Identifiers
Applies to: 
Databricks SQL Databricks Runtime
An identifier is a string used to identify an object such as a table, view, schema, or column. Databricks supports non-delimited (regular) identifiers and delimited identifiers, which are enclosed within backticks.
Identifiers are case-insensitive when referenced.
For identifiers persisted with a metastore and data source the permitted characters permitted can be restricted.
See Names for details on the specific usage of identifiers.
Non delimited identifiers
Syntax
{ letter | '_' } [ letter | digit | '_' ] [ ... ]
Parameters
- letter: Any ASCII letter from
A-Zora-z. - digit: Any ASCII numeral from
0to9.
note
In Databricks Runtime, if spark.sql.ansi.enabled and spark.sql.ansi.enforceReservedKeywords are set to true, you cannot use an ANSI SQL reserved keyword as a non-delimited identifier. For details, see ANSI Compliance.
Delimited identifiers
Syntax
`c [ ... ]`
Parameters
- c: Any character from the Unicode character set. Use
`to escape`itself.
Examples
SQL
-- This statement fails because the undelimited identifier uses a non-ASCII letter.
> DESCRIBE SELECT 5 AS Ä;
INVALID_IDENTIFIER
-- You can delimit the identifier to use a non-ASCII letter
> DESCRIBE SELECT 5 AS `Ä`;
Ä
-- An identifier with dash needs to be delimited
> DESCRIBE SELECT 5 AS `a-b`;
a-b
-- An identifier with a space needs to be delimited
> DESCRIBE SELECT 5 AS `a b`;
a b
-- An identifier with a special character needs to be delimited
> DESCRIBE SELECT 5 AS `a@b`;
a@b
-- An identifier with a Chinese character needs to be delimited
> DESCRIBE SELECT 5 AS `a中b`;
a中b
-- An identifier with a backtick needs to be delimited and escaped.
> DESCRIBE SELECT 5 AS `a``b`;
a`b