什麼是自動補全
隨用戶輸入，給與提示信息，如下圖：
ES實現原理
⽤戶每輸⼊⼀個 字符，就需要即時發送⼀個查詢請求到後段查找匹配項。
對性能要求⽐較苛刻。Elasticsearch 採⽤FST，FST 會被 ES 整個加載進內存， 速度很快。
實現方式：
Completion Suggester 實現
1.定義 Mapping，使⽤ “completion” type
2.索引數據
3.運⾏ “suggest” 查詢，得到搜索建議。
例子
1.定義 Mapping，使⽤ “completion” type
DELETE inputcompletion
# 設置mapping
PUT inputcompletion
{
"mappings": {
"_doc": {
"properties": {
"input_completion": {
"type": "completion"
}
}
}
}
}
2.索引數據
POST inputcompletion/_doc/_bulk
{ "index" : { } }
{ "input_completion": "elasticsearch 教程"}
{ "index" : { } }
{ "input_completion": "elasticsearch api 中文"}
{ "index" : { } }
{ "input_completion": "elasticsearch"}
3.運⾏ “suggest” 查詢，得到搜索建議。
POST inputcompletion/_doc/_search?pretty
{
"size": 0,
"suggest": {
"input-suggester": {
"prefix": "ela",
"completion": {
"field": "input_completion"
}
}
}
}
結果
{
"took": 3,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"skipped": 0,
"failed": 0
},
"hits": {
"total": 0,
"max_score": 0,
"hits": []
},
"suggest": {
"input-suggester": [
{
"text": "ela",
"offset": 0,
"length": 3,
"options": [
{
"text": "elasticsearch",
"_index": "inputcompletion",
"_type": "_doc",
"_id": "ezqix3cBMB6-gr9-ORLt",
"_score": 1,
"_source": {
"input_completion": "elasticsearch"
}
},
{
"text": "elasticsearch api 中文",
"_index": "inputcompletion",
"_type": "_doc",
"_id": "ejqix3cBMB6-gr9-ORLt",
"_score": 1,
"_source": {
"input_completion": "elasticsearch api 中文"
}
},
{
"text": "elasticsearch 教程",
"_index": "inputcompletion",
"_type": "_doc",
"_id": "eTqix3cBMB6-gr9-ORLt",
"_score": 1,
"_source": {
"input_completion": "elasticsearch 教程"
}
}
]
}
]
}
}
Context Suggester帶上下文的推薦
例如：再手機品類下搜索，小米提示小米手機，再食品下，提示真空小米
定義兩種類型的 Context
Category – 任意的字符串
Geo – 地理位置信息
實現 Context Suggester 的具體步驟
1.定製⼀個 Mapping 索引數據
2.並且爲每個⽂檔加⼊ Context 信息
3.結合 Context 進⾏ Suggestion 查詢
例子
1. 定製⼀個 Mapping 索引數據
DELETE inputcompletion
PUT inputcompletion
{
"mappings": {
"_doc": {
"properties": {
"input_completion": {
"type": "completion",
"contexts":[{
"type":"category",
"name":"goods_category"
}]
}
}
}
}
}
2.並且爲每個⽂檔加⼊ Context 信息
POST inputcompletion/_doc
{
"comment":"小米手機",
"input_completion":{
"input":["小米"],
"contexts":{
"goods_category":"手機"
}
}
}
POST inputcompletion/_doc
{
"comment":"真空小米",
"input_completion":{
"input":["小米"],
"contexts":{
"goods_category":"食品"
}
}
}
3. 結合 Context 進⾏ Suggestion 查詢
POST inputcompletion/_search
{
"suggest": {
"MY_SUGGESTION": {
"prefix": "小米",
"completion":{
"field":"input_completion",
"contexts":{
"goods_category":"食品"
}
}
}
}
}
結果
{
"took": 3,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"skipped": 0,
"failed": 0
},
"hits": {
"total": 0,
"max_score": 0,
"hits": []
},
"suggest": {
"MY_SUGGESTION": [
{
"text": "小米",
"offset": 0,
"length": 2,
"options": [
{
"text": "小米",
"_index": "inputcompletion",
"_type": "_doc",
"_id": "bjqgx3cBMB6-gr9-vBL2",
"_score": 1,
"_source": {
"comment": "真空小米",
"input_completion": {
"input": [
"小米"
],
"contexts": {
"goods_category": "食品"
}
}
},
"contexts": {
"goods_category": [
"食品"
]
}
}
]
}
]
}
}
參考
https://www.elastic.co/guide/en/elasticsearch/reference/7.1/search-suggesters-completion.html
https://www.elastic.co/guide/en/elasticsearch/reference/7.1/suggester-context.html