Issue Summary

When task report macro is sorted by user name ("assignee"), a new join is added to SQL query:

left join cwd_user cu on um.username = cu.user_name 

As both um.username and cu.user_name do not have indexes, the database has to scan tables (it leads to performance degradation).

Full SQL query statement:

SELECT DISTINCT t.GLOBAL_ID, t.CONTENT_ID,CASE WHEN cu.lower_display_name IS NULL THEN 0 ELSE 1 END CU_LOWER_DISPLAY_NAME,
cu.lower_display_name,t.CREATE_DATE
FROM AO_BAF3AA_AOINLINE_TASK t
JOIN CONTENT c ON t.CONTENT_ID = c.CONTENTID
JOIN SPACES s ON c.SPACEID = s.SPACEID
LEFT JOIN user_mapping um ON t.ASSIGNEE_USER_KEY = um.user_key
LEFT JOIN cwd_user cu on um.username = cu.user_name
WHERE s.SPACESTATUS <> 'ARCHIVED' AND c.CONTENT_STATUS <> 'deleted' AND t.TASK_STATUS = 'UNCHECKED'
ORDER BY CU_LOWER_DISPLAY_NAME DESC , cu.lower_display_name ASC , t.CREATE_DATE DESC;  

Steps to Reproduce

1. Prepare a Confluence instance with a lot of inline tasks and pages
2. Create a page and add a task report macro sorted by "assignee"
3. Enable SQL logging in Confluence
4. Open a page with task report macro
5. Retrieve the SQL query from logs and check its execution plan (by using DB tools)

Expected Results

SQL query does not perform full scan on cwd_user and user_mapping tables

Actual Results

SQL query performs full scan on either cwd_user or user_mapping tables (or both)

Workaround

Do not sort tasks by "assignee".