Can you help me in understan开发者_开发技巧ding of yield
keyword in asp .NET(C#)
.
Yield return automatically creates an enumerator for you.
http://msdn.microsoft.com/en-us/library/9k7k7cf0.aspx
So you can do something like
//pseudo code:
while(get_next_record_from_database)
{
yield return your_next_record;
}
It allows you to quickly create an object collection (an Enumerator) that you can loop through and retrieve records. The yield return statement handles all the of the code needed to create an enumerator for you.
The big part of the yield return statement is that you don't have to load all the of the items in a collection before returning the collection to the calling method. It allows lazy loading of the collection, so you don't pay the access penalty all at once.
When to use Yield Return.
Yield is much more than syntatic sugar or easy ways to create IEnumerables.
For more information I'd check out Justin Etherage's blog which has a great article explaining more advanced usages of yield.
yield
is used as syntactic sugar to return an IEnumerable<T>
or IEnumerator<T>
object from a method without having to implement your own class implementing these interfaces.
I wrote a simple sample showing how the yield block is invoked during a visiting cycle of the collection. Check it out here.
yield
allows you to emit an IEnumerable where you'd normally return a more concrete type (like an IList).
This is a pretty good example of how it can simplify your code and clarify your intent. As for where you'd use it, anywhere on your pages that you need to iterate over a collection you could potentially use a method that returns an IEnumerable in place of a List/Dictionary/etc.
I don't think the "Benefits" of using Stop and Continue pattern (AKA yield/enumerators), has been properly expounded upon. So let me try.
Say you have an app that needs to return 1 million records from the database. You have a few common practices:
- You could return a collection with 1 million objects like a List or Array, but that would lead to a huge amount of memory pressure and potentially the dreaded OutOfMemoryException.
- You could paginate 1,000 records at a time, but this too has its draw backs as you now have to write pagination logic
- You could serialize these objects to a file like JSON or other (this has to be the worst idea I've ever seen implemented)
- You should use the "yield" pattern; which in contrast to the other options is more like streaming objects
By using the yield pattern, you only hydrate one object at a time into memory. Furthermore, the consumption of objects is controlled by the code iterating through the IEnumerator/IEnumerable code. This should be a typical foreach code block.
Here is an example to contrast the code differences
/// Bad Deadpool...
List<MyDbRecord> GetData(int page, int pageSize) {
using (var con = new DbContext()) {
// Option 1: Straight object grabbing
// return con.MyDbRecordSet.ToList();
// Option 2: Pagination Example
return con.MyDbRecordSet.Skip(page * pageSize).Take(pageSize).ToList();
// Option 3: DON'T TRY THIS AT HOME!
// var allTheRecords = con.MyDbRecordSet.ToList(); // Memory pressure
// System.IO.File.WriteAllText(
// "C:\\Windows\\Temp\\temp.json",
// JsonConvert.SerializeObject(allTheRecords)
// );// Large JSON Object dropped
}
}
/// Bad Deadpool...
IEnumerable<MyDbRecord> GetData(int page, int pageSize) {
using (var con = new DbContext()) {
// Option 1: Straight object grabbing
// return con.MyDbRecordSet.ToList();
// Option 2: Pagination Example
//return con.MyDbRecordSet.Skip(page * pageSize).Take(pageSize).ToList();
// Option 3: DON'T TRY THIS AT HOME!
// var allTheRecords = con.MyDbRecordSet.ToList(); // Memory pressure
// System.IO.File.WriteAllText(
// "C:\\Windows\\Temp\\temp.json",
// JsonConvert.SerializeObject(allTheRecords)
// );// Large JSON Object dropped
foreach (var i in con.MyDbRecordSet.AsNoTracking().AsQueryable()) {
yield return i; // Stream it vs buffering
}
}
}
精彩评论