Skip to main content

SUBSTRING_INDEX

Description

The SUBSTRING_INDEX function is used to extract a portion of a string based on a specified delimiter. By specifying the number of delimiter occurrences, extraction can be performed from either the left or right side.

Syntax

SUBSTRING_INDEX(<content>, <delimiter>, <field>)

Parameters

ParameterDescription
<content>The source string to extract from. Type: VARCHAR
<delimiter>The delimiter string, case-sensitive and supports multi-byte characters. Type: VARCHAR
<field>Number of delimiter occurrences. Positive counts from left, negative counts from right, 0 returns empty string. Type: INT

Return Value

Returns VARCHAR type, representing the extracted substring.

Extraction rules:

  • When field > 0: returns content before the field-th delimiter from the left
  • When field < 0: returns content after the |field|-th delimiter from the right
  • When field = 0: returns empty string (returns NULL when content is NULL)
  • Case-sensitive exact delimiter matching

Special cases:

  • If any parameter is NULL, returns NULL
  • If delimiter doesn't exist in string, returns original string
  • If specified count exceeds actual delimiter occurrences, returns maximum extractable portion
  • If delimiter is empty string, returns empty string
  • If source string is empty, returns empty string

Examples

  1. Basic left extraction
SELECT SUBSTRING_INDEX('hello world', ' ', 1), SUBSTRING_INDEX('one,two,three', ',', 2);
+----------------------------------------+------------------------------------------+
| SUBSTRING_INDEX('hello world', ' ', 1) | SUBSTRING_INDEX('one,two,three', ',', 2) |
+----------------------------------------+------------------------------------------+
| hello | one,two |
+----------------------------------------+------------------------------------------+
  1. Right extraction (negative count)
SELECT SUBSTRING_INDEX('hello world', ' ', -1), SUBSTRING_INDEX('one,two,three', ',', -1);
+-----------------------------------------+-------------------------------------------+
| SUBSTRING_INDEX('hello world', ' ', -1) | SUBSTRING_INDEX('one,two,three', ',', -1) |
+-----------------------------------------+-------------------------------------------+
| world | three |
+-----------------------------------------+-------------------------------------------+
  1. NULL value handling
SELECT SUBSTRING_INDEX(NULL, ',', 1), SUBSTRING_INDEX('test', NULL, 1);
+--------------------------------+------------------------------------+
| SUBSTRING_INDEX(NULL, ',', 1) | SUBSTRING_INDEX('test', NULL, 1) |
+--------------------------------+------------------------------------+
| NULL | NULL |
+--------------------------------+------------------------------------+
  1. Zero count handling
SELECT SUBSTRING_INDEX('hello world', ' ', 0), SUBSTRING_INDEX('a,b,c', ',', 0);
+----------------------------------------+----------------------------------+
| SUBSTRING_INDEX('hello world', ' ', 0) | SUBSTRING_INDEX('a,b,c', ',', 0) |
+----------------------------------------+----------------------------------+
| | |
+----------------------------------------+----------------------------------+
  1. Delimiter doesn't exist
SELECT SUBSTRING_INDEX('hello world', ',', 1), SUBSTRING_INDEX('no-delimiter', '|', -1);
+----------------------------------------+------------------------------------------+
| SUBSTRING_INDEX('hello world', ',', 1) | SUBSTRING_INDEX('no-delimiter', '|', -1) |
+----------------------------------------+------------------------------------------+
| hello world | no-delimiter |
+----------------------------------------+------------------------------------------+
  1. Count exceeds delimiter occurrences
SELECT SUBSTRING_INDEX('a,b,c', ',', 5), SUBSTRING_INDEX('a,b,c', ',', -5);
+----------------------------------+-----------------------------------+
| SUBSTRING_INDEX('a,b,c', ',', 5) | SUBSTRING_INDEX('a,b,c', ',', -5) |
+----------------------------------+-----------------------------------+
| a,b,c | a,b,c |
+----------------------------------+-----------------------------------+
  1. UTF-8 multi-byte character delimiter
SELECT SUBSTRING_INDEX('ṭṛì→ḍḍumai→hello', '→', 1), SUBSTRING_INDEX('ṭṛì→ḍḍumai→hello', '→', -1);
+-----------------------------------------------+------------------------------------------------+
| SUBSTRING_INDEX('ṭṛì→ḍḍumai→hello', '→', 1) | SUBSTRING_INDEX('ṭṛì→ḍḍumai→hello', '→', -1) |
+-----------------------------------------------+------------------------------------------------+
| ṭṛì | hello |
+-----------------------------------------------+------------------------------------------------+
  1. Multi-character delimiter
SELECT SUBSTRING_INDEX('data::field::value', '::', 2), SUBSTRING_INDEX('data::field::value', '::', -1);
+---------------------------------------------+----------------------------------------------+
| SUBSTRING_INDEX('data::field::value', '::', 2) | SUBSTRING_INDEX('data::field::value', '::', -1) |
+---------------------------------------------+----------------------------------------------+
| data::field | value |
+---------------------------------------------+----------------------------------------------+
  1. Empty source string
SELECT SUBSTRING_INDEX('', ' ', 1);
+-----------------------------+
| SUBSTRING_INDEX('', ' ', 1) |
+-----------------------------+
| |
+-----------------------------+

Keywords

SUBSTRING_INDEX, SUBSTRING