Move limit in getFilesAndIndicesForDimensions to time dimension

[mgrunbauer]

In #361, a PostgreSQL query in the function getFilesAndIndicesForDimensions was rewritten to this format:

select distinct t1.path,member,dimmember,height,dimheight,time,dimtime from
    t20240301t095153031_xkq9pp0nv3kqmonttvhf t1
inner join t20240301t095153038_hqug9msl1wgxpew4yozf t2
    ON t1.path = t2.path
inner join t20240301t095153043_oe2pyeutb8pdrvb7gu7r t3
    ON t1.path = t3.path
where
    member = 'member4' and
    t1.adaguctilinglevel != -1 and
    abs(5000 - height) = (select min(abs(5000 - height)) from t20240301t095153038_hqug9msl1wgxpew4yozf) and
    time = '2017-01-01T00:15:00Z'
order by member,height,time
limit 512

The original goal of having a limit at all was to prevent Adaguc from returning data for a very large time period. However the limit 512 now limits all the results from the query.

If we now perform an EDR cube call on a custom dimension (Ensemble data with members=* or statistical data with percentile=*), EDR will respond with a limited set.

The result of the query is time steps * custom dimension steps:

• With 144 timesteps, you can only query 3 percentiles, because 144 * 3 = 432, which fits inside 512.
• Querying 4 percentiles results in reduced timesteps, because 144 * 4 = 576, which is more than 512.

We should rewrite the query, for example by using a subquery on the time dimension, with the limit, in the inner join. Something like the following (untested):

select distinct t1.path,member,dimmember,height,dimheight,time,dimtime from
    t20240301t095153031_xkq9pp0nv3kqmonttvhf t1
inner join (SELECT * FROM t20240301t095153038_hqug9msl1wgxpew4yozf t2 WHERE ... LIMIT 512)
    ON t1.path = t2.path
inner join t20240301t095153043_oe2pyeutb8pdrvb7gu7r t3
    ON t1.path = t3.path
where
    member = 'member4' and
    t1.adaguctilinglevel != -1 and
    abs(5000 - height) = (select min(abs(5000 - height)) from t20240301t095153038_hqug9msl1wgxpew4yozf) and
    time = '2017-01-01T00:15:00Z'
order by member,height,time

Testing and verifying this functionality is easiest after custom dimension support is added to EDR cube.