I cannot figure out how to search the documents in my index by the MatchPhrasePrefix. I want to match on the entire search term phrase, but allow prefixes on the last term.
This LINQ construction is giving me back everything with a score of 1.0. How would I construct this? On a side note, is there anyway to see the raw query that NEST is constructing? That would be very helpful!
var search = new SearchDescriptor<ObservationElasticModel>();
search = search
.Query(q => q.MatchPhrase(m => m.Query(term)));
var response = _client.Search<ObservationElasticModel>(search);
You can use the MatchPhrasePrefix method on the QueryDescriptor. You may also need to set the Operator to AND.
_client.Search<ObservationElasticModel>(s => s
.Query(q => q
.MatchPhrasePrefix(m => m
.Operator(Operator.And)
.Query(term)));
You can log raw queries to the debugging console by calling EnableTrace() and ExposeRawResponse() on the ConnectionSettings instance during setup.
var settings = new ConnectionSettings(new Uri(searchUri));
settings.EnableTrace();
settings.ExposeRawResponse();
_client = new ElasticClient(settings);
Related
I have a list of strings like this
"Users": [
"usrName|Fullname|False|0|False|False",
"usrName|Fullname|False|0|False|False",
"usrName|Fullname|False|0|False|False",
"usrName|Fullname|False|0|False|False",
"usrName|Fullname|False|0|False|False",
]
In my episerver/optimizely code I want to match items. I have written this line of code
searchResult.Filter(x => x.Users.MatchContained(k=> k.Split('|')[3], "0"));
I am trying to get all Users where after splitiing on 'pipe' I get 0 at index 3. I am not getting the result, infact I get an exception.
What am I doing wrong here?
Since you left out A LOT of details these are my assumptions
Users is an IList<string> Users
You are trying to get all Users within that list/array where index 3 equals 0
What we don't know is, among others
Is there more than one page instance that have the Users instance filled?
Anyway, this cannot be solved using any Find API with the current system design. Instead you need to rely on linq to parse the result, but then the Find implementation may not be necessary.
var searchClient = SearchClient.Instance;
var search = searchClient.Search<BlogPage>();
var result = search.GetContentResult();
var usersResult = result
.SelectMany(x => x.Users)
.Where(x => x.Split('|')[3].Equals("0"));
This would create and return an object similar to this, since I'm using SelectMany the array would contain all users throughout the systems where there are matched pages from the search result.
If you for whatever reason would like to keep or see some page properties there are alternative approaches where you construct a new object within a select and remove any object where there where not matched users
var searchClient = SearchClient.Instance;
var search = searchClient.Search<BlogPage>();
var result = search.GetContentResult();
var usersResult = result
.Select(p => new
{
PageName = p.Name,
ContentLink = p.ContentLink.ToString(),
Users = p.Users.Where(x => x.Split('|')[3].Equals("0"))
})
.Where(x => x.Users.Any());
If you like to keep using Find for this kind of implementations you must store the data in a better way than strings with delimiters.
If one is seraching several indexes at the same time, is there any way to say that if searching index A, then add this filter and if searching index B then add a different filter.
For example:
var filters = new List<Func<QueryContainerDescriptor<PropertySearchResult>, QueryContainer>>();
filters.Add(fq => fq.Term(t => t.Field(f => f.PromoterId).Value(user.Id)));
filters.Add(fq => fq.Term(t => t.Field(f => f.SubscriptionId).Value(subscriptionId)));
string indicies = String.Join(",", Utils.SupportedCountries.Select(c => c.Key.ToLower()).ToArray());
var result = await ElasticSearchConfig.GetClient().DeleteByQueryAsync<PropertySearchResult>(u => u
.Index(indicies)
.Query(q => q
.Bool(bq => bq.Filter(filters))));
at the moment, all indices will be subject to the same filters but I would like to vary the filters based on which index is being searched.
Add(with &&) a term query to each of your filters
.Term("_index", A)
Check this link
https://www.elastic.co/guide/en/elasticsearch/reference/current/mapping-index-field.html
I have this C# code which is expected to match 2 fields using multi-match Elastic Search type. I am using NEST package.
var response = await _elasticClient.SearchAsync<FileDocument>(
s => s.Query(q => q.MultiMatch(c => c
.Fields(f => f.Field(p => p.FileName).Field(query))
.Fields(f => f.Field(p => p.Metadata).Field(query))
)));
Problem is no matter what text I passed in, it returns all the result. Anything I miss out?
In order to efficiently debug these types of issue, you need to inspect the HTTP request going to Elasticsearch, ultimately your query builder will be converted to search JSON and will be executed against Elasticsearch.
I am not aware of nest but have written the answer for Java code, which prints the Elasticsearch query in JSON format.
Although my guess is that, you are not sending the correct HTTP method which should be POST and you might be sending it with GET, which is causing ES to ignore your search query and return all documents.
Solved after adding .Query(query)
var response = await _elasticClient.SearchAsync<FileDocument>(
s => s.Query(q => q.MultiMatch(c => c
.Fields(f => f.Field(p => p.FileName).Field(p=>p.Metadata))
.Query(query)
))
);
Reference - https://www.elastic.co/guide/en/elasticsearch/client/net-api/current/multi-match-usage.html
I have got unique document ids (across all types) and I would like to check which document already exists in elasticsearch index. I try to search
var duplicateCheck = _elasticClient
.MultiGet(m => m.GetMany<object>(notices.Select(s => s.Id)).Fields("Id"));
but it returns wrong result - every document has set found property to false.
update
there is workaround here
var exisitngDocIds = _elasticClient.Search<ZPBase>(s => s
.AllTypes()
.Query(q => q.Ids(notices.Select(z=>z.Id)))
.Fields("Id")
.Take(notices.Count)
);
notices = notices.Where(q => !exisitngDocIds.Hits.Any(s => s.Id == q.Id)).ToList();
From the Multi Get API documentation I realized that you can use something similar to the following code to solve your problem:
var response = _elasticClient.MultiGet(m => m
.Index(MyIndex)
.Type("")
.GetMany<ZPBase>(noticeIds));
Note the empty string passed as the Type.
I have a logstash/elasticsearch/kibana system set up and structured json type logs are getting into elastic search (not from a C# application) and visible in kibana.
I am using NEST because I would like to slice the data from logstash.
The following code is producing "hits" results that I can inspect in the debugger.
ElasticClient client = new ElasticClient(settings);
var searchResults = client.Search( s => s
.From(0)
.Size(100)
.AllIndices()
.SortDescending("#timestamp")
);
However, if I try to expand the search to include something I believe to be present in the log (visible in Kibana), I get now results.
var searchResults = client.Search( s => s
.From(0)
.Size(100)
.AllIndices()
.Query(q => q
.Term("MySpecialFieldName", "ValueThatAppears")
)
.SortDescending("#timestamp")
);
I would also like to take advantage of the type safety and other mechanisms shown in the samples. I am not sure if that is expected to be related.
( I am working on figuring that out separately: Adding a class like "client.Search( to the search seems to prevent the results as well, I am assuming that something about the class is not aligned with the data and is therefore unable to deserialize or is otherwise filtering...)
Found the correct settings to make this work:
The logstash system puts this into a type (log or logs I think) when it gets indexed by elastic search. Also the term needs to be suffixed with ".raw". So the working example ends up being:
var searchResults = client.Search<MyClass>( s => s
.From(0)
.Size(100)
.AllIndices()
.AllTypes()
.Query(q => q
.Term("MySpecialFieldName.raw", "ValueThatAppears")
)
.SortDescending("#timestamp")
);