急切加载 Paperclip 的 has_attached_file 数据？
Product has_many :assets
Asset own_to :product
Asset has_attached_file :photo 使用回形针（包含所有标准回形针选项；一些样式、S3 存储） 。
如果我形成像 p = Product.includes(:assets) 这样的 Product 查询，我可以调用 Asset 的每个实际属性，而不会产生任何额外的数据库查询，例如：
p = Product.includes(:assets)
p.each { |a| print a.assets.first.title }
titleAsset（这是一个数据库列）上的 > 属性打印，不进行任何查询。
要获取由 Paperclip 生成的网址：
p = Product.includes(:assets)
p.each { |a| print a.assets.first.photo.url }
会对每个 Product 进行单独的附加查询：
Product Load (0.3ms) SELECT "products".* FROM "products" WHERE "products"."id" = 2 LIMIT 1
根据 这个 G.Groups 帖子，我不应该在每次循环时都访问数据库，但我确实这样做了。
有没有一种方法可以避免每次迭代产生额外的数据库命中，而是一次性收集所有数据？我是否忽略了一些简单的事情？
Rails 3.0.9、REE 1.8.7、回形针 2.3.11。
已更新，已修复
问题是在我的回形针设置中，我将 :product_id 作为 :path 的一部分：:attachment/ :product_id/:filename-:style.:extension，这会导致循环中每次迭代进行额外的 Product 查询。
通过将查询更改为 p = Product.includes(:assets => [:product]).all，它删除了附加查询。
Product has_many :assets
Asset belongs_to :product
Asset has_attached_file :photo using Paperclip (with all your standard Paperclip options; a few styles, S3 storage).
If I form a Product query like p = Product.includes(:assets) I can call each of the actual attributes of Asset without incurring any additional database queries, like:
p = Product.includes(:assets)
p.each { |a| print a.assets.first.title }
The title attribute on Asset (which is a database column) prints, no queries made.
To get the URL generated by Paperclip:
p = Product.includes(:assets)
p.each { |a| print a.assets.first.photo.url }
causes a separate additional query for each Product:
Product Load (0.3ms) SELECT "products".* FROM "products" WHERE "products"."id" = 2 LIMIT 1
According to this G.Groups posting, I shouldn't be hitting the database with each pass through the loop, but I am.
Is there a way to not incur the additional database hit for each iteration, but gather all the data all at once? Am I overlooking something simple?
Rails 3.0.9, REE 1.8.7, Paperclip 2.3.11.
updated, fixed
The issue is that in my Paperclip settings, I have :product_id as part of the :path: :attachment/:product_id/:filename-:style.:extension, which is causing the additional Product query for each iteration through the loop.
By changing the query to p = Product.includes(:assets => [:product]).all, it removed the additional query.
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（2）
如果您使用查询分析器插件之一尝试查询审查器，或者如果您使用 newrelic，那么我们可以更快地找到问题的根源，开发模式包括此功能。您将获得每个查询的堆栈跟踪。我不认为它与急切加载有任何关系 - 如果确实如此，您将获得额外的资产加载而不是产品加载。
We can get to the root of the problem much faster if you use one of the query analyzer plugins Try query-reviewer, or if you're using newrelic, the dev mode includes this. You'll get a stack trace for every query. I don't think it has anything to do with eager loading - if it did, you'd be getting extra Asset loads instead of Product loads.
我认为您忽略了一些简单的事情：
includes设置关系以立即加载关联，all执行实际查询。I think you're overlooking something simple:
includessets the relation to eager load the association,alldoes the actual query.