1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
|
/* ****************************************************************************
*
* Copyright (c) Microsoft Corporation.
*
* This source code is subject to terms and conditions of the Microsoft Public License. A
* copy of the license can be found in the License.html file at the root of this distribution. If
* you cannot locate the Microsoft Public License, please send an email to
* dlr@microsoft.com. By using this source code in any fashion, you are agreeing to be bound
* by the terms of the Microsoft Public License.
*
* You must not remove this notice, or any other, from this software.
*
*
* ***************************************************************************/
using System;
using System.Collections.Generic;
using System.Text;
#if CLR2
namespace Microsoft.Scripting.Ast {
using Microsoft.Scripting.Utils;
#else
namespace System.Linq.Expressions {
#endif
/// <summary>
/// Provides an internal interface for accessing the arguments that multiple tree
/// nodes (DynamicExpression, ElementInit, MethodCallExpression, InvocationExpression, NewExpression,
/// and InexExpression).
///
/// This enables two optimizations which reduce the size of the trees. The first is it enables
/// the nodes to hold onto an IList of T instead of a ReadOnlyCollection. This saves the cost
/// of allocating the ReadOnlyCollection for each node. The second is that it enables specialized
/// subclasses to be created which hold onto a specific number of arguments. For example Block2,
/// Block3, Block4. These nodes can therefore avoid allocating both a ReadOnlyCollection and an
/// array for storing their elements saving 32 bytes per node.
///
/// Meanwhile the nodes can continue to expose the original LINQ properties of ReadOnlyCollections. They
/// do this by re-using 1 field for storing both the array or an element that would normally be stored
/// in the array.
///
/// For the array case the collection is typed to IList of T instead of ReadOnlyCollection of T.
/// When the node is initially constructed it is an array. When the compiler accesses the members it
/// uses this interface. If a user accesses the members the array is promoted to a ReadOnlyCollection.
///
/// For the object case we store the 1st argument in a field typed to object and when the node is initially
/// constructed this holds directly onto the Expression. When the compiler accesses the members
/// it again uses this interface and the accessor for the 1st argument uses Expression.ReturnObject to
/// return the object which handles the Expression or ReadOnlyCollection case. When the user accesses
/// the ReadOnlyCollection then the object field is updated to hold directly onto the ReadOnlyCollection.
///
/// It is important that the Expressions consistently return the same ReadOnlyCollection otherwise the
/// re-writer will be broken and it would be a breaking change from LINQ v1. The problem is that currently
/// users can rely on object identity to tell if the node has changed. Storing the readonly collection in
/// an overloaded field enables us to both reduce memory usage as well as maintain compatibility and an
/// easy to use external API.
/// </summary>
internal interface IArgumentProvider {
Expression GetArgument(int index);
int ArgumentCount {
get;
}
}
static class ArgumentProviderOps {
internal static T[] Map<T>(this IArgumentProvider collection, Func<Expression, T> select) {
int count = collection.ArgumentCount;
T[] result = new T[count];
count = 0;
for (int i = 0; i < count; i++) {
result[i] = select(collection.GetArgument(i));
}
return result;
}
}
}
|
