The fields.definitions
parameter looks like:
{
"field_name": {
"standard": "field_name",
"precise": "field_name.precise",
"reverse_precise": "field_name.reverse_precise",
"prefix_precise": "field_name.prefix"
}
}
Where field_name
is the name of the field you are defining.
All fields in the object under field_name
are optional. They default to
the format above but they are only used if the fields are actually defined in
the mapping. That means that if you define a field that "looks like" the
precise field or looks like the reverse_precise field it'll automatically be
where appropriate.
You can also add an ngram
to configure regular expression support. Note
that regular expressions still have to be turned on by configuring a
regex settings object. Just configuring a field without that
will do nothing. Here is an example for a field rigged for regular expression
matching:
{
"title": {
"ngram": {
"name": "title.trigram",
"gram_size": 3
}
}
}
name
is required. gram_size
defaults to 3.