July 29, 2010 at 2:35 pm
I'm no .NET programmer but i did a function that manipulates string inputs and if i set IsDeterministic=true on the CLR function it consumes around 40% cpu and also takes 40% longer to complete compared to not setting the IsDeterministic flag.
As im no CLR Integration expert, anyone cares to explain this?
--
Thiago Dantas
@DantHimself
July 29, 2010 at 11:07 pm
Can't help in the CLR area but if you demonstrate the string manipulation you're trying to do in a CLR, I can try to beat it in T-SQL. I've been successful at it a couple of times. Like anything else, "It Depends". 😉
--Jeff Moden
July 30, 2010 at 7:19 am
alright, here it goes.
T-SQL code (var names and common words are in portuguese)
CREATE FUNCTION [dbo].[UF_TRATASTRING]
(
@PALAVRA NVARCHAR(3000)
)
RETURNS NVARCHAR(3000)
AS BEGIN
DECLARE @INDICE INT
SET @INDICE = 0
DECLARE @P NVARCHAR(3000)
SET @P = ''
DECLARE @TEMP NVARCHAR(3000)
SET @TEMP = ''
DECLARE @ACERTA NVARCHAR(3000)
SET @ACERTA = @PALAVRA
--trim the word
SET @ACERTA = LTRIM(RTRIM(LOWER(@ACERTA)))
IF LEN(@ACERTA) > 1
BEGIN
--capitalize first character
SET @ACERTA = UPPER(SUBSTRING(@ACERTA, 1, 1)) + SUBSTRING(@ACERTA, 2, LEN(@ACERTA) - 1)
--remove excess dots
WHILE CHARINDEX('..', @ACERTA) > 0
BEGIN
SET @ACERTA = REPLACE(REPLACE(@ACERTA, '..', ''), '.', '')
END
--space after dot
SET @ACERTA = REPLACE(@ACERTA, '.', '. ')
-- remove excess spaces
WHILE CHARINDEX(' ', @ACERTA) > 0
BEGIN
SET @ACERTA = REPLACE(@ACERTA, ' ', ' ')
END
--remove invalid characters
SET @INDICE = 1
WHILE @INDICE <= DATALENGTH(@ACERTA)
BEGIN
IF ( ASCII(SUBSTRING(@ACERTA, @INDICE, 1)) < 32
OR ASCII(SUBSTRING(@ACERTA, @INDICE, 1)) = 63
OR ASCII(SUBSTRING(@ACERTA, @INDICE, 1)) = 39
)
AND ASCII(SUBSTRING(@ACERTA, @INDICE, 1)) <> 10
AND ASCII(SUBSTRING(@ACERTA, @INDICE, 1)) <> 13
SET @ACERTA = REPLACE(@ACERTA,
CHAR(ASCII(SUBSTRING(@ACERTA, @INDICE, 1))),
' ')
SET @INDICE = @INDICE + 1
END
--capitalize initials
SET @P = @ACERTA
SET @INDICE = 0
WHILE CHARINDEX(' ', @P) > 0
BEGIN
SET @INDICE = @INDICE + CHARINDEX(' ', @P)
SET @ACERTA = REPLACE(@ACERTA,
SUBSTRING(@ACERTA, @INDICE, 2),
UPPER(SUBSTRING(@ACERTA, @INDICE, 2)))
SET @TEMP = ''
SET @TEMP = SUBSTRING(@P, CHARINDEX(' ', @P) + 1, LEN(@P))
SET @P = @TEMP
END
-- change common words
SET @ACERTA = REPLACE(@ACERTA, ' DA ', ' da ')
SET @ACERTA = REPLACE(@ACERTA, ' DE ', ' de ')
SET @ACERTA = REPLACE(@ACERTA, ' DO ', ' do ')
SET @ACERTA = REPLACE(@ACERTA, ' E ', ' e ')
SET @ACERTA = REPLACE(@ACERTA, ' DAS ', ' das ')
SET @ACERTA = REPLACE(@ACERTA, ' DOS ', ' dos ')
SET @ACERTA = REPLACE(@ACERTA, ' COM ', ' com ')
SET @ACERTA = REPLACE(@ACERTA, ' UM ', ' um ')
END
ELSE
SET @ACERTA = UPPER(SUBSTRING(@ACERTA, 1, 1))
RETURN ISNULL(@ACERTA, '')
END
and the VB.NET code (note, i didnt write any, the VB one is used inside our web app and the T-SQL was used on the ETL process and inside some SPs, i myself don't understand very well what the VB code is doing, all i know is it returns the same as above but faster)
<Microsoft.SqlServer.Server.SqlFunction(Name:="UF_TRATASTRING")> _
Public Shared Function TRATASTRING(<SqlFacet(MaxSize:=3000)> ByVal strTexto As String) As <SqlFacet(MaxSize:=3000)> String
Dim strCharAnterior As String = "."
Dim strChar As String
Dim x As Integer = 0
Dim intPosicaoPrimeiroEspaco As Integer = 0
Dim strRetorno As New StringBuilder
Dim strPrimeiraPalavra As String = ""
If strTexto = Nothing Then
strTexto = ""
Else
strTexto = strTexto.ToLower
For Each strChar In strTexto
If Asc(strChar) < 32 AndAlso Asc(strChar) <> 13 AndAlso Asc(strChar) <> 10 Then
strChar = " "
End If
If Asc(strCharAnterior) = 13 OrElse Asc(strCharAnterior) = 10 Then
strCharAnterior = "."
End If
If strCharAnterior = " " Then
If intPosicaoPrimeiroEspaco = 0 Then
intPosicaoPrimeiroEspaco = x
End If
If strChar <> " " Then
If strTexto.Substring(x).Length >= 2 Then
Select Case strTexto.Substring(x, 2)
Case "e "
strRetorno.Append(strChar)
Case Else
If strTexto.Substring(x).Length >= 3 Then
Select Case strTexto.Substring(x, 3)
Case "da "
strRetorno.Append(strChar)
Case "de "
strRetorno.Append(strChar)
Case "do "
strRetorno.Append(strChar)
Case Else
If strTexto.Substring(x).Length >= 4 Then
Select Case strTexto.Substring(x, 4)
Case "das "
strRetorno.Append(strChar)
Case "dos "
strRetorno.Append(strChar)
Case Else
strRetorno.Append(strChar.ToUpper)
End Select
Else
strRetorno.Append(strChar.ToUpper)
End If
End Select
Else
strRetorno.Append(strChar.ToUpper)
End If
End Select
Else
If x < strTexto.Length - 1 AndAlso strTexto.Substring(x, 1) = strTexto.Substring(x, 1).ToUpper AndAlso strTexto.Substring(x + 1, 1) = strTexto.Substring(x + 1, 1).ToLower Then
strRetorno.Append(strChar.ToUpper)
Else
strRetorno.Append(strChar)
End If
End If
End If
ElseIf strCharAnterior <> "." Then
strRetorno.Append(strChar)
Else
If strChar = " " Then
strChar = "."
If strRetorno.Length > 0 AndAlso strCharAnterior <> "." Then
strRetorno.Append(" ")
End If
strRetorno.Append(strChar.ToUpper)
Else
If x < strTexto.Length - 1 AndAlso strTexto.Substring(x, 1) = strTexto.Substring(x, 1).ToUpper AndAlso strTexto.Substring(x + 1, 1) = strTexto.Substring(x + 1, 1).ToLower Then
strRetorno.Append(strChar.ToUpper)
Else
strRetorno.Append(strChar)
End If
End If
End If
strCharAnterior = strChar
x += 1
Next
End If
Return New SqlString(strRetorno.ToString().Replace(".", ". ").Trim)
End Function
--
Thiago Dantas
@DantHimself
July 31, 2010 at 1:53 pm
Try this, Thiango... it may even give your .Net code a run for it's money. It's 32 to 64 times faster than the T-SQL Function you wrote. If you didn't have the requirement to do "initial caps", it would be faster still and we could turn it into an iTVF instead of a relatively slow scalar function.
(Side bar... your function didn't work correctly... it would get rid of all periods if ANY two were adjacent)
Although it looks very wierd, you have to understand that nested REPLACE's are the fastest thing this side of a bullet. Let me know if you have any questions. Here's the function... it's heavily documented, as well...
CREATE FUNCTION dbo.UF_TRATASTRING
(@PALAVRA NVARCHAR(4000))
--===== Created by Jeff Moden
RETURNS NVARCHAR(4000) AS
BEGIN
SELECT @PALAVRA =
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE( --Dedupe Spaces (Good past 8k).
REPLACE( --Add space after periods.
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE( --Dedupe Periods (Good past 8k).
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE( --Remove unwanted characters
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE( --Remove unwanted characters
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE( --Remove unwanted characters
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE( --Remove unwanted characters
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE( --Remove unwanted characters
REPLACE(REPLACE(REPLACE(REPLACE(REPLACE( --Remove unwanted characters
REPLACE(REPLACE( --Remove unwanted characters
LTRIM(RTRIM(LOWER(@PALAVRA COLLATE LATIN1_GENERAL_BIN ))) --Trim and change to lower case.
,NCHAR(0),'') ,NCHAR(1),'') ,NCHAR(2),'') ,NCHAR(3),'') ,NCHAR(4),'') --Remove unwanted characters
,NCHAR(5),'') ,NCHAR(6),'') ,NCHAR(7),'') ,NCHAR(8),'') ,NCHAR(9),'') --Remove unwanted characters
,NCHAR(11),''),NCHAR(12),''),NCHAR(14),''),NCHAR(15),''),NCHAR(16),'') --Remove unwanted characters
,NCHAR(17),''),NCHAR(18),''),NCHAR(19),''),NCHAR(20),''),NCHAR(21),'') --Remove unwanted characters
,NCHAR(22),''),NCHAR(23),''),NCHAR(24),''),NCHAR(25),''),NCHAR(26),'') --Remove unwanted characters
,NCHAR(27),''),NCHAR(28),''),NCHAR(29),''),NCHAR(30),''),NCHAR(31),'') --Remove unwanted characters
,NCHAR(39),''),NCHAR(63),'') --Remove unwanted characters
,N'.................................',N'.') --32+1 Dedupe Periods
,N'.................',N'.') --16+1
,N'.........',N'.') --8+1
,N'.....',N'.') --4+1
,N'...',N'.') --2+1
,N'..',N'.') --1+1
,N'.',N'. ') --Add space after periods
,N' ',N' ') --32+1 Dedupe Spaces
,N' ',N' ') --16+1
,N' ',N' ') --8+1
,N' ',N' ') --4+1
,N' ',N' ') --2+1
,N' ',N' ') --1+1
--===== Do the initial caps thing.
-- The cascaded CTE's build an on-the-fly-memory-only-tally "table".
;WITH
E1(N) AS ( --=== Create eight 1's
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 --8
),
E2(N) AS (SELECT 1 FROM E1 a, E1 b), --64
E4(N) AS (SELECT 1 FROM E2 a, E2 b), --4,096
cteTally(N) AS (SELECT ROW_NUMBER() OVER (ORDER BY (SELECT NULL)) FROM E4)
SELECT @PALAVRA = STUFF(@PALAVRA,t.N,1,UPPER(SUBSTRING(@PALAVRA,t.N,1)))
FROM cteTally t
WHERE t.N <= LEN(@PALAVRA COLLATE LATIN1_GENERAL_BIN )
AND SUBSTRING(@PALAVRA COLLATE LATIN1_GENERAL_BIN,t.N-1,2) LIKE N'[ ][^ ]' --Space followed by nonblank
;
--===== Change common words back to lower case
SELECT @PALAVRA = REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(REPLACE(
@PALAVRA COLLATE LATIN1_GENERAL_BIN
,N' DA ' , N' da ' )
,N' DE ' , N' de ' )
,N' DO ' , N' do ' )
,N' E ' , N' e ' )
,N' DAS ', N' das ')
,N' DOS ', N' dos ')
,N' COM ', N' com ')
,N' UM ' , N' um ' )
--==== First character Init and Null conversion
RETURN ISNULL(STUFF(@PALAVRA,1,1,UPPER(SUBSTRING(@PALAVRA,1,1))),N'')
END
--Jeff Moden
July 31, 2010 at 9:50 pm
wow Jeff, thats a nice load of replaces lol. ill test it first thing on monday and post the results against the old T-SQL and the CLR functions. Thanks for the help
--
Thiago Dantas
@DantHimself
August 1, 2010 at 12:00 pm
dant12 (7/31/2010)
wow Jeff, thats a nice load of replaces lol. ill test it first thing on monday and post the results against the old T-SQL and the CLR functions. Thanks for the help
Heh... and don't think for a minute that I actually typed all that stuff. 😛 Looking forward to your tests and thanks for the feedback.
--Jeff Moden
August 1, 2010 at 2:54 pm
Jeff Moden (8/1/2010)
dant12 (7/31/2010)
wow Jeff, thats a nice load of replaces lol. ill test it first thing on monday and post the results against the old T-SQL and the CLR functions. Thanks for the helpHeh... and don't think for a minute that I actually typed all that stuff. 😛 Looking forward to your tests and thanks for the feedback.
how'd u do it then?
--
Thiago Dantas
@DantHimself
August 1, 2010 at 4:56 pm
dant12 (8/1/2010)
Jeff Moden (8/1/2010)
dant12 (7/31/2010)
wow Jeff, thats a nice load of replaces lol. ill test it first thing on monday and post the results against the old T-SQL and the CLR functions. Thanks for the helpHeh... and don't think for a minute that I actually typed all that stuff. 😛 Looking forward to your tests and thanks for the feedback.
how'd u do it then?
I let the machine do it. Here's an example for a project I'm working on right now. The code is really easy to convert from a basic bit of boiler plate code... the addition of a CR here and there is all that needs to be done to make the output exactly the way I want it...
DECLARE @SQL1 VARCHAR(8000),
@SQL2 VARCHAR(8000),
@SQL3 VARCHAR(8000)
SELECT @SQL1 = ISNULL(@SQL1,'') + 'REPLACE(' + CASE WHEN (t.n-1)%5 = 4 THEN CHAR(10) ELSE '' END
FROM dbo.Tally t
WHERE t.N BETWEEN 1 AND 128
AND CHAR(t.N-1) COLLATE LATIN1_GENERAL_BIN NOT LIKE '[a-zA-Z0-9. ,]' COLLATE LATIN1_GENERAL_BIN
SELECT @SQL2 = 'DirtyString COLLATE LATIN1_GENERAL_BIN'
SELECT @SQL3 = ISNULL(@SQL3,'')+ ',CHAR('+CAST(t.N-1 AS VARCHAR(3))+'),'''')'+ CASE WHEN (t.n-1)%5 = 4 THEN CHAR(10) ELSE '' END
FROM dbo.Tally t
WHERE t.N BETWEEN 1 AND 128
AND CHAR(t.N-1) COLLATE LATIN1_GENERAL_BIN NOT LIKE '[a-zA-Z0-9. ,]' COLLATE LATIN1_GENERAL_BIN
PRINT @SQL1
PRINT @SQL2
PRINT @SQL3
--Jeff Moden
August 2, 2010 at 9:03 am
dunno whats wrong with your function but it performed way slower than the old one
DECLARE @a DATETIME = GETDATE() -- CLR
SELECT DBO.UF_TRATASTRING(NOME_DEV) FROM TB_DEVEDOR
SELECT DATEDIFF(MS,@A,GETDATE())
GO
DECLARE @a DATETIME = GETDATE() -- JEFF T-SQL
SELECT DBO.UF_TRATASTRING2(NOME_DEV) FROM TB_DEVEDOR
SELECT DATEDIFF(MS,@A,GETDATE())
GO
DECLARE @a DATETIME = GETDATE() -- OLD SQL
SELECT DBO.UF_TRATASTRING3(NOME_DEV) FROM TB_DEVEDOR
SELECT DATEDIFF(MS,@A,GETDATE())
GO
/*
FIRST RUN
CLR: CPU time = 5678 ms, elapsed time = 6110 ms.
JEFF TSQL: CPU time = 788539 ms, elapsed time = 1647436 ms.
OLD TSQL: CPU time = 103164 ms, elapsed time = 119757 ms.
SECOND RUN
CLR: CPU time = 5709 ms, elapsed time = 6146 ms.
JEFF TSQL: CPU time = 792173 ms, elapsed time = 1553814 ms.
OLD TSQL: CPU time = 103101 ms, elapsed time = 119565 ms.
*/
that table has 646184 rows and the column is of variable lenght. those results are from my development machine, ill test it a better one and post the results again
ps: this post has derailed greatly lol. I'm still curious about my deterministic vs non-deterministic inquiry
///EDIT
here goes the results from our production-like server (table with 381519 records, same query as above)
/*
FIRST RUN
CLR: CPU time = 2641 ms, elapsed time = 9114 ms.
JEFF TSQL: CPU time = 16844 ms, elapsed time = 17604 ms.
OLD TSQL: CPU time = 66484 ms, elapsed time = 72546 ms.
SECOND RUN
CLR: CPU time = 2562 ms, elapsed time = 13587 ms.
JEFF TSQL: CPU time = 17094 ms, elapsed time = 21931 ms.
OLD TSQL: CPU time = 66484 ms, elapsed time = 72956 ms.
*/
now im really puzzled, why on my dev machine it got really really slow
///EDIT 2
ran the queries again on the same server as above but on a different database on a table with 7559955 records. the results were
/*
FIRST RUN
CLR: CPU time = 58829 ms, elapsed time = 440024 ms.
JEFF TSQL: CPU time = 395968 ms, elapsed time = 776934 ms.
OLD TSQL: CPU time = 1607407 ms, elapsed time = 1755791 ms.
SECOND RUN
CLR: CPU time = 60657 ms, elapsed time = 213006 ms.
JEFF TSQL: CPU time = 385547 ms, elapsed time = 408962 ms.
OLD TSQL: CPU time = 1607187 ms, elapsed time = 1764218 ms.
*/
sorry Jeff, seems the CLR beat you this time
--
Thiago Dantas
@DantHimself
August 2, 2010 at 4:51 pm
dant12 (8/2/2010)
dunno whats wrong with your function but it performed way slower than the old one
Heh... dunno what's wrong with your machine, db, or table, but the function I wrote was just like I said... 32 to 64 times faster depending on whether or not the heartbeat of the machine kicked in during the test run.
I'll set it all up again tonight to see what the difference is... any chance of you providing the table definition and some readily consumable sample data? See the first link in my sginature line below for how to do that.
--Jeff Moden
August 2, 2010 at 8:26 pm
yeah sure ill send you a .csv if your willing to play with it, but i wouldnt bother so much. you already gave me lots of ideas with both ur sample codes
--
Thiago Dantas
@DantHimself
August 3, 2010 at 6:38 am
dant12 (8/2/2010)
yeah sure ill send you a .csv if your willing to play with it, but i wouldnt bother so much. you already gave me lots of ideas with both ur sample codes
That would be very cool. The reason I'm asking is because I want to try to find what went wrong. On 8K rows, your function took about 16 seconds to run and mine took 1/4 of a second. Yup... I've got an older slower machine but that a huge difference from what you posted as time and I wanted to find out why.
Thanks.
--Jeff Moden
August 3, 2010 at 7:22 am
Jeff Moden (8/3/2010)
dant12 (8/2/2010)
yeah sure ill send you a .csv if your willing to play with it, but i wouldnt bother so much. you already gave me lots of ideas with both ur sample codesThat would be very cool. The reason I'm asking is because I want to try to find what went wrong. On 8K rows, your function took about 16 seconds to run and mine took 1/4 of a second. Yup... I've got an older slower machine but that a huge difference from what you posted as time and I wanted to find out why.
Thanks.
http://www.mediafire.com/?phfjo9005ug2wl2
thats the whole 676k sample of the entire table along with the generated create table script
--
Thiago Dantas
@DantHimself
August 3, 2010 at 12:31 pm
dant12 (8/3/2010)
Jeff Moden (8/3/2010)
dant12 (8/2/2010)
yeah sure ill send you a .csv if your willing to play with it, but i wouldnt bother so much. you already gave me lots of ideas with both ur sample codesThat would be very cool. The reason I'm asking is because I want to try to find what went wrong. On 8K rows, your function took about 16 seconds to run and mine took 1/4 of a second. Yup... I've got an older slower machine but that a huge difference from what you posted as time and I wanted to find out why.
Thanks.
http://www.mediafire.com/?phfjo9005ug2wl2
thats the whole 676k sample of the entire table along with the generated create table script
Lunch time, thought I'd give it a go. Can't access your data as it is blocked. Could you upload as a text file a small sample with the create table and insert already done so I can test what I wrote? Also, expected results based on the small sample would help me validate my process.
August 3, 2010 at 4:03 pm
Lynn Pettis (8/3/2010)
dant12 (8/3/2010)
Jeff Moden (8/3/2010)
dant12 (8/2/2010)
yeah sure ill send you a .csv if your willing to play with it, but i wouldnt bother so much. you already gave me lots of ideas with both ur sample codesThat would be very cool. The reason I'm asking is because I want to try to find what went wrong. On 8K rows, your function took about 16 seconds to run and mine took 1/4 of a second. Yup... I've got an older slower machine but that a huge difference from what you posted as time and I wanted to find out why.
Thanks.
http://www.mediafire.com/?phfjo9005ug2wl2
thats the whole 676k sample of the entire table along with the generated create table script
Lunch time, thought I'd give it a go. Can't access your data as it is blocked. Could you upload as a text file a small sample with the create table and insert already done so I can test what I wrote? Also, expected results based on the small sample would help me validate my process.
tested the mediafire link and worked with no problems, you have winrar installed?
ill provide a small sample either way
IF OBJECT_ID('TEMP_TAB') IS NOT NULL
DROP TABLE TEMP_TAB
CREATE TABLE TEMP_TAB (NAME VARCHAR(60))
INSERT INTO TEMP_TAB
SELECT 'Ponto A - Edicao e Impressao de Prod' UNION ALL
SELECT '& e S Comercio de Representacoes de Moto' UNION ALL
SELECT '. Eucelma do Carmo de O Martins' UNION ALL
SELECT '. Irlane Valquiria de Sena Gomes' UNION ALL
SELECT '. Lessandro Pinheiro de Oliveira' UNION ALL
SELECT '00071rosangela Greco Moura Fonseca' UNION ALL
SELECT '007maria Aprecida Fidelis da S Conceicao' UNION ALL
SELECT '1 Oficio do Registro Civil e Anexos' UNION ALL
SELECT '10 Igreja Evangelica Pentecostal O Brasi' UNION ALL
SELECT '100 /, Milho Verde e Pamonharia Ltda-me' UNION ALL
SELECT '1000 Razoes Comercio de Artigos do Vestu' UNION ALL
SELECT '1144851sonia Maria Mendes da Silva' UNION ALL
SELECT '14 Bis Materiais Para Construcao Ltda -' UNION ALL
SELECT '1a. Igreja Batista Em Navirai-ms' UNION ALL
SELECT '2 Andrades Treinamento e Consultoria Ltd' UNION ALL
SELECT '2 Irmaos Prestadora de Servicos de Telec' UNION ALL
SELECT '206 Fashion, Ws Couture e San Philipo Co' UNION ALL
SELECT '21 de Abril Associacao Unica das Pracas' UNION ALL
SELECT '27maria de Lourdes Bertoldo dos Santos' UNION ALL
SELECT '2831 Estudio Comunicacao e Design Ltda' UNION ALL
SELECT '2dulcicleia Fernandes dos Santos' UNION ALL
SELECT '2-m Restaurante e Choperia Ltda Me' UNION ALL
SELECT '2mm Eletro Telecomunicacoes Comercio Rep' UNION ALL
SELECT '3 M Industria e Comercio de Cafe Ltda.' UNION ALL
SELECT '3 Mmm G Comercio e Representacoes Ltda' UNION ALL
SELECT '3. Milenio Operadora de Sorteio de Bingo' UNION ALL
SELECT '3a - Servicos de Filmagens e Locacoes De' UNION ALL
SELECT '3r Ferraro Importacao e Exportacao Ltda' UNION ALL
SELECT '4 Rodas Auto Pecas e Vidros Ltda' UNION ALL
SELECT '4-in Comunicacao e Design Ltda Me' UNION ALL
SELECT '4u Network - Comercio de Equipamentos E' UNION ALL
SELECT '87226921katia Maria Andrade Lima' UNION ALL
SELECT '888782849josue Moreira da Silva' UNION ALL
SELECT '92322370mirella Maria da C de Carvalho' UNION ALL
SELECT 'A & J Informatica Brasilia Ltda' UNION ALL
SELECT 'A & T Comercio de Equipamentos Eletroele' UNION ALL
SELECT 'A & T Telecom Solucoes Empresariais e Co' UNION ALL
SELECT 'A . Comercial de Aluminios Braziliense L' UNION ALL
SELECT 'A + Design Criatividade Grafica e Digita' UNION ALL
SELECT 'A A Agata Divisorias, Forros, e Pisos Lt' UNION ALL
SELECT 'A A Francischini & Francischini Ltda Me' UNION ALL
SELECT 'A Anti Pragas Desinfestacoes Ltda' UNION ALL
SELECT 'A Associacao DOS Moradores DA Vila Uniao' UNION ALL
SELECT 'A C D Assessoria Empresarial Ltda Me' UNION ALL
SELECT 'A C G Compra e Venda de Bens Imoveis Ltd' UNION ALL
SELECT 'A C N Eventos e Promocoes Artisticas Ltd' UNION ALL
SELECT 'A C Silva Transportadora Ltda Me' UNION ALL
SELECT 'A Casa do Hospital Comercio de Produtos' UNION ALL
SELECT 'A Construlimp Materiais de Construcao E' UNION ALL
SELECT 'A D de Alcantara Neto-multivendas'
SELECT * FROM TEMP_TAB
--
Thiago Dantas
@DantHimself
Viewing 15 posts - 1 through 15 (of 23 total)
You must be logged in to reply to this topic. Login to reply