一直想写个评论系统,看了下多说,网易,简书的评论,想了下自己该实现怎样的评论系统。
评论系统关键是嵌套层数以及数据库表设计。嵌套层数多,表结构复杂,呈现也麻烦,最后决定实现一个二级评论。系统由maven构建,springboot快速搭建spring环境。前台采用angularjs+bootstrap,后台使用springmvc+mybatis,数据库采用MySQL.前台请求后台API操作评论。
目录结构
数据库表设计
##说说表或者文章表 create table saying ( saying_id INT NOT NULL AUTO_INCREMENT PRIMARY KEY, sayingContent VARCHAR(500) NOT NULL, author VARCHAR(50) NOT NULL, sayingAvatar VARCHAR(50) NOT NULL, likes VARCHAR(500) NOT NULL, createTime datetime NOT NULL ) ENGINE=InnoDB DEFAULT CHARSET=utf8; ##一级评论表 create table firstLevelComment ( flc_id INT NOT NULL AUTO_INCREMENT PRIMARY KEY, sayingId INT NOT NULL, commenter VARCHAR(50) NOT NULL, commenterAvatar VARCHAR(50) NOT NULL, commentContent VARCHAR(500) NOT NULL, commentTime datetime NOT NULL ) ENGINE=InnoDB DEFAULT CHARSET=utf8; ##二级评论表 create table secondLevelComment ( slc_id INT NOT NULL AUTO_INCREMENT PRIMARY KEY, sayingId INT NOT NULL, flcId INT NOT NULL, replier VARCHAR(50) NOT NULL, toCommenter VARCHAR(50) NOT NULL, replyContent VARCHAR(50) NOT NULL, replyTime datetime NOT NULL ) ENGINE=InnoDB DEFAULT CHARSET=utf8;
获取评论的mapper(关键)
<?xml version="1.0" encoding="UTF-8" ?> <!DOCTYPE mapper PUBLIC "-//mybatis.org//DTD Mapper 3.0//EN" "http://mybatis.org/dtd/mybatis-3-mapper.dtd" > <mapper namespace="personal.timeless.cms.mapper.SayingMapper" > <resultMap id="SayingMap" type="saying" > <id column="saying_id" property="id" jdbcType="INTEGER" /> <result column="sayingContent" property="sayingContent" jdbcType="INTEGER" /> <result column="author" property="author" jdbcType="VARCHAR" /> <result column="sayingAvatar" property="avatar" jdbcType="VARCHAR" /> <result column="likes" property="likes" jdbcType="VARCHAR" /> <result column="createTime" property="createTime" jdbcType="TIMESTAMP" /> <collection property="flcs" ofType="firstLevelComment" column="sayingId"> <id column="flc_id" property="id" jdbcType="INTEGER" /> <result column="sayingId" property="sayingId" jdbcType="INTEGER" /> <result column="commenter" property="commenter"/> <result column="commenterAvatar" property="avatar"/> <result column="commentContent" property="commentContent"/> <result column="commentTime" property="commentTime" jdbcType="TIMESTAMP" /> <collection property="slcs" ofType="secondLevelComment" column="flcId"> <id column="slc_id" property="id" jdbcType="INTEGER" /> <result column="flcId" property="flcId" jdbcType="INTEGER" /> <result column="replier" property="replier"/> <result column="toCommenter" property="toCommenter"/> <result column="replyContent" property="replyContent"/> <result column="replyTime" property="replyTime" jdbcType="TIMESTAMP" /> </collection> </collection> </resultMap> <select id="selectOneById" resultMap="SayingMap" parameterType="int" > select * from (select * from saying s left join firstLevelComment fc on s.saying_id=fc.sayingId where s.saying_id=#{id}) tmp left join secondLevelComment sc on tmp.flc_id = sc.flcId </select> <select id="updateLikesById"> update saying set likes = #{likes} where saying_id = #{id} </select> </mapper>
页面展示
以上就是本文的全部内容,希望对大家的学习有所帮助,也希望大家多多支持呐喊教程。
声明:本文内容来源于网络,版权归原作者所有,内容由互联网用户自发贡献自行上传,本网站不拥有所有权,未作人工编辑处理,也不承担相关法律责任。如果您发现有涉嫌版权的内容,欢迎发送邮件至:notice#nhooo.com(发邮件时,请将#更换为@)进行举报,并提供相关证据,一经查实,本站将立刻删除涉嫌侵权内容。