Question

I have two tables which share primary keys. I designed poorly, and it turns out that I need to ensure that every record in a1 has a corresponding record in a4

Query:

SELECT a1.id1, a4.id4
FROM  `a1` 
LEFT JOIN  `a4` ON a1.id1 = a4.id4

Results:

a1.id1.............a4.id4  
00000001    ......NULL  
00000002    ......NULL  
00001001    ......00001001  
00001002    ......00001002  

What is the best way to INSERT rows in a4 with a corresponding key to a1? In the example above, I need to insert the records 00000001 and 00000002 into a4. 00001001 and 00001002 should be left alone because they already exist in both a1 and a4

Database schema:

CREATE TABLE `a1` (
  `id1` int(8) unsigned zerofill NOT NULL auto_increment,
  `Shrt_Desc` varchar(200) default NULL,
  `ptype` int(5) NOT NULL,
  `userid` tinyint(5) NOT NULL,
  `submit_id` int(11) NOT NULL,
  `submit_time` int(11) NOT NULL,
  `update_id` int(11) NOT NULL,
  `update_time` int(11) NOT NULL,
  `pub` tinyint(1) default '1',
  `plate` tinyint(1) NOT NULL,
  `item` varchar(11) default NULL,
  PRIMARY KEY  (`id1`),
  KEY `fb_groupbyorder_Shrt_Desc_INDEX` (`Shrt_Desc`)
) ENGINE=MyISAM  DEFAULT CHARSET=utf8 AUTO_INCREMENT=124106 ;

CREATE TABLE `a4` (
  `id4` int(8) unsigned zerofill NOT NULL,
  `Water` decimal(10,2) default NULL,
  PRIMARY KEY  (`id4`)
) ENGINE=MyISAM DEFAULT CHARSET=utf8;

Answer 1

INSERT INTO a4 (id4)
SELECT a1.id1
FROM  `a1` 
LEFT JOIN  `a4` ON a1.id1 = a4.id4
WHERE a4.id4 IS NULL

http://sqlfiddle.com/#!2/91dc1/8