Infix search sometimes produces unexpected results (or maybe I'm cooking it wrong). I have a string field, and I have two documents with the following values of this field:

"SO007462"

and

"002746"

. Infix is on in the schema. When I do an infix search by

"00746"

,

"002746"

has an extremely high score in comparison with

"SO007462"

. This is unexpected because

"SO*00746*2"

contains a direct match of

"00746"

in the middle while

"002746"

is a fuzzy match. `"SO007462"`:

"text_match": 100,
"text_match_info": {
	"best_field_score": "0",
	"best_field_weight": 12,
	"fields_matched": 4,
	"score": "100",
	"tokens_matched": 0
}

`"002746"`:

"text_match": 578730054645710969,
"text_match_info": {
	"best_field_score": "1108057784320",
	"best_field_weight": 15,
	"fields_matched": 1,
	"score": "578730054645710969",
	"tokens_matched": 1
}

The second issue is about the highlighting behavior. For

"002746"

, there is no highlighting at all, which is quite confusing for the user (something is found, but it is not clear. why it is found). For

"SO007462"

, it is the opposit; the whole value is highlighted, despite the fact that the match is in the middle:

"highlight": {
	"order_nbr": {
		"matched_tokens": [
			"SO007462"
		],
		"snippet": "<mark>SO007462</mark>",
		"value": "<mark>SO007462</mark>"
	}
},

I would expect that only the matching part is highlighted:

"SO<mark>00746</mark>2"