edismax query with edismax subqueries

[Danilo Tomasoni]

Hello,

I'm trying to perform the following query with edismax parser and
sub-queries of edismax type.

f.f3.qf=...

f.f4.qf=...

f.f5.qf=...


((_query_:{!edismax qf="abstracts abstract_background abstract_objective
abstract_methods abstract_results abstract_conclusions abstracts_names
abstract_background_names abstract_objective_names
abstract_methods_names abstract_results_names
abstract_conclusions_names" q.op=OR mm=1 v=*$subquery1*}) AND
(_query_:{!edismax qf="abstracts abstract_background abstract_objective
abstract_methods abstract_results abstract_conclusions abstracts_names
abstract_background_names abstract_objective_names
abstract_methods_names abstract_results_names
abstract_conclusions_names" q.op=OR mm=1 v=*$subquery2*}) AND
(f3:("human" "patient" "participant" "volunteer" "subject" "elderly"
"man" "men" "woman" "women") f4:("humans")) AND (f5:("diet" "dietary"
"intake" "oral" "consumption" "supplementation")))



It happens that if subquery1 is of type

'subquery1': '("butter oil" "dairy" "ghee" "butter" "processed cheese"
"cheese" "cream" "sour cream" "whipped cream" .... )'


and the list of OR'ed statements is < 1024 then solr outputs ~118K documents


otherwise, if subquery1 contains a list of OR'ed statements > 1024 (also
1025, also with non-unique words)

then solr outputs ~24MLN documents.


Can someone explain why this happens?

Thank you

Danilo