OpenSearch/docs/reference/sql/functions/like-rlike.asciidoc

105 lines
4.5 KiB
Plaintext

[role="xpack"]
[testenv="basic"]
[[sql-like-rlike-operators]]
=== LIKE and RLIKE Operators
`LIKE` and `RLIKE` operators are commonly used to filter data based on string patterns. They usually act on a field placed on the left-hand side of
the operator, but can also act on a constant (literal) expression. The right-hand side of the operator represents the pattern.
Both can be used in the `WHERE` clause of the `SELECT` statement, but `LIKE` can also be used in other places, such as defining an
<<sql-index-patterns, index pattern>> or across various <<sql-commands, SHOW commands>>.
This section covers only the `SELECT ... WHERE ...` usage.
NOTE: One significant difference between `LIKE`/`RLIKE` and the <<sql-functions-search, full-text search predicates>> is that the former
act on <<sql-multi-field, exact fields>> while the latter also work on <<text, analyzed>> fields. If the field used with `LIKE`/`RLIKE` doesn't
have an exact not-normalized sub-field (of <<keyword, keyword>> type) {es-sql} will not be able to run the query. If the field is either exact
or has an exact sub-field, it will use it as is, or it will automatically use the exact sub-field even if it wasn't explicitly specified in the statement.
[[sql-like-operator]]
==== `LIKE`
.Synopsis:
[source, sql]
--------------------------------------------------
expression <1>
LIKE constant_exp <2>
--------------------------------------------------
<1> typically a field, or a constant expression
<2> pattern
.Description:
The SQL `LIKE` operator is used to compare a value to similar values using wildcard operators. There are two wildcards used in conjunction
with the `LIKE` operator:
* The percent sign (%)
* The underscore (_)
The percent sign represents zero, one or multiple characters. The underscore represents a single number or character. These symbols can be
used in combinations.
[source, sql]
----
include-tagged::{sql-specs}/docs/docs.csv-spec[simpleLike]
----
There is, also, the possibility of using an escape character if one needs to match the wildcard characters themselves. This can be done
by using the `ESCAPE [escape_character]` statement after the `LIKE ...` operator:
SELECT name, author FROM library WHERE name LIKE 'Dune/%' ESCAPE '/';
In the example above `/` is defined as an escape character which needs to be placed before the `%` or `_` characters if one needs to
match those characters in the pattern specifically. By default, there is no escape character defined.
IMPORTANT: Even though `LIKE` is a valid option when searching or filtering in {es-sql}, full-text search predicates
`MATCH` and `QUERY` are <<sql-like-prefer-full-text, faster and much more powerful and are the preferred alternative>>.
[[sql-rlike-operator]]
==== `RLIKE`
.Synopsis:
[source, sql]
--------------------------------------------------
expression <1>
RLIKE constant_exp <2>
--------------------------------------------------
<1> typically a field, or a constant expression
<2> pattern
.Description:
This operator is similar to `LIKE`, but the user is not limited to search for a string based on a fixed pattern with the percent sign (`%`)
and underscore (`_`); the pattern in this case is a regular expression which allows the construction of more flexible patterns.
For more details about the regular expressions syntax, https://docs.oracle.com/en/java/javase/11/docs/api/java.base/java/util/regex/Pattern.html[Java's Pattern class javadoc]
is a good starting point.
[source, sql]
----
include-tagged::{sql-specs}/docs/docs.csv-spec[simpleRLike]
----
IMPORTANT: Even though `RLIKE` is a valid option when searching or filtering in {es-sql}, full-text search predicates
`MATCH` and `QUERY` are <<sql-like-prefer-full-text, faster and much more powerful and are the preferred alternative>>.
[[sql-like-prefer-full-text]]
==== Prefer full-text search predicates
When using `LIKE`/`RLIKE`, do consider using <<sql-functions-search, full-text search predicates>> which are faster, much more powerful
and offer the option of sorting by relevancy (results can be returned based on how well they matched).
For example:
[cols="<m,<m"]
|===
^s|LIKE/RLIKE ^s|QUERY/MATCH
|`foo LIKE 'bar'` |`MATCH(foo, 'bar')`
|`foo LIKE 'bar' AND tar LIKE 'goo'` |`MATCH('foo^2, tar^5', 'bar goo', 'operator=and')`
|`foo LIKE 'barr'` |`QUERY('foo: bar~')`
|`foo LIKE 'bar' AND tar LIKE 'goo'` |`QUERY('foo: bar AND tar: goo')`
|`foo RLIKE 'ba.*'` |`MATCH(foo, 'ba', 'fuzziness=AUTO:1,5')`
|`foo RLIKE 'b.{1}r'` |`MATCH(foo, 'br', 'fuzziness=1')`
|===