带条件的订单清单选择运行时 [英] order List with criteria choose to runtime

问题描述

如果我有一个具有更多字段的泛型列表，例如：

if i have a generics list with more field for example:

PMyList = record 
  Field1, Field2, ... FieldN : Integer;
end;
TMyList = List<PMyList>;

要订购具有条件的列表，请选择运行时（例如：field2然后field2，或者：field3然后field1然后field2等）有某种解决方案，或者我需要为我想要的订单的所有可能组合做一个比较构造？

For order the list with criteria choose to runtime (for example: field2 then field2, or: field3 then field1 then field2 etc) there is some solution or i need to do a compare construct for all combination possibile of order that i want?

我的想法是，如果记录为N字段，我认为是这样定义的数组：

Mine idea was, if record is N field, i have thinked to a array so defined:

MyArray = array [1..n] of Integer;

并将一个累进值分配给确定sord条件的数组元素，例如，如果MyArray如下：

and assign a progressive value to elements of array that determine criteria of sord for example if MyArray is as:

MyArray = (5, 1, 3, 4, 2)

表示我的列表需要首先对field5进行排序，然后对field1，然后对field3，再对于field4，再对field2进行排序。
我的问题是：我可以使用一个唯一的结构比较列表吗？

mean that my list need to be sort first for field5, then for field1, then for field3, then for field4, then for field2. Mine question then is: can i to do it using one only construct compare for my list?

非常感谢您的帮助。

推荐答案

我将基于您的上一个问题。我还要将 MyArray 重命名为 FieldPriority 。

I'm going to base the notation on your previous question. I'm also going to rename MyArray as FieldPriority.

因此， FieldPriority [1] 标识主要比较字段， FieldPriority [2] 标识辅助比较字段，

So, FieldPriority[1] identifies the primary comparison field, FieldPriority[2] identifies the secondary comparison field and so on.

在此位置，您的比较函数如下所示：

With this in place your compare function looks like this:

type
  TMyRecord  = record
    Value: array [1..5] of Integer;
  end;

function Compare(const Left, Right: TMyRecord): Integer;
var
  i, Field: Integer;
begin
  for i := 1 to 5 do
  begin
    Field := FieldPriority[i];
    Result := CompareInt(Left.Value[Field], Right.Value[Field]);
    if Result<>0 then
      exit;
  end;
end;

如果将记录中的整数声明为数组而不是单独声明，则效果会更好。

It all works much better if the integers in your record are declared as an array rather than individually. That allows you to index them as I do here.

自然地，所有这些都可以泛化为处理任意大小的数组。

Naturally this could all be generalised to handle arbitrary sized arrays.

这篇关于带条件的订单清单选择运行时的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！