闲话iBatis

  1. 云栖社区>
  2. 博客>
  3. 正文

闲话iBatis

技术小阿哥 2017-11-13 22:43:00 浏览697
展开阅读全文
使用iBatis开发已经有一年了,这一年来不管愿意不愿意,喜欢不喜欢,都得用,想用Hibernate不行。
 
随着对iBatis的熟悉,渐渐发现初次接触的人因为不能很好的掌握iBatis使用的方法而对其有一些偏激的看法。
 
1、iBatis的是用来自由写SQL的,对SQL语句的简单的封装,相当于在JDBC上加了层外壳。
我认为,iBatis的目标是简化JDBC的操作,对SQL的查询结果做了很好的映射封装,这样查询的结果就形成了一个java对象、集合、或者Map等。这样的数据结构相比JDBC结果集的处理是一个质的飞跃。
 
iBatis虽然支持随意的SQL查询,并且对查询结果集也没有什么特别限制,最懒惰的是定义查询结果集类型为map。但是,这并不是iBatis设计者的本意,我认为,在使用iBatis的时候,尽可能写通用的增删改差sql,然后通过通过这些sql所对应DAO方法的组合,来实现各种业务,而不是写很多特殊的sql、甚至存储过程。当你在更换数据库的时候,移植这些特殊sql和过程就是一件很困难的事情。
 
2、iBatis映射太麻烦,表字段和实体(Entity,也有人称POJO)对应比较麻烦,不够直观,开发困难。
其实,最好的做法就是使用数据库表字段名作为实体的属性名,这样完全对应起来,问题就解决了。并且,可以写个工具,从数据库生成POJO和SQLMap、DAO、Service等等。这样可以极大提高效率,我就这么做的。
 
3、iBatis分页是内存分页,假分页
的确如此,对于小数据量是够用了,对于大数据量,这是不可接受的。因此可以通过动态SqL来拼装分页查询Sql。
                <dynamic prepend=""> 
                        <isNotNull property="_start"> 
                                <isNotNull property="_size"> 
                                        limit #_start#, #_size# 
                                </isNotNull> 
                        </isNotNull> 
                </dynamic> 
 
4、iBatis无法做成Hibernate那样的关连查询、级联保存、级联删除。
iBatis可以做到,而且实现很灵活方便。但是比Hibernate操作更容易,也很好理解。这里面对结果集映射类型也很讲究。
        <resultMap id="result_base" class="Channel"> 
                <result property="id" column="id"/> 
                <result property="stationid" column="stationid"/> 
                <result property="remark" column="remark"/> 
                <result property="status" column="status"/> 
        </resultMap> 
        <resultMap id="result" class="Channel" extends="channel.result_base"> 
                <result property="cfgList" column="id" select="videoregxp.findByChannelId"/> 
        </resultMap>
 
  <resultMap id="result_base" class="Videoregxp"> 
    <result property="id" column="id"/> 
    <result property="lmid" column="lmid"/> 
    <result property="name" column="name"/> 
    <result property="cname" column="cname"/> 
    <result property="sortnum" column="sortnum"/> 
    <result property="remark" column="remark"/> 
  </resultMap> 
        <resultMap id="result" class="Videoregxp" extends="videoregxp.result_base"> 
                <result property="regex" column="id" select="regexpar.findByVideoregxpId"/> 
        </resultMap>
 
        /** 
         * 保存一个视频信息,级联保存相关的剧集和视频分段文件 
         * 
         * @param videoinfo 视频信息对象 
         * @return 返回保存后的对象 
         */
 
        public Videoinfo saveVideoinfoWithCascade(Videoinfo videoinfo) { 
                videoinfoDAO.saveVideoinfo(videoinfo); 
                for (Juji juji : videoinfo.getJujiList()) { 
                        juji.setVideoid(videoinfo.getId()); 
                        jujiDAO.saveJuji(juji); 
                        for (Secfiles file : juji.getSecfileList()) { 
                                file.setSegmentid(juji.getId()); 
                                secfilesDAO.saveSecfiles(file); 
                        } 
                } 
                return videoinfo; 
        }
 
        /** 
         * 删除指定标识的一个栏目 
         * 
         * @param id 栏目标识 
         */
 
        public void deleteChannelById(Long id) { 
                //todo:做可行性校验,以及级联删除等操作,并做业务特殊需求校验 
                Channel ch = channelDAO.findChannelById(id); 
                for (Videoregxp v : ch.getCfgList()) { 
                        regexparDAO.deleteRegexparByCfgId(v.getId()); 
                } 
                videoregxpDAO.deleteVideoregxpByLmId(ch.getId()); 
                channelDAO.deleteChannelById(id); 
        }
 
5、iBatis延迟加载好像有问题,本人现在不能确信。
 
6、iBatis用熟悉了和Hibernate一样容易,通过自己开发工具,可以极大提高开发效率。我会在附件贡献出我的代码生成工具。
 
7、iBatis的cache缓存没感觉到有多大优势,也没感觉性能的提升。
 
8、iBatis的动态SQL功能强大,Map参数更能获取最大的灵活性,这是Hibernate和其他持久化框架所无法相比的。
        <select id="findByParams" parameterClass="map" resultMap="channel.result"> 
                <include refid="sql_select"/> 
                <include refid="sql_where"/> 
        </select> 
 
        /** 
         * 根据动态条件查询栏目 
         * 
         * @param map 参数map 
         * @return 查询结果集 
         */
 
        public List<Channel> findChannelByParams(Map map) { 
                return getSqlMapClientTemplate().queryForList("channel.findByParams", map); 
        }
 
但是map参数也缺乏明确性,使用者无法清楚知道map参数里面究竟可以传什么类型,但是使用者可以通过查看sqlMap进一步获得对参数的详细认识,比如上面的完整SQLMap片段是:
 
        <!-- 动态条件分页查询 --> 
        <sql id="sql_count"> 
                select count(*) 
        </sql> 
        <sql id="sql_select"> 
                select * 
        </sql> 
        <sql id="sql_where"> 
                from channel 
                <dynamic prepend="where"> 
                        <isNotNull prepend="and" property="stationid"> 
                                stationid = #stationid# 
                        </isNotNull> 
                        <isNotEmpty prepend="and" property="name"> 
                                name like '%$name$%' 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="url"> 
                                url like '%$url$%' 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="regxp"> 
                                regxp like '$regxp$%' 
                        </isNotEmpty> 
                        <isNotNull prepend="and" property="encoding"> 
                                encoding = #encoding# 
                        </isNotNull> 
                        <isNotEmpty prepend="and" property="_timeout_ge"> 
                                <![CDATA[ 
                                timeout >= #_timeout_ge# 
                        ]]> 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="_timeout_lt"> 
                                <![CDATA[ 
                                timeout < #_timeout_lt# 
                        ]]> 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="_maxrow_ge"> 
                                <![CDATA[ 
                                maxrow >= #_maxrow_ge# 
                        ]]> 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="_maxrow_lt"> 
                                <![CDATA[ 
                                maxrow < #_maxrow_lt# 
                        ]]> 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="_start_ge"> 
                                <![CDATA[ 
                                start >= #_start_ge# 
                        ]]> 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="_start_lt"> 
                                <![CDATA[ 
                                start < #_start_lt# 
                        ]]> 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="pagereg"> 
                                pagereg like '%$pagereg$%' 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="kind"> 
                                kind = #kind# 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="mapname"> 
                                mapname = #mapname# 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="replaceb"> 
                                replaceb = #replaceb# 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="replacea"> 
                                replacea = #replacea# 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="ifredo"> 
                                ifredo = #ifredo# 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="fenbianlv"> 
                                fenbianlv = #fenbianlv# 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="_size_ge"> 
                                <![CDATA[ 
                                size >= #_size_ge# 
                        ]]> 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="_size_lt"> 
                                <![CDATA[ 
                                size < #_size_lt# 
                        ]]> 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="remark"> 
                                remark like '%$remark$%' 
                        </isNotEmpty> 
                        <isNotEmpty prepend="and" property="status"> 
                                status = #status# 
                        </isNotEmpty> 
                </dynamic> 
                <dynamic prepend=""> 
                        <isNotNull property="_start"> 
                                <isNotNull property="_size"> 
                                        limit #_start#, #_size# 
                                </isNotNull> 
                        </isNotNull> 
                </dynamic> 
        </sql> 
        <select id="findByParamsForCount" parameterClass="map" resultClass="int"> 
                <include refid="sql_count"/> 
                <include refid="sql_where"/> 
        </select> 
        <select id="findByParams" parameterClass="map" resultMap="channel.result"> 
                <include refid="sql_select"/> 
                <include refid="sql_where"/> 
        </select>
 
通过上面的映射,我们可以很清楚的看到,SQL的片段得到了重用,分页很优雅,统计和全功能查询也非常强大,相比hibernate,有过之而无不及,当然如果你不会这么用,那就麻烦多了。
 
9、我认为,iBatis应该尽量使用单表增删改差业务来组合功能更为强大的业务操作。但并不排斥使用多表关联进行查询,多表关联查询,结果集往往需要定义新的JavaBean来进行存储,但也可以简单的返回一个Map来存储结果集。
        <!-- 动态条件分页查询 --> 
        <sql id="sql_count"> 
                select count(a.*) 
        </sql> 
        <sql id="sql_select"> 
                select a.id                vid, 
                a.img             imgurl, 
                a.img_s         imgfile, 
                b.vfilename vfilename, 
                c.id                sid, 
                c.url             url, 
                c.filename    filename, 
                c.status        status 
        </sql> 
        <sql id="sql_where"> 
                From secfiles c, juji b, videoinfo a 
                where 
                a.id = b. videoid 
                and b.id = c.segmentid 
                and c.status = 0 
                <dynamic prepend=""> 
                        <isNotNull property="_start"> 
                                <isNotNull property="_size"> 
                                        limit #_start#, #_size# 
                                </isNotNull> 
                        </isNotNull> 
                </dynamic> 
        </sql> 
        <!-- 返回没有下载的记录总数 --> 
        <select id="getUndownFilesForCount" parameterClass="map" resultClass="int"> 
                <include refid="sql_count"/> 
                <include refid="sql_where"/> 
        </select> 
        <!-- 返回没有下载的记录 --> 
        <select id="getUndownFiles" parameterClass="map" resultClass="java.util.HashMap"> 
                <include refid="sql_select"/> 
                <include refid="sql_where"/> 
        </select>
 
在上面,我们的结果集返回为resultClass="java.util.HashMap",这时候,需要根据字
段的名称来取值,值类型为Object,key类型为String,这点要注意了:
        public List<Map<String,Object>> findUndownFiles(Map map) { 
                return getSqlMapClientTemplate().queryForList("secfiles.getUndownFiles", map); 
        }
 
        public void test_findUndownFiles() { 
                List<Map<String, Object>> co = ser.findUndownFiles(new HashMap()); 
                StringBuilder s = new StringBuilder(); 
                for (Map<String, Object> map : co) { 
                        System.out.println("---------------------------"); 
                        for (Map.Entry<String, Object> entry : map.entrySet()) { 
                                System.out.print(entry.getKey()+"\t"); 
                                System.out.println(entry.getValue()); 
                        } 
                } 
        }
 
打印结果:
--------------------------- 
sid  1 
vfilename  200905252009235799 
url  http://d18.v.iask.com/f/1/f47817a394730dc682e660b943e84cc41006606.flv 
status  0 
filename  200905252009235799-00.flv 
imgfile  200905252009234399.jpg 
vid  1 
imgurl  http://p4.v.iask.com/95/595/1757503_1.jpg 
--------------------------- 
sid  2130 
vfilename  2009062615063867492 
url  http://lz.dhot.v.iask.com/f/1/0ee2ae8b973988f6a93c071c8045ca5217266409.mp4 
status  0 
filename  2009062615063867492-00.mp4 
imgfile  2009062615063825434.jpg 
vid  93 
imgurl  http://cache.mars.sina.com.cn/nd/movievideo//thumb/2/1502_120160.jpg 
--------------------------- 
sid  2131 
vfilename  2009062615064184076 
url  http://lz5.dhot.v.iask.com/f/1/36d3dadacb8d6bda434a58e7418ad3cc19037464.flv 
status  0 
filename  2009062615064184076-00.flv 
imgfile  2009062615064136733.jpg 
vid  94 
imgurl  http://cache.mars.sina.com.cn/nd/movievideo//thumb/6/2106_120160.jpg
 
10、iBatis在开发效率上并没有比Hibernate差,如果你觉得太难用,那说明你还不够精,两者各有优缺点,应该扬长避短,最终发现:好东西都很简单!希望在这篇文章能为iBatis正名。
 

本文转自 leizhimin 51CTO博客,原文链接:http://blog.51cto.com/lavasoft/170160,如需转载请自行联系原作者

网友评论

登录后评论
0/500
评论
技术小阿哥
+ 关注