Linq中的合并结果 [英] Coalescing results in Linq

问题描述

浏览了SO上的许多帖子，但没有找到解决此问题的任何帖子.请注意，此处提供的所有代码均已简化，但代表了真实代码.我有一个数据表，该数据表描述了保险计划的某些属性.找回最匹配项的查询如下所示:

have looked through many posts here on SO, and haven't found any that address this. Just a note that all code presented here is simplified but representative of the real code. I have a data table that describes some properties of coverage plans. The query to bring back the best match looks something like this:

select coalesce
(
(select c.PercentOfCoverageA from CoveragePlans c
where c.coverage = :COVERAGE
and c.plancode = :PLANCODE
and c.statecode = :STATECODE),

(select c.PercentOfCoverageA from CoveragePlans c
where c.coverage = :COVERAGE
and c.plancode = :DEFAULTPLANCODE
and c.statecode = :STATECODE),

(select c.PercentOfCoverageA from CoveragePlans c
where c.coverage = :COVERAGE
and c.plancode = :DEFAULTPLANCODE
and c.statecode = :COUNTRYWIDE)
) as PercentOfCoverageA
from dual

这是一个很小的表(几十行)，被击中的次数很少，并且不经常更改，因此我想将其存储到内存中，并使用Linq选择数据以加快速度.

This is a small table (a few dozen rows) that gets hit a lot and changes infrequently, so I want to bring it into memory and use Linq to select the data to speed this up.

我有此函数，该函数完全按照我的要求返回第一个匹配项:

I have this function which returns the first match exactly as I want it to:

decimal GetCoveragePercentage(string coverage, string planCode, string stateCode)
{
    IEnumerable<CoveragePlan> result = Coverages
        .Where(x => x.Coverage == coverage && x.PlanCode == planCode && x.StateCode == stateCode)
        .Select(x => x);

    if (!result.Any())
    {
        result = Coverages
            .Where(x => x.Coverage == coverage && x.PlanCode == defaultPlanCode && x.StateCode == stateCode)
            .Select(x => x);
    }

    if (!result.Any())
    {
        result = Coverages
            .Where(x => x.Coverage == coverage && x.PlanCode == defaultPlanCode && x.StateCode == countryWide)
            .Select(x => x);
    }

    return result.First().PercentOfCoverageA;
}

我的问题是:是否有更好的方法(更快，更少的代码，更少的重复)来执行此Linq查询?

My question is: Is there a better way (faster, less code, less repetition) to do this Linq query?

更新:我最终用此功能代替了

Update: I ended up with this function as a replacement:

decimal GetCoveragePercentage(string coverage, string planCode, string stateCode)
{
    return Coverages.Where(x => x.Equals(coverage, planCode, stateCode))
        .DefaultIfEmpty(Coverages.Where(x => x.Equals(coverage, defaultPlanCode, stateCode)).FirstOrDefault()
            ?? Coverages.Where(x => x.Equals(coverage, defaultPlanCode, defaultStateCode)).First())
        .First().PercentOfCoverageA;
}

DefaultIfEmpty需要一个实例，而不是实例的IEnumeration.这导致我在回退子查询上添加First/FirstOrDefault，事实证明，如果为它提供null，则DefaultIfEmpty会讨厌它，因此我使用了null合并运算符来汇总回退级别.

The DefaultIfEmpty wants an instance instead of an IEnumeration of instances. This lead me to adding the First/FirstOrDefault on the fallback subqueries, and it turns out that DefaultIfEmpty hates it if you give it a null, so I used the null coalescing operater to roll up the fallback levels.

我不确定他们为什么不给您带IEnumeration的DefaultIfEmpty，就是这样:

I'm not sure why they don't give you a DefaultIfEmpty that takes an IEnumeration, it would just be this:

public static IEnumerable<TSource> DefaultIfEmpty<TSource>(this IEnumerable<TSource> source, IEnumerable<TSource> defaultValue)
{
    return (source != null && source.Any()) ? source : defaultValue;
}

实际上，我想我将使用该扩展方法，并使我的功能如下:

Actually, I think I'll use that extension method, and have my function be this:

decimal GetCoveragePercentage(string coverage, string planCode, string stateCode)
{
    return Coverages.Where(x => x.Equals(coverage, planCode, stateCode))
        .DefaultIfEmpty(Coverages.Where(x => x.Equals(coverage, defaultPlanCode, stateCode)))
        .DefaultIfEmpty(Coverages.Where(x => x.Equals(coverage, defaultPlanCode, defaultStateCode)))
        .First().PercentOfCoverageA;
}

推荐答案

我相信.Select(x => x);实际上不执行任何操作.这样就可以删除了.您可以使用并集函数加入查询.至于是否检查结果，则可以使用此功能DefaultIfEmpty()进行调查.

I believe that .Select(x => x); actually does nothing. So that can be removed. You can join your queries using a union function. As for the if no results checks, you could investigate using this function DefaultIfEmpty().

我还建议使用harsharp，以提供有关优化LINQ的建议.

I would also recommend resharper at being helpful with suggestions on optimizing LINQ.

我还认为您应该遵守DRY原则，并且不要使用以下代码行:

I also think you should abide by DRY principals and not have this line of code:

x.Coverage == coverage && x.PlanCode == defaultPlanCode && x.StateCode == stateCode

而不是将其替换为:

x.Equals(coverage,defaultPlanCode,stateCode)

我建议您的方法的linq看起来像这样(确保同时添加等于方法的优化):

I propose your linq for your method would look like this (make sure you add that equals method optimization as well as this):

decimal GetCoveragePercentage(string coverage, string planCode, string stateCode)
{
    return Coverages
        .Where(x => x.Coverage == coverage && x.PlanCode == planCode && x.StateCode == stateCode)
        .DefaultIfEmpty(Coverages.Where(x => x.Coverage == coverage && x.PlanCode == defaultPlanCode && x.StateCode == stateCode))
        .DefaultIfEmpty(Coverages.Where(x => x.Coverage == coverage && x.PlanCode == defaultPlanCode && x.StateCode == countryWide))First().PercentOfCoverageA;

}

这篇关于Linq中的合并结果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！