ElasticSearch入门（三）

ElasticSearch入门（三）

为了方便接下来的例子演示，我们新创建一个索引名为compony，type为person的文档，里面添加四个人，还是使用postman进行请求

{"name": "zhangsan",
 "age":24,
 "address":"beijing",
 "word":"i like computer music",
 "interests":["football","dance"]
}
{"name": "lisi",
 "age":25,
 "address":"shanghai",
 "word":"i like computer game",
 "interests":["sing","dance"]
}
{"name": "wangwu",
 "age":26,
 "address":"guangzhou",
 "word":"i like read book",
 "interests":["poker","game"]
}
{"name": "zhaoliu",
 "age":27,
 "address":"chengdu",
 "word":"i like game",
 "interests":["forestry","game"]
}

一、全文搜索

查找所有喜欢computer game的人：

localhost:9200/compony/person/_search

{
    "query" : {
        "match" : {
            "word" : "computer game"
        }
    }
}

得到的结果：

{
    "took": 225,
    "timed_out": false,
    "_shards": {
        "total": 1,
        "successful": 1,
        "skipped": 0,
        "failed": 0
    },
    "hits": {
        "total": {
            "value": 3,
            "relation": "eq"
        },
        "max_score": 1.3494902,
        "hits": [
            {
                "_index": "compony",
                "_type": "person",
                "_id": "2",
                "_score": 1.3494902,
                "_source": {
                    "name": "lisi",
                    "age": 25,
                    "address": "shanghai",
                    "word": "i like computer game",
                    "interests": [
                        "sing",
                        "dance"
                    ]
                }
            },
            {
                "_index": "compony",
                "_type": "person",
                "_id": "4",
                "_score": 0.7549127,
                "_source": {
                    "name": "zhaoliu",
                    "age": 27,
                    "address": "chengdu",
                    "word": "i like game",
                    "interests": [
                        "forestry",
                        "game"
                    ]
                }
            },
            {
                "_index": "compony",
                "_type": "person",
                "_id": "1",
                "_score": 0.6747451,
                "_source": {
                    "name": "zhangsan",
                    "age": 24,
                    "address": "beijing",
                    "word": "i like computer music",
                    "interests": [
                        "football",
                        "dance"
                    ]
                }
            }
        ]
    }
}

其中的_score代表相关性得分，Elasticsearch 默认按照相关性得分排序，即每个文档跟查询的匹配程度。我们这里看到即使是zhaoliu的爱好里面没有computer，但是结果也返回了，因为它里面有game，所以也返回了，只是相关性变低了。这也看出来区别于关系型数据库的概念，数据库中的一条记录要么匹配要么不匹配。

二、短语搜索

找出一个属性中的独立单词是没有问题的，但有时候想要精确匹配一系列单词或者短语 。 比如， 我们想执行这样一个查询，仅匹配同时包含 “computer” 和 “game” ，并且 二者以短语 “computer game” 的形式紧挨着的人员记录。

localhost:9200/compony/person/_search
{
    "query" : {
        "match_phrase" : {
            "word" : "computer game"
        }
    }
}

注意这里与全文搜索变化的地方只有match变为了match_phrase，返回结果：

{
    "took": 4,
    "timed_out": false,
    "_shards": {
        "total": 1,
        "successful": 1,
        "skipped": 0,
        "failed": 0
    },
    "hits": {
        "total": {
            "value": 1,
            "relation": "eq"
        },
        "max_score": 1.3494902,
        "hits": [
            {
                "_index": "compony",
                "_type": "person",
                "_id": "2",
                "_score": 1.3494902,
                "_source": {
                    "name": "lisi",
                    "age": 25,
                    "address": "shanghai",
                    "word": "i like computer game",
                    "interests": [
                        "sing",
                        "dance"
                    ]
                }
            }
        ]
    }
}

这样的查询，只匹配到了lisi这个人

三、高亮搜索

许多应用都倾向于在每个搜索结果中 高亮 部分文本片段，以便让用户知道为何该文档符合查询条件。在 Elasticsearch 中检索出高亮片段也很容易。

再次执行前面的查询，并增加一个新的 highlight 参数：

{
    "query" : {
        "match_phrase" : {
            "word" : "computer game"
        }
    },
    "highlight":{
        "fields":{
            "word":{}
        }
    }
}

查询结果:

{
    "took": 31,
    "timed_out": false,
    "_shards": {
        "total": 1,
        "successful": 1,
        "skipped": 0,
        "failed": 0
    },
    "hits": {
        "total": {
            "value": 1,
            "relation": "eq"
        },
        "max_score": 1.3494902,
        "hits": [
            {
                "_index": "compony",
                "_type": "person",
                "_id": "2",
                "_score": 1.3494902,
                "_source": {
                    "name": "lisi",
                    "age": 25,
                    "address": "shanghai",
                    "word": "i like computer game",
                    "interests": [
                        "sing",
                        "dance"
                    ]
                },
                "highlight": {
                    "word": [
                        "i like <em>computer</em> <em>game</em>"
                    ]
                }
            }
        ]
    }
}

这样的效果可能与百度搜索那种类似，这只是我的猜测：

官网给出的基本入门也就这么多了，其它的后续看情况要不要做几个demo，现目前计划后续整合springboot在代码层面来展示