Parallel.For 方法
定義
重要
部分資訊涉及發行前產品,在發行之前可能會有大幅修改。 Microsoft 對此處提供的資訊,不做任何明確或隱含的瑕疵擔保。
執行可平行執行反覆項目的 for
迴圈。
多載
For(Int32, Int32, Action<Int32,ParallelLoopState>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
執行 for
迴圈,其中可以平行執行反覆項目,還可以監控和操作迴圈的狀態。
public:
static System::Threading::Tasks::ParallelLoopResult For(int fromInclusive, int toExclusive, Action<int, System::Threading::Tasks::ParallelLoopState ^> ^ body);
public static System.Threading.Tasks.ParallelLoopResult For (int fromInclusive, int toExclusive, Action<int,System.Threading.Tasks.ParallelLoopState> body);
static member For : int * int * Action<int, System.Threading.Tasks.ParallelLoopState> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For (fromInclusive As Integer, toExclusive As Integer, body As Action(Of Integer, ParallelLoopState)) As ParallelLoopResult
參數
- fromInclusive
- Int32
起始索引 (含)。
- toExclusive
- Int32
結束索引 (不含)。
- body
- Action<Int32,ParallelLoopState>
每個反覆項目都會叫用一次的委派。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
body
引數為 null
。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
範例
下列範例會平行執行迴圈最多 100 次反復專案。 每次反覆運算都會暫停 1 到 1,000 毫秒的隨機間隔。 隨機產生的值會決定呼叫 方法之迴圈 ParallelLoopState.Break 的反復專案。 如範例的輸出所示,索引大於 ParallelLoopState.LowestBreakIteration 屬性值的反復專案,會在呼叫 ParallelLoopState.Break 方法之後開始。
using System;
using System.Threading;
using System.Threading.Tasks;
public class Example
{
public static void Main()
{
var rnd = new Random();
int breakIndex = rnd.Next(1, 11);
Console.WriteLine($"Will call Break at iteration {breakIndex}\n");
var result = Parallel.For(1, 101, (i, state) =>
{
Console.WriteLine($"Beginning iteration {i}");
int delay;
lock (rnd)
delay = rnd.Next(1, 1001);
Thread.Sleep(delay);
if (state.ShouldExitCurrentIteration)
{
if (state.LowestBreakIteration < i)
return;
}
if (i == breakIndex)
{
Console.WriteLine($"Break in iteration {i}");
state.Break();
}
Console.WriteLine($"Completed iteration {i}");
});
if (result.LowestBreakIteration.HasValue)
Console.WriteLine($"\nLowest Break Iteration: {result.LowestBreakIteration}");
else
Console.WriteLine($"\nNo lowest break iteration.");
}
}
// The example displays output like the following:
// Will call Break at iteration 8
//
// Beginning iteration 1
// Beginning iteration 13
// Beginning iteration 97
// Beginning iteration 25
// Beginning iteration 49
// Beginning iteration 37
// Beginning iteration 85
// Beginning iteration 73
// Beginning iteration 61
// Completed iteration 85
// Beginning iteration 86
// Completed iteration 61
// Beginning iteration 62
// Completed iteration 86
// Beginning iteration 87
// Completed iteration 37
// Beginning iteration 38
// Completed iteration 38
// Beginning iteration 39
// Completed iteration 25
// Beginning iteration 26
// Completed iteration 26
// Beginning iteration 27
// Completed iteration 73
// Beginning iteration 74
// Completed iteration 62
// Beginning iteration 63
// Completed iteration 39
// Beginning iteration 40
// Completed iteration 40
// Beginning iteration 41
// Completed iteration 13
// Beginning iteration 14
// Completed iteration 1
// Beginning iteration 2
// Completed iteration 97
// Beginning iteration 98
// Completed iteration 49
// Beginning iteration 50
// Completed iteration 87
// Completed iteration 27
// Beginning iteration 28
// Completed iteration 50
// Beginning iteration 51
// Beginning iteration 88
// Completed iteration 14
// Beginning iteration 15
// Completed iteration 15
// Completed iteration 2
// Beginning iteration 3
// Beginning iteration 16
// Completed iteration 63
// Beginning iteration 64
// Completed iteration 74
// Beginning iteration 75
// Completed iteration 41
// Beginning iteration 42
// Completed iteration 28
// Beginning iteration 29
// Completed iteration 29
// Beginning iteration 30
// Completed iteration 98
// Beginning iteration 99
// Completed iteration 64
// Beginning iteration 65
// Completed iteration 42
// Beginning iteration 43
// Completed iteration 88
// Beginning iteration 89
// Completed iteration 51
// Beginning iteration 52
// Completed iteration 16
// Beginning iteration 17
// Completed iteration 43
// Beginning iteration 44
// Completed iteration 44
// Beginning iteration 45
// Completed iteration 99
// Beginning iteration 4
// Completed iteration 3
// Beginning iteration 8
// Completed iteration 4
// Beginning iteration 5
// Completed iteration 52
// Beginning iteration 53
// Completed iteration 75
// Beginning iteration 76
// Completed iteration 76
// Beginning iteration 77
// Completed iteration 65
// Beginning iteration 66
// Completed iteration 5
// Beginning iteration 6
// Completed iteration 89
// Beginning iteration 90
// Completed iteration 30
// Beginning iteration 31
// Break in iteration 8
// Completed iteration 8
// Completed iteration 6
// Beginning iteration 7
// Completed iteration 7
//
// Lowest Break Iteration: 8
Imports System.Threading
Imports System.Threading.Tasks
Module Example
Public Sub Main()
Dim rnd As New Random()
Dim breakIndex As Integer = rnd.Next(1, 11)
Dim lowest As New Nullable(Of Long)()
Console.WriteLine("Will call Break at iteration {0}", breakIndex)
Console.WriteLine()
Dim result = Parallel.For(1, 101, Sub(i, state)
Console.WriteLine("Beginning iteration {0}", i)
Dim delay As Integer
Monitor.Enter(rnd)
delay = rnd.Next(1, 1001)
Monitor.Exit(rnd)
Thread.Sleep(delay)
If state.ShouldExitCurrentIteration Then
If state.LowestBreakIteration < i Then
Return
End If
End If
If i = breakIndex Then
Console.WriteLine("Break in iteration {0}", i)
state.Break()
If state.LowestBreakIteration.HasValue Then
If lowest < state.LowestBreakIteration Then
lowest = state.LowestBreakIteration
Else
lowest = state.LowestBreakIteration
End If
End If
End If
Console.WriteLine("Completed iteration {0}", i)
End Sub )
Console.WriteLine()
If lowest.HasValue Then
Console.WriteLine("Lowest Break Iteration: {0}", lowest)
Else
Console.WriteLine("No lowest break iteration.")
End If
End Sub
End Module
' The example displays output like the following:
' Will call Break at iteration 8
'
' Beginning iteration 1
' Beginning iteration 13
' Beginning iteration 97
' Beginning iteration 25
' Beginning iteration 49
' Beginning iteration 37
' Beginning iteration 85
' Beginning iteration 73
' Beginning iteration 61
' Completed iteration 85
' Beginning iteration 86
' Completed iteration 61
' Beginning iteration 62
' Completed iteration 86
' Beginning iteration 87
' Completed iteration 37
' Beginning iteration 38
' Completed iteration 38
' Beginning iteration 39
' Completed iteration 25
' Beginning iteration 26
' Completed iteration 26
' Beginning iteration 27
' Completed iteration 73
' Beginning iteration 74
' Completed iteration 62
' Beginning iteration 63
' Completed iteration 39
' Beginning iteration 40
' Completed iteration 40
' Beginning iteration 41
' Completed iteration 13
' Beginning iteration 14
' Completed iteration 1
' Beginning iteration 2
' Completed iteration 97
' Beginning iteration 98
' Completed iteration 49
' Beginning iteration 50
' Completed iteration 87
' Completed iteration 27
' Beginning iteration 28
' Completed iteration 50
' Beginning iteration 51
' Beginning iteration 88
' Completed iteration 14
' Beginning iteration 15
' Completed iteration 15
' Completed iteration 2
' Beginning iteration 3
' Beginning iteration 16
' Completed iteration 63
' Beginning iteration 64
' Completed iteration 74
' Beginning iteration 75
' Completed iteration 41
' Beginning iteration 42
' Completed iteration 28
' Beginning iteration 29
' Completed iteration 29
' Beginning iteration 30
' Completed iteration 98
' Beginning iteration 99
' Completed iteration 64
' Beginning iteration 65
' Completed iteration 42
' Beginning iteration 43
' Completed iteration 88
' Beginning iteration 89
' Completed iteration 51
' Beginning iteration 52
' Completed iteration 16
' Beginning iteration 17
' Completed iteration 43
' Beginning iteration 44
' Completed iteration 44
' Beginning iteration 45
' Completed iteration 99
' Beginning iteration 4
' Completed iteration 3
' Beginning iteration 8
' Completed iteration 4
' Beginning iteration 5
' Completed iteration 52
' Beginning iteration 53
' Completed iteration 75
' Beginning iteration 76
' Completed iteration 76
' Beginning iteration 77
' Completed iteration 65
' Beginning iteration 66
' Completed iteration 5
' Beginning iteration 6
' Completed iteration 89
' Beginning iteration 90
' Completed iteration 30
' Beginning iteration 31
' Break in iteration 8
' Completed iteration 8
' Completed iteration 6
' Beginning iteration 7
' Completed iteration 7
'
' Lowest Break Iteration: 8
因為呼叫 方法時 ParallelLoopState.Break ,迴圈的反復專案仍可能正在執行,所以每個反復專案都會呼叫 ParallelLoopState.ShouldExitCurrentIteration 屬性,以檢查另一個反復專案是否已呼叫 ParallelLoopState.Break 方法。 如果屬性值為 true
,則反復專案會檢查屬性的值 ParallelLoopState.LowestBreakIteration ,如果大於目前反復專案的索引值,則會立即傳回 。
備註
系統會 body
針對反覆運算範圍中的每個值叫用委派一次, (fromInclusive
、 toExclusive
) 。 它提供兩個引數:
Int32值,表示反復專案計數。
ParallelLoopState可用來提前中斷迴圈的 實例。 物件 ParallelLoopState 是由編譯器所建立,無法在使用者程式碼中具現化。
Break呼叫 方法會 for
通知作業,目前的反復專案之後不需要執行。 不過,如果目前反復專案尚未執行,則所有反復專案仍必須執行。
因此,呼叫 Break 類似于在 C# 之類的語言中使用傳統 for
迴圈內的中斷作業,但不是完美的替代方式:例如,目前迴圈之後的反復專案絕對不會執行。
如果不需要在目前反復專案之前執行所有反復專案,請使用 Stop 方法,而不是使用 Break 。 呼叫 Stop 會 for
通知迴圈,不論它們是否在目前的反復專案之前或之後,它都可能會放棄所有剩餘的反復專案,因為所有必要的工作都已經完成。 不過,與 一 Break 樣,不會保證哪些其他反復專案不會執行。
如果迴圈提前結束, ParallelLoopResult 傳回的結構將會包含迴圈完成的相關資訊。
如果 fromInclusive
大於或等於 toExclusive
,方法會立即傳回,而不會執行任何反復專案。
另請參閱
適用於
For(Int32, Int32, Action<Int32>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
執行可平行執行反覆項目的 for
迴圈。
public:
static System::Threading::Tasks::ParallelLoopResult For(int fromInclusive, int toExclusive, Action<int> ^ body);
public static System.Threading.Tasks.ParallelLoopResult For (int fromInclusive, int toExclusive, Action<int> body);
static member For : int * int * Action<int> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For (fromInclusive As Integer, toExclusive As Integer, body As Action(Of Integer)) As ParallelLoopResult
參數
- fromInclusive
- Int32
起始索引 (含)。
- toExclusive
- Int32
結束索引 (不含)。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
body
引數為 null
。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
範例
下列範例會 For 針對產生隨機位元組值並計算其總和之委派的 100 個調用使用 方法。
using System;
using System.Threading.Tasks;
public class Example
{
public static void Main()
{
ParallelLoopResult result = Parallel.For(0, 100, ctr => { Random rnd = new Random(ctr * 100000);
Byte[] bytes = new Byte[100];
rnd.NextBytes(bytes);
int sum = 0;
foreach(var byt in bytes)
sum += byt;
Console.WriteLine("Iteration {0,2}: {1:N0}", ctr, sum);
});
Console.WriteLine("Result: {0}", result.IsCompleted ? "Completed Normally" :
String.Format("Completed to {0}", result.LowestBreakIteration));
}
}
// The following is a portion of the output displayed by the example:
// Iteration 0: 12,509
// Iteration 50: 12,823
// Iteration 51: 11,275
// Iteration 52: 12,531
// Iteration 1: 13,007
// Iteration 53: 13,799
// Iteration 4: 12,945
// Iteration 2: 13,246
// Iteration 54: 13,008
// Iteration 55: 12,727
// Iteration 56: 13,223
// Iteration 57: 13,717
// Iteration 5: 12,679
// Iteration 3: 12,455
// Iteration 58: 12,669
// Iteration 59: 11,882
// Iteration 6: 13,167
// ...
// Iteration 92: 12,275
// Iteration 93: 13,282
// Iteration 94: 12,745
// Iteration 95: 11,957
// Iteration 96: 12,455
// Result: Completed Normally
Imports System.Threading.Tasks
Module Example
Public Sub Main()
Dim result As ParallelLoopResult = Parallel.For(0, 100, Sub(ctr)
Dim rnd As New Random(ctr * 100000)
Dim bytes(99) As Byte
rnd.NextBytes(bytes)
Dim sum As Integer
For Each byt In bytes
sum += byt
Next
Console.WriteLine("Iteration {0,2}: {1:N0}", ctr, sum)
End Sub)
Console.WriteLine("Result: {0}", If(result.IsCompleted, "Completed Normally",
String.Format("Completed to {0}", result.LowestBreakIteration)))
End Sub
End Module
' The following is a portion of the output displayed by the example:
' Iteration 0: 12,509
' Iteration 50: 12,823
' Iteration 51: 11,275
' Iteration 52: 12,531
' Iteration 1: 13,007
' Iteration 53: 13,799
' Iteration 4: 12,945
' Iteration 2: 13,246
' Iteration 54: 13,008
' Iteration 55: 12,727
' Iteration 56: 13,223
' Iteration 57: 13,717
' Iteration 5: 12,679
' Iteration 3: 12,455
' Iteration 58: 12,669
' Iteration 59: 11,882
' Iteration 6: 13,167
' ...
' Iteration 92: 12,275
' Iteration 93: 13,282
' Iteration 94: 12,745
' Iteration 95: 11,957
' Iteration 96: 12,455
' Result: Completed Normally
備註
系統會 body
針對反覆運算範圍中的每個值叫用委派一次, (fromInclusive
、 toExclusive
) 。 其提供反復專案計數 (Int32) 做為參數。
如果 fromInclusive
大於或等於 toExclusive
,方法會立即傳回,而不會執行任何反復專案。
另請參閱
適用於
For(Int64, Int64, Action<Int64,ParallelLoopState>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
使用 64 位元索引執行 for
迴圈,其中可以平行執行反覆項目,還可以監控和操作迴圈的狀態。
public:
static System::Threading::Tasks::ParallelLoopResult For(long fromInclusive, long toExclusive, Action<long, System::Threading::Tasks::ParallelLoopState ^> ^ body);
public static System.Threading.Tasks.ParallelLoopResult For (long fromInclusive, long toExclusive, Action<long,System.Threading.Tasks.ParallelLoopState> body);
static member For : int64 * int64 * Action<int64, System.Threading.Tasks.ParallelLoopState> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For (fromInclusive As Long, toExclusive As Long, body As Action(Of Long, ParallelLoopState)) As ParallelLoopResult
參數
- fromInclusive
- Int64
起始索引 (含)。
- toExclusive
- Int64
結束索引 (不含)。
- body
- Action<Int64,ParallelLoopState>
每個反覆項目都會叫用一次的委派。
傳回
ParallelLoopResult 結構,包含迴圈已完成部分的相關資訊。
例外狀況
body
引數為 null
。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
備註
系統會 body
針對反覆運算範圍中的每個值叫用委派一次, (fromInclusive
、 toExclusive
) 。 它隨附下列參數:反復專案計數 (Int64) ,以及 ParallelLoopState 可能用來提前中斷迴圈的 實例。
Break呼叫 方法會 for
通知作業,目前的反復專案之後不需要執行,而是目前反復專案之前的所有反復專案。
因此,呼叫 Break 類似于在 C# 之類的語言中使用傳統 for
迴圈內的中斷作業,但不是完美的替代方法:例如,目前迴圈之後的反復專案絕對不會執行。
如果不需要在目前反復專案之前執行所有反復專案,請使用 Stop 方法,而不是使用 Break 。 呼叫 Stop 會 for
通知迴圈,不論它們是否在目前的反復專案之前或之後,它都可能會放棄所有剩餘的反復專案,因為所有必要的工作都已經完成。 不過,與 一 Break 樣,不會保證哪些其他反復專案不會執行。
如果迴圈提前結束, ParallelLoopResult 傳回的結構將會包含迴圈完成的相關資訊。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不需要執行任何反復專案。
另請參閱
適用於
For(Int64, Int64, Action<Int64>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
使用 64 位元索引執行 for
迴圈,其中可以平行執行反覆項目。
public:
static System::Threading::Tasks::ParallelLoopResult For(long fromInclusive, long toExclusive, Action<long> ^ body);
public static System.Threading.Tasks.ParallelLoopResult For (long fromInclusive, long toExclusive, Action<long> body);
static member For : int64 * int64 * Action<int64> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For (fromInclusive As Long, toExclusive As Long, body As Action(Of Long)) As ParallelLoopResult
參數
- fromInclusive
- Int64
起始索引 (含)。
- toExclusive
- Int64
結束索引 (不含)。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
body
引數為 null
。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
備註
系統會 body
針對反覆運算範圍中的每個值叫用委派一次, (fromInclusive
、 toExclusive
) 。 其提供反復專案計數 (Int64) 做為參數。
如果 fromInclusive
大於或等於 toExclusive
,方法會立即傳回,而不會執行任何反復專案。
另請參閱
適用於
For(Int32, Int32, ParallelOptions, Action<Int32,ParallelLoopState>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
執行 for
迴圈,其中可以平行執行反覆項目、可以設定迴圈選項,還可以監控和操作迴圈的狀態。
public:
static System::Threading::Tasks::ParallelLoopResult For(int fromInclusive, int toExclusive, System::Threading::Tasks::ParallelOptions ^ parallelOptions, Action<int, System::Threading::Tasks::ParallelLoopState ^> ^ body);
public static System.Threading.Tasks.ParallelLoopResult For (int fromInclusive, int toExclusive, System.Threading.Tasks.ParallelOptions parallelOptions, Action<int,System.Threading.Tasks.ParallelLoopState> body);
static member For : int * int * System.Threading.Tasks.ParallelOptions * Action<int, System.Threading.Tasks.ParallelLoopState> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For (fromInclusive As Integer, toExclusive As Integer, parallelOptions As ParallelOptions, body As Action(Of Integer, ParallelLoopState)) As ParallelLoopResult
參數
- fromInclusive
- Int32
起始索引 (含)。
- toExclusive
- Int32
結束索引 (不含)。
- parallelOptions
- ParallelOptions
物件,設定這個作業的行為。
- body
- Action<Int32,ParallelLoopState>
每個反覆項目都會叫用一次的委派。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
parallelOptions
引數中的 CancellationToken 已取消。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
已處置與 parallelOptions
中的 CancellationTokenSource 相關聯的 CancellationToken。
備註
反覆運算 body
範圍中的每個值都會叫用委派一次, fromInclusive
() toExclusive
。 它隨附下列參數:反復專案計數 (Int32) ,以及 ParallelLoopState 可能用來提前中斷迴圈的實例。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不會執行任何反復專案。
另請參閱
適用於
For(Int32, Int32, ParallelOptions, Action<Int32>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
執行 for
迴圈,其中可以平行執行反覆項目,還可以設定迴圈選項。
public:
static System::Threading::Tasks::ParallelLoopResult For(int fromInclusive, int toExclusive, System::Threading::Tasks::ParallelOptions ^ parallelOptions, Action<int> ^ body);
public static System.Threading.Tasks.ParallelLoopResult For (int fromInclusive, int toExclusive, System.Threading.Tasks.ParallelOptions parallelOptions, Action<int> body);
static member For : int * int * System.Threading.Tasks.ParallelOptions * Action<int> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For (fromInclusive As Integer, toExclusive As Integer, parallelOptions As ParallelOptions, body As Action(Of Integer)) As ParallelLoopResult
參數
- fromInclusive
- Int32
起始索引 (含)。
- toExclusive
- Int32
結束索引 (不含)。
- parallelOptions
- ParallelOptions
物件,設定這個作業的行為。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
parallelOptions
引數中的 CancellationToken 已取消。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
已處置與 parallelOptions
中的 CancellationTokenSource 相關聯的 CancellationToken。
範例
下列範例示範如何取消平行迴圈:
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading;
using System.Threading.Tasks;
class ParallelForCancellation
{
// Demonstrated features:
// CancellationTokenSource
// Parallel.For()
// ParallelOptions
// ParallelLoopResult
// Expected results:
// An iteration for each argument value (0, 1, 2, 3, 4, 5, 6, 7, 8, 9) is executed.
// The order of execution of the iterations is undefined.
// The iteration when i=2 cancels the loop.
// Some iterations may bail out or not start at all; because they are temporally executed in unpredictable order,
// it is impossible to say which will start/complete and which won't.
// At the end, an OperationCancelledException is surfaced.
// Documentation:
// http://msdn.microsoft.com/library/system.threading.cancellationtokensource(VS.100).aspx
static void CancelDemo()
{
CancellationTokenSource cancellationSource = new CancellationTokenSource();
ParallelOptions options = new ParallelOptions();
options.CancellationToken = cancellationSource.Token;
try
{
ParallelLoopResult loopResult = Parallel.For(
0,
10,
options,
(i, loopState) =>
{
Console.WriteLine("Start Thread={0}, i={1}", Thread.CurrentThread.ManagedThreadId, i);
// Simulate a cancellation of the loop when i=2
if (i == 2)
{
cancellationSource.Cancel();
}
// Simulates a long execution
for (int j = 0; j < 10; j++)
{
Thread.Sleep(1 * 200);
// check to see whether or not to continue
if (loopState.ShouldExitCurrentIteration) return;
}
Console.WriteLine("Finish Thread={0}, i={1}", Thread.CurrentThread.ManagedThreadId, i);
}
);
if (loopResult.IsCompleted)
{
Console.WriteLine("All iterations completed successfully. THIS WAS NOT EXPECTED.");
}
}
// No exception is expected in this example, but if one is still thrown from a task,
// it will be wrapped in AggregateException and propagated to the main thread.
catch (AggregateException e)
{
Console.WriteLine("Parallel.For has thrown an AggregateException. THIS WAS NOT EXPECTED.\n{0}", e);
}
// Catching the cancellation exception
catch (OperationCanceledException e)
{
Console.WriteLine("An iteration has triggered a cancellation. THIS WAS EXPECTED.\n{0}", e.ToString());
}
finally
{
cancellationSource.Dispose();
}
}
}
Imports System.Threading
Imports System.Threading.Tasks
Module LoopCancellation
' Demonstrated features:
' CancellationTokenSource
' Parallel.For()
' ParallelOptions
' ParallelLoopResult
' Expected results:
' An iteration for each argument value (0, 1, 2, 3, 4, 5, 6, 7, 8, 9) is executed.
' The order of execution of the iterations is undefined.
' The iteration when i=2 cancels the loop.
' Some iterations may bail out or not start at all; because they are temporally executed in unpredictable order,
' it is impossible to say which will start/complete and which won't.
' At the end, an OperationCancelledException is surfaced.
' Documentation:
' http://msdn.microsoft.com/library/system.threading.cancellationtokensource(VS.100).aspx
Private Sub Main()
Dim cancellationSource As New CancellationTokenSource()
Dim options As New ParallelOptions()
options.CancellationToken = cancellationSource.Token
Try
Dim loopResult As ParallelLoopResult = _
Parallel.For(0, 10, options, Sub(i, loopState)
Console.WriteLine("Start Thread={0}, i={1}", Thread.CurrentThread.ManagedThreadId, i)
' Simulate a cancellation of the loop when i=2
If i = 2 Then
cancellationSource.Cancel()
End If
' Simulates a long execution
For j As Integer = 0 To 9
Thread.Sleep(1 * 200)
' check to see whether or not to continue
If loopState.ShouldExitCurrentIteration Then
Exit Sub
End If
Next
Console.WriteLine("Finish Thread={0}, i={1}", Thread.CurrentThread.ManagedThreadId, i)
End Sub)
If loopResult.IsCompleted Then
Console.WriteLine("All iterations completed successfully. THIS WAS NOT EXPECTED.")
End If
Catch e As AggregateException
' No exception is expected in this example, but if one is still thrown from a task,
' it will be wrapped in AggregateException and propagated to the main thread.
Console.WriteLine("An action has thrown an AggregateException. THIS WAS NOT EXPECTED." & vbLf & "{0}", e)
Catch e As OperationCanceledException
' Catching the cancellation exception
Console.WriteLine("An iteration has triggered a cancellation. THIS WAS EXPECTED." & vbLf & "{0}", e)
Finally
cancellationSource.Dispose()
End Try
End Sub
End Module
備註
反覆運算 body
範圍中的每個值都會叫用委派一次, fromInclusive
() toExclusive
。 它會以參數的形式提供反復專案計數 (Int32) 。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不需要執行任何反復專案。
另請參閱
適用於
For(Int64, Int64, ParallelOptions, Action<Int64,ParallelLoopState>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
使用 64 位元索引執行 for
迴圈,其中可以平行執行反覆項目、可以設定迴圈選項,還可以監控和操作迴圈的狀態。
public:
static System::Threading::Tasks::ParallelLoopResult For(long fromInclusive, long toExclusive, System::Threading::Tasks::ParallelOptions ^ parallelOptions, Action<long, System::Threading::Tasks::ParallelLoopState ^> ^ body);
public static System.Threading.Tasks.ParallelLoopResult For (long fromInclusive, long toExclusive, System.Threading.Tasks.ParallelOptions parallelOptions, Action<long,System.Threading.Tasks.ParallelLoopState> body);
static member For : int64 * int64 * System.Threading.Tasks.ParallelOptions * Action<int64, System.Threading.Tasks.ParallelLoopState> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For (fromInclusive As Long, toExclusive As Long, parallelOptions As ParallelOptions, body As Action(Of Long, ParallelLoopState)) As ParallelLoopResult
參數
- fromInclusive
- Int64
起始索引 (含)。
- toExclusive
- Int64
結束索引 (不含)。
- parallelOptions
- ParallelOptions
物件,設定這個作業的行為。
- body
- Action<Int64,ParallelLoopState>
每個反覆項目都會叫用一次的委派。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
parallelOptions
引數中的 CancellationToken 已取消。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
已處置與 parallelOptions
中的 CancellationTokenSource 相關聯的 CancellationToken。
範例
下列範例示範如何搭配 ParallelOptions 物件使用 Parallel.For 方法:
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading;
using System.Threading.Tasks;
class ParallelOptionsDemo
{
// Demonstrated features:
// Parallel.For()
// ParallelOptions
// Expected results:
// An iteration for each argument value (0, 1, 2, 3, 4, 5, 6, 7, 8, 9) is executed.
// The order of execution of the iterations is undefined.
// Verify that no more than two threads have been used for the iterations.
// Documentation:
// http://msdn.microsoft.com/library/system.threading.tasks.parallel.for(VS.100).aspx
static void Main()
{
ParallelOptions options = new ParallelOptions();
options.MaxDegreeOfParallelism = 2; // -1 is for unlimited. 1 is for sequential.
try
{
Parallel.For(
0,
9,
options,
(i) =>
{
Console.WriteLine("Thread={0}, i={1}", Thread.CurrentThread.ManagedThreadId, i);
}
);
}
// No exception is expected in this example, but if one is still thrown from a task,
// it will be wrapped in AggregateException and propagated to the main thread.
catch (AggregateException e)
{
Console.WriteLine("Parallel.For has thrown the following (unexpected) exception:\n{0}", e);
}
}
}
Imports System.Threading
Imports System.Threading.Tasks
Module ParallelForDemo
' Demonstrated features:
' Parallel.For()
' ParallelOptions
' Expected results:
' An iteration for each argument value (0, 1, 2, 3, 4, 5, 6, 7, 8, 9) is executed.
' The order of execution of the iterations is undefined.
' Verify that no more than two threads have been used for the iterations.
' Documentation:
' http://msdn.microsoft.com/library/system.threading.tasks.parallel.for(VS.100).aspx
Sub Main()
Dim options As New ParallelOptions()
options.MaxDegreeOfParallelism = 2 ' -1 is for unlimited. 1 is for sequential.
Try
Parallel.For(0, 9, options, Sub(i)
Console.WriteLine("Thread={0}, i={1}", Thread.CurrentThread.ManagedThreadId, i)
End Sub)
Catch e As AggregateException
' No exception is expected in this example, but if one is still thrown from a task,
' it will be wrapped in AggregateException and propagated to the main thread.
Console.WriteLine("Parallel.For has thrown the following (unexpected) exception:" & vbLf & "{0}", e)
End Try
End Sub
End Module
備註
反覆運算 body
範圍中的每個值都會叫用委派一次, fromInclusive
() toExclusive
。 它隨附下列參數:反復專案計數 (Int64) ,以及 ParallelLoopState 可能用來提前中斷迴圈的實例。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不會執行任何反復專案。
另請參閱
適用於
For(Int64, Int64, ParallelOptions, Action<Int64>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
使用 64 位元索引執行 for
迴圈,其中可以平行執行反覆項目,還可以設定迴圈選項。
public:
static System::Threading::Tasks::ParallelLoopResult For(long fromInclusive, long toExclusive, System::Threading::Tasks::ParallelOptions ^ parallelOptions, Action<long> ^ body);
public static System.Threading.Tasks.ParallelLoopResult For (long fromInclusive, long toExclusive, System.Threading.Tasks.ParallelOptions parallelOptions, Action<long> body);
static member For : int64 * int64 * System.Threading.Tasks.ParallelOptions * Action<int64> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For (fromInclusive As Long, toExclusive As Long, parallelOptions As ParallelOptions, body As Action(Of Long)) As ParallelLoopResult
參數
- fromInclusive
- Int64
起始索引 (含)。
- toExclusive
- Int64
結束索引 (不含)。
- parallelOptions
- ParallelOptions
物件,設定這個作業的行為。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
parallelOptions
引數中的 CancellationToken 已取消。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
已處置與 parallelOptions
中的 CancellationTokenSource 相關聯的 CancellationToken。
範例
下列範例示範如何使用 ParallelOptions 來指定自訂工作排程器:
using System;
using System.Collections.Concurrent;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading;
using System.Threading.Tasks;
class ParallelSchedulerDemo2
{
// Demonstrated features:
// TaskScheduler
// BlockingCollection
// Parallel.For()
// ParallelOptions
// Expected results:
// An iteration for each argument value (0, 1, 2, 3, 4, 5, 6, 7, 8, 9) is executed.
// The TwoThreadTaskScheduler employs 2 threads on which iterations may be executed in a random order.
// Thus a scheduler thread may execute multiple iterations.
// Documentation:
// http://msdn.microsoft.com/library/system.threading.tasks.taskscheduler(VS.100).aspx
// http://msdn.microsoft.com/library/dd997413(VS.100).aspx
// More information:
// http://blogs.msdn.com/pfxteam/archive/2009/09/22/9898090.aspx
static void Main()
{
ParallelOptions options = new ParallelOptions();
// Construct and associate a custom task scheduler
options.TaskScheduler = new TwoThreadTaskScheduler();
try
{
Parallel.For(
0,
10,
options,
(i, localState) =>
{
Console.WriteLine("i={0}, Task={1}, Thread={2}", i, Task.CurrentId, Thread.CurrentThread.ManagedThreadId);
}
);
}
// No exception is expected in this example, but if one is still thrown from a task,
// it will be wrapped in AggregateException and propagated to the main thread.
catch (AggregateException e)
{
Console.WriteLine("An iteration has thrown an exception. THIS WAS NOT EXPECTED.\n{0}", e);
}
}
// This scheduler schedules all tasks on (at most) two threads
sealed class TwoThreadTaskScheduler : TaskScheduler, IDisposable
{
// The runtime decides how many tasks to create for the given set of iterations, loop options, and scheduler's max concurrency level.
// Tasks will be queued in this collection
private BlockingCollection<Task> _tasks = new BlockingCollection<Task>();
// Maintain an array of threads. (Feel free to bump up _n.)
private readonly int _n = 2;
private Thread[] _threads;
public TwoThreadTaskScheduler()
{
_threads = new Thread[_n];
// Create unstarted threads based on the same inline delegate
for (int i = 0; i < _n; i++)
{
_threads[i] = new Thread(() =>
{
// The following loop blocks until items become available in the blocking collection.
// Then one thread is unblocked to consume that item.
foreach (var task in _tasks.GetConsumingEnumerable())
{
TryExecuteTask(task);
}
});
// Start each thread
_threads[i].IsBackground = true;
_threads[i].Start();
}
}
// This method is invoked by the runtime to schedule a task
protected override void QueueTask(Task task)
{
_tasks.Add(task);
}
// The runtime will probe if a task can be executed in the current thread.
// By returning false, we direct all tasks to be queued up.
protected override bool TryExecuteTaskInline(Task task, bool taskWasPreviouslyQueued)
{
return false;
}
public override int MaximumConcurrencyLevel { get { return _n; } }
protected override IEnumerable<Task> GetScheduledTasks()
{
return _tasks.ToArray();
}
// Dispose is not thread-safe with other members.
// It may only be used when no more tasks will be queued
// to the scheduler. This implementation will block
// until all previously queued tasks have completed.
public void Dispose()
{
if (_threads != null)
{
_tasks.CompleteAdding();
for (int i = 0; i < _n; i++)
{
_threads[i].Join();
_threads[i] = null;
}
_threads = null;
_tasks.Dispose();
_tasks = null;
}
}
}
}
Imports System.Collections.Concurrent
Imports System.Threading
Imports System.Threading.Tasks
Module SchedulerDemo
' Demonstrated features:
' TaskScheduler
' BlockingCollection
' Parallel.For()
' ParallelOptions
' Expected results:
' An iteration for each argument value (0, 1, 2, 3, 4, 5, 6, 7, 8, 9) is executed.
' The TwoThreadTaskScheduler employs 2 threads on which iterations may be executed in a random order.
' A task is internally created for each thread of the task scheduler (plus an aditional internal task).
' Thus a scheduler thread may execute multiple iterations.
' Documentation:
' http://msdn.microsoft.com/library/system.threading.tasks.taskscheduler(VS.100).aspx
' http://msdn.microsoft.com/library/dd997413(VS.100).aspx
' More information:
' http://blogs.msdn.com/pfxteam/archive/2009/09/22/9898090.aspx
Sub Main()
Dim options As New ParallelOptions()
' Construct and associate a custom task scheduler
options.TaskScheduler = New TwoThreadTaskScheduler()
Try
Parallel.For(0, 10, options, Sub(i, localState)
Console.WriteLine("i={0}, Task={1}, Thread={2}", i, Task.CurrentId, Thread.CurrentThread.ManagedThreadId)
End Sub)
Catch e As AggregateException
' No exception is expected in this example, but if one is still thrown from a task,
' it will be wrapped in AggregateException and propagated to the main thread.
Console.WriteLine("An iteration has thrown an exception. THIS WAS NOT EXPECTED." & vbLf & "{0}", e)
End Try
End Sub
' This scheduler schedules all tasks on (at most) two threads
Private NotInheritable Class TwoThreadTaskScheduler
Inherits TaskScheduler
Implements IDisposable
' The runtime decides how many tasks to create for the given set of iterations, loop options, and scheduler's max concurrency level.
' Tasks will be queued in this collection
Private _tasks As New BlockingCollection(Of Task)()
' Maintain an array of threads. (Feel free to bump up _n.)
Private ReadOnly _n As Integer = 2
Private _threads As Thread()
Public Sub New()
_threads = New Thread(_n - 1) {}
' Create unstarted threads based on the same inline delegate
For i As Integer = 0 To _n - 1
_threads(i) = New Thread(Sub()
' The following loop blocks until items become available in the blocking collection.
' Then one thread is unblocked to consume that item.
For Each task In _tasks.GetConsumingEnumerable()
TryExecuteTask(task)
Next
End Sub)
' Start each thread
_threads(i).IsBackground = True
_threads(i).Start()
Next
End Sub
' This method is invoked by the runtime to schedule a task
Protected Overloads Overrides Sub QueueTask(ByVal task As Task)
_tasks.Add(task)
End Sub
' The runtime will probe if a task can be executed in the current thread.
' By returning false, we direct all tasks to be queued up.
Protected Overloads Overrides Function TryExecuteTaskInline(ByVal task As Task, ByVal taskWasPreviouslyQueued As Boolean) As Boolean
Return False
End Function
Public Overloads Overrides ReadOnly Property MaximumConcurrencyLevel() As Integer
Get
Return _n
End Get
End Property
Protected Overloads Overrides Function GetScheduledTasks() As IEnumerable(Of Task)
Return _tasks.ToArray()
End Function
' Dispose is not thread-safe with other members.
' It may only be used when no more tasks will be queued
' to the scheduler. This implementation will block
' until all previously queued tasks have completed.
Public Sub Dispose() Implements IDisposable.Dispose
If _threads IsNot Nothing Then
_tasks.CompleteAdding()
For i As Integer = 0 To _n - 1
_threads(i).Join()
_threads(i) = Nothing
Next
_threads = Nothing
_tasks.Dispose()
_tasks = Nothing
End If
End Sub
End Class
End Module
備註
支援 64 位索引。 反覆運算 body
範圍中的每個值都會叫用委派一次, fromInclusive
() toExclusive
。 它會以參數的形式提供反復專案計數 (Int64) 。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不需要執行任何反復專案。
另請參閱
適用於
For<TLocal>(Int32, Int32, Func<TLocal>, Func<Int32,ParallelLoopState,TLocal,TLocal>, Action<TLocal>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
使用執行緒區域資料執行 for
迴圈,其中可以平行執行反覆項目,還可以監控和操作迴圈的狀態。
public:
generic <typename TLocal>
static System::Threading::Tasks::ParallelLoopResult For(int fromInclusive, int toExclusive, Func<TLocal> ^ localInit, Func<int, System::Threading::Tasks::ParallelLoopState ^, TLocal, TLocal> ^ body, Action<TLocal> ^ localFinally);
public static System.Threading.Tasks.ParallelLoopResult For<TLocal> (int fromInclusive, int toExclusive, Func<TLocal> localInit, Func<int,System.Threading.Tasks.ParallelLoopState,TLocal,TLocal> body, Action<TLocal> localFinally);
static member For : int * int * Func<'Local> * Func<int, System.Threading.Tasks.ParallelLoopState, 'Local, 'Local> * Action<'Local> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For(Of TLocal) (fromInclusive As Integer, toExclusive As Integer, localInit As Func(Of TLocal), body As Func(Of Integer, ParallelLoopState, TLocal, TLocal), localFinally As Action(Of TLocal)) As ParallelLoopResult
類型參數
- TLocal
執行緒區域資料的類型。
參數
- fromInclusive
- Int32
起始索引 (含)。
- toExclusive
- Int32
結束索引 (不含)。
- localInit
- Func<TLocal>
函式委派,會針對每項工作傳回本機資料的初始狀態。
- body
- Func<Int32,ParallelLoopState,TLocal,TLocal>
每個反覆項目都會叫用一次的委派。
- localFinally
- Action<TLocal>
委派,會針對每項工作的本機狀態執行最終動作。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
備註
反覆運算 body
範圍中的每個值都會叫用委派一次, fromInclusive
() toExclusive
。 它隨附下列參數:反復專案計數 (Int32) 、 ParallelLoopState 可用來提前中斷迴圈的實例,以及可能會在相同執行緒上執行的反復專案之間共用的一些本機狀態。
每個 localInit
參與迴圈執行的工作都會叫用委派一次,並傳回每個工作的初始本機狀態。 這些初始狀態會傳遞至每個工作上的第一個 body
調用。 然後,每個後續的主體調用都會傳回可能修改的狀態值,這個值會傳遞至下一個主體調用。 最後,每個工作的最後一個主體調用會傳回傳遞至 localFinally
委派的狀態值。 每個 localFinally
工作都會叫用委派一次,以在每個工作的本機狀態上執行最終動作。 此委派可能會在多個工作上同時叫用;因此,您必須同步存取任何共用變數。
方法 Parallel.For 在執行期間可能會使用比執行緒更多的工作,因為現有的工作已完成,而且會由新工作取代。 這可讓基礎 TaskScheduler 物件有機會新增、變更或移除服務迴圈的執行緒。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不需要執行任何反復專案。
如需使用此方法的範例,請參閱 How to: Write a Parallel.For Loop with Thread-Local Variables。
另請參閱
適用於
For<TLocal>(Int64, Int64, Func<TLocal>, Func<Int64,ParallelLoopState,TLocal,TLocal>, Action<TLocal>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
使用 64 位元索引和執行緒區域資料執行 for
迴圈,其中可以平行執行反覆項目,還可以監控和操作迴圈的狀態。
public:
generic <typename TLocal>
static System::Threading::Tasks::ParallelLoopResult For(long fromInclusive, long toExclusive, Func<TLocal> ^ localInit, Func<long, System::Threading::Tasks::ParallelLoopState ^, TLocal, TLocal> ^ body, Action<TLocal> ^ localFinally);
public static System.Threading.Tasks.ParallelLoopResult For<TLocal> (long fromInclusive, long toExclusive, Func<TLocal> localInit, Func<long,System.Threading.Tasks.ParallelLoopState,TLocal,TLocal> body, Action<TLocal> localFinally);
static member For : int64 * int64 * Func<'Local> * Func<int64, System.Threading.Tasks.ParallelLoopState, 'Local, 'Local> * Action<'Local> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For(Of TLocal) (fromInclusive As Long, toExclusive As Long, localInit As Func(Of TLocal), body As Func(Of Long, ParallelLoopState, TLocal, TLocal), localFinally As Action(Of TLocal)) As ParallelLoopResult
類型參數
- TLocal
執行緒區域資料的類型。
參數
- fromInclusive
- Int64
起始索引 (含)。
- toExclusive
- Int64
結束索引 (不含)。
- localInit
- Func<TLocal>
函式委派,會針對每項工作傳回本機資料的初始狀態。
- body
- Func<Int64,ParallelLoopState,TLocal,TLocal>
每個反覆項目都會叫用一次的委派。
- localFinally
- Action<TLocal>
委派,會針對每項工作的本機狀態執行最終動作。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
備註
系統會 body
針對反覆運算範圍中的每個值叫用委派一次, (fromInclusive
、 toExclusive
) 。 它隨附下列參數:反復專案計數 (Int64) 、 ParallelLoopState 可能用來提前中斷迴圈的實例,以及在相同工作上執行的反復專案之間共用的一些本機狀態。
localInit
委派會針對參與迴圈執行的每個工作叫用一次,並傳回每個工作的初始本機狀態。 這些初始狀態會傳遞至每個工作上的第一個 body
調用。 然後,每個後續的主體調用都會傳回可能修改的狀態值,該值會傳遞至下一個主體調用。 最後,每個工作的最後一個主體調用會傳回傳遞至 localFinally
委派的狀態值。 每個 localFinally
工作都會叫用委派一次,以在每個工作的本機狀態上執行最終動作。 此委派可能會在多個工作上同時叫用;因此,您必須同步存取任何共用變數。
當現有工作完成且由新工作取代時,方法 Parallel.For 可能會在執行期間使用比執行緒更多的工作。 這可讓基礎 TaskScheduler 物件有機會新增、變更或移除服務迴圈的執行緒。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不需要執行任何反復專案。
如需使用此方法的範例,請參閱 How to: Write a Parallel.For Loop with Thread-Local Variables。
另請參閱
適用於
For<TLocal>(Int32, Int32, ParallelOptions, Func<TLocal>, Func<Int32,ParallelLoopState,TLocal,TLocal>, Action<TLocal>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
使用執行緒區域資料執行 for
迴圈,其中可以平行執行反覆項目、可以設定迴圈選項,還可以監控和操作迴圈的狀態。
public:
generic <typename TLocal>
static System::Threading::Tasks::ParallelLoopResult For(int fromInclusive, int toExclusive, System::Threading::Tasks::ParallelOptions ^ parallelOptions, Func<TLocal> ^ localInit, Func<int, System::Threading::Tasks::ParallelLoopState ^, TLocal, TLocal> ^ body, Action<TLocal> ^ localFinally);
public static System.Threading.Tasks.ParallelLoopResult For<TLocal> (int fromInclusive, int toExclusive, System.Threading.Tasks.ParallelOptions parallelOptions, Func<TLocal> localInit, Func<int,System.Threading.Tasks.ParallelLoopState,TLocal,TLocal> body, Action<TLocal> localFinally);
static member For : int * int * System.Threading.Tasks.ParallelOptions * Func<'Local> * Func<int, System.Threading.Tasks.ParallelLoopState, 'Local, 'Local> * Action<'Local> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For(Of TLocal) (fromInclusive As Integer, toExclusive As Integer, parallelOptions As ParallelOptions, localInit As Func(Of TLocal), body As Func(Of Integer, ParallelLoopState, TLocal, TLocal), localFinally As Action(Of TLocal)) As ParallelLoopResult
類型參數
- TLocal
執行緒區域資料的類型。
參數
- fromInclusive
- Int32
起始索引 (含)。
- toExclusive
- Int32
結束索引 (不含)。
- parallelOptions
- ParallelOptions
物件,設定這個作業的行為。
- localInit
- Func<TLocal>
函式委派,會針對每項工作傳回本機資料的初始狀態。
- body
- Func<Int32,ParallelLoopState,TLocal,TLocal>
每個反覆項目都會叫用一次的委派。
- localFinally
- Action<TLocal>
委派,會針對每項工作的本機狀態執行最終動作。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
body
引數為 null
。
-或-
localInit
引數為 null
。
-或-
localFinally
引數為 null
。
-或-
parallelOptions
引數為 null
。
parallelOptions
引數中的 CancellationToken 已取消。
已處置與 parallelOptions
中的 CancellationTokenSource 相關聯的 CancellationToken。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
範例
下列範例會使用執行緒區域變數來計算許多冗長作業的結果總和。 本範例會將平行處理原則的程度限制為四。
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading;
using System.Threading.Tasks;
class ThreadLocalForWithOptions
{
// The number of parallel iterations to perform.
const int N = 1000000;
static void Main()
{
// The result of all thread-local computations.
int result = 0;
// This example limits the degree of parallelism to four.
// You might limit the degree of parallelism when your algorithm
// does not scale beyond a certain number of cores or when you
// enforce a particular quality of service in your application.
Parallel.For(0, N, new ParallelOptions { MaxDegreeOfParallelism = 4 },
// Initialize the local states
() => 0,
// Accumulate the thread-local computations in the loop body
(i, loop, localState) =>
{
return localState + Compute(i);
},
// Combine all local states
localState => Interlocked.Add(ref result, localState)
);
// Print the actual and expected results.
Console.WriteLine("Actual result: {0}. Expected 1000000.", result);
}
// Simulates a lengthy operation.
private static int Compute(int n)
{
for (int i = 0; i < 10000; i++) ;
return 1;
}
}
Imports System.Threading
Imports System.Threading.Tasks
Module ThreadLocalForWithOptions
' The number of parallel iterations to perform.
Const N As Integer = 1000000
Sub Main()
' The result of all thread-local computations.
Dim result As Integer = 0
' This example limits the degree of parallelism to four.
' You might limit the degree of parallelism when your algorithm
' does not scale beyond a certain number of cores or when you
' enforce a particular quality of service in your application.
Parallel.For(0, N, New ParallelOptions With {.MaxDegreeOfParallelism = 4},
Function()
' Initialize the local states
Return 0
End Function,
Function(i, loopState, localState)
' Accumulate the thread-local computations in the loop body
Return localState + Compute(i)
End Function,
Sub(localState)
' Combine all local states
Interlocked.Add(result, localState)
End Sub
)
' Print the actual and expected results.
Console.WriteLine("Actual result: {0}. Expected 1000000.", result)
End Sub
' Simulates a lengthy operation.
Function Compute(ByVal n As Integer) As Integer
For i As Integer = 0 To 10000
Next
Return 1
End Function
End Module
備註
系統會 body
針對反覆運算範圍中的每個值叫用委派一次, (fromInclusive
、 toExclusive
) 。 它隨附下列參數:反復專案計數 (Int32) 、 ParallelLoopState 可能用來提前中斷迴圈的實例,以及在相同工作上執行的反復專案之間共用的一些本機狀態。
localInit
委派會針對參與迴圈執行的每個工作叫用一次,並傳回每個工作的初始本機狀態。 這些初始狀態會傳遞至每個工作上的第一個 body
調用。 然後,每個後續的主體調用都會傳回可能修改的狀態值,該值會傳遞至下一個主體調用。 最後,每個工作的最後一個主體調用會傳回傳遞至 localFinally
委派的狀態值。 每個 localFinally
工作都會叫用委派一次,以在每個工作的本機狀態上執行最終動作。 此委派可能會在多個執行緒上同時叫用;因此,您必須同步存取任何共用變數。
當現有工作完成且由新工作取代時,方法 Parallel.For 可能會在執行期間使用比執行緒更多的工作。 這可讓基礎 TaskScheduler 物件有機會新增、變更或移除服務迴圈的執行緒。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不需要執行任何反復專案。
另請參閱
適用於
For<TLocal>(Int64, Int64, ParallelOptions, Func<TLocal>, Func<Int64,ParallelLoopState,TLocal,TLocal>, Action<TLocal>)
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
- 來源:
- Parallel.cs
使用 64 位元索引和執行緒區域資料執行 for
迴圈,其中可以平行執行反覆項目、可以設定迴圈選項,還可以監控和操作迴圈的狀態。
public:
generic <typename TLocal>
static System::Threading::Tasks::ParallelLoopResult For(long fromInclusive, long toExclusive, System::Threading::Tasks::ParallelOptions ^ parallelOptions, Func<TLocal> ^ localInit, Func<long, System::Threading::Tasks::ParallelLoopState ^, TLocal, TLocal> ^ body, Action<TLocal> ^ localFinally);
public static System.Threading.Tasks.ParallelLoopResult For<TLocal> (long fromInclusive, long toExclusive, System.Threading.Tasks.ParallelOptions parallelOptions, Func<TLocal> localInit, Func<long,System.Threading.Tasks.ParallelLoopState,TLocal,TLocal> body, Action<TLocal> localFinally);
static member For : int64 * int64 * System.Threading.Tasks.ParallelOptions * Func<'Local> * Func<int64, System.Threading.Tasks.ParallelLoopState, 'Local, 'Local> * Action<'Local> -> System.Threading.Tasks.ParallelLoopResult
Public Shared Function For(Of TLocal) (fromInclusive As Long, toExclusive As Long, parallelOptions As ParallelOptions, localInit As Func(Of TLocal), body As Func(Of Long, ParallelLoopState, TLocal, TLocal), localFinally As Action(Of TLocal)) As ParallelLoopResult
類型參數
- TLocal
執行緒區域資料的類型。
參數
- fromInclusive
- Int64
起始索引 (含)。
- toExclusive
- Int64
結束索引 (不含)。
- parallelOptions
- ParallelOptions
物件,設定這個作業的行為。
- localInit
- Func<TLocal>
函式委派,會針對每個執行緒傳回本機資料的初始狀態。
- body
- Func<Int64,ParallelLoopState,TLocal,TLocal>
每個反覆項目都會叫用一次的委派。
- localFinally
- Action<TLocal>
委派,會針對每個執行緒的本機狀態執行最終動作。
傳回
包含迴圈已完成部分之相關資訊的的結構。
例外狀況
body
引數為 null
。
-或-
localInit
引數為 null
。
-或-
localFinally
引數為 null
。
-或-
parallelOptions
引數為 null
。
parallelOptions
引數中的 CancellationToken 已取消。
已處置與 parallelOptions
中的 CancellationTokenSource 相關聯的 CancellationToken。
在所有執行緒擲回包含所有個別例外狀況的例外狀況。
備註
系統會 body
針對反覆運算範圍中的每個值叫用委派一次, (fromInclusive
、 toExclusive
) 。 它提供下列參數:反復專案計數 (Int64) 、 ParallelLoopState 可能用來提前中斷迴圈的實例,以及在相同執行緒上執行的反復專案之間可能共用的一些本機狀態。
localInit
委派會針對參與迴圈執行的每個執行緒叫用一次,並傳回每個執行緒的初始本機狀態。 這些初始狀態會傳遞至每個執行緒上的第一個 body
調用。 然後,每個後續的主體調用都會傳回可能修改的狀態值,該值會傳遞至下一個主體調用。 最後,每個執行緒上的最後一個主體調用會傳回傳遞至 localFinally
委派的狀態值。 每個 localFinally
執行緒都會叫用委派一次,以在每個執行緒的本機狀態上執行最終動作。 此委派可能會在多個執行緒上同時叫用;因此,您必須同步存取任何共用變數。
當現有工作完成且由新工作取代時,方法 Parallel.For 可能會在執行期間使用比執行緒更多的工作。 這可讓基礎 TaskScheduler 物件有機會新增、變更或移除服務迴圈的執行緒。
如果 fromInclusive
大於或等於 toExclusive
,則方法會立即傳回,而不需要執行任何反復專案。