要随机查询数据，一开始我们可能想到最简单的order by rand() 来操作，如果生成的页面很多，效率不敢恭维啊，怎么做才能提高效率，现提供代码给各位鞋童参考一哈
如果指定了栏目，一般最简单的做法是
{pc:content action="lists" catid="2" num="10" order="rand()"}
{loop $data $r}
<li><a href="{$r[url]}" target="_blank"{title_style($r[style])}>{$r[title]}</a> {date('Y-m-d H:i:s',$r[inputtime])}</li>
{/loop}
{/pc}
或者
{pc:get sql="select * from phpcms_news where status='99' catid in(1,2,3) order by rand()" num="10"}
刚刚说了，这样的效率很低
高效率代码：（以文章模型为例）
{pc:get sql="select * from phpcms_news as t1 join (select round(rand() * ((select max(id) from phpcms_news)-(select min(id) from phpcms_news))+(select min(id) from phpcms_news)) as id) as t2 where t1.id >= t2.id and t1.status=99 order by t1.id" num="10"}
如果要指定栏目，则在 “order by t1.id” 前面加上 and t1.catid in(1,2,3)
{pc:get sql="select * from phpcms_news as t1 join (select round(rand() * ((select max(id) from phpcms_news)-(select min(id) from phpcms_news))+(select min(id) from phpcms_news)) as id) as t2 where t1.id >= t2.id and t1.status=99 and t1.catid in(1,2,3) order by t1.id" num="10"}
如果需要调用当前栏目（包含子栏目）
则
{php $bcatid = $CATEGORYS[$catid]['arrchildid'];}
{pc:get sql="select * from phpcms_news as t1 join (select round(rand() * ((select max(id) from phpcms_news)-(select min(id) from phpcms_news))+(select min(id) from phpcms_news)) as id) as t2 where t1.id >= t2.id and t1.status=99 and t1.catid in($bcatid) order by t1.id" num="10"}
这里的数据库前缀 phpcms_ 不用管，系统会自动替换为你当前的数据库前缀
如果需要随机调用 下载 模型的，则把 phpcms_news 改为 phpcms_download 即可