RankQueryQueryFormatter generates incorrect JSON

[gkinsman]

NEST/Elasticsearch.Net version:
NEST 7.5.1

Elasticsearch version:
7.6.0

Description of the problem including expected versus actual behavior:
If no function is selected, then the JSON formatter will still add a comma after the field name:

elasticsearch-net/src/Nest/QueryDsl/Specialized/RankFeature/RankFeatureQuery.cs

Lines 189 to 206 in 453a608

	writer.WritePropertyName("field");
	var fieldFormatter = formatterResolver.GetFormatter<Field>();
	fieldFormatter.Serialize(ref writer, value.Field, formatterResolver);
	writer.WriteValueSeparator();
	switch (value.Function)
	{
	case IRankFeatureSigmoidFunction sigmoid:
	SerializeScoreFunction(ref writer, "sigmoid", sigmoid, formatterResolver);
	break;
	case IRankFeatureSaturationFunction saturation:
	SerializeScoreFunction(ref writer, "saturation", saturation, formatterResolver);
	break;
	case IRankFeatureLogarithmFunction log:
	SerializeScoreFunction(ref writer, "log", log, formatterResolver);
	break;
	}
	writer.WriteEndObject();

Steps to reproduce:

1. Create an index with a rank_feature field
2. Add some docs to the index with field values
3. Query for a single field without a Function parameter.
4. Query generated will look like: {"query":{"rank_feature":{"field":"field.prop",}}} - note the extra comma.

[russcam]

Thanks @gkinsman, will address this in the next release.