[FTS] UI generated mapping verbosity / "where's the beef?"

Converting email from Will to JIRA ticket...

...UI creates and passes very verbose JSON. It’s hard for a user to see the parts that are important amid the settings that presumably are the same as the cbft defaults. Those could be passed if (1) they are set explicitly or (2) are needed for some reason, like there’s no way to default them.

I’m guessing they are built like this to simplify generation from the UI and parsing. But given that index mapping is already confusing, having to pass, for example, an empty “analysis” structure just makes it hard to find “the beef”, in the old Wendy’s commercial sense of the term.

Open to hearing feedback on this, I’m not sure if this is something we can do relatively easily with a proper formal definition or whether it would be better to think about what a complete index mapping redesign would look like after the DP goes out.

{
"mapping": {
"types": {
"airline": {
"enabled": true,
"dynamic": false,
"fields": [],
"properties": {
"name": {
"enabled": true,
"dynamic": false,
"properties": {},
"fields": [

{ "name": "name", "type": "text", "analyzer": "simple", "store": false, "index": true, "include_term_vectors": true, "include_in_all": true, "date_format": null, "display_order": "0" }

]
}
},
"display_order": "0"
}
},
"default_mapping": {
"enabled": false,
"dynamic": true,
"fields": [],
"default_analyzer": "en",
"properties": {},
"display_order": "1"
},
"type_field": "_type",
"default_type": "_default",
"default_analyzer": "standard",
"default_datetime_parser": "dateTimeOptional",
"default_field": "_all",
"byte_array_converter": "json",
"analysis": {
"analyzers": {},
"char_filters": {},
"tokenizers": {},
"token_filters": {},
"token_maps": {}
}
},
"store":

{ "kvStoreName": "forestdb" }

}