trwiki:Most frequent unexisting links to Mirada articles

SQL

USE trwiki_p;
SELECT pl_title, COUNT(pl_title) AS counting, arc.ar_len, GROUP_CONCAT(DISTINCT COALESCE(IF(r.ar_parent_id = 0, COALESCE(CAST(r.ar_len AS INT),0),CAST(r.ar_len AS INT) - COALESCE(CAST(rp.ar_len AS INT),0))) SEPARATOR "; ") AS mirada_bytes
FROM pagelinks
JOIN page ON pl_from = page_id AND pl_from_namespace = 0 AND pl_namespace = 0
JOIN archive AS arc ON 
ar_title=pl_title 
AND ar_timestamp <= (SELECT TIMESTAMP("2017-03-29 16:26:55")) 
AND ar_id=(SELECT MAX(ar_id) FROM archive AS a1 WHERE a1.ar_title=arc.ar_title AND ar_namespace=0 AND ar_timestamp <= (SELECT TIMESTAMP("2017-03-29 16:26:55")))
AND ar_namespace=0
JOIN archive r ON arc.ar_title=r.ar_title AND r.ar_actor=2732 AND r.ar_namespace=0 AND r.ar_timestamp <= arc.ar_timestamp
LEFT JOIN archive rp ON 
  r.ar_parent_id = rp.ar_id  AND r.ar_namespace=0
WHERE 
NOT EXISTS
(SELECT p1.page_title FROM page AS p1
WHERE p1.page_title = pl_title
AND pl_namespace = 0
AND p1.page_namespace = 0)
AND pl_title IN (SELECT pl_title FROM pagelinks 
                 WHERE pl_from = 2219492 AND pl_from_namespace = 4 AND pl_namespace = 0)
GROUP BY pl_title
#HAVING counting > 19
ORDER BY counting DESC, pl_title ASC, page_title ASC;
SQL

Checking query status...
