Question

while implementing an inline search function for our local file archive I’ve come up with a serious problem I have found no answer for. We have two tables:

 fild_id | file_name
---------------------
       1 | this_file
       2 | that_file
       3 | new_file

 file_archive_id | file_archive_version | file_id
--------------------------------------------------
               1 |                    1 |       1
               2 |                    2 |       1
               3 |                    1 |       2
               4 |                    1 |       3
               5 |                    3 |       1

I want to join both tables via file_id, selecting only the one file_archive row with the biggest file_archive_version:

 fild_id | file_name | file_archive_id | file_archive_version
--------------------------------------------------------------
       1 | this_file |               5 |                    3
       2 | that_file |               3 |                    1
       3 | new_file  |               4 |                    1

Is there any possibility to do this via a single select statement?

Solution:

SELECT df.*,
       (
        SELECT dfa.file_archive_id
          FROM dca_file_archive dfa
          WHERE df.file_id = dfa.file_id
          ORDER BY dfa.file_archive_version desc LIMIT 1
       ) as file_archive_id,
       (
        SELECT dfa.file_archive_version
          FROM dca_file_archive dfa
          WHERE df.file_id = dfa.file_id
          ORDER BY dfa.file_archive_version desc LIMIT 1
       ) as file_archive_version
FROM dca_file df

Both tables having ~16k rows, this statement takes 0.9 seconds to perform, which is 120x faster than the first join solution.

Answer 1

Solution (without altering the indexes on my tables):

SELECT df.*,
       (
        SELECT dfa.file_archive_id
          FROM dca_file_archive dfa
          WHERE df.file_id = dfa.file_id
          ORDER BY dfa.file_archive_version desc LIMIT 1
       ) as file_archive_id,
       (
        SELECT dfa.file_archive_version
          FROM dca_file_archive dfa
          WHERE df.file_id = dfa.file_id
          ORDER BY dfa.file_archive_version desc LIMIT 1
       ) as file_archive_version
FROM dca_file df

Both tables having ~16k rows, this statement takes 0.9 seconds to perform, which is 120x faster than the first join solution.

I know this is not the finest you can do with SQL