Can anyone please shorten this query? - mysql

Question

This query generates next datafile name for a existing schema.
For e.g. id last datafile name is test_schema_05.dbf
then this query gives test_schema_06.dbf
I need to shorten this query .
Is this possible?
SELECT CONCAT
(SUBSTR
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/'))) + 1
)
),
1,
INSTR
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/'))) + 1
)
),
'_',
1,
(LENGTH
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/'))) + 1
)
)
) - LENGTH
(REPLACE
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/'))) + 1
)
),
'_'
)
)
)
)
),
CONCAT
('0',
SUBSTR
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/' ))) + 1
)
),
INSTR
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/' ))) + 1
)
),
'_',
1,
(LENGTH
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/'))) + 1
)
)
) - LENGTH
(REPLACE
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/'))) + 1
)
),
'_'
)
)
)
) + 1,
INSTR
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/'))) + 1
)
),
'.',
1
) - INSTR
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/' ))) + 1
)
),
'_',
1,
(LENGTH
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/' ))) + 1
)
)
) - LENGTH
(REPLACE
(MAX
(SUBSTR
(file_name,
INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/' ))) + 1
)
),
'_'
)
)
)
) - 1
) + 1
)
)
|| '.dbf' AS data_file_name
FROM dba_data_files
WHERE tablespace_name =
(SELECT default_tablespace
FROM dba_users
WHERE username = 'schema_name'
);

Answer 1

this almost made me cry. Also, it looks like your query is just somewhat wrong, as your MAX functions appear to only contain a single value, which is a string
You have SO MANY repeated lines that are clogging up your query, and your brain when you try to read it. Extract that out to a variable! Assuming this is all inside a sproc somewhere, you end up with something like this:
DECLARE i_intLocation INTEGER;
SET i_intLocation = INSTR (file_name, '/', 1, LENGTH (file_name) - LENGTH (REPLACE (file_name, '/'))) + 1;
DECLARE i_strSubstr VARCHAR;
SET i_strSubstr = (SUBSTR (file_name, i_intLocation));
SELECT
CONCAT(
SUBSTR(
MAX(i_strSubstr),
1,
INSTR(
MAX(i_strSubstr),
'_',
1,
(
LENGTH(MAX(i_strSubstr))
- LENGTH(
REPLACE(
MAX(i_strSubstr),
'_'
)
)
)
)
),
CONCAT(
'0',
SUBSTR(
MAX(i_strSubstr),
INSTR(
MAX(i_strSubstr),
'_',
1,
(
LENGTH(MAX(i_strSubstr))
- LENGTH(
REPLACE(
MAX(i_strSubstr),
'_'
)
)
)
) + 1,
INSTR(
MAX(i_strSubstr),
'.',
1
)
- INSTR(
MAX(i_strSubstr),
'_',
1,
(
LENGTH(MAX(i_strSubstr))
- LENGTH(
REPLACE(
MAX(i_strSubstr),
'_'
)
)
)
) - 1
) + 1
)
)
|| '.dbf' AS data_file_name
FROM dba_data_files
WHERE tablespace_name =
(SELECT default_tablespace
FROM dba_users
WHERE username = 'schema_name'
);

Answer 2

Using a sub query to get the highest file name to shorten the code it would come out something like this, but I suspect that this is no more efficient (and might well be less efficient).
SELECT CONCAT(SUBSTR(sub1.derived_bit,1,INSTR(sub1.derived_bit,'_',1,(LENGTH(sub1.derived_bit)- LENGTH(REPLACE(sub1.derived_bit,'_'))))),
CONCAT('0',SUBSTR(sub1.derived_bit,INSTR(sub1.derived_bit,'_',1,(LENGTH(sub1.derived_bit)- LENGTH(REPLACE(sub1.derived_bit,'_'))))+ 1,INSTR
(sub1.derived_bit,'.',1)- INSTR(sub1.derived_bit,'_',1,(LENGTH(sub1.derived_bit)- LENGTH(REPLACE(sub1.derived_bit, '_' ))))- 1)+ 1), '.dbf') AS data_file_name
FROM
(
SELECT MAX(SUBSTR(file_name, INSTR(file_name, '/', 1, LENGTH(file_name) - LENGTH(REPLACE(file_name, '/'))) + 1)) as derived_bit
FROM dba_data_files
INNER JOIN dba_users
ON dba_data_files.tablespace_name = dba_users.default_tablespace AND username = 'schema_name'
) sub1
However this is just based on your code which has syntax errors (which I have just repeated in my code above). For example your code is using INSTR with 4 parameters (should only have 2). Similarly you are using REPLACE with only 2 parameters. Your functions are looking for '/' while the file names you give as example do not contain any '/' characters.
How fixed is the format of the file name? Is there only ever 1 . in the file name? Is the number to increment always immediatly before the .dbf?