Connecting tables when querying in MySQL - php

Question

I have 3 tables: tco_articles, tco_module_eostext and tco_articles_modules. My tco_articles has unique id key. One for each article. My tco_module_eostext has unique instance_id that belongs to each article.
My tco_articles_modules contains all article_ids, but have 9 times as much instance_ids that are used in other tables.
So I can have article_id with instance_id that when you query in the tco_module_eostext will return empty.
I'd like to make a query that will return correct body text for the correct article.
So far I have:
global $wpdb;
$posts = array();
$ids = $wpdb->get_results('SELECT DISTINCT instance_id, article_id FROM tco_articles_modules', ARRAY_A);
This returns array with all the instances and ids like:
Array
(
[0] => Array
(
[instance_id] => 928615
[article_id] => 129396
)
[1] => Array
(
[instance_id] => 928616
[article_id] => 129396
)
[2] => Array
(
[instance_id] => 928617
[article_id] => 129396
)
[3] => Array
(
[instance_id] => 928618
[article_id] => 129396
)
You can see that the article_ids are the same but instance_id. When you put
$wpdb->get_results('SELECT body FROM tco_module_eostext WHERE instance_id=928617 ', ARRAY_A);
You may get empty, but for
$wpdb->get_results('SELECT body FROM tco_module_eostext WHERE instance_id=928618 ', ARRAY_A);
You could have some body text.
This is my problem. I need to go through all of them and filter out the not empty ones and assign them correct article. I managed to output the articles
foreach ($ids as $key => $value) {
$instance_ID = $value['instance_id'];
$article_ID = $value['article_id'];
$article_out = $wpdb->get_results('SELECT * FROM tco_articles WHERE id='.$article_ID.' ', ARRAY_A);
$posts[$article_ID] = $article_out[0];
}
Which returns something like:
Array
(
[129396] => Array
(
[id] => 129396
[headline] => Bla bla bla title
[intro] => This is cool article intro
[needs_review] => 0
[published] => 2014-12-16 09:17:00
[unpublished] =>
[hidden] => 0
[no_single_page] => 0
[permalink] => link-perma-link-here
[internal_slug] =>
[type_id] => 3
[thread_id] => 0
[news_id] => 0
[header_game_id] => 0
[puff_hh_id] => 0
[puff_title] =>
[hdrcol_id] => 900
[review_queued] =>
[lock_timeout] => 0
[created] => 2014-12-16 09:17:00
[updated] => 2015-01-16 13:51:30
[created_by] => 84142
[updated_by] => 84142
)
...
Now I'd like to append the body text from the tco_module_eostext table.
Is there a query I can use to do this automatically or to do this one at the time and then append to the $posts array?
The foreach method of querying is kinda slow when you have 180000+ posts.
Any help is appreciated.

Answer 1

If you are sure that there is always only one row in tco_module_eostext against each article_id, you can use JOIN (inner join), which will only show one row for each article_id.
SELECT a.*, t.body
FROM tco_articles a
JOIN tco_articles_modules m ON m.article_id = a.id
JOIN tco_module_eostext t ON m.instance_id = t.instance_id
//WHERE .....
But, this will not show any row of some articles if there is no entry in other two tables for that article_id. But there is still way to solve this. We can use LEFT OUTER JOIN and then make sure we only make the join if there is any row in tco_module_eostext for any instace_id. This will make sure you get at least the article info from tco_articles table when there is no data in other tables.
SELECT a.*, t.body
FROM tco_articles a
LEFT OUTER JOIN tco_articles_modules m ON m.article_id = a.id AND EXISTS ( SELECT NULL FROM tco_module_eostext WHERE instance_id = m.instance_id )
LEFT OUTER JOIN tco_module_eostext t ON m.instance_id = t.instance_id
//WHERE .....

Answer 2

Why not use a query with join?
Not tested!:
SELECT
article.*,
text.body as bodytext
FROM
tco_articles_modules AS modules LEFT OUTER JOIN
tco_module_eostext AS text ON
modules.instance_id = text.instance_id LEFT OUTER JOIN
tco_articles AS article ON
article.id = modules.article_id
it should get all articles with the assigned article_id from tco_articles_modules
Have a look at OUTER Join - you may want to replace this with an INNER JOIN for faster Queries. Also you ma want an WHERE condition for filtering in the query. Watch also out for the right indexing in mysql table - each joining column should be indexed - this will get much more faster results.