strange thing I just ran into, not sure if this is a bug or what ... but pretty annoying.
In MS SQLServer 2000 :
SELECT (ISNUMERIC('0E010101'))
returns "1"
but
SELECT CAST ('0E010101' AS numeric)
returns "Error converting data type varchar to numeric"
any idea?
hi,
i am migrating data from a legacy system with a not nice front-end.
as a result, i have all sorts of garbage stored on the tables.
i
am trying to convert values from varchar(12) to float, but i have an
error during selecting data that says that data can not be converted
eventhough i am using the ISNUMERIC() function to check.
case when
isNumeric( myCol01 ) = 0 then null
else
convert( float , myCol01 )
end
but my error occours when ISNUMERIC() encounters the value '. ' ; that is a dot with spaces after it.
try the expression below;
SELECT
myValue = '. '
, is_numeric = ISNUMERIC('. ')
, converted = CONVERT( FLOAT , '. ')
has anyone got any idea how to work around this?
nicolas
Hi,
I have a table described as follows:
TableA
GrpDate DateTime;
grpMiscError varchar(1);
Ex Data:
2/1/2008 1
2/1/2008 1
2/1/2008 0
2/1/2008 x
2/1/2008 0
The grpMiscError can contain 0, 1 or x only. I need to sum up this column for all the zeros by a particular date.
I have the following but doesn't work:
SELECT
SUM(CASE ISNUMERIC(grpMiscError) WHEN 0 THEN 1 ELSE 0 END)AS MiscError
FROM TableA
WHERE GrpDate = '2/1/2008'
I get back an answer of 1 MiscError instead of 2
What am I doing wrong here?
Thanks,
J
Hi all,
I have never seen this problem in SQL Server 7. The isnumeric() returns true for non-numeric data, as show in the following example:
declare @x as varchar(5)
select @x = '00d01'
select isnumeric(@x)
I would expect the isnumeric() function to return false. Can anyone give a reason why this should occur.
Does SQL Server think that this is a hex value and is performing an implicit conversion? If so how horrible...
Nick
The isnumeric function returns 1 in examples where it should not:
select isnumeric('1D9') returns 1.
Can anyone explain this?
Hello,
I would like to perform a delete statement on one of my tables.
I have a code column that has data in the form 011-234-12
and at some point in this column the data is like R0 or D1
I want to delete the rows where the code data is R0 or D1, E3, etc...
I would like to know how to create an SQL Command for Sql Server 7 that would delete from Soumission_detail where:
first left char is not Numeric in the code column.
Thanks.
HiHere's the problem:I need to search a postcode database by the first one or two letters.Problems occur for example when i want to search north London postcodes (N) when using:postcode LIKE @postcode + '%' As this picks up everything beginning with N, eg, NG for Nottingham, or NE for Newcastle. So i need a like statement which searches for the first one or two digits followed by a number!I've found the ISNUMERIC() function but not sure what the best way to use it with the like statement - or even if there is a better way altogether - can you use regular expressions in MSSQL?thanks
View 4 Replies View Relatedsneaky, sneaky, sneaky
ISNUMERIC returns 1 when the input expression evaluates to a valid integer, floating point number, money or decimal type; otherwise it returns 0. A return value of 1 guarantees that expression can be converted to one of these numeric types.
thanks, but which one??
numeric as far as float is concerned, is not the same thing as numeric as far as money is concerned
create table isnumerics
( id integer not null identity
, txtfld varchar(11)
)
insert into isnumerics (txtfld) values ( '1' )
insert into isnumerics (txtfld) values ( '937' )
insert into isnumerics (txtfld) values ( '937.0' )
insert into isnumerics (txtfld) values ( '$937' )
insert into isnumerics (txtfld) values ( '$937.00' )
insert into isnumerics (txtfld) values ( 'free' )
insert into isnumerics (txtfld) values ( '.50' )
insert into isnumerics (txtfld) values ( '1,000' )
insert into isnumerics (txtfld) values ( '' )
select id
, txtfld
, isnumeric(txtfld)
from isnumerics
111
29371
3937.01
4$9371
5$937.001
6free0
7.501
81,0001
90
select id
, txtfld
, isnumeric(txtfld)
, case when isnumeric(txtfld) = 1
then cast(txtfld as money)
else cast(null as money)
end as case1
from isnumerics
1111.0000
29371937.0000
3937.01937.0000
4$9371937.0000
5$937.001937.0000
6free0
7.501.5000
81,00011000.0000
90
select id
, txtfld
, isnumeric(txtfld)
, case isnumeric(txtfld)
when 1
then cast(txtfld as money)
else cast(null as money)
end as case2
from isnumerics
1111.0000
29371937.0000
3937.01937.0000
4$9371937.0000
5$937.001937.0000
6free0
7.501.5000
81,00011000.0000
90
select id
, txtfld
, isnumeric(txtfld)
, case isnumeric(txtfld)
when 1
then cast(txtfld as float)
else cast(null as float)
end as case2
from isnumerics
1111.0
29371937.0
3937.01937.0
6free0
7.5010.5
the others got "Error converting data type varchar to float"
no, there wasn't a question here, but yes, i'd love to hear your comments
Hi,
I'm casting a varchar field to a decimal field using the format
CASE ISNUMERIC(GrossMktCapGbp)
WHEN 1 THEN CONVERT(DECIMAL(18,6),GrossMktCapGbp)
ELSE NULL
end
Thinking this would ensure that any spurious rows got set to null.
However I had a problem with some values that were set to '.', it seems that isnumeric thinks these are numbers but casting them to decimal produces an error.
SELECT ISNUMERIC('.')
SELECT CAST('.' AS DECIMAL(18,6))
Should I have been doing something different in my check possibly.
Sean
=iif(
IsNumeric(Fields!Accreditation.Value),
Int(Fields!Accreditation.Value),
Fields!Accreditation.Value
)
The above expression seems to work fine if Fields!Accreditation.Value is a number. However, if Fields!Accreditation.Value is not a number, it gives an #Error. Why is the true part evaluated when the expression is false?
Been meaning to post this for a while. It does a very limited job of only allowing [0-9], but could be extended to allow negative numbers, or numeric values that are suitable for numeric types other than INT, but avoiding the pitfalls of IsNumeric() which might allow through data not suitable for some of the numeric datatypes
IF EXISTS (SELECT * FROM dbo.sysobjects WHERE id = object_id(N'[dbo].[kk_fn_UTIL_IsINT]') AND xtype IN (N'FN', N'IF', N'TF'))
DROP FUNCTION dbo.kk_fn_UTIL_IsINT
GO
CREATE FUNCTION dbo.kk_fn_UTIL_IsINT
(
-- String to be tested - Must only contain [0-9], any spaces are trimmed
@strINTvarchar(8000)
)
RETURNS int-- NULL = Bad INT encountered, else cleanedup INT returned
/* WITH ENCRYPTION */
AS
/*
* kk_fn_UTIL_IsINTCheck that a String is a valid INT
*SELECT dbo.kk_fn_UTIL_IsINT(MyINTColumn)
*IF dbo.kk_fn_UTIL_IsINT(MyINTColumn) IS NULL ... Bad INT
*
* Returns:
*
*int valueValid integer
*NULLBad parameter passed
*
* HISTORY:
*
* 30-Sep-2005 Started
*/
BEGIN
DECLARE@intValueint
SELECT@strINT = LTRIM(RTRIM(@strINT)),
@intValue = CASE WHEN @strINT NOT LIKE '%[^0-9]%'
THEN CONVERT(int, @strINT)
ELSE NULL
END
RETURN @intValue
/** TEST RIG
SELECTdbo.kk_fn_UTIL_IsINT('123'),IsNumeric('123')
SELECTdbo.kk_fn_UTIL_IsINT(' 123 '),IsNumeric(' 123 ')
SELECTdbo.kk_fn_UTIL_IsINT('123.'),IsNumeric('123.')
SELECTdbo.kk_fn_UTIL_IsINT('123e2'),IsNumeric('123e2')
SELECTdbo.kk_fn_UTIL_IsINT('XYZ'),IsNumeric('XYZ')
SELECTdbo.kk_fn_UTIL_IsINT('-123'),IsNumeric('-123')
SELECTdbo.kk_fn_UTIL_IsINT('-'),IsNumeric('-')
**/
--==================== kk_fn_UTIL_IsINT ====================--
END
GO
Kristen
I would like to validate datatype using Derived Column.My data type are such as numeric(X,X),datetime,int, and varchar.How do I do this using Derived Column.Example if row does not qualify as ISNUMERIC()...throw it in ERROR table else send it to SUCCESS table.Any Idea ?
View 4 Replies View RelatedHello,
I have searched the forum, and have discovered that the DTS method using IsNumeric to check for numierc values (ActiveX) is not valid in SSIS. Most of what I have seen prescribes using the script component to handle this.
So formerly, I checked to see if a column was numeric. If it was, then I needed to use the numeric value as is, or in some cases, I needed to perform a calculation on the value and use the result. If the value was not numeric, then whatever the value was needed to be changed to zero.
Here is an example of how I would use the current value, or set the value to zero:
If IsNumeric(DTSSource("Col003")) Then DTSDestination("ADepTrnx") = CLng(DTSSource("Col003")) Else DTSDestination("ADepTrnx") = 0 End If
This is an example of how I would use the current value in a calculation, or set the value to zero:
If IsNumeric(DTSSource("Col012")) Then DTSDestination("AlliStdFee") = CLng(DTSSource("Col012"))/100 Else DTSDestination("AlliStdFee") = 0 End If
Does anyone have an example of how I would handle both situations in a script component?
Thank you for your help!
cdun2
Hello all!
I've wrote a small query for SQL 2005 and it's doesn't seem to work.
I have a table that contains two columns (X and Y), X is an int and Y is an nvarchar(50). I've populated this table with some data where Y contains numbers and some strings (e.g. "1", "2", "foo", etc). I've then got a view which only returns the rows where Y is numeric - now, I then query this table stating I only want numbers greater than 0 (i've casted the column) but this throws an error stating "foo" can't be casted. This is strange because the view doesn't return that.
What's going on? All of this works fine in SQL 2000 but not in SQL 2005 - looks like it's looking at the underlying table rather than the view. Sample code below to help you all out: -
Create Table
============
CREATE TABLE [dbo].[tblTest](
[X] [int] NOT NULL,
[nvarchar](50) NOT NULL
) ON [PRIMARY]
Insert Data
===========
INSERT INTO tblTest(X, Y) VALUES(1, '1')
INSERT INTO tblTest(X, Y) VALUES(1, '2')
INSERT INTO tblTest(X, Y) VALUES(2, 'foo')
INSERT INTO tblTest(X, Y) VALUES(2, 'bar')
Create View
===========
CREATE VIEW [dbo].[vwTest]
AS
SELECT X, Y
FROM dbo.tblTest
WHERE (ISNUMERIC(Y) = 1)
Finally
=======
SELECT X, Y
FROM dbo.vwTest
WHERE (CONVERT(int, Y) >= 0)
Msg 245, Level 16, State 1, Line 1
Conversion failed when converting the nvarchar value 'foo' to data type int.
Thanks in advance!
Does anyone else get the following result when running this query?
Select isnumeric('4D7')
-----------
1
(1 row(s) affected)
Does anyone know why this would return true for numeric?
Thanks,
Ray
I have a task (Derived Column Task) and I want to write something like this :
IsNumeric(aColumnOfString) == true ? "All numbers" : "there are some characters"
Here aColumnOfString can be something like "123a5" or 12345". I do not want to simply check if the left-most character is a number or not. I want to check the entire expression and return me a TRUE or false.
A TRUE is returned if the entire expression contains ONLY numbers, and FALSE otherwise.
I read some posting using regular expression. But that is not a solution for this situation.
Anyone knows how to accomplish this, please help!
Hi - Please excuse me if this is really simple, but I'm fairly new to this lark.
My (made up) code is below... I'd be grateful for any pointers.
insert into [tblInvoices]
(full_period,
supplier_no,
account_code,
tran_amount,
function)
select
full_period,
supplier_no,
account_code,
tran_amount
case
when substring(account_code,1,2) = 'FY' then '-'
when isNumeric(account_code) then left(account_code, 2)
when not isNumeric(substring(account_code,1,2)) then left(account_code, 1)
else 'oops'
end as function,
from
tblLoadMSV900_i
end
Is this even close?
I'm using a stored proc to insert the data from tblLoadMSV900_i into tblInvoices and at the same time insert some data into the function field.
In plain english I want make sure that:
If the first 2 chars of account_code are 'FY' then function='-',
If the first char of account_code is numeric then function=left(account_code, 2),
If the the first char is not numeric (and if first two chars are not 'FY' i.e. first char could be 'F') then function=left(account_code, 1)
And there's plenty more where this came from! But if I can crack this with your help then I should have a better idea about the rest of the proc.
Thanks
Sara
There is a MSSQL function that check the value. Like ISNULL(), ISDATE() and ISNUMERIC(). I don't see a function that check for decimal. If there isn't any then is there an user-defined function for it? I need to be able to validate the string value for decimal before it get assigned to a decimal datatype or T-SQL will run into an error.
I'm using MS-SQL 2000...
Thanks...
I found this to work:
SELECT uri, evFieldUri, evFieldVal
, CAST(evFieldVal
AS BIGINT)
FROM TSEXFIELDV
[Code] ....
It Returns:
uri
evFieldUri
evFieldVal
(No column name)
224016 3267
+000089243829 89243829
224019 2717
+000089243825 89243825
224472 3333
+000000000000000000000017 17
225052 3267
+000089243829 89243829
225055 2717
+000089243825 89243825
So, then I went back to:
SELECT uri, evFieldUri, evFieldVal
, CAST(evFieldVal
AS BIGINT)
FROM TSEXFIELDV
[Code] ....
And it returns this error:
Msg 8114, Level 16, State 5, Line 1
Error converting data type varchar to bigint.
So, I tried again, and this worked…
SELECT uri, evFieldUri, evFieldVal,CAST(evFieldVal
AS BIGINT),
ISNUMERIC(evFieldVal)
FROM TSEXFIELDV WHERE URI
> 0 AND evFieldUri
IN ( SELECT URI
FROM TSEXFIELD WHERE exFieldFormat
IN (1,11))
I logged out and came back and tried again, and it still worked. So then I tried…
SELECT uri, evFieldUri, evFieldVal,CAST(evFieldVal
AS BIGINT)
FROM TSEXFIELDV
WHERE URI
> 0
[Code] ...
And it fails.
Hi,
I have another issue. I have an excel file that I pipe through a "data conversion" task. I have set all the column data types to strings, because there's no way to know beforehand if a particular column will be number or text because the file is very non-standard (it looks more like a formatted report).
After the data conversion, I send all the rows to a script task. In the script task, I do a check on the numeric fields.
for example:
If Not IsNumeric(Row.Price) Then
Row.Price_IsNull = True
End If
However, this check fails each and every time, even if the field contains a number! I don't have this problem when using flat file sources.
So, none of my numeric fields are getting loaded to my ole db destination.
Help, is there a way around this? Or am I forced to just skip this number check altogether? I'd prefer not to.
Thanks