nandi/glean · fd2d54c

nandi/gleanpublic⑂ Fork 0

⑂ fd2d54c

Optimize recommendation query and remove unnecessary migrationUnverified

Julien Robert committed 2026-04-20T12:52:18+02:00 Browse files

fd2d54c parent: 5641d4d

modified internal/cluster/jaccard.go +5 -7 Viewed

@@ -24,21 +24,19 @@ func (e *Engine) ComputeArticleRecommendations(ctx context.Context) error {
24	24
25	25	_, err = tx.ExecContext(ctx, `
26	26	INSERT INTO user_article_recommendations (user_did, feed_url, article_url, score)
27		- SELECT target, l.feed_url, l.article_url, SUM(us.jaccard) AS score
	27	+ SELECT targets.target, l.feed_url, l.article_url, SUM(targets.jaccard) AS score
28	28	FROM (
29		- SELECT us.user_a AS target, s.user_did AS peer
	29	+ SELECT us.user_a AS target, us.user_b AS peer, us.jaccard
30	30	FROM user_similarity us
31	31	WHERE us.jaccard > 0.2
32	32	UNION ALL
33		- SELECT us.user_b AS target, s.user_did AS peer
	33	+ SELECT us.user_b AS target, us.user_a AS peer, us.jaccard
34	34	FROM user_similarity us
35	35	WHERE us.jaccard > 0.2
36	36	) targets
37	37	JOIN likes l ON l.author_did = targets.peer
38		- WHERE l.article_url NOT IN (
39		- SELECT a.url FROM articles a
40		- JOIN subscriptions s ON a.feed_url = s.feed_url AND s.user_did = targets.target
41		- LEFT JOIN read_state r ON r.user_did = targets.target AND r.article_id = a.id
	38	+ WHERE NOT EXISTS (
	39	+ SELECT 1 FROM subscriptions sub WHERE sub.user_did = targets.target AND sub.feed_url = l.feed_url
42	40	)
43	41	AND NOT EXISTS (
44	42	SELECT 1 FROM likes ul WHERE ul.author_did = targets.target AND ul.feed_url = l.feed_url AND ul.article_url = l.article_url

  	_, err = tx.ExecContext(ctx, `
  		INSERT INTO user_article_recommendations (user_did, feed_url, article_url, score)
--		SELECT target, l.feed_url, l.article_url, SUM(us.jaccard) AS score
++		SELECT targets.target, l.feed_url, l.article_url, SUM(targets.jaccard) AS score
  		FROM (
--			SELECT us.user_a AS target, s.user_did AS peer
++			SELECT us.user_a AS target, us.user_b AS peer, us.jaccard
  			FROM user_similarity us
  			WHERE us.jaccard > 0.2
  			UNION ALL
--			SELECT us.user_b AS target, s.user_did AS peer
++			SELECT us.user_b AS target, us.user_a AS peer, us.jaccard
  			FROM user_similarity us
  			WHERE us.jaccard > 0.2
  		) targets
  		JOIN likes l ON l.author_did = targets.peer
--		WHERE l.article_url NOT IN (
++		WHERE NOT EXISTS (
--			SELECT a.url FROM articles a
++			SELECT 1 FROM subscriptions sub WHERE sub.user_did = targets.target AND sub.feed_url = l.feed_url
--			JOIN subscriptions s ON a.feed_url = s.feed_url AND s.user_did = targets.target
--			LEFT JOIN read_state r ON r.user_did = targets.target AND r.article_id = a.id
  		)
  		AND NOT EXISTS (
  			SELECT 1 FROM likes ul WHERE ul.author_did = targets.target AND ul.feed_url = l.feed_url AND ul.article_url = l.article_url

modified internal/db/db.go +0 -7 Viewed

@@ -161,12 +161,5 @@ func migrate(db *sql.DB) error {
161	161	}
162	162	}
163	163
164		- migrations := []string{
165		- `ALTER TABLE feeds ADD COLUMN favicon_url TEXT`,
166		- }
167		- for _, m := range migrations {
168		- tx.Exec(m)
169		- }
170		-
171	164	return tx.Commit()
172	165	}

  		}
  	}
--	migrations := []string{
--		`ALTER TABLE feeds ADD COLUMN favicon_url TEXT`,
--	}
--	for _, m := range migrations {
--		tx.Exec(m)
--	}
--
  	return tx.Commit()
  }