Elasticsearch(3) show me some examples

Let's clear up that confusion right now. Seeing the actual query structure makes a world of difference.

To make this completely clear, let's pretend we have an index called store_products. Before we search, let's look at the type of data we are working with. Imagine our index has documents that look like this:

json 复制代码
{
  "name": "Wireless Noise-Canceling Headphones",
  "brand": "Sony",
  "category": "Electronics",
  "price": 199.99,
  "in_stock": true,
  "tags": ["audio", "wireless", "gadget"]
}

Now, let's look at three detailed, real-world search examples, ranging from simple to advanced.


Example 1: The "Bool" Query (Combining Queries and Filters)

In the real world, you rarely just search for a keyword. Usually, a user types a word, and then clicks some checkboxes to filter the results. In Elasticsearch, we do this using a bool (Boolean) query.

Inside a bool query, we use two main clauses:

  • must: The results must match this text search (calculates a relevance score).
  • filter: The results must match this exact criteria (fast, cached, does not affect the score).

The Scenario

A user searches your store for the word "wireless" , but they only want items in the "Electronics" category that cost under $250.

json 复制代码
GET /store_products/_search
{
  "query": {
    "bool": {
      "must": [
        {
          "match": {
            "name": "wireless"
          }
        }
      ],
      "filter": [
        {
          "term": {
            "category.keyword": "Electronics"
          }
        },
        {
          "range": {
            "price": {
              "lt": 250.00
            }
          }
        }
      ]
    }
  }
}

💡 Crucial Detail: What is .keyword?

Notice that for the category filter, I wrote "category.keyword" instead of just "category".

  • category (Text field) is broken down into lowercase tokens for searching (e.g., "electronics").
  • category.keyword (Keyword field) treats the entire string as one exact unit ("Electronics"). When doing exact filters, always use the .keyword version of a text field.

Example 2: Aggregations (Getting Analytics Data)

Aggregations don't just find documents; they calculate data about your documents. Think of it like a GROUP BY and AVG() in SQL.

The Scenario

You want to build a dashboard sidebar. You need Elasticsearch to look at all your products and tell you:

  1. How many products are in each category? (Bucket Aggregation)
  2. What is the average_price of the products in each of those categories? (Metric Aggregation)
json 复制代码
GET /store_products/_search
{
  "size": 0, 
  "aggs": {
    "group_by_category": {
      "terms": {
        "field": "category.keyword"
      },
      "aggs": {
        "average_price": {
          "avg": {
            "field": "price"
          }
        }
      }
    }
  }
}

💡 Crucial Detail: Why "size": 0?

By default, a search returns the top 10 matching documents plus the aggregation data. By setting "size": 0, we tell Elasticsearch: "I don't want to see individual products. I only want the analytical totals." This makes the query incredibly fast.


Example 3: Putting It All Together (Search, Filter, and Aggregate)

This is the ultimate intermediate query. You can combine everything we just learned into a single request.

The Scenario

A user searches for the word "gadget" . You want to show them the matching items that are in stock , but you also want to show them a breakdown of which brands make those gadgets and how much they cost on average.

json 复制代码
GET /store_products/_search
{
  "size": 5,
  "query": {
    "bool": {
      "must": [
        { "match": { "tags": "gadget" } }
      ],
      "filter": [
        { "term": { "in_stock": true } }
      ]
    }
  },
  "aggs": {
    "brands_breakdown": {
      "terms": {
        "field": "brand.keyword"
      },
      "aggs": {
        "avg_brand_price": {
          "avg": { "field": "price" }
        }
      }
    }
  }
}

What Elasticsearch will return:

  1. The top 5 documents ("size": 5) that have the tag "gadget" and are currently in stock, ranked by how relevant they are to the search.
  2. An aggregation block that only looks at those specific matching gadgets, groups them by brand, and calculates the average price per brand.

If you paste these templates into Elasticvue or Kibana, you can easily change the field names to match your own data!

相关推荐
大大大大晴天1 小时前
浅聊维度建模中的星型模型与雪花模型
大数据
数字新视界5 小时前
能耗管理系统提升企业能源效率的实践与展望
大数据·人工智能·数据中心·机房管理·模块化机房
大大大大晴天6 小时前
别让金额越算越多:事实表与维度表设计避坑指南
大数据
如夏倾城7 小时前
入侵检测系统(IDS)原理、部署模式与 IPS/防火墙对比
大数据·网络·人工智能·网络安全·云原生·架构·kubernetes
wuyk5558 小时前
Spider-第20讲-综合实战可上线爬虫
大数据·爬虫·python·elasticsearch
培培说证8 小时前
大数据与会计专业性价比高的证书有哪些?
大数据
伞伞悦读10 小时前
【第90期】KNN分类回归流水线与K值对照的实验与排查
大数据·人工智能·python·深度学习·机器学习
老余说AI10 小时前
跨境电商、课程出海:六款多语言视频翻译工具选型指南
大数据·运维·人工智能
starzy199010 小时前
Structured Streaming 模块与编程模型:引擎内部到底由什么组成
大数据·hadoop·hdfs
Lansonli11 小时前
大数据Spark(九十二):Spark WebUI介绍
大数据·分布式·spark