STRING_AGG (Transact-SQL)
Applies to:
SQL Server 2017 (14.x) and later
Azure SQL Database
Azure SQL Managed Instance
Azure Synapse Analytics
SQL analytics endpoint in Microsoft Fabric
Warehouse in Microsoft Fabric
Concatenates the values of string expressions and places separator values between them. The separator isn't added at the end of string.
Transact-SQL syntax conventions
Syntax
STRING_AGG ( expression, separator ) [ <order_clause> ]
<order_clause> ::=
WITHIN GROUP ( ORDER BY <order_by_expression_list> [ ASC | DESC ] )
Note
To view Transact-SQL syntax for SQL Server 2014 (12.x) and earlier versions, see Previous versions documentation.
Arguments
expression
Is an expression of any type. Expressions are converted to NVARCHAR
or VARCHAR
types during concatenation. Non-string types are converted to NVARCHAR
type.
separator
Is an expression of NVARCHAR
or VARCHAR
type that is used as separator for concatenated strings. It can be literal or variable.
<order_clause>
Optionally specify order of concatenated results using WITHIN GROUP
clause:
WITHIN GROUP ( ORDER BY <order_by_expression_list> [ ASC | DESC ] )
<order_by_expression_list>
A list of non-constant expressions that can be used for sorting results. Only one order_by_expression
is allowed per query. The default sort order is ascending.
Return Types
Return type depends on first argument (expression). If input argument is string type (NVARCHAR
, VARCHAR
), result type will be same as input type. The following table lists automatic conversions:
Input expression type | Result |
---|---|
NVARCHAR(MAX) | NVARCHAR(MAX) |
VARCHAR(MAX) | VARCHAR(MAX) |
NVARCHAR(1...4000) | NVARCHAR(4000) |
VARCHAR(1...8000) | VARCHAR(8000) |
int, bigint, smallint, tinyint, numeric, float, real, bit, decimal, smallmoney, money, datetime, datetime2, | NVARCHAR(4000) |
Remarks
STRING_AGG
is an aggregate function that takes all expressions from rows and concatenates them into a single string. Expression values are implicitly converted to string types and then concatenated. The implicit conversion to strings follows the existing rules for data type conversions. For more information about data type conversions, see CAST and CONVERT (Transact-SQL).
If the input expression is type VARCHAR
, the separator cannot be type NVARCHAR
.
Null values are ignored and the corresponding separator is not added. To return a place holder for null values, use the ISNULL
function as demonstrated in example B.
STRING_AGG
is available in any compatibility level.
Note
<order_clause>
is available with database compatibility level 110 and above.
Examples
Most of the examples in this article reference the AdventureWorks sample databases.
A. Generate list of names separated in new lines
The following example produces a list of names in a single result cell, separated with carriage returns.
USE AdventureWorks2022;
GO
SELECT STRING_AGG (CONVERT(NVARCHAR(max),FirstName), CHAR(13)) AS csv
FROM Person.Person;
GO
Here is the result set.
csv |
---|
Syed Catherine Kim Kim Kim Hazem ... |
NULL
values found in name
cells aren't returned in the result.
Note
If using the SQL Server Management Studio Query Editor, the Results to Grid option cannot implement the carriage return. Switch to Results to Text to see the result set properly.
Results to Text are truncated to 256 characters by default. To increase this limit, change the Maximum number of characters displayed in each column option.
B. Generate list of names separated with comma without NULL values
The following example replaces null values with 'N/A' and returns the names separated by commas in a single result cell.
USE AdventureWorks2022;
GO
SELECT STRING_AGG(CONVERT(NVARCHAR(max), ISNULL(FirstName,'N/A')), ',') AS csv
FROM Person.Person;
GO
Here is the result set.
Note
Results are shown trimmed.
csv |
---|
Syed,Catherine,Kim,Kim,Kim,Hazem,Sam,Humberto,Gustavo,Pilar,Pilar, ... |
C. Generate comma-separated values
USE AdventureWorks2022;
GO
SELECT STRING_AGG(CONVERT(NVARCHAR(max), CONCAT(FirstName, ' ', LastName, '(', ModifiedDate, ')')), CHAR(13)) AS names
FROM Person.Person;
GO
Here is the result set.
Note
Results are shown trimmed.
names |
---|
Ken Sánchez (Feb 8 2003 12:00AM) Terri Duffy (Feb 24 2002 12:00AM) Roberto Tamburello (Dec 5 2001 12:00AM) Rob Walters (Dec 29 2001 12:00AM) ... |
Note
If using the Management Studio Query Editor, the Results to Grid option cannot implement the carriage return. Switch to Results to Text to see the result set properly.
D. Return news articles with related tags
Imagine a database where articles and their tags are separated into different tables. A developer wants to return one row per each article with all associated tags. The following query achieves this result:
SELECT a.articleId, title, STRING_AGG (tag, ',') as tags
FROM dbo.Article AS a
LEFT JOIN dbo.ArticleTag AS t
ON a.ArticleId = t.ArticleId
GROUP BY a.articleId, title;
GO
Here is the result set.
articleId | title | tags |
---|---|---|
172 | Polls indicate close election results | politics,polls,city council |
176 | New highway expected to reduce congestion | NULL |
177 | Dogs continue to be more popular than cats | polls,animals |
Note
The GROUP BY
clause is required if the STRING_AGG
function isn't the only item in the SELECT
list.
E. Generate list of emails per towns
The following query finds the email addresses of employees and groups them by city:
USE AdventureWorks2022;
GO
SELECT TOP 10 City, STRING_AGG(CONVERT(NVARCHAR(max), EmailAddress), ';') AS emails
FROM Person.BusinessEntityAddress AS BEA
INNER JOIN Person.Address AS A ON BEA.AddressID = A.AddressID
INNER JOIN Person.EmailAddress AS EA ON BEA.BusinessEntityID = EA.BusinessEntityID
GROUP BY City;
GO
Here is the result set.
Note
Results are shown trimmed.
City | emails |
---|---|
Ballard | paige28@adventure-works.com;joshua24@adventure-works.com;javier12@adventure-works.com;... |
Baltimore | gilbert9@adventure-works.com |
Barstow | kristen4@adventure-works.com |
Basingstoke Hants | dale10@adventure-works.com;heidi9@adventure-works.com |
Baytown | kelvin15@adventure-works.com |
Beaverton | billy6@adventure-works.com;dalton35@adventure-works.com;lawrence1@adventure-works.com;... |
Bell Gardens | christy8@adventure-works.com |
Bellevue | min0@adventure-works.com;gigi0@adventure-works.com;terry18@adventure-works.com;... |
Bellflower | philip0@adventure-works.com;emma34@adventure-works.com;jorge8@adventure-works.com;... |
Bellingham | christopher23@adventure-works.com;frederick7@adventure-works.com;omar0@adventure-works.com;... |
Emails returned in the emails column can be directly used to send emails to group of people working in some particular cities.
F. Generate a sorted list of emails per towns
Similar to the previous example, the following query finds the email addresses of employees, groups them by city, and sorts the emails alphabetically:
USE AdventureWorks2022;
GO
SELECT TOP 10 City, STRING_AGG(CONVERT(NVARCHAR(max), EmailAddress), ';') WITHIN GROUP (ORDER BY EmailAddress ASC) AS Emails
FROM Person.BusinessEntityAddress AS BEA
INNER JOIN Person.Address AS A ON BEA.AddressID = A.AddressID
INNER JOIN Person.EmailAddress AS EA ON BEA.BusinessEntityID = EA.BusinessEntityID
GROUP BY City;
GO
Here is the result set.
Note
Results are shown trimmed.
City | Emails |
---|---|
Barstow | kristen4@adventure-works.com |
Basingstoke Hants | dale10@adventure-works.com;heidi9@adventure-works.com |
Braintree | mindy20@adventure-works.com |
Bell Gardens | christy8@adventure-works.com |
Byron | louis37@adventure-works.com |
Bordeaux | ranjit0@adventure-works.com |
Carnation | don0@adventure-works.com;douglas0@adventure-works.com;george0@adventure-works.com;... |
Boulogne-Billancourt | allen12@adventure-works.com;bethany15@adventure-works.com;carl5@adventure-works.com;... |
Berkshire | barbara41@adventure-works.com;brenda4@adventure-works.com;carrie14@adventure-works.com;... |
Berks | adriana6@adventure-works.com;alisha13@adventure-works.com;arthur19@adventure-works.com;... |
Next steps
Learn more about Transact-SQL functions in the following articles: