ICU_NORMALIZE - Firebolt Documentation

Transliterate a string using a specified ICU transliterate ID.

Syntax

ICU_NORMALIZE(<expression>, <transliterate_id>)

Parameter	Description	Supported input types
`<expression>`	An input string to transliterate.	`TEXT`
`<transliterate_id>`	A valid ICU library transliterate ID string.	`TEXT`

The ICU_NORMALIZE function returns a result of type TEXT.

If <transliterate_id> is invalid, an error is thrown.

The following example normalizes the word ‘München’ using the Latin-ASCII transliterate ID:

SELECT ICU_NORMALIZE('München', 'Latin-ASCII') AS normalized_word;

normalized_word text
Munchen

Rows: 1Execution time: 5.18ms

The following example applies a similar operation to the UPPER function:

SELECT ICU_NORMALIZE('AabB', 'Any-Upper[^xyzXYZ]') AS normalized_word;

normalized_word text
AABB

Rows: 1Execution time: 6.03ms

The function only works for a valid ICU transliterate ID:

SELECT ICU_NORMALIZE('', 'X');

Fails with the following error:

Line 1, Column 8: Unsupported locale: X
select ICU_NORMALIZE('','X')
       ^

⌘I