Can you split/explode a field in a MySQL query?

Question

I have to create a report on some student completions. The students each belong to one client. Here are the tables (simplified for this question).

CREATE TAB         


        

Answer 1

Until now, I wanted to keep those comma separated lists in my SQL db - well aware of all warnings!

I kept thinking that they have benefits over lookup tables (which provide a way to a normalized data base). After some days of refusing, I've seen the light:

• Using lookup tables is NOT causing more code than those ugly string operations when using comma separated values in one field.
• The lookup table allows for native number formats and is thus NOT bigger than those csv fields. It is SMALLER though.
• The involved string operations are slim in high level language code (SQL and PHP), but expensive compared to using arrays of integers.
• Databases are not meant to be human readable, and it is mostly stupid to try to stick to structures due to their readability / direct editability, as I did.

In short, there is a reason why there is no native SPLIT() function in MySQL.

Answer 2

SELECT
  tab1.std_name, tab1.stdCode, tab1.payment,
  SUBSTRING_INDEX(tab1.payment, '|', 1) as rupees,
  SUBSTRING(tab1.payment, LENGTH(SUBSTRING_INDEX(tab1.payment, '|', 1)) + 2,LENGTH(SUBSTRING_INDEX(tab1.payment, '|', 2))) as date
FROM (
  SELECT DISTINCT
    si.std_name, hfc.stdCode,
    if(isnull(hfc.payDate), concat(hfc.coutionMoneyIn,'|', year(hfc.startDtae), '-',  monthname(hfc.startDtae)), concat(hfc.payMoney, '|', monthname(hfc.payDate), '-', year(hfc.payDate))) AS payment
  FROM hostelfeescollection hfc
  INNER JOIN hostelfeecollectmode hfm ON hfc.tranId = hfm.tranId
  INNER JOIN student_info_1 si ON si.std_code = hfc.stdCode
  WHERE hfc.tranId = 'TRAN-AZZZY69454'
) AS tab1

Answer 3

I used the above logic but modified it slightly. My input is of format : "apple:100|pinapple:200|orange:300" stored in a variable @updtAdvanceKeyVal

Here is the function block :

set @res = "";

set @i = 1;
set @updtAdvanceKeyVal = updtAdvanceKeyVal;

REPEAT


 -- set r =  replace(SUBSTRING(SUBSTRING_INDEX(@updtAdvanceKeyVal, "|", @i),
 --  LENGTH(SUBSTRING_INDEX(@updtAdvanceKeyVal, "|", @i -1)) + 1),"|","");

-- wrapping the function in "replace" function as above causes to cut off a character from
 -- the 2nd splitted value if the value is more than 3 characters. Writing it in 2 lines causes no such problem and the output is as expected
-- sample output by executing the above function :
-- orange:100
-- pi apple:200    !!!!!!!!strange output!!!!!!!!
-- tomato:500

      set @r =  SUBSTRING(SUBSTRING_INDEX(@updtAdvanceKeyVal, "|", @i),
                  LENGTH(SUBSTRING_INDEX(@updtAdvanceKeyVal, "|", @i -1)) + 1);

      set @r = replace(@r,"|","");

      if @r <> "" then

              set @key = SUBSTRING_INDEX(@r, ":",1);
              set @val = SUBSTRING_INDEX(@r, ":",-1);

              select @key, @val;
      end if;

      set @i = @i + 1;

     until @r = ""
END REPEAT;

Answer 4

MySQL's only string-splitting function is SUBSTRING_INDEX(str, delim, count). You can use this, to, for example:

• Return the item before the first separator in a string:

  mysql> SELECT SUBSTRING_INDEX('foo#bar#baz#qux', '#', 1);
  +--------------------------------------------+
  | SUBSTRING_INDEX('foo#bar#baz#qux', '#', 1) |
  +--------------------------------------------+
  | foo                                        |
  +--------------------------------------------+
  1 row in set (0.00 sec)
  

• Return the item after the last separator in a string:

  mysql> SELECT SUBSTRING_INDEX('foo#bar#baz#qux', '#', -1);
  +---------------------------------------------+
  | SUBSTRING_INDEX('foo#bar#baz#qux', '#', -1) |
  +---------------------------------------------+
  | qux                                         |
  +---------------------------------------------+
  1 row in set (0.00 sec)
  

• Return everything before the third separator in a string:

  mysql> SELECT SUBSTRING_INDEX('foo#bar#baz#qux', '#', 3);
  +--------------------------------------------+
  | SUBSTRING_INDEX('foo#bar#baz#qux', '#', 3) |
  +--------------------------------------------+
  | foo#bar#baz                                |
  +--------------------------------------------+
  1 row in set (0.00 sec)
  

• Return the second item in a string, by chaining two calls:

  mysql> SELECT SUBSTRING_INDEX(SUBSTRING_INDEX('foo#bar#baz#qux', '#', 2), '#', -1);
  +----------------------------------------------------------------------+
  | SUBSTRING_INDEX(SUBSTRING_INDEX('foo#bar#baz#qux', '#', 2), '#', -1) |
  +----------------------------------------------------------------------+
  | bar                                                                  |
  +----------------------------------------------------------------------+
  1 row in set (0.00 sec)
  

In general, a simple way to get the nth element of a #-separated string (assuming that you know it definitely has at least n elements) is to do:

SUBSTRING_INDEX(SUBSTRING_INDEX(your_string, '#', n), '#', -1);

The inner SUBSTRING_INDEX call discards the nth separator and everything after it, and then the outer SUBSTRING_INDEX call discards everything except the final element that remains.

If you want a more robust solution that returns NULL if you ask for an element that doesn't exist (for instance, asking for the 5th element of 'a#b#c#d'), then you can count the delimiters using REPLACE and then conditionally return NULL using IF():

IF(
    LENGTH(your_string) - LENGTH(REPLACE(your_string, '#', '')) / LENGTH('#') < n - 1,
    NULL,
    SUBSTRING_INDEX(SUBSTRING_INDEX(your_string, '#', n), '#', -1)
)

Of course, this is pretty ugly and hard to understand! So you might want to wrap it in a function:

CREATE FUNCTION split(string TEXT, delimiter TEXT, n INT)
RETURNS TEXT DETERMINISTIC
RETURN IF(
    (LENGTH(string) - LENGTH(REPLACE(string, delimiter, ''))) / LENGTH(delimiter) < n - 1,
    NULL,
    SUBSTRING_INDEX(SUBSTRING_INDEX(string, delimiter, n), delimiter, -1)
);

You can then use the function like this:

mysql> SELECT SPLIT('foo,bar,baz,qux', ',', 3);
+----------------------------------+
| SPLIT('foo,bar,baz,qux', ',', 3) |
+----------------------------------+
| baz                              |
+----------------------------------+
1 row in set (0.00 sec)

mysql> SELECT SPLIT('foo,bar,baz,qux', ',', 5);
+----------------------------------+
| SPLIT('foo,bar,baz,qux', ',', 5) |
+----------------------------------+
| NULL                             |
+----------------------------------+
1 row in set (0.00 sec)

mysql> SELECT SPLIT('foo###bar###baz###qux', '###', 2);
+------------------------------------------+
| SPLIT('foo###bar###baz###qux', '###', 2) |
+------------------------------------------+
| bar                                      |
+------------------------------------------+
1 row in set (0.00 sec)

Answer 5

Well, nothing I used worked, so I decided creating a real simple split function, hope it helps:

    DECLARE inipos INTEGER;
    DECLARE endpos INTEGER;
    DECLARE maxlen INTEGER;
    DECLARE item VARCHAR(100);
    DECLARE delim VARCHAR(1);

    SET delim = '|';
    SET inipos = 1;
    SET fullstr = CONCAT(fullstr, delim);
    SET maxlen = LENGTH(fullstr);

    REPEAT
        SET endpos = LOCATE(delim, fullstr, inipos);
        SET item =  SUBSTR(fullstr, inipos, endpos - inipos);

        IF item <> '' AND item IS NOT NULL THEN           
            USE_THE_ITEM_STRING;
        END IF;
        SET inipos = endpos + 1;
    UNTIL inipos >= maxlen END REPEAT;