I want to support queries mixing exact (phrase) and more lenient matching without parsing the user query myself. The simple_query_string query with its quote_field_suffix parameter seemed like a good candidate for this. However, I noticed that unlike the query_string query it does not support passing a quote_analyzer, so I can only supply a single analyzer that is applied to both, the phrase subqueries and the simple term subqueries. This is not helpful as I would like to apply some stemming and character removal on the inexact part(s) of the query while I'd like to preserve the exact tokens for the phrase match part(s).
How is the simple_query_string query supposed to work if you can define a different field (using quote_field_suffix) but not a different (search) analyzer for phrase subqueries?