开发者_JAVA技巧
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 5 years ago.
Improve this questionC# 2 and VB.Net 8 introduced a new feature called iterators, which were designed to make it easier to return enumerables and enumerators.
However, iterators are actually a limited form of coroutines, and can be used to do many useful things that have nothing to do with collections of objects.
What non-standard uses of iterators have you seen in real code?
I used them to write a system in ASP.NET for creating a series of linked page interactions. If you imagine a user's conversation with a website as a series of requests and responses, you can model an interaction as an IEnumerable
. Conceptually, like this;
IEnumerable<PageResponse> SignupProcess(FormValues form)
{
// signup starts with a welcome page, asking
// the user to accept the license.
yield return new WelcomePageResponse();
// if they don't accept the terms, direct
// them to a 'thanks anyway' screen
if (!form["userAcceptsTerms"])
{
yield return new ThanksForYourTimePageResponse();
yield break;
}
// On the second page, we gather their email;
yield new EmailCapturePage("");
while(!IsValid(form["address"]))
{
// loop until we get a valid address.
yield return new EmailCapturePage("The email address is incorrect. Please fix.");
}
}
You can store the iterator in session state, so that when the user returns to the site you just pull the iterator out, move the iterator onto the next page, and yield it back for rendering. Complex site interactions are coded in a single place.
To start things off:
- Jeffrey Richter wrote a powerful threading system called AsyncEnumerator using iterators. It's described in MSDN Magazine, parts one and two.
- Iterators can also be used to wait for UI interaction within a method without blocking the UI thread, as I described here.
- In a similar vein, I used iterators to create an IE-based web scraper, with scraping methods that return IEnumerators of
WebAction
s which call back into the enumerator when ready. (Typically, when the document finishes loading).
If people are interested, I can post it here.
I used it to recursively iterate over the files contained in a folder, its subfolders and so on. For every file I had to perform a specific action. A recursive function with "yield return" statements was simple for everyone's else understanding.
I wrote this function before I found out about the lazy yield operator. This recursively builds a massive iterator and returns it. It is not exactly efficient but I think it is clever.
static member generatePrimeNumbers max =
let rec generate number numberSequence =
if number * number > max then numberSequence else
let filteredNumbers = numberSequence |> Seq.filter (fun v -> v = number || v % number <> 0L)
let newNumberSequence = seq { for i in filteredNumbers -> i }
let newNumber = newNumberSequence |> Seq.find (fun x -> x > number)
generate newNumber newNumberSequence
generate 2L (seq { for i in 2L..max -> i })
Rhino.ETL makes heavy use of it to combine transformations over a sequence of input
for example 3 operations which can be combined, reused
public IEnumerable<Row> Execute(IEnumerable<Row> rows)
{
foreach(var line in File.EnumerateLines())
{
var row = new Row();
row["key"] = int.Parse(line.Substring(1));
yield return row;
}
}
public IEnumerable<Row> Execute(IEnumerable<Row> rows)
{
foreach(var row in rows)
{
var value = (int)row["key"];
row["key"] = value + 2;
yield return row;
}
}
public IEnumerable<Row> Execute(IEnumerable<Row> rows)
{
using (var file = new Streamwriter(filename))
{
foreach(var row in rows)
{
file.WriteLine(row["key"]);
yield return row;
}
}
}
精彩评论