Compliance score query mixes columns

[carlosms]

This is the query for the Overview dashboard, Compliance score chart:

SELECT 
    SUM(changelog != 0) / COUNT(changelog)  as changelog, 
    SUM(readme != 0)  / COUNT(changelog) as readme, 
    SUM(notice != 0)  / COUNT(changelog) as notice, 
    SUM(license != 0) / COUNT(changelog) as license,
    SUM(contributing != 0) / COUNT(changelog) as contributing,
    SUM(travis != 0) / COUNT(changelog) as travis, 
    SUM(ci != 0) / COUNT(changelog) as ci, 
    SUM(dockerfile != 0) / COUNT(changelog) as dockerfile 
FROM (
    SELECT 
        SUM(changelog) as changelog, 
        SUM(readme) as readme, 
        SUM(notice) as notice, 
        SUM(license) as license,
        SUM(contributing) as contributing,
        SUM(travis) as travis,
        SUM(ci) as ci,
        SUM(dockerfile) as dockerfile
    FROM (
        SELECT 
            refs.repository_id, 
            CASE WHEN file_path LIKE '(?i)^CHANGELOG(|\.MD)' THEN 1 ELSE 0 END AS changelog,
            CASE WHEN file_path LIKE '(?i)^README(|\.MD)' THEN 1 ELSE 0 END AS readme,
            CASE WHEN file_path LIKE '(?i)^NOTICE(|\.MD)' THEN 1 ELSE 0 END AS notice,
            CASE WHEN file_path LIKE '(?i)^LICENSE(|\.MD)' THEN 1 ELSE 0 END AS license,
            CASE WHEN file_path LIKE '(?i)^CONTRIBUTING(|\.MD)' THEN 1 ELSE 0 END AS contributing,
            CASE WHEN file_path LIKE '(?i)^\.TRAVIS\.YML' THEN 1 ELSE 0 END AS travis,
            CASE WHEN file_path LIKE '(?i)^CI' THEN 1 ELSE 0 END AS ci,
            CASE WHEN file_path LIKE '(?i)^DOCKERFILE$' THEN 1 ELSE 0 END AS dockerfile
        FROM refs
        JOIN commit_files 
        ON refs.commit_hash = commit_files.commit_hash 
        WHERE ref_name = 'refs/heads/master' 
    ) q
    GROUP BY repository_id
) q2

I assume these lines:

    SUM(changelog != 0) / COUNT(changelog)  as changelog, 
    SUM(readme != 0)  / COUNT(changelog) as readme, 
    SUM(notice != 0)  / COUNT(changelog) as notice, 
....

Should be like this instead:

    SUM(changelog != 0) / COUNT(changelog)  as changelog, 
    SUM(readme != 0)  / COUNT(readme) as readme, 
    SUM(notice != 0)  / COUNT(notice) as notice, 
...

[se7entyse7en]

Since it just COUNTs I don't think it makes any difference in the end, but for sure for clarity, it should be as you said.