# Keep Types Token Filter（保留指定類型過濾器） 原文鏈接 : [https://www.elastic.co/guide/en/elasticsearch/reference/5.4/analysis-keep-types-tokenfilter.html](https://www.elastic.co/guide/en/elasticsearch/reference/5.4/analysis-keep-types-tokenfilter.html) 譯文鏈接 : [http://www.apache.wiki/display/Elasticsearch](http://www.apache.wiki/display/Elasticsearch) 貢獻者 : [李亞運](/display/~liyayun)，[ApacheCN](/display/~apachecn)，[Apache中文網](/display/~apachechina) ## 簡述 當type為`keep_types時`，過濾器將只保留包含在預定義集合中的token。 ## 參數 | types | 要保留的類型列表 | ## 示例 你可以這樣設置： ``` PUT /keep_types_example { "settings" : { "analysis" : { "analyzer" : { "my_analyzer" : { "tokenizer" : "standard", "filter" : ["standard", "lowercase", "extract_numbers"] } }, "filter" : { "extract_numbers" : { "type" : "keep_types", "types" : [ "<NUM>" ] } } } } } ``` 并用下述文本進行測試： ``` POST /keep_types_example/_analyze { "analyzer" : "my_analyzer", "text" : "this is just 1 a test" } ``` 返回結果： ``` { "tokens": [ { "token": "1", "start_offset": 13, "end_offset": 14, "type": "<NUM>", "position": 3 } ] } ``` 注意輸出中只有`&lt;NUM&gt;token（只有數字1,其他文本都被忽略了）`。