Change long to float for Weight #1087
Conversation
|
ty for bringing this to our attention @virgilxie this PR has some issues though The following unit tests fail: We can't change the property in 1.x releases since it would break backwards compatibility this PR will need to target our and lastly can you sign the elasticsearch cla: http://www.elasticsearch.org/contributor-agreement/ so that we may pull your bits in ? 😄 Please let us know if you would like to take this on or would like us to do so (we'll still credit you in our changelog) |
|
Hi Martijn, Thank you so much for your reply. I have signed CLA. About the test (FunctionScoreQueryWithJustWeight and etc) : I don't really understand why it's will break the backwards compatibility Thank you for your great work on nest client! 2014-12-05 6:09 GMT-08:00 Martijn Laarman [email protected]:
Fei (Virgil) Xie Software Engineer @zazzle Alumni, Tsinghua University, Carnegie Mellon University |
Enable Functions in FunctionScoreQueryDescriptor accept FunctionScoreFunction<T>[] functions. So when you add multiple functions with different filters, you don't need to have FieldValueFactor for each of them. In the other words, I think we can only make the following query with this modification. Let me know if I was wrong.
"query": {
"function_score": {
"filter": {
"term": { "city": "Barcelona" }
},
"functions": [
{
"field_value_factor": {
"field": "score_US"
}
},
{
"filter": { "term": { "features": "wifi" }},
"weight": 1
},
{
"filter": { "term": { "features": "garden" }},
"weight": 1
},
{
"filter": { "term": { "features": "pool" }},
"weight": 2
}
],
"score_mode": "sum",
}
}
|
@virgilxie it's not a backwards compatibility issue with Elasticsearch, but with NEST. By changing the property from |
Elasitcsearch itself accepts float type weight