Say I have a table in a database called TestDB with a table called Table1 which contains two columns
ID
Name
In this table ,there is two row,
ID 1 and Name 'Row1'
ID 2 and Name 'Row2'
If this is my code
var currObj = Table1.Where(o => o.Name.Contains("Row1")).FirstOrDefault();
currObj.Name = "Row1a";
var a = Table1.Where(o => o.Name.Contains("Row1a")).FirstOrDefault();
var b = Table1.Local.Where(o => o.Name.Contains("Row1a")).FirstOrDefault();
a is going to return null, whereas b is going to return a value.
if I do this though
var c = Table1.Where(o => o.Name.Contains("Row2")).FirstOrDefault();
var d = Table1.Local.Where(o => o.Name.Contains("Row2")).FirstOrDefault();
then c will return something, but d will not.
For me, this seems unintuitive because there is two different places the data could be. For every query I do, I have to look in the database, and the Local object and merge them together. Like if it’s changed in the local object, I have to take that into account. Does Entity Framework have any mechanism for this, which would consider BOTH the data in the database AND the local at the same time?
So if I went
var e = Table1.AMagicSolution.Where(o => o.Name.Contains("Row1a")).FirstOrDefault();
var f = Table1.AMagicSolution.Where(o => o.Name.Contains("Row2")).FirstOrDefault();
then e and f would both return something (f from the database and e from the Local object)
Your expectation is little bit strange. EF context works like a unit of work = it is used to process single logical transaction and because of that you should try to design your application in the way that it knows if the record is in the local storage or not.
If for any reason your application doesn’t know if the record is in the local storage you should always separate queries to the local storage and to the database! Reasons are performance and unnecessary roundtrips to the database. Use helper like in following example to query the local storage first and only if record is not present in the local storage query the database: