Can this be done without a loop

  • Hi folks. I needed to write a stored proc that takes generates a string of random characters @minlen to @maxlen characters long for testing some functions. What I came up with is posted below. I couldn't do it without a loop. You'll see that I used temp tables; when I attempted to use that logic in a CTE it screwed up the "length" (number of rows). Just curious if anyone has done something like this without without a loop.

    Thanks.

    CREATE PROC dbo.RandomStrings

    (

    -- include some default values

    @rows int = 10000,

    @minlen int = 2,

    @maxlen int = 6 -- max 4225

    )

    /****************************************************************************************

    To get 50,000 rows of random string 5 to 20 character long:

    EXEC dbo.RandomStrings 50000, 5, 20

    ****************************************************************************************/

    AS

    -- prep temp tables

    DECLARE

    @strings TABLE (chrn int NOT NULL);

    DECLARE

    @stringsFinal TABLE (stringID int identity primary key, string varchar(8000) NOT NULL)

    DECLARE

    @chars TABLE (charid int primary key, charn int NOT NULL);

    -- Insert ascii values characters that match this pattern: [!#$0-9,a-z,A-Z] into @chars

    WITH

    chars AS

    (

    SELECT rn = ROW_NUMBER() OVER (ORDER BY (SELECT NULL)), N

    FROM (VALUES

    (33),(35),(36),(48),(49),(50),(51),(52),(53),(54),(55),(56),(57),(65),(66),(67),(68),(69),

    (70),(71),(72),(73),(74),(75),(76),(77),(78),(79),(80),(81),(82),(83),(84),(85),(86),(87),

    (88),(89),(90),(97),(98),(99),(100),(101),(102),(103),(104),(105),(106),(107),(108),(109),

    (110),(111),(112),(113),(114),(115),(116),(117),(118),(119),(120),(121),(122)) N(N)

    )

    INSERT @chars

    SELECT * FROM chars;

    DECLARE @i int = 1

    WHILE @i <= @rows

    BEGIN

    DELETE FROM @strings; --clear it out for new characters

    INSERT @strings

    SELECT TOP (ABS(CHECKSUM(newid())%(@maxlen-(@minlen-1)))+@minlen) -- gets me the numbers @min-2 through @maxlen

    c = ABS(CHECKSUM(newid())%65)+1 -- get me 1 through 65

    FROM @chars a, @chars b;

    INSERT @stringsFinal (string)

    SELECT SUBSTRING((SELECT REPLICATE

    (

    CHAR(c.charn),

    CASE CHECKSUM(newid())%3 WHEN 2 THEN 2 ELSE 1 END -- 1 in 6

    )

    FROM @strings s

    JOIN @chars c ON s.chrn = c.charid

    FOR XML PATH(''), TYPE).value('.','varchar(8000)'),1,@maxlen);

    SET @i = @i+1

    END

    -- Output results

    SELECT * FROM @stringsFinal;

    GO

    "I cant stress enough the importance of switching from a sequential files mindset to set-based thinking. After you make the switch, you can spend your time tuning and optimizing your queries instead of maintaining lengthy, poor-performing code."

    -- Itzik Ben-Gan 2001

  • Alan.B (10/28/2015)


    Hi folks. I needed to write a stored proc that takes generates a string of random characters @minlen to @maxlen characters long for testing some functions. What I came up with is posted below. I couldn't do it without a loop. You'll see that I used temp tables; when I attempted to use that logic in a CTE it screwed up the "length" (number of rows). Just curious if anyone has done something like this without without a loop.

    Thanks.

    CREATE PROC dbo.RandomStrings

    (

    -- include some default values

    @rows int = 10000,

    @minlen int = 2,

    @maxlen int = 6 -- max 4225

    )

    /****************************************************************************************

    To get 50,000 rows of random string 5 to 20 character long:

    EXEC dbo.RandomStrings 50000, 5, 20

    ****************************************************************************************/

    AS

    -- prep temp tables

    DECLARE

    @strings TABLE (chrn int NOT NULL);

    DECLARE

    @stringsFinal TABLE (stringID int identity primary key, string varchar(8000) NOT NULL)

    DECLARE

    @chars TABLE (charid int primary key, charn int NOT NULL);

    -- Insert ascii values characters that match this pattern: [!#$0-9,a-z,A-Z] into @chars

    WITH

    chars AS

    (

    SELECT rn = ROW_NUMBER() OVER (ORDER BY (SELECT NULL)), N

    FROM (VALUES

    (33),(35),(36),(48),(49),(50),(51),(52),(53),(54),(55),(56),(57),(65),(66),(67),(68),(69),

    (70),(71),(72),(73),(74),(75),(76),(77),(78),(79),(80),(81),(82),(83),(84),(85),(86),(87),

    (88),(89),(90),(97),(98),(99),(100),(101),(102),(103),(104),(105),(106),(107),(108),(109),

    (110),(111),(112),(113),(114),(115),(116),(117),(118),(119),(120),(121),(122)) N(N)

    )

    INSERT @chars

    SELECT * FROM chars;

    DECLARE @i int = 1

    WHILE @i <= @rows

    BEGIN

    DELETE FROM @strings; --clear it out for new characters

    INSERT @strings

    SELECT TOP (ABS(CHECKSUM(newid())%(@maxlen-(@minlen-1)))+@minlen) -- gets me the numbers @min-2 through @maxlen

    c = ABS(CHECKSUM(newid())%65)+1 -- get me 1 through 65

    FROM @chars a, @chars b;

    INSERT @stringsFinal (string)

    SELECT SUBSTRING((SELECT REPLICATE

    (

    CHAR(c.charn),

    CASE CHECKSUM(newid())%3 WHEN 2 THEN 2 ELSE 1 END -- 1 in 6

    )

    FROM @strings s

    JOIN @chars c ON s.chrn = c.charid

    FOR XML PATH(''), TYPE).value('.','varchar(8000)'),1,@maxlen);

    SET @i = @i+1

    END

    -- Output results

    SELECT * FROM @stringsFinal;

    GO

    Here's something I've borrowed and adapted from one of Jeff's posts. Doesn't scale well to high values of maxlen and you'll have to play with the character selection to get the ones you want.

    declare @minlen int = 2

    ,@maxlen int = 6;

    if object_id('dbo.JBMTest', 'U') is not null

    drop table dbo.JBMTest;

    select top 50000

    SomeID = identity( int,1,1

    ),SomeLetters2 = left(ch1.c + ch2.c + ch3.c + ch4.c + ch5.c + ch6.c + ch7.c + ch8.c + ch9.c + ch10.c, lth.Ln)

    ,Ln = (abs(checksum(newid()) % (@maxlen - (@minlen - 1))) + @minlen)

    into dbo.JBMTest

    from sys.all_columns ac1

    cross join sys.all_columns ac2

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch1

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch2

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch3

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch4

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch5

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch6

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch7

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch8

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch9

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch10

    cross apply (select Ln = (abs(checksum(newid()) % (@maxlen - (@minlen - 1))) + @minlen)

    ) lth;

    alter table dbo.JBMTest

    add constraint PK_JBMTest primary key clustered (SomeID) with fillfactor = 90;

    select *

    from dbo.JBMTest jt;

    The absence of evidence is not evidence of absence
    - Martin Rees
    The absence of consumable DDL, sample data and desired results is, however, evidence of the absence of my response
    - Phil Parkin

  • see this article and read the comments from Jeff at the bottom of the article

    https://www.mssqltips.com/sqlservertip/3576/generate-random-strings-with-high-performance-with-a-sql-clr-function

    ________________________________________________________________
    you can lead a user to data....but you cannot make them think
    and remember....every day is a school day

  • J Livingston SQL (10/29/2015)


    see this article and read the comments from Jeff at the bottom of the article

    https://www.mssqltips.com/sqlservertip/3576/generate-random-strings-with-high-performance-with-a-sql-clr-function

    Excellent link and way too many Jeffs getting involved there!

    The absence of evidence is not evidence of absence
    - Martin Rees
    The absence of consumable DDL, sample data and desired results is, however, evidence of the absence of my response
    - Phil Parkin

  • Phil Parkin (10/29/2015)


    J Livingston SQL (10/29/2015)


    see this article and read the comments from Jeff at the bottom of the article

    https://www.mssqltips.com/sqlservertip/3576/generate-random-strings-with-high-performance-with-a-sql-clr-function

    Excellent link and way too many Jeffs getting involved there!

    you are absolutely correct Phil.....didnt see that:-)

    ________________________________________________________________
    you can lead a user to data....but you cannot make them think
    and remember....every day is a school day

  • Phil Parkin (10/29/2015)


    Alan.B (10/28/2015)


    Hi folks. I needed to write a stored proc that takes generates a string of random characters @minlen to @maxlen characters long for testing some functions. What I came up with is posted below. I couldn't do it without a loop. You'll see that I used temp tables; when I attempted to use that logic in a CTE it screwed up the "length" (number of rows). Just curious if anyone has done something like this without without a loop.

    Thanks.

    CREATE PROC dbo.RandomStrings

    (

    -- include some default values

    @rows int = 10000,

    @minlen int = 2,

    @maxlen int = 6 -- max 4225

    )

    /****************************************************************************************

    To get 50,000 rows of random string 5 to 20 character long:

    EXEC dbo.RandomStrings 50000, 5, 20

    ****************************************************************************************/

    AS

    -- prep temp tables

    DECLARE

    @strings TABLE (chrn int NOT NULL);

    DECLARE

    @stringsFinal TABLE (stringID int identity primary key, string varchar(8000) NOT NULL)

    DECLARE

    @chars TABLE (charid int primary key, charn int NOT NULL);

    -- Insert ascii values characters that match this pattern: [!#$0-9,a-z,A-Z] into @chars

    WITH

    chars AS

    (

    SELECT rn = ROW_NUMBER() OVER (ORDER BY (SELECT NULL)), N

    FROM (VALUES

    (33),(35),(36),(48),(49),(50),(51),(52),(53),(54),(55),(56),(57),(65),(66),(67),(68),(69),

    (70),(71),(72),(73),(74),(75),(76),(77),(78),(79),(80),(81),(82),(83),(84),(85),(86),(87),

    (88),(89),(90),(97),(98),(99),(100),(101),(102),(103),(104),(105),(106),(107),(108),(109),

    (110),(111),(112),(113),(114),(115),(116),(117),(118),(119),(120),(121),(122)) N(N)

    )

    INSERT @chars

    SELECT * FROM chars;

    DECLARE @i int = 1

    WHILE @i <= @rows

    BEGIN

    DELETE FROM @strings; --clear it out for new characters

    INSERT @strings

    SELECT TOP (ABS(CHECKSUM(newid())%(@maxlen-(@minlen-1)))+@minlen) -- gets me the numbers @min-2 through @maxlen

    c = ABS(CHECKSUM(newid())%65)+1 -- get me 1 through 65

    FROM @chars a, @chars b;

    INSERT @stringsFinal (string)

    SELECT SUBSTRING((SELECT REPLICATE

    (

    CHAR(c.charn),

    CASE CHECKSUM(newid())%3 WHEN 2 THEN 2 ELSE 1 END -- 1 in 6

    )

    FROM @strings s

    JOIN @chars c ON s.chrn = c.charid

    FOR XML PATH(''), TYPE).value('.','varchar(8000)'),1,@maxlen);

    SET @i = @i+1

    END

    -- Output results

    SELECT * FROM @stringsFinal;

    GO

    Here's something I've borrowed and adapted from one of Jeff's posts. Doesn't scale well to high values of maxlen and you'll have to play with the character selection to get the ones you want.

    declare @minlen int = 2

    ,@maxlen int = 6;

    if object_id('dbo.JBMTest', 'U') is not null

    drop table dbo.JBMTest;

    select top 50000

    SomeID = identity( int,1,1

    ),SomeLetters2 = left(ch1.c + ch2.c + ch3.c + ch4.c + ch5.c + ch6.c + ch7.c + ch8.c + ch9.c + ch10.c, lth.Ln)

    ,Ln = (abs(checksum(newid()) % (@maxlen - (@minlen - 1))) + @minlen)

    into dbo.JBMTest

    from sys.all_columns ac1

    cross join sys.all_columns ac2

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch1

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch2

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch3

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch4

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch5

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch6

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch7

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch8

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch9

    cross apply (select c = char(abs(checksum(newid())) % (ascii('Z') - ascii('A') + 1) + ascii('A'))

    ) ch10

    cross apply (select Ln = (abs(checksum(newid()) % (@maxlen - (@minlen - 1))) + @minlen)

    ) lth;

    alter table dbo.JBMTest

    add constraint PK_JBMTest primary key clustered (SomeID) with fillfactor = 90;

    select *

    from dbo.JBMTest jt;

    Thanks Phil! Excellent, I just played around with this, that's what I was looking for. Much faster and cleaner.

    Sorry for the late reply - I spent that day dealing with SQL emergencies.

    "I cant stress enough the importance of switching from a sequential files mindset to set-based thinking. After you make the switch, you can spend your time tuning and optimizing your queries instead of maintaining lengthy, poor-performing code."

    -- Itzik Ben-Gan 2001

  • J Livingston SQL (10/29/2015)


    see this article and read the comments from Jeff at the bottom of the article

    https://www.mssqltips.com/sqlservertip/3576/generate-random-strings-with-high-performance-with-a-sql-clr-function

    Thanks Mr. Livingston. I just installed and played around with both of Jeff Yao's CLRs that are in that assembly, excellent stuff. Thanks for mentioning the comments too - I have been looking for Jeff's fnTally function for awhile (I saw him mention it on a forum some months back but it's impossible to find where).

    Thanks again for your help sir!

    "I cant stress enough the importance of switching from a sequential files mindset to set-based thinking. After you make the switch, you can spend your time tuning and optimizing your queries instead of maintaining lengthy, poor-performing code."

    -- Itzik Ben-Gan 2001

Viewing 7 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic. Login to reply