String functions - e6data Docs

ReferenceApplies to: Unified engine deployments

String functions operate on text expressions. The common functions below return NULL for a NULL required argument unless the row documents a different behavior. Character counts and byte counts are different for multibyte text.

Case, length, and trimming

Function	Returns	Behavior
`lower(text)`, `lcase(text)`	STRING	Lowercase text
`upper(text)`, `ucase(text)`	STRING	Uppercase text
`character_length(text)`	Integer	Character count; `char_length` is an alias
`octet_length(text)`	Integer	Byte count
`trim(text)`	STRING	Remove surrounding trim characters, ordinarily spaces
`ltrim(text)`, `rtrim(text)`	STRING	Trim one side

SELECT c_customer_id, LOWER(c_first_name) AS lower_name,
       UPPER(c_last_name) AS upper_name,
       CHARACTER_LENGTH(c_first_name) AS name_length,
       LTRIM(CONCAT(' ', c_first_name)) AS trimmed_name
FROM demo.tpcds.customer
ORDER BY c_customer_sk
LIMIT 5;

The Native translation of two-sided TRIM currently emits an unavailable btrim function. The live example uses LTRIM.

Combining and extracting text

Function	Returns	Behavior
`concat(value, ...)`	STRING	Concatenate arguments; null behavior follows the resolved function
`concat_ws(separator, value, ...)`	STRING	Join values with a separator; null value arguments are skipped
`substr(text, start[, length])`	STRING	Substring using SQL’s one-based position
`replace(text, from, to)`	STRING	Replace matching text occurrences
`reverse(text)`	STRING	Reverse character order
`starts_with(text, prefix)`	BOOLEAN	Test a prefix
`ends_with(text, suffix)`	BOOLEAN	Test a suffix
`split_part(text, delimiter, position)`	STRING	Select a delimited text part

SELECT c_customer_id,
       CONCAT_WS(' ', c_first_name, c_last_name) AS full_name,
       SUBSTR(c_customer_id, 1, 4) AS code_prefix,
       REPLACE(c_customer_id, 'A', 'X') AS replaced_code
FROM demo.tpcds.customer
ORDER BY c_customer_sk
LIMIT 5;

Regular expressions

regexp_like(text, pattern) tests a regular expression. regexp_replace(text, pattern, replacement) substitutes matching text. Supported flags and replacement syntax depend on the resolved function and backend; validate patterns before moving a script from another dialect.

SELECT c_customer_id,
       REGEXP_LIKE(c_customer_id, '^AAAA') AS has_prefix,
       REGEXP_REPLACE(c_customer_id, 'A', 'x') AS changed_code
FROM demo.tpcds.customer
ORDER BY c_customer_sk
LIMIT 5;

Use LIKE for simple wildcard matches. The SQL operator SIMILAR TO is unavailable under Native execution because its translation would change semantics.

Last modified on November 4, 2026