流利Nhibernate加载不会延缓加载？ （孙子实体） [英] Fluent Nhibernate loading not falling back on lazy loading? (grandchild entities)

问题描述

当查询nhibernate时，我看到一些奇怪的行为。当我写这样的查询时，

<

（$ = $ x code> Repository.QueryOver< Entity>（
.Fetch（x => x.Child1）.Eager
.Fetch（x => x.child2） .eager

它会急切地抓住child1和child2的实体，但有child1和child2的孙子懒洋洋地装着。我有点困惑如何实现这一点。

在我的nhibernate映射，似乎没有影响的懒惰或孙子孙子，我至少需要为了避免N + 1查询问题，一些实体被急切地加载。

我也想知道如何在我的原始实体下急切地加载孙子实体。

任何帮助或想法都会被赞赏！

解决方案

我会建议使用<批量抓取。正如这里所讨论的，流利的语法是：

<1> 设置

  HasMany< MyEntity>（x => x.Entities）
。批量大小（100）; 
  

<2> class 级别设置

  public MyEntityMap（）
 {
 Id（x => x .... 
 ... 
 BatchSize（100）; 
  

这个设置应该适用于每个集合和每个类。流利的 - 我们可以使用 惯例 - 看更多例如这里

'IClassConvention ' - 用来改变你的ClassMaps值，你不能用这个约定来改变属性，集合等，只能改变 Lazy 和 BatchSize 。

When querying nhibernate, I'm seeing some odd behavior

When I write a query like this

Repository.QueryOver<Entity>()
    .Fetch(x => x.Child1).Eager
    .Fetch(x => x.child2).Eager

It will eagerly grab child1 and child2 entities, but there are grandchildren for child1 and child2 that aren't lazily loaded. I'm a bit confused on how to achieve this.

In my nhibernate mappings, it seems to have no affect on the laziness or eagerness of grandchildren and I require at least some entities be eagerly loaded to avoid the N+1 query problem.

I'm also wondering how I could eagerly load grandchildren entities under my original entity.

Any help or ideas are appreciated!

解决方案

I would suggest to use the batch-fetching. As discussed here, the fluent syntax is:

1) the collection setting

HasMany<MyEntity>(x => x.Entities)
  .BatchSize(100);

2) the class level setting

public MyEntityMap()
{
    Id(x => x....
    ...
    BatchSize(100);

This setting should be applied on every collection and every class. To do that with fluent - we can use Conventions - see more e.g. here

'IClassConvention' - Use to alter your ClassMaps values. You can't change properties, collections, etc with this convention, only alter the settings such as Lazy and BatchSize.

这篇关于流利Nhibernate加载不会延缓加载？ （孙子实体）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！