Include similarity in index settings (#2924) (#2966)

Non binary breaking back port of 648c3cf

Similarity settings are nested under settings>index
Add ClassicSimilarity
Add DFISimilarity
Add similarity settings to create index integration tests
Add XML comments for each similarity
Refactor SimilarityJsonConverter to not look up the NEST ISimilarity type with Type.GetType() but to infer it from the type property value directly.
Add Similarity to the updatable index settings

Closes #2890

(cherry picked from commit bf78582)

Author: russcam

docs/aggregations/writing-aggregations.asciidoc

@@ -172,6 +172,7 @@ return s => s
     );
 ----
 <1> a list of aggregation functions to apply
+
 <2> Using LINQ's `Aggregate()` function to accumulate/apply all of the aggregation functions
 
 Combining multiple `AggregationDescriptor` is also possible using the bitwise `&&` operator
@@ -237,5 +238,6 @@ var maxPerChild = childAggregation.Max("max_per_child");
 maxPerChild.Should().NotBeNull(); <2>
 ----
 <1> Do something with the average per child. Here we just assert it's not null
+
 <2> Do something with the max per child. Here we just assert it's not null
 

docs/client-concepts/certificates/working-with-certificates.asciidoc

@@ -144,8 +144,11 @@ public class PkiCluster : CertgenCaCluster
 }
 ----
 <1> Set the client certificate on `ConnectionSettings`
+
 <2> The path to the `.cer` file
+
 <3> The path to the `.key` file
+
 <4> The password for the private key
 
 Or per request on `RequestConfiguration` which will take precedence over the ones defined on `ConnectionConfiguration`

docs/client-concepts/connection-pooling/exceptions/unexpected-exceptions.asciidoc

@@ -60,8 +60,11 @@ audit = await audit.TraceUnexpectedException(
 );
 ----
 <1> set up a cluster with 10 nodes
+
 <2> where node 2 on port 9201 always throws an exception
+
 <3> The first call to 9200 returns a healthy response
+
 <4> ...but the second call, to 9201, returns a bad response
 
 Sometimes, an unexpected exception happens further down in the pipeline. In this scenario, we
@@ -100,7 +103,9 @@ audit = await audit.TraceUnexpectedException(
 );
 ----
 <1> calls on 9200 set up to throw a `WebException`
+
 <2> calls on 9201 set up to throw an `Exception`
+
 <3> Assert that the audit trail for the client call includes the bad response from 9200 and 9201
 
 An unexpected hard exception on ping and sniff is something we *do* try to recover from and failover to retrying on the next node.
@@ -145,6 +150,8 @@ audit = await audit.TraceUnexpectedException(
 );
 ----
 <1> `InnerException` is the exception that brought the request down
+
 <2> The hard exception that happened on ping is still available though
+
 <3> An exception can be hard to relate back to a point in time, so the exception is also available on the audit trail
 

docs/client-concepts/connection-pooling/exceptions/unrecoverable-exceptions.asciidoc

@@ -68,6 +68,7 @@ var audit = new Auditor(() => Framework.Cluster
 );
 ----
 <1> Always succeed on ping
+
 <2> ...but always fail on calls with a 401 Bad Authentication response
 
 Now, let's make a client call. We'll see that the first audit event is a successful ping
@@ -88,7 +89,9 @@ audit = await audit.TraceElasticsearchException(
 );
 ----
 <1> First call results in a successful ping
+
 <2> Second call results in a bad response
+
 <3> The reason for the bad response is Bad Authentication
 
 When a bad authentication response occurs, the client does not attempt to deserialize the response body returned;
@@ -122,6 +125,7 @@ audit = await audit.TraceElasticsearchException(
 );
 ----
 <1> Always return a 401 bad response with a HTML response on client calls
+
 <2> Assert that the response body bytes are null
 
 Now in this example, by turning on `DisableDirectStreaming()` on `ConnectionSettings`, we see the same behaviour exhibited
@@ -154,5 +158,6 @@ audit = await audit.TraceElasticsearchException(
 );
 ----
 <1> Response bytes are set on the response
+
 <2> Assert that the response contains `"nginx/"`
 

docs/client-concepts/connection-pooling/request-overrides/disable-sniff-ping-per-request.asciidoc

@@ -67,8 +67,11 @@ audit = await audit.TraceCalls(
 );
 ----
 <1> disable sniffing
+
 <2> first call is a successful ping
+
 <3> sniff on startup call happens here, on the second call
+
 <4> No sniff on startup again
 
 Now, let's disable pinging on the request 
@@ -92,6 +95,7 @@ audit = await audit.TraceCall(
 );
 ----
 <1> disable ping
+
 <2> No ping after sniffing
 
 Finally, let's demonstrate disabling both sniff and ping on the request 
@@ -113,5 +117,6 @@ audit = await audit.TraceCall(
 );
 ----
 <1> diable ping and sniff
+
 <2> no ping or sniff before the call
 

docs/client-concepts/connection-pooling/round-robin/skip-dead-nodes.asciidoc

@@ -142,7 +142,9 @@ await audit.TraceCalls(
 );
 ----
 <1> The first call goes to 9200 which succeeds
+
 <2> The 2nd call does a ping on 9201 because its used for the first time. It fails so we wrap over to node 9202
+
 <3> The next call goes to 9203 which fails so we should wrap over
 
 A cluster with 2 nodes where the second node fails on ping 
@@ -192,5 +194,6 @@ await audit.TraceCalls(
 );
 ----
 <1> All the calls fail
+
 <2> After all our registered nodes are marked dead we want to sample a single dead node each time to quickly see if the cluster is back up. We do not want to retry all 4 nodes
 

docs/client-concepts/high-level/analysis/writing-analyzers.asciidoc

@@ -102,6 +102,7 @@ var createIndexResponse = client.CreateIndex("my-index", c => c
 );
 ----
 <1> Pre-defined list of English stopwords within Elasticsearch
+
 <2> Use the `standard_english` analyzer configured
 
 [source,javascript]
@@ -267,6 +268,7 @@ var createIndexResponse = client.CreateIndex("questions", c => c
 );
 ----
 <1> Use an analyzer at index time that strips HTML tags
+
 <2> Use an analyzer at search time that does not strip HTML tags
 
 With this in place, the text of a question body will be analyzed with the `index_question` analyzer

docs/client-concepts/high-level/getting-started.asciidoc

@@ -109,6 +109,7 @@ var indexResponse = client.Index(person); <1>
 var asyncIndexResponse = await client.IndexAsync(person); <2>
 ----
 <1> synchronous method that returns an `IIndexResponse`
+
 <2> asynchronous method that returns a `Task<IIndexResponse>` that can be awaited
 
 NOTE: All methods available within NEST are exposed as both synchronous and asynchronous versions,

docs/client-concepts/high-level/inference/field-inference.asciidoc

@@ -473,9 +473,13 @@ class Precedence
 }
 ----
 <1> Even though this property has a NEST property mapping _and_ a `JsonProperty` attribute, We are going to provide a hard rename for it on ConnectionSettings later that should win.
+
 <2> This property has both a NEST attribute and a `JsonProperty`, NEST should win.
+
 <3> We should take the json property into account by itself
+
 <4> This property we are going to special case in our custom serializer to resolve to ask
+
 <5> We are going to register a DefaultFieldNameInferrer on ConnectionSettings that will uppercase all properties.
 
 Here we create a custom serializer that renames any property named `AskSerializer` to `ask`

docs/client-concepts/high-level/inference/indices-paths.asciidoc

@@ -90,6 +90,7 @@ singleIndexFromIndexName.Match(
 );
 ----
 <1> `_all` will override any specific index names here
+
 <2> The `Project` type has been mapped to a specific index name using <<index-name-type-mapping,`.InferMappingFor<Project>`>>
 
 [[nest-indices]]
@@ -120,7 +121,9 @@ ISearchRequest singleTypedRequest = new SearchDescriptor<Project>().Index(single
 var invalidSingleString = Nest.Indices.Index("name1, name2"); <3>
 ----
 <1> specifying a single index using a string
+
 <2> specifying a single index using a type
+
 <3> an **invalid** single index name
 
 ===== Multiple indices
@@ -145,7 +148,9 @@ manyStringRequest = new SearchDescriptor<Project>().Type(new[] { "name1", "name2
 ((IUrlParameter)manyStringRequest.Type).GetString(client.ConnectionSettings).Should().Be("name1,name2");
 ----
 <1> specifying multiple indices using strings
+
 <2> specifying multiple indices using types
+
 <3> The index names here come from the Connection Settings passed to `TestClient`. See the documentation on <<index-name-inference, Index Name Inference>> for more details.
 
 ===== All Indices