How to format a numeric column as phone number in SQL - sql

Question

I have table in the database with a phone number column. The numbers look like this:
123456789
I want to format that to look like this:
123-456-789

Answer 1

This should do it:
UPDATE TheTable
SET PhoneNumber = SUBSTRING(PhoneNumber, 1, 3) + '-' +
SUBSTRING(PhoneNumber, 4, 3) + '-' +
SUBSTRING(PhoneNumber, 7, 4)
Incorporated Kane's suggestion, you can compute the phone number's formatting at runtime. One possible approach would be to use scalar functions for this purpose (works in SQL Server):
CREATE FUNCTION FormatPhoneNumber(#phoneNumber VARCHAR(10))
RETURNS VARCHAR(12)
BEGIN
RETURN SUBSTRING(#phoneNumber, 1, 3) + '-' +
SUBSTRING(#phoneNumber, 4, 3) + '-' +
SUBSTRING(#phoneNumber, 7, 4)
END

Answer 2

I'd generally recommend you leave the formatting up to your front-end code and just return the data as-is from SQL. However, to do it in SQL, I'd recommend you create a user-defined function to format it. Something like this:
CREATE FUNCTION [dbo].[fnFormatPhoneNumber](#PhoneNo VARCHAR(20))
RETURNS VARCHAR(25)
AS
BEGIN
DECLARE #Formatted VARCHAR(25)
IF (LEN(#PhoneNo) <> 10)
SET #Formatted = #PhoneNo
ELSE
SET #Formatted = LEFT(#PhoneNo, 3) + '-' + SUBSTRING(#PhoneNo, 4, 3) + '-' + SUBSTRING(#PhoneNo, 7, 4)
RETURN #Formatted
END
GO
Which you can then use like this:
SELECT [dbo].[fnFormatPhoneNumber](PhoneNumber) AS PhoneNumber
FROM SomeTable
It has a safeguard in, in case the phone number stored isn't the expected number of digits long, is blank, null etc - it won't error.
EDIT: Just clocked on you want to update your existing data. The main bit that's relevant from my answer then is that you need to protect against "dodgy"/incomplete data (i.e. what if some existing values are only 5 characters long)

Answer 3

Above users mentioned, those solutions are very basic and they won't work if the database has different phone formats like:
[phone]
123-456-4564
1234567989
etc
Here is a more complex solution that will work with ANY input given:
CREATE FUNCTION [dbo].[ufn_FormatPhone] (#PhoneNumber VARCHAR(32))
RETURNS VARCHAR(32)
AS
BEGIN
DECLARE #Phone CHAR(32)
SET #Phone = #PhoneNumber
-- cleanse phone number string
WHILE PATINDEX('%[^0-9]%', #PhoneNumber) > 0
SET #PhoneNumber = REPLACE(#PhoneNumber, SUBSTRING(#PhoneNumber, PATINDEX('%[^0-9]%', #PhoneNumber), 1), '')
-- skip foreign phones
IF (
SUBSTRING(#PhoneNumber, 1, 1) = '1'
OR SUBSTRING(#PhoneNumber, 1, 1) = '+'
OR SUBSTRING(#PhoneNumber, 1, 1) = '0'
)
AND LEN(#PhoneNumber) > 11
RETURN #Phone
-- build US standard phone number
SET #Phone = #PhoneNumber
SET #PhoneNumber = '(' + SUBSTRING(#PhoneNumber, 1, 3) + ') ' + SUBSTRING(#PhoneNumber, 4, 3) + '-' + SUBSTRING(#PhoneNumber, 7, 4)
IF LEN(#Phone) - 10 > 1
SET #PhoneNumber = #PhoneNumber + ' X' + SUBSTRING(#Phone, 11, LEN(#Phone) - 10)
RETURN #PhoneNumber
END

Answer 4

I do not recommend keeping bad data in the database and then only correcting it on the output. We have a database where phone numbers are entered in variously as :
[phone]
555+555+5555
555.555.5555
[phone]
5555555555
Different people in an organization may write various retrieval functions and updates to the database, and therefore it would be harder to set in place formatting and retrieval rules. I am therefore correcting the data in the database first and foremost and then setting in place rules and form validations that protect the integrity of this database going forward.
I see no justification for keeping bad data unless as suggested a duplicate column be added with corrected formatting and the original data kept around for redundancy and reference, and YES I consider badly formatted data as BAD data.

Answer 5

Solutions that use SUBSTRING and concatenation + are nearly independent of RDBMS. Here is a short solution that is specific to SQL Server:
declare #x int = 123456789
select stuff(stuff(#x, 4, 0, '-'), 8, 0, '-')

Answer 6

You can also try this:
CREATE function [dbo].[fn_FormatPhone](#Phone varchar(30))
returns varchar(30)
As
Begin
declare #FormattedPhone varchar(30)
set #Phone = replace(#Phone, '.', '-') --alot of entries use periods instead of dashes
set #FormattedPhone =
Case
When isNumeric(#Phone) = 1 Then
case
when len(#Phone) = 10 then '('+substring(#Phone, 1, 3)+')'+ ' ' +substring(#Phone, 4, 3)+ '-' +substring(#Phone, 7, 4)
when len(#Phone) = 7 then substring(#Phone, 1, 3)+ '-' +substring(#Phone, 4, 4)
else #Phone
end
When #phone like '[0-9][0-9][0-9]-[0-9][0-9][0-9][0-9][0-9][0-9][0-9]' Then '('+substring(#Phone, 1, 3)+')'+ ' ' +substring(#Phone, 5, 3)+ '-' +substring(#Phone, 8, 4)
When #phone like '[0-9][0-9][0-9] [0-9][0-9][0-9] [0-9][0-9][0-9][0-9]' Then '('+substring(#Phone, 1, 3)+')'+ ' ' +substring(#Phone, 5, 3)+ '-' +substring(#Phone, 9, 4)
When #phone like '[0-9][0-9][0-9]-[0-9][0-9][0-9]-[0-9][0-9][0-9][0-9]' Then '('+substring(#Phone, 1, 3)+')'+ ' ' +substring(#Phone, 5, 3)+ '-' +substring(#Phone, 9, 4)
Else #Phone
End
return #FormattedPhone
end
use on it select
(SELECT [dbo].[fn_FormatPhone](f.coffphone)) as 'Phone'
Output will be

Answer 7

You Can Use FORMAT if you column is a number Syntax like
FORMAT ( value, format [, culture ] ) In use like
FORMAT ( #d, 'D', 'en-US' ) or FORMAT(123456789,'###-##-####') (But This works for only SQL SERVER 2012 And After)
In Use Like
UPDATE TABLE_NAME SET COLUMN_NAME = FORMAT(COLUMN_NAME ,'###-##-####')
And
if your column is Varchar Or Nvarchar use do like this
CONCAT(SUBSTRING(CELLPHONE,0,4),' ',SUBSTRING(CELLPHONE,4,3),' ',SUBSTRING(CELLPHONE,7,2) ,' ',SUBSTRING(CELLPHONE,9,2) )
You can always get help from
https://msdn.microsoft.com/en-us/library/hh213505.aspx

Answer 8

I found that this works if wanting in a (123) - 456-7890 format.
UPDATE table
SET Phone_number = '(' +
SUBSTRING(Phone_number, 1, 3)
+ ') '
+ '- ' +
SUBSTRING(Phone_number, 4, 3)
+ '-' +
SUBSTRING(Phone_number, 7, 4)

Answer 9

Updated #sqiller's function for my purposes
CREATE FUNCTION [toolbox].[FormatPhoneNumber] (
#PhoneNumber VARCHAR(50),
#DefaultIfUnknown VARCHAR(50)
)
RETURNS VARCHAR(50)
AS
BEGIN
-- remove any extension
IF CHARINDEX('x', #PhoneNumber, 1) > 0
SET #PhoneNumber = SUBSTRING(#PhoneNumber, 1, CHARINDEX('x', #PhoneNumber, 1) - 1)
-- cleanse phone number string
WHILE PATINDEX('%[^0-9]%',#PhoneNumber) > 0
SET #PhoneNumber = REPLACE(#PhoneNumber,
SUBSTRING(#PhoneNumber,PATINDEX('%[^0-9]%',#PhoneNumber),1),'')
-- Remove US international code if exists, i.e. 12345678900
IF SUBSTRING(#PhoneNumber,1,1) = '1' AND LEN(#PhoneNumber) = 11
SET #PhoneNumber = SUBSTRING(#PhoneNumber, 2, 10)
-- any phone numbers without 10 characters are set to default
IF LEN(#PhoneNumber) <> 10
RETURN #DefaultIfUnknown
-- build US standard phone number
SET #PhoneNumber = '(' + SUBSTRING(#PhoneNumber,1,3) + ') ' +
SUBSTRING(#PhoneNumber,4,3) + '-' + SUBSTRING(#PhoneNumber,7,4)
RETURN #PhoneNumber
END

Answer 10

If you want to just format the output no need to create a new table or a function. In this scenario the area code was on a separate fields. I use field1, field2 just to illustrate you can select other fields in the same query:
area phone
213 8962102
Select statement:
Select field1, field2,areacode,phone,SUBSTR(tablename.areacode,1,3) + '-' + SUBSTR(tablename.phone,1,3) + '-' + SUBSTR(tablename.areacode,4,4) as Formatted Phone from tablename
Sample OUTPUT:
columns: FIELD1, FIELD2, AREA, PHONE, FORMATTED PHONE
data: Field1, Field2, 213, 8962102, 213-896-2102