.Field<string>("Product")));
if (channelFilter != null && channelFilter.Count > 0)
query = query.Where(zz => channelFilter.Contains(zz.Field<string>("Channel")));
// Once finished with filtering, do the ordering
query = query.OrderBy(zz => zz.Field<int>("ID"));
am a newbie in linq.. am stuck with one scenario. ie,
i have to sort the search results based on user input.
user inputs are Last Name, First Name and Title. for input 3 drop downs are there and i have to sort result based on the values selected.
i tried
order = Request["orders"].Split(',');
var param = order[0];
var p1 = typeof(Test).GetProperty(param);
param = order[1];
var p2 = typeof(Test).GetProperty(param);
param = order[2];
var p3 = typeof(Test).GetProperty(param);
model.Test = (from tests in model.Test
select tests).
OrderBy(x => p1.GetValue(x, null)).
ThenBy(x => p2.GetValue(x, null)).
ThenBy(x => p3.GetValue(x, null));
but it doesn't works.
i want qry like this
from tests in model.Test
select tests).OrderBy(x => x.lastname).
ThenBy(x => x.firstname).ThenBy(x => x.Title);
order[0]== lastname but how can i use it in the place of OrderBy(x => x.order[0])..?
Thanks in advance.
i solved my case as follows
// list of columns to be used for sorting
List<string>order = Request["orders"].Split(',').ToList();
//map the column string to property
var mapp = new Dictionary<string, Func<Test, string>>
{
{"FirstName", x => x.FirstName},
{"LastName", x => x.LastName},
{"SimpleTitle", x => x.SimpleTitle}
};
//user inputted order
var paras = new List<Func<Test, string>>();
foreach (var para in order)
{
if(!string.IsNullOrEmpty(para))
paras.Add(mapp[para]);
}
//sorting
model.Test= model.Test.OrderBy(paras[0]).ThenBy(paras[1]).ThenBy(paras[2]);
Thanks all,
Actually you are looking for dynamic linq query than you can try out Dynamic LINQ (Part 1: Using the LINQ Dynamic Query Library)
which allow to do like this
it means you can dynamically pass string propertyname to short you collection in orderby function
You can also read about : Dynamic query with Linq
You can compose the expression (any Expression) manually from pieces and then append it to the previous part of query. You can find more info, with example in "Sorting in IQueryable using string as column name".
I have the following code
nodes = data.Descendants(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}Results")).Nodes();
System.Collections.Generic.IEnumerable<Result> res = new List<Result>();
if (nodes.Count() > 0)
{
var results = from uris in nodes
select new Result
{
URL =
((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}Url")).Value,
Title =
((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}Title")).Value,
Description =
((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}Description")).Value,
DateTime =
((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}DateTime")).Value,
};
res = results;
}
Where Results is a object who has those URL, Title, Description, and DateTime variables defined.
This all works fine normally, but when a 'node' in nodes doesnt contain a Description element (or at least I think thats whats throwing it) the program hits the "res = results;"
line of code and throws a 'object reference not set to...' error and highlights the whole section right after "select new Results"..
How do I fix this?
The simplest way is to cast to string instead of using the Value property. That way you'll end up with a null reference for the Description instead.
However, your code can also be made a lot nicer:
XNamespace ns = "http://schemas.microsoft.com/LiveSearch/2008/04/XML/web";
var results = data.Descendants(ns + "Results")
.Elements()
.Select(x => new Result
{
URL = (string) x.Element(ns + "Url"),
Title = (string) x.Element(ns + "Title"),
Description = (string) x.Element(ns + "Description"),
DateTime = (string) x.Element(ns + "DateTime")
})
.ToList();
See how much simpler that is? Techiques used:
Calling ToList() on an empty sequence gives you a list anyway
This way you'll only ever perform the query once; before you were calling Count() which would potentially have iterated over each node. In general, use Any() instead of Count() > 0) - but this time just making the list unconditional is simpler.
Use the Elements() method to get child elements, rather than casting multiple times. (Your previous code would have thrown an exception if it had encountered any non-element nodes)
Use the implicit conversion from string to XNamespace
Use the +(XNamespace, string) operator to get an XName
If the Description element is not included you should test if this
((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}Description"))
is not null before using Value. Try this code:
var results = from uris in nodes let des = ((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}Description"))
select new Result
{
URL = ((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}Url")).Value,
Title = ((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}Title")).Value,
Description = (des != null) ? des.Value : string.Empty,
DateTime = ((XElement)uris).Element(XName.Get("{http://schemas.microsoft.com/LiveSearch/2008/04/XML/web}DateTime")).Value,
};
I'm very new to Linq, I can find multi-line data reading examples everywhere (by using foreach()), but what is the correct way of reading a single line of data? Like a classic Product Detail page.
Below is what I tried:
var q = from c in db.Products
where c.ProductId == ProductId
select new { c.ProductName, c.ProductDescription, c.ProductPrice, c.ProductDate };
string strProductName = q.First().ProductName.ToString();
string strProductDescription = q.First().ProductDescription.ToString();
string strProductPrice = q.First().ProductPrice.ToString();
string strProductDate = q.First().ProductDate.ToString();
The code looks good to me, but when I see the actual SQL expressions generated by using SQL Profiler, it makes me scared! The program executed four Sql expressions and they are exactly the same!
Because I'm reading four columns from a single line. I think I must did something wrong, so I was wondering what is the right way of doing this?
Thanks!
Using the First() extension method would throw the System.InvalidOperationException when no element in a sequence satisfies a specified condition.
If you use the FirstOrDefault() extension method, you can test against the returned object to see if it's null or not.
FirstOrDefault returns the first element of a sequence, or a default value if the sequence contains no elements; in this case the default value of a Product should be null. Attempting to access the properties on this null object will throw ArgumentNullException
var q = (from c in db.Products
where c.ProductId == ProductId
select new { c.ProductName, c.ProductDescription, c.ProductPrice, c.ProductDate }).FirstOrDefault();
if (q != null)
{
string strProductName = q.ProductName;
string strProductDescription = q.ProductDescription;
string strProductPrice = q.ProductPrice;
string strProductDate = q.ProductDate;
}
Also, you shouldn't have to cast each Property ToString() if you're object model is setup correctly. ProductName, ProductDescription, etc.. should already be a string.
The reason you're getting 4 separate sql queries, is because each time you call q.First().<PropertyHere> linq is generating a new Query.
var q = (from c in db.Products
where c.ProductId == ProductId
select new { c.ProductName, c.ProductDescription, c.ProductPrice, c.ProductDate }
).First ();
string strProductName = q.ProductName.ToString();
string strProductDescription = q.ProductDescription.ToString();
string strProductPrice = q.ProductPrice.ToString();
string strProductDate = q.ProductDate.ToString();
I have the following LINQ query:
DataClassesDataContext dc = new DataClassesDataContext();
var query = from contact in dc.Contacts
select new
{
ContactId = contact.ContactId,
LastName = contact.LastName,
FirstName = contact.FirstName,
Addresses = contact.Addresses,
Phones = contact.Phones,
DOB = contact.BirthDate,
LastNote = contact.Notes.Max(n => n.Created), //this line causes the error
Status = contact.ContactStatus.ContactStatusName,
EmailAddress = contact.Emails
};
The line where I get the maximum created date for the notes collection causes the following exception to be thrown:
Exception: The null value cannot be assigned to a
member with type System.DateTime which
is a non-nullable value type.
How do I write the query to allow null values into the LastNote field? The DOB field is defined as DateTime? and has not problem handling nulls.
Think I figured it out.
If I cast the maximum note value to a nullable DateTime it seems to eliminate the exception. The following change worked for me:
LastNote = (Nullable<DateTime>)contact.Notes.Max(n => n.Created)
As others have pointed out, it can also be written using the shorthand notation for a nullable DateTime as follows:
LastNote = (DateTime?) contact.Notes.Max(n => n.Created)
Rewrite that line as:
LastNote = (DateTime?) contact.Notes.Max(n => n.Created),
LastNote = contact.Notes.Max(n => (DateTime?)n.Created)
Couldn't find this on the net so i hope this helps others.
In VB is something like:
LastNote = CType(contact.Notes.Max(n => n.Created), Global.System.Nullable(Of Date))
I think...
You could do that, or you could alter your database schema so that the column 'Created' does not allow nulls.
The scenario is arising because one of the rows comes back with a null value for Created.
If the db didn't allow nulls, the scenario would never occur.