spring-data-elasticsearch使用笔记

一、定义document

@Data
@EqualsAndHashCode(callSuper = false)
@Document(indexName = "syni_sjb_encyclopedia_entry")
public class SjbEncyclopediaEntryDocument implements Serializable {

    private static final long serialVersionUID = 1L;

    @Id
    private String id;

    /**
     * 词条名称
     */
    @Field(type = FieldType.Text, analyzer = "ik_max_word", searchAnalyzer="ik_max_word")
    private String name;

    /**
     * 词条解释，富文本类型
     */
    @Field(type = FieldType.Text, analyzer = "ik_max_word", searchAnalyzer="ik_max_word")
    private String entryExplain;

    /**
     * 初始浏览人数
     */
    @Field(type = FieldType.Integer)
    private Integer initPeopleCount;

    /**
     * 补增浏览人数
     */
    @Field(type = FieldType.Integer)
    private Integer complementPeopleCount;

    /**
     * 真实浏览人数，非实时
     */
    @Field(type = FieldType.Integer)
    private Integer realPeopleCount;

    /**
     * 初始人数+补增人数+真实浏览人数
     */
    @Field(type = FieldType.Integer)
    private Integer sumPeopleCount;

    /**
     * 词条生成的h5文件url
     */
    @Field(type = FieldType.Keyword)
    private String link;

    /**
     * 来源
     */
    @Field(type = FieldType.Keyword)
    private String createWay;

    /**
     * 创建人
     */
    @Field(type = FieldType.Keyword)
    private String creator;

    /**
     * 创建时间
     */
    @Field(type = FieldType.Date, format = DateFormat.date_hour_minute_second)
    @JsonFormat(shape = JsonFormat.Shape.STRING, pattern ="yyyy-MM-dd HH:mm:ss",timezone="GMT+8")
    private LocalDateTime createTime;

    /**
     * 更新时间
     */
    @Field(type = FieldType.Date, format = DateFormat.date_hour_minute_second)
    @JsonFormat(shape = JsonFormat.Shape.STRING, pattern ="yyyy-MM-dd HH:mm:ss",timezone="GMT+8")
    private LocalDateTime updateTime;


}

这里有个坑：时间字段使用
@Field(type = FieldType.Date, format = DateFormat.custom, pattern = “yyyy-MM-dd HH:mm:ss”)
调用find时时间会反序列化失败，不知道是不是版本问题
我的版本是2.3.4 RELEASE

二、手动put es mapping（类似于建表，定义表结构）

put http://xxx.xxx.xx:9200/syni_sjb_encyclopedia_entry
{
    "mappings": {
        "properties": {
            "_class": {
                "type": "text",
                "fields": {
                    "keyword": {
                        "type": "keyword",
                        "ignore_above": 256
                    }
                }
            },
            "complementPeopleCount": {
                "type": "integer"
            },
            "createTime": {
                "type": "date",
                "format": "yyyy-MM-dd'T'HH:mm:ss||yyyy-MM-dd||epoch_millis"
            },
            "createWay": {
                "type": "keyword"
            },
            "creator": {
                "type": "keyword"
            },
            "entryExplain": {
                "type": "text",
                "analyzer": "ik_max_word"
            },
            "initPeopleCount": {
                "type": "integer"
            },
            "link": {
                "type": "keyword"
            },
            "name": {
                "type": "text",
                "analyzer": "ik_max_word"
            },
            "realPeopleCount": {
                "type": "integer"
            },
            "sumPeopleCount": {
                "type": "integer"
            },
            "updateTime": {
                "type": "date",
                "format": "yyyy-MM-dd'T'HH:mm:ss||yyyy-MM-dd||epoch_millis"
            }
        }
    }
}

ES自带有自动建表的功能，那为什么还要手动put?
因为字段的类型可能不会合你意，比如时间格式只有yyyy-MM-dd’T’HH:mm:ss，但实际上还可能需要yyyy-MM-dd

三、创建respository，并定义方法（可省略）

public interface SjbEncyclopediaEntryRepository extends ElasticsearchRepository<SjbEncyclopediaEntryDocument,String> {

    List<SjbEncyclopediaEntryDocument> findByNameAndIdNot(String name, String id);
}

方法定义类似JPA（虽然我也没用过JPA，大概意思就是按照一定规范定义方法，框架就自动帮你实现，比如findByName）
这个不太好用，因为用这个就不知道怎么分页，怎么排序了。

四、ElasticsearchOperations使用及与Mybatis-plus方法的转换

我的需求是把mysql的数据转移到ES，功能不变。所以需要把Mybatis-plus的方法转化成ES

- 增删改

增：xxxService.save(entity) => elasticsearchOperations.save(entity)
改：xxxService.update(entity) => elasticsearchOperations.save(entity)
删：xxxService.removeById(id) => elasticsearchOperations.deleteById(id)

增和改同样是使用save，只要id一样就会自动覆盖。

- 查

查比较复杂，上一段代码大家看看前后对比

    public Page<SjbEncyclopediaEntryVO> queryByParam(SjbEncyclopediaEntryQueryDTO dto) {
    	// 分页
        Page<SjbEncyclopediaEntry> page = new Page<>(dto.getPageNo(), dto.getPageSize());
        QueryWrapper<SjbEncyclopediaEntry> queryWrapper = new QueryWrapper<>();
        if(StringUtils.isNotBlank(dto.getName())) {
            queryWrapper.lambda().eq(SjbEncyclopediaEntry::getName, dto.getName());
        }
        if(dto.getStartDate() != null && dto.getEndDate() != null) {
            queryWrapper.lambda().between(SjbEncyclopediaEntry::getCreateTime, dto.getStartDate(), dto.getEndDate().plusDays(1));
        }
        // 排序
        queryWrapper.lambda().orderByDesc(SjbEncyclopediaEntry::getRealPeopleCount);
        page = page(page, queryWrapper);
        Page<SjbEncyclopediaEntryVO> result = new Page<>(dto.getPageNo(), dto.getPageSize());
        BeanUtils.copyProperties(page, result);
        result.setRecords(page.getRecords().stream().map(sjbEncyclopediaEntry -> {
            SjbEncyclopediaEntryVO vo = new SjbEncyclopediaEntryVO();
            BeanUtils.copyProperties(sjbEncyclopediaEntry, vo);
            vo.setRealPeopleCount(getRedisRealPeopleCount(sjbEncyclopediaEntry.getId()));
            return vo;
        }).collect(Collectors.toList()));
        return result;
    }

    public Page<SjbEncyclopediaEntryVO> queryByParam(SjbEncyclopediaEntryQueryDTO dto) {
        Pageable pageable = PageRequest.of(dto.getPageNo()-1, dto.getPageSize());
        BoolQueryBuilder boolQueryBuilder = QueryBuilders.boolQuery();
        if(StringUtils.isNotBlank(dto.getName())) {
            boolQueryBuilder.must(QueryBuilders.termsQuery("name", dto.getName()));
        }
        if(dto.getStartDate() != null && dto.getEndDate() != null) {
            boolQueryBuilder.filter(QueryBuilders.rangeQuery("createTime").from(dto.getStartDate()).to(dto.getEndDate().plusDays(1)));
        }
        SortBuilder sortBuilder = SortBuilders.fieldSort("createTime").order(SortOrder.DESC);
        NativeSearchQuery build = new NativeSearchQueryBuilder()
                .withQuery(boolQueryBuilder)
                .withPageable(pageable)
                .withSort(sortBuilder)
                .build();
        SearchHits<SjbEncyclopediaEntryDocument> search = elasticsearchOperations.search(build, SjbEncyclopediaEntryDocument.class);
        Page<SjbEncyclopediaEntryVO> result = new Page<>(dto.getPageNo(), dto.getPageSize());
        result.setTotal(search.getTotalHits());

        List<SearchHit<SjbEncyclopediaEntryDocument>> collect = search.get().collect(Collectors.toList());
        result.setRecords(collect.stream().map(sjbEncyclopediaEntry -> {
            SjbEncyclopediaEntryVO vo = new SjbEncyclopediaEntryVO();
            BeanUtils.copyProperties(sjbEncyclopediaEntry.getContent(), vo);
            vo.setRealPeopleCount(getRedisRealPeopleCount(sjbEncyclopediaEntry.getId()));
            return vo;
        }).collect(Collectors.toList()));
        return result;
    }

可以大致理解mybatis-plus与es的实体与api对应关系为：

QueryWrapper<?> => NativeSearchQuery
Page<?> => Pageable
wrapper.orderByDesc => SortBuilder
Page<?> result => List<SearchHit<?>>
page(page, queryWrapper); => elasticsearchOperations.search(build, SjbEncyclopediaEntryDocument.class);

五、一些坑

boolQueryBuilder.filter(QueryBuilders.rangeQuery(“publishTime”).lt(LocalDateTime.now()));
这句话又报错了，估计是lt里的时间值转字符串时多了毫秒值，试了mapping里加毫秒格式也不行。
解决方案：直接用字符串
String format = LocalDateTime.now().format(DateTimeFormatter.ofPattern(“yyyy-MM-dd’T’HH:mm:ss”));
boolQueryBuilder.filter(QueryBuilders.rangeQuery(“publishTime”).lt(format));