Elasticsearch script脚本计算及统计

项目场景：

Elasticsearch使用scriptj脚本计算和统计，通过DSL和java API两种方式

---

解决方案：

一、DSL方式 :

1、计算结束时间减去开始时间时间差

{
  "script" : {
    "source": "(doc['completeTime'].value.toInstant().toEpochMilli() - doc['createTime'].value.toInstant().toEpochMilli())"
  }

2、统计大于20岁用户访问次数

"aggs": {
    "total": {
      "sum": {
        "script": {
          "source": "if(doc['age'].value < 30) { return 0 ; } else { return doc['number'].value}",
           "lang": "painless"
        }
      }
    }
  }

 3、统计在朝阳区的用户人数

"aggs": {
    "oh_total_count": {
      "cardinality": {
        "script": {
          "source": "def address = doc['address'].value; if(address.indexOf(params['keyword']) == -1) { return null ; } else { return doc['_id'].value}",
           "lang": "painless",
           "params": {
             "keyword": "朝阳区"
           }
        }
      }
    }
  }

 二、java API方式

AvgAggregationBuilder avgAggregationBuilder = AggregationBuilders
                            .avg("avg_grade")
                            .script(new Script("(doc['completeTime'].value.toInstant().toEpochMilli() - doc['createTime'].value.toInstant().toEpochMilli())"));