Identifiers

Identifiers are used when referencing database objects.

Case-Sensitivity

Identifiers in a SQL query are case-insensitive.

All of the following queries behave exactly the same:

SELECT * FROM CITIES;
SELECT * FROM cities;
SELECT * FROM Cities;

Identifiers wrapped in a double quote (") retain their original casing.

Both of these queries are referencing different tables:

SELECT * FROM "Cities";
SELECT * FROM cities;

Case-insensitivity applies to all database objects, including functions.

Both queries are equivalent:

SELECT abs(-1.0);
SELECT ABS(-1.0);

When querying columns, unquoted identifiers match column names ignoring case. For example, if we have a parquet file with columns EmployeeName and Department, all of the below queries behave the same:

SELECT EmployeeName, Department FROM './employees.parquet';
SELECT employeename, department FROM './employees.parquet';
SELECT "EmployeeName", "Department" FROM './employees.parquet';

Note that quoting a column name forces case sensitive matching. Using the same parquet example, the below query will error:

-- Error: Missing column for reference: employeename
SELECT "employeename", "department" FROM './employees.parquet';

Generated Column Names

When no explicit alias is provided for an expression in the SELECT list, an column name will be generated using the following rules:

• If an expression is a column reference, the column name is used.
• If an expression is a function, the function name is used.
• If none of the above rules apply, the identifier ?column? is used.