1. 业务开发
基本步骤：
- 配置字典
ES索引增加索引配置 - 使用注解，在需要同步es的方法上面使用注解
- 创建模型Bean，创建业务es文档模型
- 创建Service，实现对es的curd
- 创建拦截Aop，创建业务es同步aop
- 创建搜索action，创建控制层对外部提供查询接口
[!tip]
下面代码片段截取自政务版本的内容搜索
1.1. 字典配置
通过配置字典ES索引来增加es索引，特别注意字典的数据值的json格式
{"name":"GovESContentBean","url":"/gov/es/sync.do"}
name：模型bean的名称，具体开发参考模型Bean,必须与ESBaseBean的子类上的@Component名称一致;
url：es同步请求的方法,具体开发参考 同步方法
1.2. 注解
在对应的业务代码使用
1.2.1. @ESSave、@ESDelete
在对应的方法上进行注解，提供给aop进行拦截，
[!tip] 主要提供AOP拦截，在AOP中进行ES库的同步,所有涉及数据更新的地方都需要使用此注解
1.2.2. 实例
...
//保存方法
@ESSave
public ResultData save(@ModelAttribute ContentEntity content) {
contentBiz.saveEntity(_product);
return ResultData.build().success(_product);
}
//更新方法
@ESSave
public ResultData update(@ModelAttribute ContentEntity content) {
contentBiz.updateEntity(_product);
return ResultData.build().success(_product);
}
//删除方法
@ESDelete
public ResultData delete(@RequestBody List<ContentEntity> contents,HttpServletResponse response, HttpServletRequest request) {
int[] ids = new int[contents.size()];
for(int i = 0;i<contents.size();i++){
ids[i] =Integer.parseInt(contents.get(i).getId()) ;
}
contentBiz.delete(ids);
return ResultData.build().success();
}
...
1.3. 模型Bean
创建业务中需要的ES文档模型 public class *Bean extends ESBaseBean，业务Bean继承ESBaseBean
注解参考
es文档定义，indexName可以定义不同的es库,注意：高版本es容器已经废弃了type的设置
@Document(indexName = "cms",type = "content")
es关键id注解
@Id
es关键字定义，例如：我是中国人，只能通过我是中国人搜索到，
@Field(type = FieldType.Keyword)
es关键字定义，分词搜索，例如：我是中国人，可以通过我是、中国人、中国、都能搜索到，
@Field(type = FieldType.Keyword, analyzer = "ik_max_word"）
es长文本关键字定义，推荐加上analyzer = "ik_max_word"
@Field(type = FieldType.Text, analyzer = "ik_max_word")
es数值定义
@Field(type = FieldType.Integer )
es日期定义
@Field(type = FieldType.Date)
1.3.1. 实例
以文章为例，创建es文档模型bean，只需要定义需要存入es库的字段
@Document(indexName = "cms-gov")
@Component("GovESContentBean")
public class ESContentBean extends ESBaseBean {
/**
* 存储文章作者
*/
@Field(type = FieldType.Text, analyzer = "ik_max_word", searchAnalyzer="ik_smart")
private String author;
/**
* 存储自定义扩张模型
*/
@Field(type = FieldType.Object)
private Map<String, Object> MDiyModel;
@Field(type = FieldType.Keyword)
private String typeId;
@Field(type = FieldType.Text)
private String litPic;
@Field(type = FieldType.Keyword)
private String flag;
/**
* 存储文章发布到
*/
@Field(type = FieldType.Text, analyzer = "ik_max_word", searchAnalyzer="ik_smart")
private String styles;
/**
* 存储类型
*/
private String type;
/**
* 存储栏目父ID集合
*/
@Field(type = FieldType.Text, analyzer = "ik_max_word", searchAnalyzer="ik_smart")
private String parentIds;
/**
* 储存站点ID
*/
@Field(type = FieldType.Keyword)
private String appId;
/**
* 存储栏目标题
*/
@Field(type = FieldType.Text)
private String typeTitle;
/**
* 自定义排序
*/
@Field(type = FieldType.Integer)
private Integer sort;
@Field(type = FieldType.Text)
private String descrip;
/**
* 文章审批进度
*/
@Field(type = FieldType.Keyword)
private String progressStatus;
@Field(type = FieldType.Date,format = DateFormat.custom, pattern="yyyy-MM-dd HH:mm:ss")
@JSONField(format = "yyyy-MM-dd HH:mm:ss")
@DateTimeFormat(pattern = "yyyy-MM-dd HH:mm:ss")
@JsonFormat(timezone = "GMT+8",pattern = "yyyy-MM-dd HH:mm:ss")
private Date date;
/**
* 链接地址
*/
@Field(type = FieldType.Text,index=false)
private String url;
// .. 省略 set get 方法
}
1.4. 同步方法
通过业务代码查询出所有的数据再通过 `Service` 提供的方法同步到es库
...
@Autowired
private IContentBiz contentBiz;
@Autowired
private IModelBiz modelBiz;
@Autowired
private ICategoryBiz categoryBiz;
@Autowired
private IESContentService esContentService;
/**
* 同步es文章
* @return
*/
@ApiOperation(value = "同步es文章接口")
@PostMapping("/sync")
@RequiresPermissions("es:sync")
@ResponseBody
public ResultData sync() {
// 获取栏目列表
LambdaQueryWrapper<CategoryEntity> categoryWrapper = new LambdaQueryWrapper<>();
categoryWrapper.ne(CategoryEntity::getCategoryType, CategoryTypeEnum.LINK.toString());
List<CategoryEntity> categoryList = categoryBiz.list(categoryWrapper);
for (CategoryEntity category : categoryList) {
// 获取文章列表
LambdaQueryWrapper<ContentEntity> contentWrapper = new LambdaQueryWrapper<>();
contentWrapper.eq(ContentEntity::getCategoryId, category.getId());
List<ContentEntity> contentList = contentBiz.list(contentWrapper);
boolean hasModel = false;
ModelEntity model = null;
if (StringUtils.isNotBlank(category.getMdiyModelId())) {
hasModel = true;
// 获取模型实体
model = modelBiz.getById(category.getMdiyModelId());
}
List<ESContentBean> beanList = new ArrayList<>();
for (ContentEntity content : contentList) {
ESContentBean esContentBean = new ESContentBean();
ESContentBeanUtil.fixESContentBean(esContentBean, content, category);
if (hasModel) {
//配置link_id Map
Map<String, String> modelMap = new HashMap<>(1);
modelMap.put("link_id", content.getId());
List<Map<String, Object>> modelFieldList = contentBiz.queryBySQL(
model.getModelTableName(),
null ,
modelMap, null, null, null, null, null);
// 防止NP
if (CollUtil.isNotEmpty(modelFieldList)) {
// 模型字段MAP
Map<String, Object> objectMap = modelFieldList.get(0);
for (String key : objectMap.keySet()) {
// 类型为BigInteger时需要转换为long防止es类型构造器转换错误
if (objectMap.get(key) instanceof BigInteger) {
objectMap.put(key, ((BigInteger) objectMap.get(key)).longValue());
}
}
esContentBean.setMDiyModel(objectMap);
}
}
beanList.add(esContentBean);
}
if (CollUtil.isNotEmpty(beanList)) {
try {
esContentService.saveAll(beanList);
}catch (DataAccessResourceFailureException e) {
return ResultData.build().error("未找到当前ES信息,请检查当前ES链接是否正常");
}catch (NoSuchIndexException e) {
return ResultData.build().error("未找到当前ES索引信息,请检查是否创建索引");
}
}
}
return ResultData.build().success().msg("全部同步完成!");
}
...
1.5. Service继承ElasticsearchRepository
继承后能进行简单的增删改查操作,具体可以参考CrudRepository
@Service("govIESContentService")
public interface IESContentService extends ElasticsearchRepository<ESContentBean, String> {
}
[!tip] 基于spring boot data 快速实现ES的CURD
1.6. AOP拦截
创建AOP public class *Aop extends ESBaseAop {},需要进行es全文检索的数据，通过aop的方式同步到es库，业务Aop继承ESBaseAop
1.6.1. 实例
通过拦截的方式将模型Bean的内容同步到es库
@Component("esContentAop")
@Aspect
public class ESContentAop extends ESBaseAop {
@Autowired
private IESContentService esContentService;
@Autowired
private ICategoryBiz categoryBiz;
//保存更新同步
@Override
public void save(JoinPoint joinPoint) {
try {
BaseEntity obj = this.getType(joinPoint, BaseEntity.class, true);
ESContentBean bean = new ESContentBean();
if(obj instanceof ContentEntity){
ContentEntity content = (ContentEntity)obj;
CategoryEntity category = categoryBiz.getById(content.getCategoryId());
bean.setTitle(content.getContentTitle());
bean.setContent(content.getContentDetails());
bean.setAuthor(content.getContentAuthor());
bean.setDate(content.getContentDatetime());
bean.setTypeId(content.getCategoryId());
bean.setDescrip(content.getContentDescription());
bean.setSort(content.getContentSort());
bean.setLitPic(content.getContentImg());
bean.setFlag(content.getContentType());
bean.setUrl(category.getCategoryPath();
// 有appId则存放
AppEntity app = BasicUtil.getWebsiteApp();
if (app != null) {
// 查询时会自动拼接appID,也就是当前session,所以这里可以直接设置
bean.setAppId(app.getId());
bean.setStyles(content.getContentStyle());
}
}
LOG.info("保存es库");
esContentService.save(bean);
} catch (Exception e) {
e.printStackTrace();
}
}
//删除同步
@Override
public void delete(JoinPoint joinPoint, ResultData result) {
try {
//注意！！！默认接收切入方法第一个参数
List<BaseEntity> arrayList = this.getType(joinPoint, ArrayList.class, true);
BaseEntity obj ;
obj = this.getType(joinPoint, BaseEntity.class, true);
if (arrayList != null) {
obj = arrayList.get(0);
}
// 单个删除
if(ObjectUtil.isNotEmpty(obj)){
esContentService.deleteById(obj.getId());
}
} catch (Exception e) {
e.printStackTrace();
}
}
}
[!tip] 所有涉及到文字内容变化的控制层都需要别拦截到，否则会出现es库数据不同步的问题产生
1.7. 搜索Action
创建 *Action ，根据业务编写对应的控制层，调用ESUtils的方法进行搜索
/**
* 高亮搜索，保留搜索字段keyword、order、orderBy、pageNo、pageSize，其余字段根据实际业务调整
*
* @param keyword 关键字
* @return 内容
*/
@ApiOperation(value = "高亮搜索接口")
@ApiImplicitParams({
@ApiImplicitParam(name = "keyword", value = "搜索关键字", required = true, paramType = "query"),
@ApiImplicitParam(name = "order", value = "排序方式", required = false, paramType = "query"),
@ApiImplicitParam(name = "orderBy", value = "排序的字段", required = false, paramType = "query"),
@ApiImplicitParam(name = "pageNo", value = "页数", required = false, paramType = "query", dataType = "Integer"),
@ApiImplicitParam(name = "pageSize", value = "页数大小", required = false, paramType = "query", dataType = "Integer"),
@ApiImplicitParam(name = "styles", value = "皮肤，例:outsite、insite", required = false, paramType = "query"),
@ApiImplicitParam(name = "typeId", value = "所属栏目(用于搜索指定栏目)", required = false, paramType = "query"),
@ApiImplicitParam(name = "parentIds", value = "父栏目ID(用于搜索所有子栏目)", required = false, paramType = "query"),
@ApiImplicitParam(name = "fields", value = "搜索关键字字段,值输入ES索引字段值,多个用逗号隔开 例:content,title", required = false, paramType = "query"),
@ApiImplicitParam(name = "searchMethod", value = "搜索方式,不传默认分词搜索(传值参考SearchMethodEnum)", required = false, paramType = "query"),
@ApiImplicitParam(name = "rangeField", value = "范围字段,需要传开始范围和结束范围才能生效,可以对时间进行范围筛选,值填写对应es索引的索引字段", required = false, paramType = "query"),
@ApiImplicitParam(name = "start", value = "开始范围 时间格式:yyyy-MM-dd HH:mm:ss 例:2021-08-06 12:25:36", required = false, paramType = "query"),
@ApiImplicitParam(name = "end", value = "结束范围 时间格式:yyyy-MM-dd HH:mm:ss 例:2021-08-06 12:25:36", required = false, paramType = "query"),
@ApiImplicitParam(name = "noflag", value = "不等于的文章类型,多个逗号隔开", required = false, paramType = "query"),
})
@PostMapping("/highlightSearch")
@ResponseBody
public ResultData highlightSearch(String keyword) {
ESSearchBean esSearchBean = new ESSearchBean();
String noflag = BasicUtil.getString("noflag");
Map<String, Object> notEquals = new HashMap<>();
if (StringUtils.isNotBlank(noflag)) {
List<String> collect = Arrays.stream(noflag.split(",")).collect(Collectors.toList());
// es flag字段
notEquals.put("flag", collect);
}
// 过滤条件
String filterStr = BasicUtil.getString("filter");
Map<Object,Object> filter = new HashMap<>();
if (StringUtils.isNotBlank(filterStr)) {
try {
filter = JSONUtil.toBean(filterStr, Map.class);
} catch (Exception e) {
LOG.debug("过滤条件转json失败");
e.printStackTrace();
}
}
//如果有站群插件，需要根据站群插件过滤
if (BasicUtil.getWebsiteApp() != null) {
// 站点过滤
LambdaQueryWrapper<AppEntity> wrapper = new QueryWrapper<AppEntity>().lambda();
wrapper.like(AppEntity::getAppUrl, BasicUtil.getDomain());
AppEntity appEntity = appBiz.getOne(wrapper, false);
filter.put("net", appEntity.getId());
}
// 设置基本过滤字段
String styles = BasicUtil.getString("styles");
if (StringUtils.isNotBlank(styles)) {
filter.put("styles", styles);
}
//审核过滤
filter.put("progressStatus", "终审通过");
esSearchBean.setFilterWhere(filter);
esSearchBean.setKeyword(keyword);
// 设置排序
String order = BasicUtil.getString("order", "desc");
String orderBy = BasicUtil.getString("orderBy", "sort");
esSearchBean.setOrder(order);
esSearchBean.setOrderBy(orderBy);
//设置分页
Integer pageNo = BasicUtil.getInt("pageNo", 1);
Integer pageSize = BasicUtil.getInt("pageSize", 10);
esSearchBean.setPageNo(pageNo);
esSearchBean.setPageSize(pageSize);
//设置高亮搜索配置
Map<String, Object> highlightConfig = new HashMap<>();
//设置高亮样式
highlightConfig.put("preTags", "<span style='color:red'>");
highlightConfig.put("postTags", "</span>");
//设置搜索字段
String author = BasicUtil.getString("author");
String content = BasicUtil.getString("content");
String title = BasicUtil.getString("title");
LinkedHashMap<String, Float> searchFields = new LinkedHashMap<>();
//设置搜索字段，数值f标识用于权重排序,一般多字段搜索使用,数值越大该字段的权重越大,当searchMethod等于geo_function_search生效
if (StringUtils.isNotBlank(author)) {
searchFields.put("author", 5f);
}
if (StringUtils.isNotBlank(title)) {
searchFields.put("title", 10f);
}
if (StringUtils.isNotBlank(content)) {
searchFields.put("content", 5f);
}
// 默认设置搜索标题
if (CollUtil.isEmpty(searchFields)) {
searchFields.put("title", 10f);
}
// 设置搜索方式,默认使用分词搜索
String searchMethod = BasicUtil.getString("searchMethod", "");
return esService.search(ESContentBean.class, searchFields, highlightConfig,
notEquals, esSearchBean, searchMethod);
}
[!tip] 注意代码中的
filter.put与searchFields.put根据实际业务字段调整,searchMethod=should_search可以匹配完整标题注意：fields参数里可以填写es库中已有的字段,如果传参为 title,content 表示搜索标题、文章中带有关键字的数据；styles参数为outsite或者insite
返回格式参考
{
"result": true,
"code": 200,
"data": {
"total": 1,
"rows": [ //记录列表
{
"contentKeyword": "<span style='color:red'>国</span><span style='color:red'>内</span>ms",
"sort": 0,
"id": "1346299650146922497",
"title": "[【网站】<span style='color:red'>国</span><span style='color:red'>内</span> 前6名 Java开源CMS建站系统]",
"content": "[下面我们开始分享一下开源中<span style='color:red'>国</span>中最火的Java开源CMS建站系统]",
"typeId": "1392359380113481730",
"parentIds": "1382233650262241282",
"date": "2017-08-29 20:04:31",
"author" : "",
"net" : "1",
"typeId" : "1458331507951800321",
"litPic" : """[{"url":"blob:http://192.168.0.118:8080/5e548893-1b5f-46cd-8193-fc7c48b2cc7f","name":"1637824298771.jpg","path":"/upload/1/cms/content/1637906304084.jpg","uid":1637906304036,"status":"success"}]""",
"typeTitle" : "早间新闻",
"flag" : "c",
"type" : "content",
"descrip": "-",
"url" : """/xinwen/zaojianxinwen\c1458332611057946626.html"""
}
]
}
}
1.8. 联想搜索\热词搜索
/es/suggestSearch.do 根据输入的关键字显示相关的关键词
/es/hotSearch.do 搜索指定时间访问的热词搜索及搜索量
[!tip] 根据搜索的结果，通过前端技术处理方式进行跳转
1.9. 搜索页面
可以通过swagger接口快速调试，前端页面通过异步请求处理。
代码参考：
...
<!-- 搜索表单，取自公共头文件，作用为跳转到搜索页面 -->
<div class="hd">
<div class="hd-inner">
<form id="searchDataForm" action="/mcms/search.do" method="post">
<div class="input-wrap">
<input type="text" style="display: none;" name="style" value="{ms:global.template/}">
<input id="content_title" name="content_title" value="${search.content_title}" content="${search.content_title}" type="text" class="search-input" placeholder="请输入关键字">
<a onclick="document.getElementById('searchDataForm').submit();" class="search-btn">
<img src="{ms:global.contextpath/}/{ms:global.style/}/image/search.png" alt="">
</a>
</div>
</form>
</div>
</div>
<div id="app">
<div class="search-wrap">
<#if search.content_title?? && search.content_title>
<ul class="result-list">
<!--列表-->
<template v-for="item in contentList">
<!--列表项-start-->
<li class="img">
<img :src="item.litPic" alt="" v-if="item.litPic != null">
<a :href="<#if isDo?? && isDo>'{ms:global.url/}/mcms/view.do?style=out&id=' + item.id<#else>'{ms:global.url/}' + item.url</#if>" v-html="item.title" target="_blank" class="ellipsis-2">{{ item.title}}</a>
<p class="desc ellipse">{{ item.descrip }}</p>
<span>{{ item.date }}</span>
</li>
</template>
</ul>
<!--分页-->
<div class="news-page"></div>
<#else>
<span style="text-align: center;display:block;font-size: 20px;color: red">搜索关键字不可以为空</span>
</#if>
</div>
</div>
...
...
<script>
function search(val){
ar contextPath = '{ms:global.contextpath/}';
// 过滤条件及范围查询，非必须
var rangeFields = JSON.stringify(app.rangeFields);
var filter = JSON.stringify(app.filter);
<!--es通用搜索,支持高亮搜索,不使用es的情况下将下面代码全部注释即可-->
//ms.http.post(ms.base+"/cms/es/search/highlightSearch.do",{
ms.http.post(contextPath+"/es/search.do",{
docName:'文章',
fields: 'title,content',// 匹配哪些字段，逗号分隔
keyword: '{ms:search.content_title/}',// 匹配关键字
orderBy: 'date',// 排序字段
order: 'desc',// 倒序
filter: filter,// 过滤条件
rangeFields: rangeFields,// 范围查询
pageNo: val,// 页码
pageSize: 5,// 页显示数
}).then(res => {
var content;
// 由于在es环境下无法用标签获取文章相关值，在这里预先对文章图片做处理
for (var i = 0;i<res.data.rows.length;i++){
content = res.data.rows[i];
if(content.litPic){
content.litPic = JSON.parse(content.litPic);
if(content.litPic.length > 0 ){
res.data.rows[i].litPic = content.litPic[0].path;
}
} else {
res.data.rows[i].litPic = null;
}
}
app.contentList = res.data.rows
//getData(this.contentCount);
}).catch(err => {
console.log(err)
})
}
var app = new Vue({
el: '#app',
component(){
},
data: {
// 过滤条件,如下条件为查询栏目id为123456的文章
filter: {
typeId: '123456',
},
// 范围条件，如下为查找指定时间范围内，且排序值在1-9之间的文章
rangeFields: [
{
type: 'date',
field: 'date',
start: '2022-03-26 12:12:00',
end: '2022-03-26 14:36:00',
},
{
type: 'text',
field: 'sort',
start: '1',
end: '9',
}
],
searchForm: {
keyword: ""
},
hotSearchKeywords: [], //当前页数
// 搜索数据
contentList: []
},
watch: {
},
methods: {
handleCurrentChange:function(val) {
<!--es通用搜索-->
ms.http.post(ms.base+"/es/search.do",{
docName:'文章',
fields: 'title',
keyword: '{ms:search.content_title/}',
orderBy: 'date',
order: 'asc',
pageNo: val,
pageSize: 5,
}).then(res => {
this.contentList = res.data.rows
}).catch(err => {
console.log(err)
})
},
switchShow:function(arr){
var that = this
arr.forEach(function(x){
let e = that.$el.querySelector("#key_"+x)
if(e){
e.style.display=getComputedStyle(e,null).display=='none'?'flex':'none'
}
})
},
show:function(arr){
var that = this
arr.forEach(function(x){
let e = that.$el.querySelector("#key_"+x)
if(e){
e.style.display='flex'
}
})
},
hide:function(arr){
var that = this
arr.forEach(function(x){
let e = that.$el.querySelector("#key_"+x)
if(e){
e.style.display='none'
}
})
}
},
computed(){
},
created(){
search();
}
})
</script>
...
[!tip]
可以根据本页面代码来实现其他数据的全文搜索