The OrigoDb engine can compile and cache parameterized LINQ queries passed as a strings using the following extension methods, defined in the namespace OrigoDB.Core.Linq
:
public object Engine.Execute(string linqQuery, param object[] args);
public R Engine.Execute<M, R>(string linqQuery, param object[] args);
The linqQuery must be a valid c# expression. The expression is evaluated with the provided arguments and the result is returned to the caller.
Use db
within the query to reference the model during execution.
Access the arguments in the args
param array using @arg0
, @arg1
, @arg2
etc
The type of each parameter is derived from the actual arguments on the first invocation. Subsequent invocations expect arguments that match the signature. Example:
class Customer{ public string Name; }
engine.Execute("@arg0.Name", new Customer{Name = "Homer"});
Note how the Customer.Name
property is accessed within the query.
//import extension methods
using OrigoDb.Core.Linq;
var query = "(from Customer c in db.Customers where c.Name.StartsWith(@arg0)).First()"
//call untyped
object untypedResult = engine.Execute(query, "R"); // "R" is params object[]
//call typed
string customerName = engine.Execute<MyModel, string>(query, "H");
The query is compiled the first time it is evaluated. The compilation depends on the number and types of arguments passed. Subsequent calls with the same linqQuery must pass the same number and types of arguments.
Beware of the risk of arbitrary code injection. There is a test case in the source with an example exploit.