MySQL is just the best
Unfortunately I make a living using Microsoft and Oracle products. I shouldn't say unfortunately as I don't see myself doing any other job and it beats daytime television any day.

I use this quite a lot so I thought I'd put an article here somewhere. Based on the following concept:

  1.  RowID     column_to_return_as_string    
  2.  --------- -------------------------- 
  3.  1         Me 
  4.  2         Myself 
  5.  3         I 
  6.   
  7.  -- to be returned as 
  8.  RowID     my_field_name    
  9.  --------- -------------------------- 
  10.  1         Me,Myself,I 

Note that the following queries include the nested version because I find myself needing to group concatenate more often from another table then using data from the same table.

MySQL
  1.  SELECT 
  2.       ( 
  3.  SELECT  
  4.  GROUP_CONCAT(column_to_return_as_string)  
  5.  FROM  
  6.  a_pretend_table 
  7.  WHERE 
  8.  a_condition=true 
  9.       ) AS my_field_name 
  10.  FROM 
  11.       another_pretend_table 

T-SQL
  1.  SELECT 
  2.       STUFF
  3.         (  
  4.   SELECT  
  5.  ',' + column_to_return_as_string  
  6.   FROM  
  7.  a_pretend_table  
  8.   WHERE 
  9.  a_condition=true 
  10.   FOR XML PATH ('') 
  11.         ),1,1,'' 
  12.       ) AS my_field_name 
  13.  FROM 
  14.       another_pretend_table 

Oracle
  1.  SELECT 
  2.       ( 
  3.  SELECT  
  4.  wmsys.wm_concat(column_to_return_as_string)  
  5.  FROM  
  6.  a_pretend_table 
  7.  WHERE 
  8.  a_condition=true 
  9.       ) AS my_field_name 
  10.  FROM 
  11.       another_pretend_table 

Comments   

0 Lucas Thursday, 21st June 2012, 6:42 pm
Hi Sir,

For the T-SQL, is there a way to do it with out the FOR XML PATH (''), I am trying to GROUP_CONCAT 2 cell [1111] and [2222] which all belong to ID column as
ID
1111
2222
My desired out put will be [1111,2222],rig ht now I am getting TAG_SEAL>111122 22, which have something to do with the FOR XML PATH (''). And if I get rid of it, it give me error "Subquery returned more than 1 value."

Your help is much appreciated !!!

Add comment


Send