elasticsearch中文英文拼音查询

默认分词使用的是IK分词

但对于英文，比如nike，分词结果是nike，但我们如果搜索nik就不会命中

对于英文，我使用模糊搜索来解决，添加一个字段，分词结果就是不区分大小写

下面贴上代码

{
  "index": {
    "analysis": {
      "filter": {
        "edge_ngram_filter": {
          "type": "edge_ngram",
          "min_gram": 1,
          "max_gram": 50
        },
        "pinyin_simple_filter": {
          "type": "pinyin",
          "first_letter": "prefix",
          "padding_char": " ",
          "limit_first_letter_length": 50,
          "lowercase": true
        }
      },
      "char_filter": {
        "tsconvert": {
          "type": "stconvert",
          "convert_type": "t2s"
        }
      },
      "analyzer": {
        "ikSearchAnalyzer": {
          "type": "custom",
          "tokenizer": "ik_max_word",
          "char_filter": [
            "tsconvert"
          ]
        },
        "pinyinSimpleIndexAnalyzer": {
          "tokenizer": "keyword",
          "filter": [
            "pinyin_simple_filter",
            "edge_ngram_filter",
            "lowercase"
          ]
        },
        "caseSensitive": {
          "type": "custom",
          "tokenizer": "keyword",
          "filter": "lowercase"
        }
      }
    }
  }
}

{
  "search": {
    "_all": {
      "enabled": true
    },
    "properties": {
      "id": {
        "type": "text"
      },
      "keyword": {
        "type": "text",
        "analyzer": "caseSensitive",
        "search_analyzer": "caseSensitive"
      },
      "keywordEn": {
        //        text和keyword的区别text：存储数据时候，会自动分词，并生成索引，keyword：存储数据时候，不会分词建立索引
        "type": "text",
        "analyzer": "ikSearchAnalyzer",
        "search_analyzer": "ikSearchAnalyzer",
        "fields": {
          "pinyin": {
            "type": "text",
            "analyzer": "pinyinSimpleIndexAnalyzer",
            "search_analyzer": "pinyinSimpleIndexAnalyzer"
          }
        }
      }
    }
  }
}

            // 构建查询
            NativeSearchQueryBuilder searchQuery = new NativeSearchQueryBuilder();
            // 索引查询
            searchQuery.withIndices("search-entity");
            BoolQueryBuilder boolQueryBuilder = QueryBuilders.boolQuery();
            //boost 设置权重
            //分词查询
            QueryBuilder queryBuilder = QueryBuilders.matchQuery("keywordEn", searchVo.getKeyword()).boost(2f);
            boolQueryBuilder.should(queryBuilder);
            //拼音查询
            QueryBuilder queryBuilder2 = QueryBuilders.matchQuery("keywordEn.pinyin", searchVo.getKeyword()).boost(2f);
            boolQueryBuilder.should(queryBuilder2);
            //模糊查询，不区分大小写
            QueryBuilder queryBuilder1 = QueryBuilders.wildcardQuery("keyword", "*"+searchVo.getKeyword().toLowerCase()+"*").boost(2f);
            boolQueryBuilder.should(queryBuilder1);
            boolQueryBuilder.must(QueryBuilders.matchQuery("module",searchVo.getModule()));
            //必须满足should其中一个条件
            boolQueryBuilder.minimumShouldMatch(1);
            searchQuery.withQuery(boolQueryBuilder);

使用should来关联搜索，注意，必须指定should命中的数量minimumShouldMatch，否则结果就是会出现所有must，优先should，没有should也会出现

关于高亮

拼音结果我不做高亮，只高亮了中英文分词和模糊，模糊检索高亮用正则替换来高亮

 //高亮字段（拼音不做高亮，拼音的高亮有问题，会将整个字符串高亮）
                        if (!StringUtils.isEmpty(searchHit.getHighlightFields().get("keywordEn"))) {
                            Text[] text = searchHit.getHighlightFields().get("keywordEn").getFragments();
                            model.setKeyword(text[0].toString());
                        }
                        //模糊搜索高亮,不区分大小写直接字符串替换
                        if(!model.getKeyword().contains("<high>")){
                            Pattern pattern = Pattern.compile("(?i)"+searchVo.getKeyword());
                            Matcher matcher = pattern.matcher(model.getKeyword());
                            if(matcher.find()){
                                String s = matcher.group();
                                model.setKeyword(model.getKeyword().replace(s,"<high>"+s+"</high>"));
                            }
                        }