Details
-
Bug
-
Resolution: Fixed
-
Minor
-
5.0.0
-
Untriaged
-
No
Description
Build : 4.7.0-971
Test : (not yet merged)
./testrunner -i resources/jenkins/fts.ini -t fts.stable_topology_fts.StableTopFTS.index_query_custom_mapping,items=1000,custom_map=True,num_custom_analyzers=1,multiple_filters=true,cm_id=0,num_queries=100,compare_es=True,GROUP=P0,cluster=D+F
Description : For the above test, there are 3 queries that return different hits in case of FTS vs ES in case the custom analyzer uses an edge_ngram token filter with back=true,min=3 and max=5 params. There are two cases here –
1) 2 queries using match_phrase. It seems that FTS does not correctly use edge_ngram with match_phrase type of queries. This looks like a bug to me in FTS.
2) 1 query which has should, must_not and must Boolean queries. I think FTS is correct here in not hitting a few documents.
Case 1 : Match Phrase with back_edge_ngram (min 3, max 5)
Query :
{"field": "name", "match_phrase": "Sidonia Gerónimo"}Doc not returned by FTS, but returned by ES :
{
|
"salary": 60402.95, |
"name": "Antonia Gerónimo", |
"dept": "Info-tech", |
"is_manager": false, |
"mutated": 0, |
"join_date": "1971-02-11T06:57:00", |
"languages_known": [ |
"Hindi", |
"Italian", |
"Spanish" |
],
|
"emp_id": "10000318", |
"type": "emp", |
"email": "antonia@mcdiabetes.com" |
}
|
Case 2: Use of should, must_not and must Boolean queries.
Query :
{
"should":{"disjuncts": [
,
{"field": "name", "match": "Riona Adams"}]},
"must_not": {"disjuncts": [
,
{"field": "name", "match": "Juan José"}]},
"must": {"conjuncts": [
]}
}
Sample doc not returned by FTS, but returned by ES :
{
|
"salary": 91984.77, |
"name": "Keelia Carter", |
"mutated": 0, |
"is_manager": true, |
"dept": "Support", |
"join_date": "1952-01-24T10:20:00", |
"manages": { |
"team_size": 6, |
"reports": [ |
"Ciara Ann", |
"Kala Moore", |
"Severin Hall", |
"Deandra Aarón", |
"Mia Parker X", |
"Hedda Jr." |
]
|
},
|
"languages_known": [ |
"Arabic", |
"Vietnamese", |
"Portuguese" |
],
|
"emp_id": "10000120", |
"type": "emp", |
"email": "keelia@mcdiabetes.com" |
}
|
Attaching the testrunner console output which has the index definition as well the query results.
Attachments
Issue Links
- relates to
-
MB-22793 [FTS] queries should work when using the ngram and edge_ngram analyzers
- Closed