分不多，只有100，希望大家可以帮我

A表： 关键字段     id(字段id)    attr_type（字段类别 只获取是1）
                                    15              1
                                    14              0
                                    13              1
                                    12              1             可能会有更多
B表： 关键字段     a_id(字段id)   in_id（同A表ID 关联）shop_id(获取依据字段 id)
                                    400              15                                 15
                                    399              15                                 15

                                    388              14                                 15-----
                                    387              14                                 15-----这两组不做组合 以为A表中 type是0

                                    386              13                                 15
                                    385              13                                 15

                                    384              12                                 15
                                    383              12                                 15        可能会有更多是1的属性 也可能就一组

                                    384              12                                 14
                                    383              12                                 14 
                                    382              12                                 14
                                    381              12                                 14
                                    380              12                                 14        可能会有更多
C表：关键字段     id(字段id)   shop_id（同B表shop_id 关联）attr_id(循环出来的数值)   以下是我想得到的结果
                                      1              15                                          400|386|384
                                      2              15                                          400|386|385   
                                      3              15                                          400|387|384   如果此组已存在，就跳过 如果不存在 往下自动循环出来写入

意图 我给予ID 15的指令  从AB表获取数据 生成到C表 如果原先已存在，就忽略不生成不存在的生成





比如我想循环shop_id  15的这个依据id 生成到C表 求代码
以下是我写的部分sql 后面不会了 求救

    $sql = "SELECT b.a_id ".            "FROM " .$GLOBALS['ecs']->table('B'). " AS b ".            "LEFT JOIN " .$GLOBALS['ecs']->table('A'). " AS a ".            "ON a.id IN(b.in_id) ".            "WHERE b.shop_id = '15' AND a.attr_type = '1' " .             "ORDER BY b.a_id DESC";    $row = $GLOBALS['db']->getAll($sql);

回复讨论(解决方案)

没看明白结果是怎么来的，猜测一下

create temporary table A (id int, attr_type int);insert into A values (15, 1),(14, 0),(13, 1),(12, 1); create temporary table B (a_id int, in_id int, shop_id int);insert into B values (400, 15, 15),(399, 15, 15),(388, 14, 15)      ,(387, 14, 15),(386, 13, 15),(385, 13, 15),(384, 12, 15)      ,(383, 12, 15),(384, 12, 14),(383, 12, 14)       ,(382, 12, 14),(381, 12, 14),(380, 12, 14);select shop_id, group_concat(a_id) as attr_id from A, B where A.id=B.in_id and A.attr_type=1 group by shop_id, B.in_id

shop_id attr_id 14      384,383,382,381,380 15      384,383 15      386,385 15      400,399 

结果就是 我浏览器给与 ....&shop_id=15  这个格局 php收到指令 从b表获取shop_id为15的数据 然后从b表隔离掉 attr_type 不是1的数据 
获取到的应该是 以下几组
B表： 关键字段     a_id(字段id)   in_id（同A表ID 关联）shop_id(获取依据字段 id)
                                    400              15                                 15
                                    399              15                                 15
                                    386              13                                 15
                                    385              13                                 15
                                    384              12                                 15
                                    383              12                                 15 

然后归类  以 in_id 区分

                                    400              15                                 15
                                    399              15                                 15
*********************************************************************************************************************一个分组
                                    386              13                                 15
                                    385              13                                 15
*********************************************************************************************************************一个分组
                                    384              12                                 15
                                    383              12                                 15 
*********************************************************************************************************************一个分组

然后把 前面的 a_id 字段数据做组合变化 写入到c表 形成 400|386|384 这种

C表：关键字段     id(字段id)   shop_id（同B表shop_id 关联）attr_id(循环出来的数值)   以下是我想得到的结果
                                      1              15                                          400|386|384
                                      2              15                                          400|386|385   

C表 id 为 auto属性 shop_id 为我给与的id  attr_id 就是组合出来的变化结果

结果就是 我浏览器给与 ....&shop_id=15  这个格局 php收到指令 从b表获取shop_id为15的数据 然后从b表隔离
写错了 应该是从A表帅选 
结果就是 我浏览器给与 ....&shop_id=15  这个格局 php收到指令 从b表获取shop_id为15的数据 然后从 A表隔离

select in_id, shop_id, group_concat(a_id) as attr_id from A, B where A.id=B.in_id and A.attr_type=1 and shop_id=15 group by  B.in_id

in_id shop_id attr_id 12    15      384,383 13    15      386,385 15    15      400,399 

select in_id, shop_id, group_concat(a_id) as attr_id from A, B where A.id=B.in_id and A.attr_type=1 and shop_id=15 group by  B.in_id

in_id shop_id attr_id 12    15      384,383 13    15      386,385 15    15      400,399 

版主，我已经自己写好了，谢谢你啊

大神，请您帮个忙，怎么把这个表里的数据按照我的要求隔离出来
img_id 标示为1的 goods_id 261 attr_id 区分组的依据 

最终的要求就是  输出
 a= '584,583,582,581,580,579'
 b= '578,577,576,575,574,573'

572和495不输出 因为img_id 不等于一 谢谢了

select goods_id, attr_id, groub_concat(goods_attr_id) from tbl_name where img_id=1 group by attr_id

$a = array('571','570');$b = array('569','568','567','566','565','564');$c = array('563','562','561','560','559');foreach(combineDika($a,$b) as $v){	$r[] = join('|', $v);}$attr_all =  join(',', $r);$attr_all =  explode(',', $attr_all);foreach ($attr_all as $list){        $sql = "SELECT goods_attr " .                "FROM " .$GLOBALS['ecs']->table('products') .                "WHERE goods_attr = '".$list."' ";        $check_attr = $GLOBALS['db']->getOne($sql);		$value = 'ECS000261g_p'.rand(10000,99999);		if (empty($check_attr))        {             $sql = "INSERT INTO " . $GLOBALS['ecs']->table('products') . " (goods_id, goods_attr, product_sn, product_number)  VALUES ('261', '".$list."', '$value', '10000')";             $GLOBALS['db']->query($sql);        }}

这是我后面执行的代码

用个2B方法读取出来了
谢谢版主
我现在结贴