How to merge the result of multiple SphinxQL queries, grouped by key?

Question

MySQL

On the Russian forum of the sphinx, the question hung for two weeks without an answer. I'll try to ask here.
There are two requests:

SELECT *, (10*weight()) AS qweight FROM elverIndex WHERE MATCH('ZONESPAN:(h1,h2)
ТоЧтоЯИщу') LIMIT 0,1000

and

SELECT *, (5*weight()) AS qweight FROM elverIndex WHERE MATCH('ZONESPAN:(h3,h4)
ТоЧтоЯИщу') LIMIT 0,1000

Separately, they work correctly. I need to combine their result in one query
with grouping by id_attr field, summing qweight and sorting in descending order of
total weight.
I do:

SELECT *, tweight AS SUM(`qweight`) FROM (SELECT *, (10*weight()) AS qweight FROM
elverIndex WHERE MATCH('ZONESPAN:(h1,h2) ТоЧтоЯИщу') LIMIT 0,1000 UNION SELECT *,
(5*weight()) AS qweight FROM elverIndex WHERE MATCH('ZONESPAN:(h3,h4) ТоЧтоЯИщу') LIMIT
0,1000) t GROUP BY `id_attr` ORDER BY `tweight` DESC LIMIT 0,1000;

At the output I get

ERROR 1064 (42000): sphinxql: syntax error, unexpected '(', expecting FROM or ',' near
'(`qweight`) FROM (SELECT *, (10*weight()).......

The question of what I'm doing wrong, in general, is not worth it (the obvious answer is everything). The question is how
to do it right or at least where to read about how to do it right?

Answer 1

tweight AS SUM(`qweight`)

Answer 2

SphinxQL does not know how to work with subqueries. I myself faced a similar problem, I had to implement 2 separate requests, and then process them manually.