I have a table with standard columns where I want to perform regular INSERTs.
But one of the columns is of type varchar with special semantics. It's a string that's supposed to behave as a set of strings, where the elements of the set are separated by commas.
Eg. if one row has in that varchar column the value fish,sheep,dove, and I insert the string ,fish,eagle, I want the result to be fish,sheep,dove,eagle (ie. eagle gets added to the set, but fish doesn't because it's already in the set).
I have here this Postgres code that does the "set concatenation" that I want:
SELECT string_agg(unnest, ',') AS x FROM (SELECT DISTINCT unnest(string_to_array('fish,sheep,dove' || ',fish,eagle', ','))) AS x;
But I can't figure out how to apply this logic to insertions.
What I want is something like:
CREATE TABLE IF NOT EXISTS t00(
userid int8 PRIMARY KEY,
a int8,
b varchar);
INSERT INTO t00 (userid,a,b) VALUES (0,1,'fish,sheep,dove');
INSERT INTO t00 (userid,a,b) VALUES (0,1,',fish,eagle')
ON CONFLICT (userid)
DO UPDATE SET
a = EXCLUDED.a,
b = SELECT string_agg(unnest, ',') AS x FROM (SELECT DISTINCT unnest(string_to_array(t00.b || EXCLUDED.b, ','))) AS x;
How can I achieve something like that?
Storing comma separated values is a huge mistake to begin with. But if you really want to make your life harder than it needs to be, you might want to create a function that merges two comma separated lists:
create function merge_lists(p_one text, p_two text)
returns text
as
$$
select string_agg(item, ',')
from (
select e.item
from unnest(string_to_array(p_one, ',')) as e(item)
where e.item <> '' --< necessary because of the leading , in your data
union
select t.item
from unnest(string_to_array(p_two, ',')) t(item)
where t.item <> ''
) t;
$$
language sql;
If you are using Postgres 14 or later, unnest(string_to_array(..., ',')) can be replace with string_to_table(..., ',')
Then your INSERT statement gets a bit simpler:
INSERT INTO t00 (userid,a,b) VALUES (0,1,',fish,eagle')
ON CONFLICT (userid)
DO UPDATE SET
a = EXCLUDED.a,
b = merge_lists(excluded.b, t00.b);
I think I was only missing parentheses around the SELECT statement:
INSERT INTO t00 (userid,a,b) VALUES (0,1,',fish,eagle')
ON CONFLICT (userid)
DO UPDATE SET
a = EXCLUDED.a,
b = (SELECT string_agg(unnest, ',') AS x FROM (SELECT DISTINCT unnest(string_to_array(t00.b || EXCLUDED.b, ','))) AS x);
I've got dirty data in a column with variable alpha length. I just want to strip out anything that is not 0-9.
I do not want to run a function or proc. I have a script that is similar that just grabs the numeric value after text, it looks like this:
Update TableName
set ColumntoUpdate=cast(replace(Columnofdirtydata,'Alpha #','') as int)
where Columnofdirtydata like 'Alpha #%'
And ColumntoUpdate is Null
I thought it would work pretty good until I found that some of the data fields I thought would just be in the format Alpha # 12345789 are not.
Examples of data that needs to be stripped
AB ABCDE # 123
ABCDE# 123
AB: ABC# 123
I just want the 123. It is true that all data fields do have the # prior to the number.
I tried substring and PatIndex, but I'm not quite getting the syntax correct or something. Anyone have any advice on the best way to address this?
See this blog post on extracting numbers from strings in SQL Server. Below is a sample using a string in your example:
DECLARE #textval NVARCHAR(30)
SET #textval = 'AB ABCDE # 123'
SELECT LEFT(SUBSTRING(#textval, PATINDEX('%[0-9.-]%', #textval), 8000),
PATINDEX('%[^0-9.-]%', SUBSTRING(#textval, PATINDEX('%[0-9.-]%', #textval), 8000) + 'X') -1)
Here is an elegant solution if your server supports the TRANSLATE function (on sql server it's available on sql server 2017+ and also sql azure).
First, it replaces any non numeric characters with a # character.
Then, it removes all # characters.
You may need to add additional characters that you know may be present in the second parameter of the TRANSLATE call.
select REPLACE(TRANSLATE([Col], 'abcdefghijklmnopqrstuvwxyz+()- ,#+', '##################################'), '#', '')
You can use stuff and patindex.
stuff(Col, 1, patindex('%[0-9]%', Col)-1, '')
SQL Fiddle
This works well for me:
CREATE FUNCTION [dbo].[StripNonNumerics]
(
#Temp varchar(255)
)
RETURNS varchar(255)
AS
Begin
Declare #KeepValues as varchar(50)
Set #KeepValues = '%[^0-9]%'
While PatIndex(#KeepValues, #Temp) > 0
Set #Temp = Stuff(#Temp, PatIndex(#KeepValues, #Temp), 1, '')
Return #Temp
End
Then call the function like so to see the original something next to the sanitized something:
SELECT Something, dbo.StripNonNumerics(Something) FROM TableA
In case if there are some characters possible between digits (e.g. thousands separators), you may try following:
declare #table table (DirtyCol varchar(100))
insert into #table values
('AB ABCDE # 123')
,('ABCDE# 123')
,('AB: ABC# 123')
,('AB#')
,('AB # 1 000 000')
,('AB # 1`234`567')
,('AB # (9)(876)(543)')
;with tally as (select top (100) N=row_number() over (order by ##spid) from sys.all_columns),
data as (
select DirtyCol, Col
from #table
cross apply (
select (select C + ''
from (select N, substring(DirtyCol, N, 1) C from tally where N<=datalength(DirtyCol)) [1]
where C between '0' and '9'
order by N
for xml path(''))
) p (Col)
where p.Col is not NULL
)
select DirtyCol, cast(Col as int) IntCol
from data
Output is:
DirtyCol IntCol
--------------------- -------
AB ABCDE # 123 123
ABCDE# 123 123
AB: ABC# 123 123
AB # 1 000 000 1000000
AB # 1`234`567 1234567
AB # (9)(876)(543) 9876543
For update, add ColToUpdate to select list of the data cte:
;with num as (...),
data as (
select ColToUpdate, /*DirtyCol, */Col
from ...
)
update data
set ColToUpdate = cast(Col as int)
CREATE FUNCTION FN_RemoveNonNumeric (#Input NVARCHAR(512))
RETURNS NVARCHAR(512)
AS
BEGIN
DECLARE #Trimmed NVARCHAR(512)
SELECT #Trimmed = #Input
WHILE PATINDEX('%[^0-9]%', #Trimmed) > 0
SELECT #Trimmed = REPLACE(#Trimmed, SUBSTRING(#Trimmed, PATINDEX('%[^0-9]%', #Trimmed), 1), '')
RETURN #Trimmed
END
GO
SELECT dbo.FN_RemoveNonNumeric('ABCDE# 123')
Pretty late to the party, I found the following which I though worked brilliantialy.. if anyone is still looking
SELECT
(SELECT CAST(CAST((
SELECT SUBSTRING(FieldToStrip, Number, 1)
FROM master..spt_values
WHERE Type='p' AND Number <= LEN(FieldToStrip) AND
SUBSTRING(FieldToStrip, Number, 1) LIKE '[0-9]' FOR XML Path(''))
AS xml) AS varchar(MAX)))
FROM
SourceTable
Here's a version which pulls all digits from a string; i.e. given I'm 35 years old; I was born in 1982. The average family has 2.4 children. this would return 35198224. i.e. it's good where you've got numeric data which may have been formatted as a code (e.g. #123,456,789 / 123-00005), but isn't appropriate if you're looking to pull out specific numbers (i.e. as opposed to digits / just the numeric characters) from the text. Also it only handles digits; so won't return negative signs (-) or periods .).
declare #table table (id bigint not null identity (1,1), data nvarchar(max))
insert #table (data)
values ('hello 123 its 45613 then') --outputs: 12345613
,('1 some other string 98 example 4') --outputs: 1984
,('AB ABCDE # 123') --outputs: 123
,('ABCDE# 123') --outputs: 123
,('AB: ABC# 123') --outputs: 123
; with NonNumerics as (
select id
, data original
--the below line replaces all digits with blanks
, replace(replace(replace(replace(replace(replace(replace(replace(replace(replace(data,'0',''),'1',''),'2',''),'3',''),'4',''),'5',''),'6',''),'7',''),'8',''),'9','') nonNumeric
from #table
)
--each iteration of the below CTE removes another non-numeric character from the original string, putting the result into the numerics column
, Numerics as (
select id
, replace(original, substring(nonNumeric,1,1), '') numerics
, replace(nonNumeric, substring(nonNumeric,1,1), '') charsToreplace
, len(replace(nonNumeric, substring(nonNumeric,1,1), '')) charsRemaining
from NonNumerics
union all
select id
, replace(numerics, substring(charsToreplace,1,1), '') numerics
, replace(charsToreplace, substring(charsToreplace,1,1), '') charsToreplace
, len(replace(charsToreplace, substring(charsToreplace,1,1), '')) charsRemaining
from Numerics
where charsRemaining > 0
)
--we select only those strings with `charsRemaining=0`; i.e. the rows for which all non-numeric characters have been removed; there should be 1 row returned for every 1 row in the original data set.
select * from Numerics where charsRemaining = 0
This code works by removing all the digits (i.e. the characters we want) from a the given strings by replacing them with blanks. Then it goes through the original string (which includes the digits) removing all of the characters that were left (i.e. the non-numeric characters), thus leaving only the digits.
The reason we do this in 2 steps, rather than just removing all non-numeric characters in the first place is there are only 10 digits, whilst there are a huge number of possible characters; so replacing that small list is relatively fast; then gives us a list of those non-numeric characters which actually exist in the string, so we can then replace that small set.
The method makes use of recursive SQL, using common table expressions (CTEs).
To add on to Ken's answer, this handles commas and spaces and parentheses
--Handles parentheses, commas, spaces, hyphens..
declare #table table (c varchar(256))
insert into #table
values
('This is a test 111-222-3344'),
('Some Sample Text (111)-222-3344'),
('Hello there 111222 3344 / How are you?'),
('Hello there 111 222 3344 ? How are you?'),
('Hello there 111 222 3344. How are you?')
select
replace(LEFT(SUBSTRING(replace(replace(replace(replace(replace(c,'(',''),')',''),'-',''),' ',''),',',''), PATINDEX('%[0-9.-]%', replace(replace(replace(replace(replace(c,'(',''),')',''),'-',''),' ',''),',','')), 8000),
PATINDEX('%[^0-9.-]%', SUBSTRING(replace(replace(replace(replace(replace(c,'(',''),')',''),'-',''),' ',''),',',''), PATINDEX('%[0-9.-]%', replace(replace(replace(replace(replace(c,'(',''),')',''),'-',''),' ',''),',','')), 8000) + 'X') -1),'.','')
from #table
Create function fn_GetNumbersOnly(#pn varchar(100))
Returns varchar(max)
AS
BEGIN
Declare #r varchar(max) ='', #len int ,#c char(1), #x int = 0
Select #len = len(#pn)
while #x <= #len
begin
Select #c = SUBSTRING(#pn,#x,1)
if ISNUMERIC(#c) = 1 and #c <> '-'
Select #r = #r + #c
Select #x = #x +1
end
return #r
End
In your case It seems like the # will always be after teh # symbol so using CHARINDEX() with LTRIM() and RTRIM() would probably perform the best. But here is an interesting method of getting rid of ANY non digit. It utilizes a tally table and table of digits to limit which characters are accepted then XML technique to concatenate back to a single string without the non-numeric characters. The neat thing about this technique is it could be expanded to included ANY Allowed characters and strip out anything that is not allowed.
DECLARE #ExampleData AS TABLE (Col VARCHAR(100))
INSERT INTO #ExampleData (Col) VALUES ('AB ABCDE # 123'),('ABCDE# 123'),('AB: ABC# 123')
DECLARE #Digits AS TABLE (D CHAR(1))
INSERT INTO #Digits (D) VALUES ('0'),('1'),('2'),('3'),('4'),('5'),('6'),('7'),('8'),('9')
;WITH cteTally AS (
SELECT
I = ROW_NUMBER() OVER (ORDER BY (SELECT NULL))
FROM
#Digits d10
CROSS APPLY #Digits d100
--add more cross applies to cover longer fields this handles 100
)
SELECT *
FROM
#ExampleData e
OUTER APPLY (
SELECT CleansedPhone = CAST((
SELECT TOP 100
SUBSTRING(e.Col,t.I,1)
FROM
cteTally t
INNER JOIN #Digits d
ON SUBSTRING(e.Col,t.I,1) = d.D
WHERE
I <= LEN(e.Col)
ORDER BY
t.I
FOR XML PATH('')) AS VARCHAR(100))) o
Declare #MainTable table(id int identity(1,1),TextField varchar(100))
INSERT INTO #MainTable (TextField)
VALUES
('6B32E')
declare #i int=1
Declare #originalWord varchar(100)=''
WHile #i<=(Select count(*) from #MainTable)
BEGIN
Select #originalWord=TextField from #MainTable where id=#i
Declare #r varchar(max) ='', #len int ,#c char(1), #x int = 0
Select #len = len(#originalWord)
declare #pn varchar(100)=#originalWord
while #x <= #len
begin
Select #c = SUBSTRING(#pn,#x,1)
if(#c!='')
BEGIN
if ISNUMERIC(#c) = 0 and #c <> '-'
BEGIN
Select #r = cast(#r as varchar) + cast(replace((SELECT ASCII(#c)-64),'-','') as varchar)
end
ELSE
BEGIN
Select #r = #r + #c
END
END
Select #x = #x +1
END
Select #r
Set #i=#i+1
END
I have created a function for this
Create FUNCTION RemoveCharacters (#text varchar(30))
RETURNS VARCHAR(30)
AS
BEGIN
declare #index as int
declare #newtexval as varchar(30)
set #index = (select PATINDEX('%[A-Z.-/?]%', #text))
if (#index =0)
begin
return #text
end
else
begin
set #newtexval = (select STUFF ( #text , #index , 1 , '' ))
return dbo.RemoveCharacters(#newtexval)
end
return 0
END
GO
Here is the answer:
DECLARE #t TABLE (tVal VARCHAR(100))
INSERT INTO #t VALUES('123')
INSERT INTO #t VALUES('123S')
INSERT INTO #t VALUES('A123,123')
INSERT INTO #t VALUES('a123..A123')
;WITH cte (original, tVal, n)
AS
(
SELECT t.tVal AS original,
LOWER(t.tVal) AS tVal,
65 AS n
FROM #t AS t
UNION ALL
SELECT tVal AS original,
CAST(REPLACE(LOWER(tVal), LOWER(CHAR(n)), '') AS VARCHAR(100)),
n + 1
FROM cte
WHERE n <= 90
)
SELECT t1.tVal AS OldVal,
t.tval AS NewVal
FROM (
SELECT original,
tVal,
ROW_NUMBER() OVER(PARTITION BY tVal + original ORDER BY original) AS Sl
FROM cte
WHERE PATINDEX('%[a-z]%', tVal) = 0
) t
INNER JOIN #t t1
ON t.original = t1.tVal
WHERE t.sl = 1
You can create SQL CLR scalar function in order to be able to use regular expressions like replace patterns.
Here you can find example of how to create such function.
Having such function will solve the issue with just the following lines:
SELECT [dbo].[fn_Utils_RegexReplace] ('AB ABCDE # 123', '[^0-9]', '');
SELECT [dbo].[fn_Utils_RegexReplace] ('ABCDE# 123', '[^0-9]', '');
SELECT [dbo].[fn_Utils_RegexReplace] ('AB: ABC# 123', '[^0-9]', '');
More important, you will be able to solve more complex issues as the regular expressions will bring a whole new world of options directly in your T-SQL statements.
Use this:
REPLACE(TRANSLATE(SomeString, REPLACE(TRANSLATE(SomeString, '0123456789', '##########'), '#', ''), REPLICATE('#', LEN(REPLACE(TRANSLATE(SomeString, '0123456789', '##########'), '#', '') + 'x') - 1)), '#', '')
Demo:
DROP TABLE IF EXISTS #MyTempTable;
CREATE TABLE #MyTempTable (SomeString VARCHAR(255));
INSERT INTO #MyTempTable
VALUES ('ssss123ssg99d362sdg')
, ('hey 62q&*^(n43')
, (NULL)
, ('')
, ('hi')
, ('123');
SELECT SomeString
, REPLACE(TRANSLATE(SomeString, REPLACE(TRANSLATE(SomeString, '0123456789', '##########'), '#', ''), REPLICATE('#', LEN(REPLACE(TRANSLATE(SomeString, '0123456789', '##########'), '#', '') + 'x') - 1)), '#', '')
FROM #MyTempTable;
DROP TABLE IF EXISTS #MyTempTable;
Results:
SomeString
(No column name)
ssss123ssg99d362sdg
12399362
hey62q&*^(n43
6243
NULL
NULL
hi
123
123
While the OP wanted to "strip out anything that is not 0-9", the post is also tagged with "substring" and "patindex", and the OP mentioned the concern "not quite getting the syntax correct or something". To address that the requirements note that "all data fields do have the # prior to the number" and to provide an answer that addresses the challenges with substring/patindex, consider the following:
/* A sample select */
;WITH SampleValues AS
( SELECT 'AB ABCDE # 123' [Columnofdirtydata]
UNION ALL SELECT 'AB2: ABC# 123')
SELECT
s.Columnofdirtydata,
f1.pos1,
'['+ f2.substr +']' [InspectOutput]
FROM
SampleValues s
CROSS APPLY (SELECT PATINDEX('%# %',s.Columnofdirtydata) [pos1]) f1
CROSS APPLY (SELECT SUBSTRING(s.Columnofdirtydata, f1.pos1 + LEN('#-'),LEN(s.Columnofdirtydata)) [substr]) f2
/* Using update scenario from OP */
UPDATE t1
SET t1.Columntoupdate = CAST(f2.substr AS INT)
FROM
TableName t1
CROSS APPLY (SELECT PATINDEX('%# %',t1.Columnofdirtydata) [pos1]) f1
CROSS APPLY (SELECT SUBSTRING(t1.Columnofdirtydata, f1.pos1 + LEN('#-'),LEN(t1.Columnofdirtydata)) [substr]) f2
Note that my syntax advice for patindex/substring, is to:
consider using APPLY as a way to temporarily alias results from one function for use as parameters in the next. It's not uncommon to (in ETL, for example) need to parse out parameter/position-based substrings in an updatable column of a staging table. If you need to "debug" and potentially fix some parsing logic, this style will help.
consider using LEN('PatternSample') in your substring logic, to account for reusing this pattern or adjusting it when your source data changes (instead of "+ 1"
SUBSTRING() requires a length parameter, but it can be greater than the length of the string. Therefore, if you are getting "the rest of the string" after the pattern, you can just use "The source length"
DECLARE #STR VARCHAR(400)
DECLARE #specialchars VARCHAR(50) = '%[~,#,#,$,%,&,*,(,),!^?:]%'
SET #STR = '1, 45 4,3 68.00-'
WHILE PATINDEX( #specialchars, #STR ) > 0
---Remove special characters using Replace function
SET #STR = Replace(Replace(REPLACE( #STR, SUBSTRING( #STR, PATINDEX( #specialchars, #STR ), 1 ),''),'-',''), ' ','')
SELECT #STR
SELECT REGEXP_REPLACE( col, '[^[:digit:]]', '' ) AS new_col FROM my_table
-- Given a CSV string like this:
declare #roles varchar(800)
select #roles = 'Pub,RegUser,ServiceAdmin'
-- Question: How to get roles into a table view like this:
select 'Pub'
union
select 'RegUser'
union
select 'ServiceAdmin'
After posting this, I started playing with some dynamic SQL. This seems to work, but seems like there might be some security risks by using dynamic SQL - thoughts on this?
declare #rolesSql varchar(800)
select #rolesSql = 'select ''' + replace(#roles, ',', ''' union select ''') + ''''
exec(#rolesSql)
If you're working with SQL Server compatibility level 130 then the STRING_SPLIT function is now the most succinct method available.
Reference link: https://msdn.microsoft.com/en-gb/library/mt684588.aspx
Usage:
SELECT * FROM string_split('Pub,RegUser,ServiceAdmin',',')
RESULT:
value
-----------
Pub
RegUser
ServiceAdmin
See my answer from here
But basically you would:
Create this function in your DB:
CREATE FUNCTION dbo.Split(#origString varchar(max), #Delimiter char(1))
returns #temptable TABLE (items varchar(max))
as
begin
declare #idx int
declare #split varchar(max)
select #idx = 1
if len(#origString )<1 or #origString is null return
while #idx!= 0
begin
set #idx = charindex(#Delimiter,#origString)
if #idx!=0
set #split= left(#origString,#idx - 1)
else
set #split= #origString
if(len(#split)>0)
insert into #temptable(Items) values(#split)
set #origString= right(#origString,len(#origString) - #idx)
if len(#origString) = 0 break
end
return
end
and then call the function and pass in the string you want to split.
Select * From dbo.Split(#roles, ',')
Here's a thorough discussion of your options:
Arrays and Lists in SQL Server
What i do in this case is just using some string replace to convert it to json and open the json like a table. May not be suitable for every use case but it is very simple to get running and works with strings and files. With files you just need to watch your line break character, mostly i find it to be "Char(13)+Char(10)"
declare #myCSV nvarchar(MAX)= N'"Id";"Duration";"PosX";"PosY"
"•P001";223;-30;35
"•P002";248;-28;35
"•P003";235;-26;35'
--CSV to JSON
--convert to json by replacing some stuff
declare #myJson nvarchar(MAX)= '[['+ replace(#myCSV, Char(13)+Char(10), '],[' ) +']]'
set #myJson = replace(#myJson, ';',',') -- Optional: ensure coma delimiters for json if the current delimiter differs
-- set #myJson = replace(#myJson, ',,',',null,') -- Optional: empty in between
-- set #myJson = replace(#myJson, ',]',',null]') -- Optional: empty before linebreak
SELECT
ROW_NUMBER() OVER (ORDER BY (SELECT 0))-1 AS LineNumber, *
FROM OPENJSON( #myJson )
with (
col0 varchar(255) '$[0]'
,col1 varchar(255) '$[1]'
,col2 varchar(255) '$[2]'
,col3 varchar(255) '$[3]'
,col4 varchar(255) '$[4]'
,col5 varchar(255) '$[5]'
,col6 varchar(255) '$[6]'
,col7 varchar(255) '$[7]'
,col8 varchar(255) '$[8]'
,col9 varchar(255) '$[9]'
--any name column count is possible
) csv
order by (SELECT 0) OFFSET 1 ROWS --hide header row
Using SQL Server's built in XML parsing is also an option. Of course, this glosses over all the nuances of an RFC-4180 compliant CSV.
-- Given a CSV string like this:
declare #roles varchar(800)
select #roles = 'Pub,RegUser,ServiceAdmin'
-- Here's the XML way
select split.csv.value('.', 'varchar(100)') as value
from (
select cast('<x>' + replace(#roles, ',', '</x><x>') + '</x>' as xml) as data
) as csv
cross apply data.nodes('/x') as split(csv)
If you are using SQL 2016+, using string_split is better, but this is a common way to do this prior to SQL 2016.
Using BULK INSERT you can import a csv file into your sql table -
http://blog.sqlauthority.com/2008/02/06/sql-server-import-csv-file-into-sql-server-using-bulk-insert-load-comma-delimited-file-into-sql-server/
Even the accepted answer is working fine. but I got this function much faster even for thousands of record. create below function and use.
IF EXISTS (
SELECT 1
FROM Information_schema.Routines
WHERE Specific_schema = 'dbo'
AND specific_name = 'FN_CSVToStringListTable'
AND Routine_Type = 'FUNCTION'
)
BEGIN
DROP FUNCTION [dbo].[FN_CSVToStringListTable]
END
GO
CREATE FUNCTION [dbo].[FN_CSVToStringListTable] (#InStr VARCHAR(MAX))
RETURNS #TempTab TABLE (Id NVARCHAR(max) NOT NULL)
AS
BEGIN
;-- Ensure input ends with comma
SET #InStr = REPLACE(#InStr + ',', ',,', ',')
DECLARE #SP INT
DECLARE #VALUE VARCHAR(1000)
WHILE PATINDEX('%,%', #INSTR) <> 0
BEGIN
SELECT #SP = PATINDEX('%,%', #INSTR)
SELECT #VALUE = LEFT(#INSTR, #SP - 1)
SELECT #INSTR = STUFF(#INSTR, 1, #SP, '')
INSERT INTO #TempTab (Id)
VALUES (#VALUE)
END
RETURN
END
GO
---Test like this.
declare #v as NVARCHAR(max) = N'asdf,,as34df,234df,fs,,34v,5fghwer,56gfg,';
SELECT Id FROM dbo.FN_CSVToStringListTable(#v)
I was about you use the solution mentioned in the accepted answer, but doing more research led me to use Table Value Types:
These are far more efficient and you don't need a TVF (Table valued function) just to create a table from csv. You can use it directly in your scripts or pass that to a stored procedure as a Table Value Parameter. The Type can be created as :
CREATE TYPE [UniqueIdentifiers] AS TABLE(
[Id] [varchar](20) NOT NULL
)
I've got semicolon-separated values in a column Values in my table:
Values
1;2;3;4;5
I would like to transform it in a procedure to have there values as rows:
Values
1
2
3
4
5
How could I do it in T-SQL?
Solution 1(using xml):
declare #str varchar(20)
declare #xml as xml
set #str= '1;2;3;4;5'
SET #xml = cast(('<x>'+replace(#str,';' ,'</x><x>')+'</x>') as xml)
SELECT col.value('.', 'varchar(10)') as value FROM #xml.nodes('x') as tbl(col)
Solution 2(using recursive cte)
declare #str as varchar(100)
declare #delimiter as char(1)
set #delimiter = ';'
set #str = '1;2;3;4;5' -- original data
set #str = #delimiter + #str + #delimiter
;with num_cte as
(
select 1 as rn
union all
select rn +1 as rn
from num_cte
where rn <= len(#str)
)
, get_delimiter_pos_cte as
(
select
ROW_NUMBER() OVER (ORDER BY rn) as rowid,
rn as delimiterpos
from num_cte
cross apply( select substring(#str,rn,1) AS chars) splittedchars
where chars = #delimiter
)
select substring(#str,a.delimiterpos+1 ,c2.delimiterpos - a.delimiterpos - 1) as Countries
from get_delimiter_pos_cte a
inner join get_delimiter_pos_cte c2 on c2.rowid = a.rowid+1
option(maxrecursion 0)
The thing that struck me as possibly leaving room for an additional answer, or additional improvement was that most of the answers/links given were how to split values like this for a single scalar value as opposed to how to apply that kind of splitting logic for a column of values in a table.
I include both a numbers table solution and an XML solution. The XML solution was inspired by the earlier post priyanka.sarkar. I think that a numbers table solution, using an actual numbers table instead of the CTE as in the below solution is probably the fastest, but the XML approach deserves to be developed upon because it's really nice looking.
So, here goes my attempt.
CREATE PROCEDURE PARSE_DELIMITED_VALUES
AS
WITH FIRST_NUMBERS (N) AS (
SELECT 1 UNION ALL SELECT 1
), SECOND_NUMBERS (N) AS (
SELECT E1.N
FROM FIRST_NUMBERS E1
CROSS JOIN FIRST_NUMBERS E2
), THIRD_NUMBERS (N) AS (
SELECT E1.N
FROM SECOND_NUMBERS E1
CROSS JOIN SECOND_NUMBERS E2
), FOURTH_NUMBERS (N) AS (
SELECT E1.N
FROM THIRD_NUMBERS E1
CROSS JOIN THIRD_NUMBERS E2
), FIFTH_NUMBERS (N) AS (
SELECT E1.N
FROM FOURTH_NUMBERS E1
CROSS JOIN FOURTH_NUMBERS E2
), NUMBERS (N) AS (
SELECT N
FROM NUMBERS
WHERE N <= 8000 /*adjust these as needed to come up with a max number equal to the max character length allowed in the Values column*/
/*or better yet, if you can, just remove this first...numbers... header stuff so long as you create a temp or permanent table that contains the same numbers to work with*/
)
SELECT SUBSTRING(
MYTABLE.Values,
CASE
WHEN NUMBERS.NUMBER = 1 THEN 1
ELSE NUMBERS.NUMBER + 1
END,
CASE CHARINDEX(';', MYTABLE.Values, NUMBERS.NUMBER + 1)
WHEN 0 THEN LEN('^' + MYTABLE.Values + '^') - 2 + 1
ELSE CHARINDEX(';', MYTABLE.Values, NUMBERS.NUMBER + 1)
END
- CASE
WHEN NUMBERS.NUMBER = 1 THEN 1
ELSE NUMBERS.NUMBER + 1
END
) AS PARSED_VALUE
FROM MYTABLE
INNER JOIN NUMBERS
ON NUMBERS.NUMBER <= LEN('^' + MYTABLE.Values + '^') - 2
AND (
NUMBERS.NUMBER = 1
OR SUBSTRING(MYTABLE.Values, NUMBERS.NUMBER, 1) = ';'
)
GO
-- if your values column can contain NULL values I would change the join at the end as follows:
--from INNER JOIN NUMBERS
--to LEFT OUTER JOIN NUMBERS
The above would probably be most performant if the WITH NUMBERS ... CTEs were replaced by a temporary or permanent table containing the same numeric values.
On the other hand the CTE does the job and keeps it more in one place.
CREATE PROCEDURE PARSE_DELIMITED_VALUES
AS
SELECT E.x.value('.', 'VARCHAR(MAX)') AS PARSED_VALUE
FROM (
SELECT CAST('<x>' + REPLACE(Values, ';', '</x><x>') + '</x>' AS XML) my_x
FROM MYTABLE
) TT
CROSS APPLY my_x.nodes('/x') AS E(x)
GO
-- if your values column can contain NULL values I would change the join at the end as follows:
from `CROSS APPLY`
to `OUTER APPLY`
It's not the most elegant approach, but this might be worth a try. It creates a Sql Command as a string, and at the end executes it.
DECLARE #Values VARCHAR(8000)
-- Flatten all values lists into one string
SET #Values = REPLACE(REPLACE((SELECT [Value] FROM [dbo.MyTable] FOR XML PATH('')), '<Value>', ''), '</Value>', ';')
SET #Values = SUBSTRING(#Values, 0, LEN(#Values))
DECLARE #SeparatorIndex INT
SET #SeparatorIndex = (SELECT TOP 1 PATINDEX('%[;]%', #Values))
DECLARE #InsertClause VARCHAR(50)
SET #InsertClause = 'INSERT INTO [dbo.MyTable] VALUES ('
DECLARE #SQL VARCHAR(500)
SET #SQL = #InsertClause + SUBSTRING(#Values, 0, #SeparatorIndex) + '); '
SET #Values = RIGHT(#Values, LEN(#Values) - (#SeparatorIndex - 1))
SET #SQL = REPLACE(#SQL + (SELECT (REPLACE(#Values, ';', '); ' + #InsertClause))) + ')', '; )', '')
EXEC (#SQL)
The command ends up (in Sql Server 2005) as:
INSERT INTO [dbo.MyTable] VALUES (1); INSERT INTO [dbo.MyTable] VALUES (2); INSERT INTO [dbo.MyTable] VALUES (3); INSERT INTO [dbo.MyTable] VALUES (4); INSERT INTO [dbo.MyTable] VALUES (5) ...'
Do you actually mean, "rows," as in, "tuples," (so you can insert the data into another table, one element per row) or do you mean you want the data displayed vertically?
I'd think a string Replace (look up T-SQL's String Functions) would do the trick, no? Depending on the output target, you'd replace ; with CRLF or . You could even use Replace to create dynamic SQL Insert statements that could be executed by the SP to do row inserts (if that was your intent).
For presentation purposes, this is bad practice.
If it is purely for presentation and you are permitted, I'd output everything as XML then XSLT it any way you want. Honestly, I don't remember the last time I operated directly on a recordset. I always output to XML.