我有一个类line.strip()
,其中包含一个“核心”初始化逻辑块。
FooBarSet
由FooBar
和Foo
组成。使用Bar
列表初始化类FooBarSet
。 FooBar
也可以 通过单独的并行FooBarSet
和Foo
列表进行初始化。
理想情况下,我可以像这样运行它:
Bar
这是C#,而不是Java,因此public class FooBarSet
{
private readonly List<FooBar> _list;
// Primary constructor.
public FooBarSet(List<FooBar> foobarList)
{
// Contracts and initialization...
_list = foobarList;
}
// Secondary constructor.
public FooBarSet(List<Foo> fooList, List<Bar> barList)
{
// Zip a new list of new FooBars
var zipped = fooList.Zip(barList,
(foo, bar) => new FooBar(foo, bar));
// Call primary constructor with zipped list.
this(zipped);
}
}
是非法的。常见的解决方案here是将核心初始化拉入一个通用的私有方法:
this(zipped)
但是,由于public class FooBarSet
{
private readonly List<FooBar> _list;
// Common "constructor" called from actual constructors.
private Init(List<FooBar> foobarList)
{
// Contracts and initialization...
_list = foobarList;
}
public FooBarSet(List<FooBar> foobarList)
{
Init(foobarList);
}
public FooBarSet(List<Foo> fooList, List<Bar> barList)
{
var zipped = fooList.Zip(barList,
(foo, bar) => new FooBar(foo, bar));
Init(zipped);
}
}
字段,这也不起作用。
假设readonly _list
必须是_list
,我该如何让这些构造函数共享初始化代码?
答案 0 :(得分:3)
您可以将“zipper”代码移动到静态函数并使用:
public class FooBarSet
{
private readonly List<FooBar> _list;
private static List<FooBar> Prepare( List<Foo> fooList, List<Bar> barList )
{
return fooList.Zip(barList, (foo, bar) => new FooBar(foo, bar));
}
public FooBarSet(List<Foo> fooList, List<Bar> barList) :
this( Prepare( fooList, barList ) )
{
}
public FooBarSet(List<FooBar> zippedList)
{
_list = zippedList;
}
}
答案 1 :(得分:2)
简单的答案是让Init
方法返回在构造函数中设置的值:
public class FooBarSet
{
private readonly List<FooBar> _list;
// Common "constructor" called from actual constructors.
private List<FooBar> Init(List<FooBar> foobarList)
{
// Contracts and initialization...
return whateverList;
}
public FooBarSet(List<FooBar> foobarList)
{
_list = Init(foobarList);
}
public FooBarSet(List<Foo> fooList, List<Bar> barList)
{
var zipped = fooList.Zip(barList,
(foo, bar) => new FooBar(foo, bar));
_list = Init(zipped);
}
}