Related
I am new to Web API, Entity Framework and OData. I asked a similar question in another forum but haven't gotten a relevant response.
We have a OData compliant web api service for use in Salesforce. We have a custom complex query in Oracle that we need to expose.
I am not sure how to use a custom query like we want to also allow for odata parameter filtering to occur? ($filter, $top, $skip, etc) For example, when a $filter is used i want to apply that filter to the custom query and then send it back to the database to have it return the result set. How can i do this?
The issue i seem to have is that I can see the parameters as they come in but they are not translating to the query being passed to oracle. It seems that it will fire the query returning the full result set and then apply the parameters. This is very slow as the result set is very large.
I am hoping 2 figure out 2 things
1. How can i use custom sql and apply odata parameters to the underlying query?
2. When using EF or a custom query, how can i apply odata parameters to the query so that when the query is sent to the database that the $filter parameter, for example, is included in the query? I don't want the full result returned then apply the filter.
Can anyone give me some pointers on how to make this happen?
private static ODataValidationSettings _validationSettings = new ODataValidationSettings();
//public IHttpActionResult GetName()
//{ }
// GET: odata/ShareData
[ODataRoute("Orders")]
[EnableQuery(PageSize = 50)]
public IHttpActionResult GetOrders(ODataQueryOptions<Orders> queryOptions)
{
// validate the query.
try
{
queryOptions.Validate(_validationSettings);
}
catch (ODataException ex)
{
return BadRequest(ex.Message);
}
try
{
string connectionString = ConfigurationManager.ConnectionStrings["DNATestConnectionString"].ConnectionString;
var items = GetDataItems(connectionString);
return Ok<IEnumerable<Orders>>(items);
}
catch (Exception ex)
{
return StatusCode(HttpStatusCode.InternalServerError);
}
}
#region Load Data Methods
private static List<Orders> GetDataItems(string connectionString)
{
List<Orders> items = new List<Orders>();
using (OracleConnection con = new OracleConnection(connectionString))
{
con.Open();
using (OracleCommand cmd = con.CreateCommand())
{
cmd.CommandText = "select po_header_id, segment1, vendor_id, vendor_site_id from po_headers_all where vendor_id=4993";
using (OracleDataReader rdr = cmd.ExecuteReader())
{
while (rdr.Read())
items.Add(ToOrders(rdr));
}
}
}
return items;
}
private static Orders ToOrders(OracleDataReader rdr)
{
Orders data = new Orders();
data.VENDOR_ID = ToInt32(rdr, "VENDOR_ID");
data.VENDOR_SITE_ID = ToInt32(rdr, "VENDOR_SITE_ID");
data.PO_HEADER_ID = ToInt32(rdr, "PO_HEADER_ID");
data.SEGMENT1 = Convert.ToString(rdr["SEGMENT1"]);
return data;
}
private static int ToInt32(OracleDataReader rdr, string name)
{
int index = rdr.GetOrdinal(name);
return rdr.IsDBNull(index) ? 0 : Convert.ToInt32(rdr[index]);
}
#endregion
I don't think this is possible.
How can i use custom sql and apply odata parameters to the underlying query?
As far as I'm aware, you can't. The whole point of the OData library is that it needs to work off an IQueryable. By using custom SQL in a string like you have in your example, you can't combine it with the OData parameters that are being passed in.
One approach would be to have your custom SQL in a SQL view, then add the SQL view to your EF model in the same way as you would add a table - it will be represented as a DbSet just like tables are.
You can then get an IQueryable to represent the dataset and then apply the OData parameters as follows:
public IHttpActionResult GetOrders(ODataQueryOptions<OrdersView> queryOptions)
{
IQueryable<OrdersView> allData = // ... get the DbSet from entity framework...
// this will apply the OData query to the data set and only pull the data you want from the database
var filteredResults = queryOptions.ApplyTo(allData) as IQueryable<OrdersView>;
return Ok<IQueryable<OrdersView>>(filteredResults);
}
I'm trying out Dapper with Oracle and I was trying to run a multi-resultset query but Oracle requires a dbtype of refcursor.
StringBuilder query = new StringBuilder("BEGIN ");
query.Append("OPEN :rs1 FOR SELECT * FROM Table1 where key=:KEY; ");
query.Append("OPEN :rs2 FOR SELECT * FROM Table2 where key=:KEY; ");
query.Append("END;");
Is there a way to pass an OracleParameter (maybe as DbParameter?) to Dapper? When I tried, it threw an error.
What is the advantage of using the DynamicParameter vs. using a DbParameter (assuming types are known etc.)?
A new interface was added in the most recent build that allows more control over the parameter - it was added to support TVPs in SQL server, but should work in this scenario. However, I'm also fairly content to add special casing for any types that look like dbparameter - and add them directly, which would allow you to add an oracleparameter directly.
DynamicParameters is about how many parameters to add, so it is a bit orthogonal to the value vs DbParameter discussion. At the moment, the code generally prefers to take control of adding parameters itself, so the caller just knows "an int named id with value 7" - not any ado.net details. But it could do.
Edit: if you really want to work with lists-of-parameters (i.e. List<DbParameter> etc), then you can do that with something like:
public class DbParams : Dapper.SqlMapper.IDynamicParameters,
IEnumerable<IDbDataParameter>
{
private readonly List<IDbDataParameter> parameters =
new List<IDbDataParameter>();
public IEnumerator<IDbDataParameter> GetEnumerator() {
return parameters.GetEnumerator(); }
IEnumerator IEnumerable.GetEnumerator() { return GetEnumerator(); }
public void Add(IDbDataParameter value)
{
parameters.Add(value);
}
void Dapper.SqlMapper.IDynamicParameters.AddParameters(IDbCommand command,
Dapper.SqlMapper.Identity identity)
{
foreach (IDbDataParameter parameter in parameters)
command.Parameters.Add(parameter);
}
}
with usage like:
public void TestCustomParameters()
{
var args = new DbParams {
new SqlParameter("foo", 123),
new SqlParameter("bar", "abc")
};
var result = connection.Query("select Foo=#foo, Bar=#bar", args).Single();
int foo = result.Foo;
string bar = result.Bar;
foo.IsEqualTo(123);
bar.IsEqualTo("abc");
}
which passes the test.
However, I must stress that I would prefer not to encumber the calling code with db-parameter knowledge unless it really really needs to know; I would by-far prefer:
var args = new {
foo = 123, bar = "abc"
};
which does exactly the same thing, but without dropping to ADO.NET; this can be especially important if you are using a "decorated" ADO.NET connection (for example, mini-profiler) - in which case the layer you get is not an OracleCommand / OracleConnection etc - it is abstracted. This means that forcibly adding an OracleParameter may not always work - but adding a parameter with name "foo" and value 123 - that is pretty reliable.
Note: I'm specifically not using Fluent NHibernate but am using 3.x's built-in mapping style. However, I am getting a blank recordset when I think I should be getting records returned.
I'm sure I'm doing something wrong and it's driving me up a wall. :)
Background / Setup
I have an Oracle 11g database for a product by IBM called Maximo
This product has a table called workorder which lists workorders; that table has a field called "wonum" which represents a unique work order number.
I have a "reporting" user which can access the table via the maximo schema
e.g. "select * from maximo.workorder"
I am using Oracle's Managed ODP.NET DLL to accomplish data tasks, and using it for the first time.
Things I've Tried
I created a basic console application to test this
I added the OracleManagedClientDriver.cs from the NHibernate.Driver on the master branch (it is not officially in the release I'm using).
I created a POCO called WorkorderBriefBrief, which only has a WorkorderNumber field.
I created a class map, WorkorderBriefBriefMap, which maps only that value as a read-only value.
I created a console application with console output to attempt to write the lines of work orders.
The session and transaction appear to open correct,
I tested a standard ODP.NET OracleConnection to my connection string
The Code
POCO: WorkorderBriefBrief.cs
namespace PEApps.Model.WorkorderQuery
{
public class WorkorderBriefBrief
{
public virtual string WorkorderNumber { get; set; }
}
}
Mapping: WorkorderBriefBriefMap.cs
using NHibernate.Mapping.ByCode;
using NHibernate.Mapping.ByCode.Conformist;
using PEApps.Model.WorkorderQuery;
namespace ConsoleTests
{
public class WorkorderBriefBriefMap : ClassMapping<WorkorderBriefBrief>
{
public WorkorderBriefBriefMap()
{
Schema("MAXIMO");
Table("WORKORDER");
Property(x=>x.WorkorderNumber, m =>
{
m.Access(Accessor.ReadOnly);
m.Column("WONUM");
});
}
}
}
Putting it Together: Program.cs
namespace ConsoleTests
{
class Program
{
static void Main(string[] args)
{
NHibernateProfiler.Initialize();
try
{
var cfg = new Configuration();
cfg
.DataBaseIntegration(db =>
{
db.ConnectionString = "[Redacted]";
db.Dialect<Oracle10gDialect>();
db.Driver<OracleManagedDataClientDriver>();
db.KeywordsAutoImport = Hbm2DDLKeyWords.AutoQuote;
db.BatchSize = 500;
db.LogSqlInConsole = true;
})
.AddAssembly(typeof(WorkorderBriefBriefMap).Assembly)
.SessionFactory().GenerateStatistics();
var factory = cfg.BuildSessionFactory();
List<WorkorderBriefBrief> query;
using (var session = factory.OpenSession())
{
Console.WriteLine("session opened");
Console.ReadLine();
using (var transaction = session.BeginTransaction())
{
Console.WriteLine("transaction opened");
Console.ReadLine();
query =
(from workorderbriefbrief in session.Query<WorkorderBriefBrief>() select workorderbriefbrief)
.ToList();
transaction.Commit();
Console.WriteLine("Transaction Committed");
}
}
Console.WriteLine("result length is {0}", query.Count);
Console.WriteLine("about to write WOs");
foreach (WorkorderBriefBrief wo in query)
{
Console.WriteLine("{0}", wo.WorkorderNumber);
}
Console.WriteLine("DONE!");
Console.ReadLine();
// Test a standard connection below
string constr = "[Redacted]";
OracleConnection con = new OracleConnection(constr);
con.Open();
Console.WriteLine("Connected to Oracle Database {0}, {1}", con.ServerVersion, con.DatabaseName.ToString());
con.Dispose();
Console.WriteLine("Press RETURN to exit.");
Console.ReadLine();
}
catch (Exception ex)
{
Console.WriteLine("Error : {0}", ex);
Console.ReadLine();
}
}
}
}
Thanks in advance for any help you can give!
Update
The following code (standard ADO.NET with OracleDataReader) works fine, returning the 16 workorder numbers that it should. To me, this points to my use of NHibernate more than the Oracle Managed ODP.NET. So I'm hoping it's just something stupid that I did above in the mapping or configuration.
// Test a standard connection below
string constr = "[Redacted]";
OracleConnection con = new Oracle.ManagedDataAccess.Client.OracleConnection(constr);
con.Open();
Console.WriteLine("Connected to Oracle Database {0}, {1}", con.ServerVersion, con.DatabaseName);
var cmd = new OracleCommand();
cmd.Connection = con;
cmd.CommandText = "select wonum from maximo.workorder where upper(reportedby) = 'MAXADMIN'";
cmd.CommandType = CommandType.Text;
Oracle.ManagedDataAccess.Client.OracleDataReader reader = cmd.ExecuteReader();
while (reader.Read())
{
Console.WriteLine(reader.GetString(0));
}
con.Dispose();
When configuring NHibernate, you need to tell it about your mappings.
I found the answer -- thanks to Oskar's initial suggestion, I realized it wasn't just that I hadn't added the assembly, I also needed to create a new mapper.
to do this, I added the following code to the configuration before building my session factory:
var mapper = new ModelMapper();
//define mappingType(s) -- could be an array; in my case it was just 1
var mappingType = typeof (WorkorderBriefBriefMap);
//use AddMappings instead if you're mapping an array
mapper.AddMapping(mappingType);
//add the compiled results of the mapper to the configuration
cfg.AddMapping(mapper.CompileMappingForAllExplicitlyAddedEntities());
var factory = cfg.BuildSessionFactory();
Using Linq commands and Linq To SQL datacontext, Im trying to instance an Entity called "Produccion" from my datacontext in this way:
Demo.View.Data.PRODUCCION pocoProduccion =
(
from m in db.MEDICOXPROMOTORs
join a in db.ATENCIONs on m.cmp equals a.cmp
join e in db.EXAMENXATENCIONs on a.numeroatencion equals e.numeroatencion
join c in db.CITAs on e.numerocita equals c.numerocita
where e.codigo == codigoExamenxAtencion
select new Demo.View.Data.PRODUCCION
{
cmp = a.cmp,
bonificacion = comi,
valorventa = precioEstudio,
codigoestudio = lblCodigoEstudio.Content.ToString(),
codigopaciente = Convert.ToInt32(lblCodigoPaciente.Content.ToString()),
codigoproduccion = Convert.ToInt32(lblNroInforme.Content.ToString()),
codigopromotor = m.codigopromotor,
fecha = Convert.ToDateTime(DateTime.Today.ToShortDateString()),
numeroinforme = Convert.ToInt32(lblNroInforme.Content.ToString()),
revisado = false,
codigozona = (c.codigozona.Value == null ? Convert.ToInt32(c.codigozona) : 0),
codigoclinica = Convert.ToInt32(c.codigoclinica),
codigoclase = e.codigoclase,
}
).FirstOrDefault();
While executing the above code, I'm getting the following error that the stack trace is included:
System.NotSupportedException was caught
Message="The explicit construction of the entity type 'Demo.View.Data.PRODUCCION' in a query is not allowed."
Source="System.Data.Linq"
StackTrace:
en System.Data.Linq.SqlClient.QueryConverter.VisitMemberInit(MemberInitExpression init)
en System.Data.Linq.SqlClient.QueryConverter.VisitInner(Expression node)
en System.Data.Linq.SqlClient.QueryConverter.Visit(Expression node)
en System.Data.Linq.SqlClient.QueryConverter.VisitSelect(Expression sequence, LambdaExpression selector)
en System.Data.Linq.SqlClient.QueryConverter.VisitSequenceOperatorCall(MethodCallExpression mc)
en System.Data.Linq.SqlClient.QueryConverter.VisitMethodCall(MethodCallExpression mc)
en System.Data.Linq.SqlClient.QueryConverter.VisitInner(Expression node)
en System.Data.Linq.SqlClient.QueryConverter.Visit(Expression node)
en System.Data.Linq.SqlClient.QueryConverter.VisitFirst(Expression sequence, LambdaExpression lambda, Boolean isFirst)
en System.Data.Linq.SqlClient.QueryConverter.VisitSequenceOperatorCall(MethodCallExpression mc)
en System.Data.Linq.SqlClient.QueryConverter.VisitMethodCall(MethodCallExpression mc)
en System.Data.Linq.SqlClient.QueryConverter.VisitInner(Expression node)
en System.Data.Linq.SqlClient.QueryConverter.ConvertOuter(Expression node)
en System.Data.Linq.SqlClient.SqlProvider.BuildQuery(Expression query, SqlNodeAnnotations annotations)
en System.Data.Linq.SqlClient.SqlProvider.System.Data.Linq.Provider.IProvider.Execute(Expression query)
en System.Data.Linq.DataQuery`1.System.Linq.IQueryProvider.Execute[S](Expression expression)
en System.Linq.Queryable.FirstOrDefault[TSource](IQueryable`1 source)
en Demo.View.InformeMedico.realizarProduccionInforme(Int32 codigoExamenxAtencion, Double precioEstudio, Int32 comi) en D:\cs_InformeMedico\app\InformeMedico.xaml.cs:línea 602
en Demo.View.InformeMedico.UpdateEstadoEstudio(Int32 codigo, Char state) en D:\cs_InformeMedico\app\InformeMedico.xaml.cs:línea 591
en Demo.View.InformeMedico.btnGuardar_Click(Object sender, RoutedEventArgs e) en D:\cs_InformeMedico\app\InformeMedico.xaml.cs:línea 683
InnerException:
Is that now allowed in LINQ2SQL?
Entities can be created outside of queries and inserted into the data store using a DataContext. You can then retrieve them using queries. However, you can't create entities as part of a query.
I am finding this limitation to be very annoying, and going against the common trend of not using SELECT * in queries.
Still with c# anonymous types there is a workaround, by fetching the objects into an anonymous type, and then copy it over into the correct type.
For example:
var q = from emp in employees where emp.ID !=0
select new {Name = emp.First + " " + emp.Last, EmployeeId = emp.ID }
var r = q.ToList();
List<User> users = new List<User>(r.Select(new User
{
Name = r.Name,
EmployeeId = r.EmployeeId
}));
And in the case when we deal with a single value (as in the situation described in the question) it is even easier, and we just need to copy directly the values:
var q = from emp in employees where emp.ID !=0
select new { Name = emp.First + " " + emp.Last, EmployeeId = emp.ID }
var r = q.FirstOrDefault();
User user = new User { Name = r.Name, EmployeeId = r.ID };
If the name of the properties match the database columns we can do it even simpler in the query, by doing select
var q = from emp in employees where emp.ID !=0
select new { emp.First, emp.Last, emp.ID }
One might go ahead and write a lambda expression that can copy automatically based on the property name, without needing to specify the values explictly.
Here's another workaround:
Make a class that derives from your LINQ to SQL class. I'm assuming that the L2S class that you want to return is Order:
internal class OrderView : Order { }
Now write the query this way:
var query = from o in db.Order
select new OrderView // instead of Order
{
OrderID = o.OrderID,
OrderDate = o.OrderDate,
// etc.
};
Cast the result back into Order, like this:
return query.Cast<Order>().ToList(); // or .FirstOrDefault()
(or use something more sensible, like BLToolkit / LINQ to DB)
Note: I haven't tested to see if tracking works or not; it works to retrieve data, which is what I needed.
I have found that if you do a .ToList() on the query before trying to contruct new objects it works
I just ran into the same issue.
I found a very easy solution.
var a = att as Attachment;
Func<Culture, AttachmentCulture> make =
c => new AttachmentCulture { Culture = c };
var culs = from c in dc.Cultures
let ac = c.AttachmentCultures.SingleOrDefault(
x => x.Attachment == a)
select ac == null ? make(c) : ac;
return culs;
I construct an anonymous type, use IEnumerable (which preserves deferred execution), and then re-consruct the datacontext object. Both Employee and Manager are datacontext objects:
var q = dc.Employees.Where(p => p.IsManager == 1)
.Select(p => new { Id = p.Id, Name = p.Name })
.AsEnumerable()
.Select(item => new Manager() { Id = item.Id, Name = item.Name });
Within the book "70-515 Web Applications Development with Microsoft .NET Framework 4 - Self paced training kit", page 638 has the following example to output results to a strongly typed object:
IEnumerable<User> users = from emp in employees where emp.ID !=0
select new User
{
Name = emp.First + " " + emp.Last,
EmployeeId = emp.ID
}
Mark Pecks advice appears to contradict this book - however, for me this example still displays the above error as well, leaving me somewhat confused. Is this linked to version differences? Any suggestions welcome.
I found another workaround for the problem that even lets you retain your result as IQueryale, so it doesn't actually execute the query until you want it to be executed (like it would with the ToList() method).
So linq doesn't allow you to create an entity as a part of query? You can shift that task to the database itself and create a function that will grab the data you want. After you import the function to your data context, you just need to set the result type to the one you want.
I found out about this when I had to write a piece of code that would produce a IQueryable<T> in which the items don't actually exist in the table containing T.
pbz posted a work around by creating a View class inherited from an entity class that you could be working with. I'm working with a dbml model of a table that has > 200 columns. When I try and return the whole table I get "Root Element missing" errors. I couldn't find anyone who wanted to deal with my particular issue so I was looking at rewriting my entire approach. Just creating a view class for the entitiy class worked in my case.
As pbz suggests : Create a view class that inherits from your entity class. For me this is tbCamp so :
internal class tbCampView : tbCamp
{
}
Then use the view class in your query :
using (var dc = ConnectionClass.Connect(Dev))
{
var camps = dc.tbCamps.Select(s => new tbCampView
{
active = s.active,
idCamp = s.idCamp,
campName = s.campName
});
SmartTableViewer(camps, dg1);
}
private void SmartTableViewer<T>(IEnumerable<T> allRecords)
{
// Build sorted rows back into new table
var table = new DataTable();
// Create columns based on type
if (allRecords is IEnumerable<tbCamp> tbCampRecords)
{
// Get the columns you want
table.Columns.Add("idCamp");
table.Columns.Add("campName");
foreach (var record in tbCampRecords)
{
// Make a new row
var r = table.NewRow();
// Add the contents to each column of the row
r["idCamp"] = record.idCamp;
r["campName"] = record.campName;
// Add the row to the table.
table.Rows.Add(r);
}
}
else
{
MessageBox.Show("Unhandled type. Add support for new data type in SmartTableViewer()");
return;
}
// Update table in grid
dg1.DataSource = table.DefaultView;
}
Here is what happens when you try and create an entity class object in the query.
I didn't want to have to use an anonymous type if I could help it because I wanted the type to be tbCamp. Since tbCampView is of type tbCamp the is operator works well. see Brian Hasden's answer Passing a generic List<> in C#
I'm surprised this is even an issue but with larger tables I run into this error so I thought I would just show it here :
When trying to read this table into memory I get the following error. There are < 2000 rows but the columns are > 200 for each. I don't know if that is an issue or not.
If I just want a few columns I need to create a custom class and handle that which isn't that big of a pain. With the approach pbz provided I don't have to worry about it.
Here is the entire project in case it helps someone.
public partial class Form1 : Form
{
private const bool Dev = true;
public Form1()
{
InitializeComponent();
}
private void btnGetAllCamps_Click(object sender, EventArgs e)
{
using (var dc = ConnectionClass.Connect(Dev))
{
IQueryable<tbCampView> camps = dc.tbCamps.Select(s => new tbCampView
{
// Project columns as needed.
active = s.active,
idCamp = s.idCamp,
campName = s.campName
});
// pass in as a
SmartTableViewer(camps);
}
}
private void SmartTableViewer<T>(IEnumerable<T> allRecords)
{
// Build sorted rows back into new table
var table = new DataTable();
// Create columns based on type
if (allRecords is IEnumerable<tbCamp> tbCampRecords)
{
// Get the columns you want
table.Columns.Add("idCamp");
table.Columns.Add("campName");
foreach (var record in tbCampRecords)
{
//var newRecord = record;
// Make a new row
var r = table.NewRow();
// Add the contents to each column of the row
r["idCamp"] = record.idCamp;
r["campName"] = record.campName;
// Add the row to the table.
table.Rows.Add(r);
}
}
else
{
MessageBox.Show("Unhandled type. Add support for new data type in SmartTableViewer()");
return;
}
// Update table in grid
dg1.DataSource = table.DefaultView;
}
internal class tbCampView : tbCamp
{
}
}
There's no full text search built into Linq and there don't seem to be many posts on the subject so I had a play around and came up with this method for my utlity class:
public static IEnumerable<TSource> GenericFullTextSearch<TSource>(string text, MyDataContext context)
{
//Find LINQ Table attribute
object[] info = typeof(TSource).GetCustomAttributes(typeof(System.Data.Linq.Mapping.TableAttribute), true);
//Get table name
String table = (info[0] as System.Data.Linq.Mapping.TableAttribute).Name;
//Full text search on that table
return context.ExecuteQuery<TSource>(String.Concat("SELECT * FROM ", table, " WHERE CONTAINS(*, {0})"), text);
}
And added this wrapper to each partial Linq class where there is a full text index
public static IEnumerable<Pet> FullTextSearch(string text, MyDataContext context)
{
return (LinqUtilities.GenericFullTextSearch<Pet>(text, context) as IEnumerable<Pet>);
}
So now I can do full text searches with neat stuff like
var Pets = Pet.FullTextSearch(helloimatextbox.Text, MyDataContext).Skip(10).Take(10);
I'm assuming only a very basic search is necessary at present. Can anyone improve on this? Is it possible to implement as an extension method and avoid the wrapper?
The neatest solution is to use an inline table valued function in sql and add it to your model
http://sqlblogcasts.com/blogs/simons/archive/2008/12/18/LINQ-to-SQL---Enabling-Fulltext-searching.aspx
To get it working you need to create a table valued function that does
nothing more than a CONTAINSTABLE query based on the keywords you
pass in,
create function udf_sessionSearch
(#keywords nvarchar(4000)) returns table as return (select [SessionId],[rank]
from containstable(Session,(description,title),#keywords))
You then add this function to your LINQ 2 SQL model and he presto you
can now write queries like.
var sessList = from s in DB.Sessions
join fts in DB.udf_sessionSearch(SearchText) on s.sessionId equals fts.SessionId
select s;
I was pretty frustrated with the lack of clear examples... especially when there are potentially large data sets and paging is needed. So, here's an example that hopefully encompasses everything you might need :-)
create function TS_projectResourceSearch
( #KeyStr nvarchar(4000),
#OwnId int,
#SkipN int,
#TakeN int )
returns #srch_rslt table (ProjectResourceId bigint not null, Ranking int not null )
as
begin
declare #TakeLast int
set #TakeLast = #SkipN + #TakeN
set #SkipN = #SkipN + 1
insert into #srch_rslt
select pr.ProjectResourceId, Ranking
from
(
select t.[KEY] as ProjectResourceId, t.[RANK] as Ranking, ROW_NUMBER() over (order by t.[Rank] desc) row_num
from containstable( ProjectResource,(ResourceInfo, ResourceName), #KeyStr )
as t
) as r
join ProjectResource pr on r.ProjectResourceId = pr.ProjectResourceId
where (pr.CreatorPersonId = #OwnId
or pr.ResourceAvailType < 40)
and r.row_num between #SkipN and #TakeLast
order by r.Ranking desc
return
end
go
select * from ts_projectResourceSearch(' "test*" ',1002, 0,1)
Enjoy,
Patrick
I use a little hack using Provider Wrapper techniques. I have a c# code that rewrite magic word in SQL with FTS search for MS SQL (you can adjust for any server you like).
if you have context class MyEntities, create subclass like
public class MyEntitiesWithWrappers : MyEntities
{
private IEFTraceListener listener;
public string FullTextPrefix = "-FTSPREFIX-";
public MyEntitiesWithWrappers(): this("name=MyEntities")
{
}
public MyEntitiesWithWrappers(string connectionString)
: base(EntityConnectionWrapperUtils.CreateEntityConnectionWithWrappers(connectionString,"EFTracingProvider"))
{
TracingConnection.CommandExecuting += RewriteFullTextQuery;
}
/// <summary>
/// Rewrites query that contains predefined prefix like: where n.NOTETEXT.Contains(Db.FullTextPrefix + text) with SQL server FTS
/// To be removed when EF will support FTS
/// </summary>
/// <param name="o"></param>
/// <param name="args"></param>
public void RewriteFullTextQuery(object o, CommandExecutionEventArgs args)
{
var text = args.Command.CommandText;
for (int i = 0; i < args.Command.Parameters.Count; i++)
{
DbParameter parameter = args.Command.Parameters[i];
if (parameter.DbType.In(DbType.String, DbType.AnsiString, DbType.StringFixedLength, DbType.AnsiStringFixedLength))
{
if (parameter.Value == DBNull.Value)
continue;
var value = (string) parameter.Value;
parameter.Size = 4096;
if (value.IndexOf(FullTextPrefix) >= 0)
{
value = value.Replace(FullTextPrefix, ""); // remove prefix we added n linq query
value = value.Substring(1, value.Length-2); // remove %% escaping by linq translator from string.Contains to sql LIKE
parameter.Value = value;
args.Command.CommandText = Regex.Replace(text,
string.Format(#"\(\[(\w*)\].\[(\w*)\]\s*LIKE\s*#{0}\s?(?:ESCAPE '~')\)", parameter.ParameterName),
string.Format(#"contains([$1].[$2], #{0})", parameter.ParameterName));
}
}
}
}
}
And then use it like this:
var fullTextSearch = Db.FullTextPrefix + textToSearch;
var q = Db.Notes.Where(n => !n.Private && n.NoteText.Contains(fullTextSearch));
A slighty nicer method (takes rank into effect) using CONTAINSTABLE
String pkey = context.Mapping.GetTable(typeof(TSource)).RowType.DataMembers.SingleOrDefault(x => x.IsPrimaryKey).Name;
string query = String.Concat(#"SELECT *
FROM ", table, #" AS FT_TBL INNER JOIN
CONTAINSTABLE(", table, #", *, {0}) AS KEY_TBL
ON FT_TBL.", pkey, #" = KEY_TBL.[KEY]
ORDER BY KEY_TBL.[RANK] DESC");
return context.ExecuteQuery<TSource>(query, text);
.NET Core 2.1 and above supports an extension method that allows the use of FREETEXT and FREETEXTTABLE searches
using Microsoft.EntityFrameworkCore;
var results = dbContext.MyTable
.Where(e => EF.Functions.FreeText("*", "search criteria"));
The FreeText function is defined in Microsoft.EntityFrameworkCore.SqlServer so your project must reference that package.
Documentation
I've been trying to solve the exact problem. I like to write my SQL logic in my LINQtoSQL but I needed a way to do Full Text Search. right now I'm just using SQL functions and then calling the user-defined functions inline of the linq queries. not sure if that's the most efficient way. what do you guys think?
You can just do something like this
var results = (from tags in _dataContext.View_GetDeterminationTags
where tags.TagName.Contains(TagName) ||
SqlMethods.Like(tags.TagName,TagName)
select new DeterminationTags
{
Row = tags.Row,
Record = tags.Record,
TagID = tags.TagID,
TagName = tags.TagName,
DateTagged = tags.DateTagged,
DeterminationID = tags.DeterminationID,
DeterminationMemberID = tags.DeterminationMemberID,
MemberID = tags.MemberID,
TotalTagged = tags.TotalTagged.Value
}).ToList();
Notice where TagName.Contains also the SQLMethods.Like just do a using
using System.Data.Linq.SqlClient;
to gain access to that SQLMethods.
dswatik - the reason for wanting full text search is that .contains translates to
SELECT * FROM MYTABLE WHERE COLUMNNAME LIKE '%TEXT%'
Which ignores any indexes and is horrible on a large table.