将搜索结果缩小到多个类别 [英] Narrowing search result to multiple categories

问题描述

我正在尝试将搜索结果缩小到任何给定的类别。
我的搜索body，在elasticsearch上工作2.2.0

I am trying to narrow down search results to be within any given categories. My search body that works on elasticsearch 2.2.0

{
   "filter":{
      "bool":{
         "should":[
            {
               "match":{
                  "categories":"16310211"
               }
            },
            {
               "match":{
                  "categories":"493964"
               }
            }
         ]
      }
   },
   "size":25,
   "from":0
}

然而，现在我们使用的是AWS托管服务，它们只支持1.5.6。当我运行相同的代码时，我会得到一个http 400的不良请求。如何在1.5.6上实现多个条件（一个或多个）过滤结果的相同效果？

However now we are using AWS hosted services and they only provide support for 1.5.6. When I run the same code I get back a http 400 bad request. How can I achieve the same effect of filtering results by multiple criteria (one or more) on 1.5.6?

推荐答案

如果你确认使用多个匹配查询，那么您需要将它们与 bool 查询即可。尝试这样：

If you're committed to using multiple match queries, then you will need to combine them with a bool query. Try this:

{
   "query":{        <--- use query here instead of filter
      "bool":{
         "should":[
            {
               "match":{
                  "categories":"16310211"
               }
            },
            {
               "match":{
                  "categories":"493964"
               }
            }
         ]
      }
   },
   "size":25,
   "from":0
}

但是，通常使用匹配查询进行全文搜索。它看起来好像你的类别是确切的值。我会为这种类型的数据推荐一个术语过滤器。您可能有以下更好的体验：

However, generally the match query is used for doing full text searching. It looks as if your "categories" are exact values. I would recommend a term filter for this type of data instead. You may have a better experience with the following:

{
  "query":{
    "filtered":{
      "filter":{
        "bool":{
          "should":[
            { "term":{"categories":"16310211"} },
            { "term":{"categories":"493964"} }
          ]
        }
      }
    }
  },
  "size":25,
  "from":0
}

这篇关于将搜索结果缩小到多个类别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！