This happens when using mapper-annotated-text highlighter type:
eg Query:
{
"query": {
"ids": {
"values": [
"1"
]
}
},
"_source": {
"includes": [
"test"
]
},
"highlight": {
"type": "annotated",
"highlight_query": {
"query_string": {
"query": "xxx",
"fields": [
"test"
]
}
},
"fields": {
"test": {
"number_of_fragments": 0,
"no_match_size": 2147483647
}
}
}
}
The response (when using highlight) skips the adjacent words, eg:
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 1,
"relation": "eq"
},
"max_score": 1.0,
"hits": [
{
"_index": "test_index",
"_type": "_doc",
"_id": "1",
"_score": 1.0,
"_source": {
"test": "These are the english entities [reduce](_ent_bioverb_reduce)[increase](_ent_bioverb_increase)[treat](_ent_bioverb_treat)[affect](_ent_bioverb_affect)"
},
"highlight": {
"test": [
"These are the english entities [reduce](_ent_bioverb_reduce)increase[treat](_ent_bioverb_treat)affect"
]
}
}
]
}
}
Note that highlight skipped "increase" and "affect".
If I separate the words by adding whitespace before querying, it works. This issue only happens when there is no space between entities