OpenSearch/docs/reference/analysis/tokenizers/edgengram-tokenizer.asciidoc

81 lines
2.6 KiB
Plaintext

[[analysis-edgengram-tokenizer]]
=== Edge NGram Tokenizer
A tokenizer of type `edgeNGram`.
This tokenizer is very similar to `nGram` but only keeps n-grams which
start at the beginning of a token.
The following are settings that can be set for a `edgeNGram` tokenizer
type:
[cols="<,<,<",options="header",]
|=======================================================================
|Setting |Description |Default value
|`min_gram` |Minimum size in codepoints of a single n-gram |`1`.
|`max_gram` |Maximum size in codepoints of a single n-gram |`2`.
|`token_chars` | Characters classes to keep in the
tokens, Elasticsearch will split on characters that don't belong to any
of these classes. |`[]` (Keep all characters)
|=======================================================================
`token_chars` accepts the following character classes:
[horizontal]
`letter`:: for example `a`, `b`, `ï` or `京`
`digit`:: for example `3` or `7`
`whitespace`:: for example `" "` or `"\n"`
`punctuation`:: for example `!` or `"`
`symbol`:: for example `$` or `√`
[float]
==== Example
[source,js]
--------------------------------------------------
curl -XPUT 'localhost:9200/test' -d '
{
"settings" : {
"analysis" : {
"analyzer" : {
"my_edge_ngram_analyzer" : {
"tokenizer" : "my_edge_ngram_tokenizer"
}
},
"tokenizer" : {
"my_edge_ngram_tokenizer" : {
"type" : "edgeNGram",
"min_gram" : "2",
"max_gram" : "5",
"token_chars": [ "letter", "digit" ]
}
}
}
}
}'
curl 'localhost:9200/test/_analyze?pretty=1&analyzer=my_edge_ngram_analyzer' -d 'FC Schalke 04'
# FC, Sc, Sch, Scha, Schal, 04
--------------------------------------------------
[float]
==== `side` deprecated
There used to be a `side` parameter up to `0.90.1` but it is now deprecated. In
order to emulate the behavior of `"side" : "BACK"` a
<<analysis-reverse-tokenfilter,`reverse` token filter>> should be used together
with the <<analysis-edgengram-tokenfilter,`edgeNGram` token filter>>. The
`edgeNGram` filter must be enclosed in `reverse` filters like this:
[source,js]
--------------------------------------------------
"filter" : ["reverse", "edgeNGram", "reverse"]
--------------------------------------------------
which essentially reverses the token, builds front `EdgeNGrams` and reverses
the ngram again. This has the same effect as the previous `"side" : "BACK"` setting.