elasticsearch/docs/reference/analysis/analyzers/pattern-analyzer.asciidoc

[[analysis-pattern-analyzer]]
=== Pattern Analyzer

The `pattern` analyzer uses a regular expression to split the text into terms.
The regular expression should match the *token separators*  not the tokens
themselves. The regular expression defaults to `\W+` (or all non-word characters).

[WARNING]
.Beware of Pathological Regular Expressions
========================================

The pattern analyzer uses
http://docs.oracle.com/javase/8/docs/api/java/util/regex/Pattern.html[Java Regular Expressions].

A badly written regular expression could run very slowly or even throw a
StackOverflowError and cause the node it is running on to exit suddenly.

Read more about http://www.regular-expressions.info/catastrophic.html[pathological regular expressions and how to avoid them].

========================================

[float]
=== Example output

[source,js]
---------------------------
POST _analyze
{
  "analyzer": "pattern",
  "text": "The 2 QUICK Brown-Foxes jumped over the lazy dog's bone."
}
---------------------------
// CONSOLE

/////////////////////

[source,js]
----------------------------
{
  "tokens": [
    {
      "token": "the",
      "start_offset": 0,
      "end_offset": 3,
      "type": "word",
      "position": 0
    },
    {
      "token": "2",
      "start_offset": 4,
      "end_offset": 5,
      "type": "word",
      "position": 1
    },
    {
      "token": "quick",
      "start_offset": 6,
      "end_offset": 11,
      "type": "word",
      "position": 2
    },
    {
      "token": "brown",
      "start_offset": 12,
      "end_offset": 17,
      "type": "word",
      "position": 3
    },
    {
      "token": "foxes",
      "start_offset": 18,
      "end_offset": 23,
      "type": "word",
      "position": 4
    },
    {
      "token": "jumped",
      "start_offset": 24,
      "end_offset": 30,
      "type": "word",
      "position": 5
    },
    {
      "token": "over",
      "start_offset": 31,
      "end_offset": 35,
      "type": "word",
      "position": 6
    },
    {
      "token": "the",
      "start_offset": 36,
      "end_offset": 39,
      "type": "word",
      "position": 7
    },
    {
      "token": "lazy",
      "start_offset": 40,
      "end_offset": 44,
      "type": "word",
      "position": 8
    },
    {
      "token": "dog",
      "start_offset": 45,
      "end_offset": 48,
      "type": "word",
      "position": 9
    },
    {
      "token": "s",
      "start_offset": 49,
      "end_offset": 50,
      "type": "word",
      "position": 10
    },
    {
      "token": "bone",
      "start_offset": 51,
      "end_offset": 55,
      "type": "word",
      "position": 11
    }
  ]
}
----------------------------
// TESTRESPONSE

/////////////////////


The above sentence would produce the following terms:

[source,text]
---------------------------
[ the, 2, quick, brown, foxes, jumped, over, the, lazy, dog, s, bone ]
---------------------------

[float]
=== Configuration

The `pattern` analyzer accepts the following parameters:

[horizontal]
`pattern`::

    A http://docs.oracle.com/javase/8/docs/api/java/util/regex/Pattern.html[Java regular expression], defaults to `\W+`.

`flags`::

    Java regular expression http://docs.oracle.com/javase/8/docs/api/java/util/regex/Pattern.html#field.summary[flags].
    Flags should be pipe-separated, eg `"CASE_INSENSITIVE|COMMENTS"`.

`lowercase`::

    Should terms be lowercased or not. Defaults to `true`.

`stopwords`::

    A pre-defined stop words list like `_english_` or an array  containing a
    list of stop words.  Defaults to `\_none_`.

`stopwords_path`::

    The path to a file containing stop words.

See the <<analysis-stop-tokenfilter,Stop Token Filter>> for more information
about stop word configuration.


[float]
=== Example configuration

In this example, we configure the `pattern` analyzer to split email addresses
on non-word characters or on underscores (`\W|_`), and to lower-case the result:

[source,js]
----------------------------
PUT my_index
{
  "settings": {
    "analysis": {
      "analyzer": {
        "my_email_analyzer": {
          "type":      "pattern",
          "pattern":   "\\W|_", <1>
          "lowercase": true
        }
      }
    }
  }
}

POST my_index/_analyze
{
  "analyzer": "my_email_analyzer",
  "text": "John_Smith@foo-bar.com"
}
----------------------------
// CONSOLE

<1> The backslashes in the pattern need to be escaped when specifying the
    pattern as a JSON string.

/////////////////////

[source,js]
----------------------------
{
  "tokens": [
    {
      "token": "john",
      "start_offset": 0,
      "end_offset": 4,
      "type": "word",
      "position": 0
    },
    {
      "token": "smith",
      "start_offset": 5,
      "end_offset": 10,
      "type": "word",
      "position": 1
    },
    {
      "token": "foo",
      "start_offset": 11,
      "end_offset": 14,
      "type": "word",
      "position": 2
    },
    {
      "token": "bar",
      "start_offset": 15,
      "end_offset": 18,
      "type": "word",
      "position": 3
    },
    {
      "token": "com",
      "start_offset": 19,
      "end_offset": 22,
      "type": "word",
      "position": 4
    }
  ]
}
----------------------------
// TESTRESPONSE

/////////////////////


The above example produces the following terms:

[source,text]
---------------------------
[ john, smith, foo, bar, com ]
---------------------------

[float]
==== CamelCase tokenizer

The following more complicated example splits CamelCase text into tokens:

[source,js]
--------------------------------------------------
PUT my_index
{
  "settings": {
    "analysis": {
      "analyzer": {
        "camel": {
          "type": "pattern",
          "pattern": "([^\\p{L}\\d]+)|(?<=\\D)(?=\\d)|(?<=\\d)(?=\\D)|(?<=[\\p{L}&&[^\\p{Lu}]])(?=\\p{Lu})|(?<=\\p{Lu})(?=\\p{Lu}[\\p{L}&&[^\\p{Lu}]])"
        }
      }
    }
  }
}

GET my_index/_analyze
{
  "analyzer": "camel",
  "text": "MooseX::FTPClass2_beta"
}
--------------------------------------------------
// CONSOLE

/////////////////////

[source,js]
----------------------------
{
  "tokens": [
    {
      "token": "moose",
      "start_offset": 0,
      "end_offset": 5,
      "type": "word",
      "position": 0
    },
    {
      "token": "x",
      "start_offset": 5,
      "end_offset": 6,
      "type": "word",
      "position": 1
    },
    {
      "token": "ftp",
      "start_offset": 8,
      "end_offset": 11,
      "type": "word",
      "position": 2
    },
    {
      "token": "class",
      "start_offset": 11,
      "end_offset": 16,
      "type": "word",
      "position": 3
    },
    {
      "token": "2",
      "start_offset": 16,
      "end_offset": 17,
      "type": "word",
      "position": 4
    },
    {
      "token": "beta",
      "start_offset": 18,
      "end_offset": 22,
      "type": "word",
      "position": 5
    }
  ]
}
----------------------------
// TESTRESPONSE

/////////////////////


The above example produces the following terms:

[source,text]
---------------------------
[ moose, x, ftp, class, 2, beta ]
---------------------------

The regex above is easier to understand as:

[source,regex]
--------------------------------------------------
  ([^\p{L}\d]+)                 # swallow non letters and numbers,
| (?<=\D)(?=\d)                 # or non-number followed by number,
| (?<=\d)(?=\D)                 # or number followed by non-number,
| (?<=[ \p{L} && [^\p{Lu}]])    # or lower case
  (?=\p{Lu})                    #   followed by upper case,
| (?<=\p{Lu})                   # or upper case
  (?=\p{Lu}                     #   followed by upper case
    [\p{L}&&[^\p{Lu}]]          #   then lower case
  )
--------------------------------------------------

[float]
=== Definition

The `pattern` anlayzer consists of:

Tokenizer::
* <<analysis-pattern-tokenizer,Pattern Tokenizer>>

Token Filters::
*  <<analysis-lowercase-tokenfilter,Lower Case Token Filter>>
*  <<analysis-stop-tokenfilter,Stop Token Filter>> (disabled by default)

If you need to customize the `pattern` analyzer beyond the configuration
parameters then you need to recreate it as a `custom` analyzer and modify
it, usually by adding token filters. This would recreate the built-in
`pattern` analyzer and you can use it as a starting point for further
customization:

[source,js]
----------------------------------------------------
PUT /pattern_example
{
  "settings": {
    "analysis": {
      "tokenizer": {
        "split_on_non_word": {
          "type":       "pattern",
          "pattern":    "\\W+" <1>
        }
      },
      "analyzer": {
        "rebuilt_pattern": {
          "tokenizer": "split_on_non_word",
          "filter": [
            "lowercase"       <2>
          ]
        }
      }
    }
  }
}
----------------------------------------------------
// CONSOLE
// TEST[s/\n$/\nstartyaml\n  - compare_analyzers: {index: pattern_example, first: pattern, second: rebuilt_pattern}\nendyaml\n/]
<1> The default pattern is `\W+` which splits on non-word characters
and this is where you'd change it.
<2> You'd add other token filters after `lowercase`.