I have 2 types of solr query.One uses EDisMax parser and other one uses standard query parser as default.Following are the 2 query I have used.Purpose of the query is document that doesn't contains following word ELM,PST,June.
EDisMax Query
q= content:(-EML -PST -June) OR contentenstem:(-EML -PST -June)&start=0&rows=25 &defType=edismax&fl=content filename
It produces about 12232 results; In which 4292 results of docs contains any of the given query string that EML or PST or June.So 4292 result docs are wrong.
Standard parser Query
query:q=size:[ 0 TO * ]&fq=content:( -EML -PST -June )&fq=contentenstem:( -EML -PST -June )&start=0&rows=25&fl=content file name
It produces about 7940 result.
Why EDisMax query failed for this query;How to achieve the right result with use of EDisMax parser query
If we know how EDisMax and Standard query parser parses AND & OR Boolean query on the document, we may get the answer for why different doc count occurred for different parser.
Switch OR to AND in the EDisMax query. OR and AND act differently:
OR will cause results that meet one or both of your restrictions (for content and contentstem).
AND is stricter: results must meet both the restrictions.