本系列文章主要介绍基于 Spring Data Elasticsearch 实现商品搜索的后端代码，介绍代码逻辑和代码实现。
主要实现功能：根据搜索关键字查询、条件筛选、规格过滤、价格区间搜索、搜索查询分页、搜索查询排序、高亮查询。

主要应用技术:canal，Eureka，微服务架构（Microservices Architecture），Spring Data Elasticsearch

一、 搜索分页

1 分页分析

基于spring data ElasticSearch 对于查询结果进行分页操作。
页面需要实现分页搜索，所以我们后台每次查询的时候，需要实现分页。用户页面每次会传入当前页和每页查询多少条数据，当然如果不传入每页显示多少条数据，设置默认查询条即可。

2 分页实现

分页使用PageRequest.of( pageNo- 1, pageSize);实现，第1个参数表示第N页，从0开始，第2个参数表示每页显示多少条，实现代码如下：
代码如下：

@Overridepublic Map search(Map<String, String> searchMap) throws Exception {Map<String, Object> resultMap = new HashMap<>();
​//有条件才查询Esif (null != searchMap) {//组合条件对象BoolQueryBuilder boolQuery = QueryBuilders.boolQuery();//0:关键词if (!StringUtils.isEmpty(searchMap.get("keywords"))) {boolQuery.must(QueryBuilders.matchQuery("name", searchMap.get("keywords")).operator(Operator.AND));}//1:条件 品牌if (!StringUtils.isEmpty(searchMap.get("brand"))) {boolQuery.filter(QueryBuilders.termQuery("brandName", searchMap.get("brand")));}
​//2:条件 规格for (String key : searchMap.keySet()) {if (key.startsWith("spec_")) {String value = searchMap.get(key).replace("%2B", "+");boolQuery.filter(QueryBuilders.termQuery("specMap." + key.substring(5) + ".keyword",value));}}//3:条件 价格if (!StringUtils.isEmpty(searchMap.get("price"))) {String[] p = searchMap.get("price").split("-");boolQuery.filter(QueryBuilders.rangeQuery("price").gte(p[0]));if (p.length == 2) {boolQuery.filter(QueryBuilders.rangeQuery("price").lte(p[1]));}}
​//4. 原生搜索实现类NativeSearchQueryBuilder nativeSearchQueryBuilder = new NativeSearchQueryBuilder();nativeSearchQueryBuilder.withQuery(boolQuery);
​//6. 品牌聚合(分组)查询String skuBrand = "skuBrand";nativeSearchQueryBuilder.addAggregation(AggregationBuilders.terms(skuBrand).field("brandName"));
​//7. 规格聚合(分组)查询String skuSpec = "skuSpec";nativeSearchQueryBuilder.addAggregation(AggregationBuilders.terms(skuSpec).field("spec.keyword"));
​
​String pageNum = searchMap.get("pageNum");if (null == pageNum) {pageNum = "1";}//9: 分页nativeSearchQueryBuilder.withPageable(PageRequest.of(Integer.parseInt(pageNum) - 1, Page.pageSize));
​//10: 执行查询, 返回结果对象AggregatedPage<SkuInfo> aggregatedPage = esTemplate.queryForPage(nativeSearchQueryBuilder.build(), SkuInfo.class, new SearchResultMapper() {@Overridepublic <T> AggregatedPage<T> mapResults(SearchResponse searchResponse, Class<T> aClass, Pageable pageable) {
​List<T> list = new ArrayList<>();
​SearchHits hits = searchResponse.getHits();if (null != hits) {for (SearchHit hit : hits) {SkuInfo skuInfo = JSON.parseObject(hit.getSourceAsString(), SkuInfo.class);
​list.add((T) skuInfo);}}return new AggregatedPageImpl<T>(list, pageable, hits.getTotalHits(), searchResponse.getAggregations());}});
​//11. 总条数resultMap.put("total", aggregatedPage.getTotalElements());//12. 总页数resultMap.put("totalPages", aggregatedPage.getTotalPages());//13. 查询结果集合resultMap.put("rows", aggregatedPage.getContent());
​//14. 获取品牌聚合结果StringTerms brandTerms = (StringTerms) aggregatedPage.getAggregation(skuBrand);List<String> brandList = brandTerms.getBuckets().stream().map(bucket -> bucket.getKeyAsString()).collect(Collectors.toList());resultMap.put("brandList", brandList);
​//15. 获取规格聚合结果StringTerms specTerms = (StringTerms) aggregatedPage.getAggregation(skuSpec);List<String> specList = specTerms.getBuckets().stream().map(bucket -> bucket.getKeyAsString()).collect(Collectors.toList());resultMap.put("specList", specList(specList));
​//16. 返回当前页resultMap.put("pageNum", pageNum);
​return resultMap;}
​return null;}

3测试

使用postman测试分页：

二、 搜索排序

1 排序分析

排序这里总共有根据价格排序、根据评价排序、根据新品排序、根据销量排序，排序要想实现只需要告知排序的域以及排序方式即可实现。

价格排序：只需要根据价格高低排序即可，降序价格高->低，升序价格低->高

评价排序：评价分为好评、中评、差评，可以在数据库中设计3个列，用来记录好评、中评、差评的量，每次排序的时候，好评的比例来排序，当然还要有条数限制，评价条数需要超过N条。

新品排序：直接根据商品的发布时间或者更新时间排序。

销量排序：销量排序除了销售数量外，还应该要有时间段限制。

2 排序代码实现

这里我们不单独针对某个功能实现排序，我们只需要在后台接收2个参数，分别是排序域名字和排序方式，代码如下：

   @Overridepublic Map search(Map<String, String> searchMap) throws Exception {Map<String, Object> resultMap = new HashMap<>();
​//有条件才查询Esif (null != searchMap) {//组合条件对象BoolQueryBuilder boolQuery = QueryBuilders.boolQuery();//0:关键词if (!StringUtils.isEmpty(searchMap.get("keywords"))) {boolQuery.must(QueryBuilders.matchQuery("name", searchMap.get("keywords")).operator(Operator.AND));}//1:条件 品牌if (!StringUtils.isEmpty(searchMap.get("brand"))) {boolQuery.filter(QueryBuilders.termQuery("brandName", searchMap.get("brand")));}
​//2:条件 规格for (String key : searchMap.keySet()) {if (key.startsWith("spec_")) {String value = searchMap.get(key).replace("%2B", "+");boolQuery.filter(QueryBuilders.termQuery("specMap." + key.substring(5) + ".keyword",value));}}//3:条件 价格if (!StringUtils.isEmpty(searchMap.get("price"))) {String[] p = searchMap.get("price").split("-");boolQuery.filter(QueryBuilders.rangeQuery("price").gte(p[0]));if (p.length == 2) {boolQuery.filter(QueryBuilders.rangeQuery("price").lte(p[1]));}}
​//4. 原生搜索实现类NativeSearchQueryBuilder nativeSearchQueryBuilder = new NativeSearchQueryBuilder();nativeSearchQueryBuilder.withQuery(boolQuery);
​//6. 品牌聚合(分组)查询String skuBrand = "skuBrand";nativeSearchQueryBuilder.addAggregation(AggregationBuilders.terms(skuBrand).field("brandName"));
​//7. 规格聚合(分组)查询String skuSpec = "skuSpec";nativeSearchQueryBuilder.addAggregation(AggregationBuilders.terms(skuSpec).field("spec.keyword"));
​//8: 排序if (!StringUtils.isEmpty(searchMap.get("sortField"))) {if ("ASC".equals(searchMap.get("sortRule"))) {nativeSearchQueryBuilder.withSort(SortBuilders.fieldSort(searchMap.get("sortField")).order(SortOrder.ASC));} else {
​nativeSearchQueryBuilder.withSort(SortBuilders.fieldSort(searchMap.get("sortField")).order(SortOrder.DESC));}
​}
​String pageNum = searchMap.get("pageNum");if (null == pageNum) {pageNum = "1";}//9: 分页nativeSearchQueryBuilder.withPageable(PageRequest.of(Integer.parseInt(pageNum) - 1, Page.pageSize));
​//10: 执行查询, 返回结果对象AggregatedPage<SkuInfo> aggregatedPage = esTemplate.queryForPage(nativeSearchQueryBuilder.build(), SkuInfo.class, new SearchResultMapper() {@Overridepublic <T> AggregatedPage<T> mapResults(SearchResponse searchResponse, Class<T> aClass, Pageable pageable) {
​List<T> list = new ArrayList<>();
​SearchHits hits = searchResponse.getHits();if (null != hits) {for (SearchHit hit : hits) {SkuInfo skuInfo = JSON.parseObject(hit.getSourceAsString(), SkuInfo.class);list.add((T) skuInfo);}}return new AggregatedPageImpl<T>(list, pageable, hits.getTotalHits(), searchResponse.getAggregations());}});
​//11. 总条数resultMap.put("total", aggregatedPage.getTotalElements());//12. 总页数resultMap.put("totalPages", aggregatedPage.getTotalPages());//13. 查询结果集合resultMap.put("rows", aggregatedPage.getContent());
​//14. 获取品牌聚合结果StringTerms brandTerms = (StringTerms) aggregatedPage.getAggregation(skuBrand);List<String> brandList = brandTerms.getBuckets().stream().map(bucket -> bucket.getKeyAsString()).collect(Collectors.toList());resultMap.put("brandList", brandList);
​//15. 获取规格聚合结果StringTerms specTerms = (StringTerms) aggregatedPage.getAggregation(skuSpec);List<String> specList = specTerms.getBuckets().stream().map(bucket -> bucket.getKeyAsString()).collect(Collectors.toList());resultMap.put("specList", specList(specList));
​//16. 返回当前页resultMap.put("pageNum", pageNum);
​return resultMap;}
​return null;}

3测试

使用postman测试
根据价格降序：

{"keywords":"手机","pageNum":"1","sortRule":"DESC","sortField":"price"}

根据价格升序：

{"keywords":"手机","pageNum":"1","sortRule":"ASC","sortField":"price"}