Adding space to middle of string

Question

Post reply

Adding space to middle of string

Gordon Pollokoff

SSCommitted

Points: 1930
More actions
February 9, 2005 at 11:07 am

#162638

I have a table with a column called displayname.

Values like 'John' or 'John Doe' are fine; values like 'JohnDoe' need to be converted to 'John Doe'.

The rule is simple: if an uppercase letter appears in the middle of a string, insert a single space in front of it.

Is there a way to do this other than parsing each string letter by letter and looking for an ascii value in the uppercase range? Can this be done using pattern matching on a database with a case insensitive collation?

Gordon Pollokoff

"Wile E. is my reality, Bugs Bunny is my goal" - Chuck Jones

Viewing 9 posts - 1 through 8 (of 8 total)

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

Ninja's_RGR'us SSC Guru Points: 294069 More actions · Answer 1

This is not the prettiest code I've ever written but it works...

CREATE TABLE [Numbers] (

[PkNumber] [int] IDENTITY (1, 1) NOT NULL ,

CONSTRAINT [Pk_Number] PRIMARY KEY CLUSTERED

(

[PkNumber]

) ON [PRIMARY]

GO

Declare @i as int

set @i = 0

while @i < 8000

begin

Insert into dbo.Numbers Default values

set @i = @i + 1

end

GO

Create table #test (DisplayName varchar(100))

Insert into #test (DisplayName)

Select 'John Doe'

union all

select 'JaneDoe'

union all

select 'John'

union all

select 'AFreakNameWithLotsOfNames'

Select * from #test

--Select DisplayName, PkNumber from #test cross join dbo.Numbers where PkNumber between 2 and LEN(DisplayName) AND ASCII(SUBSTRING(DisplayName, PkNumber, 1)) BETWEEN 65 and 90 AND SUBSTRING(DisplayName, PkNumber - 1, 1) ' '

while @@rowcount > 0

begin

Update T set T.DisplayName = LEFT(T.DisplayName, PkNumber - 1) + ' ' + RIGHT(T.DisplayName, LEN(T.DisplayName) - PkNumber + 1) FROM #test T inner join (Select DisplayName, PkNumber from #test cross join dbo.Numbers where PkNumber between 2 and LEN(DisplayName) AND ASCII(SUBSTRING(DisplayName, PkNumber, 1)) BETWEEN 65 and 90 AND SUBSTRING(DisplayName, PkNumber - 1, 1) ' ') dtReplaces on T.DisplayName = dtReplaces.DisplayName

end

Select * from #test

drop table #test

I had to put the Update in a while because I couldn't think of a way to update the name in one shot when more than 1 replace had to occur (besides making a UDF that would accept the word, scan it letter by letter and make the whole replace there... which would be kind of dumb since it's exactly what I'm trying to avoid with the Numbers table).

I think either way there's not gonna be any super fast solution to this problem. But hopefully it's only a one time deal.

Gordon Pollokoff SSCommitted Points: 1930 More actions · Answer 2

A quick google search turned up this solution:

select Ltrim (

Replace (

Replace (displayName

,'A', ' A')

,'B', ' B')

,'C', ' C')

,'D', ' D')

,'E', ' E')

,'F', ' F')

,'G', ' G')

,'H', ' H')

,'I', ' I')

,'J', ' J')

,'K', ' K')

,'L', ' L')

,'M', ' M')

,'N', ' N')

,'O', ' O')

,'P', ' P')

,'Q', ' Q')

,'R', ' R')

,'S', ' S')

,'T', ' T')

,'U', ' U')

,'V', ' V')

,'W', ' W')

,'X', ' X')

,'Y', ' Y')

,'Z', ' Z')

)

from mytable

where displayname not like '% %'

This query replaces every uppercase letter with ' ' and the uppercase letter and excludes records that already have the space (data specific since I know that all records are either one or two names). The LTRIM removes the space added to the beginning.

Gordon Pollokoff

"Wile E. is my reality, Bugs Bunny is my goal" - Chuck Jones

Ninja's_RGR'us SSC Guru Points: 294069 More actions · Answer 3

I've ran this on my test data and it returns incorrect results :

John Doe

JaneDoe

John

AFreakNameWithLotsOfNames

becomes

J A N E D O E

J O H N

A F R E A K N A M E W I T H L O T S O F N A M E S

Is it possible that the paste you made to post this message might have changed something in the output or is it just a collation issue?

PW-201837 SSC-Insane Points: 20805 More actions · Answer 4

I don't think this will be possible with a case-insensitive collation, even if you do use ASCII char values.

Example - uppercase D is ASCII 68. Try this, to place a space before the uppercase D:

Select replace('JodyDoe', Char(68), ' ' + char(68))

On my case-insensitive server, this matches both the 'd' and 'D' despite specifying an explicit ASCII value.

Ninja's_RGR'us SSC Guru Points: 294069 More actions · Answer 5

Ninja's_RGR'us

SSC Guru

Points: 294069

February 9, 2005 at 1:45 pm

#541458

Exactly my point ... my solution doesn't have this limit.

olavho SSChasing Mays Points: 646 More actions · Answer 6

olavho

SSChasing Mays

Points: 646

February 9, 2005 at 3:13 pm

#541472

I agree that Gordon's Google-based solution have problems when the collation is case insensitive, but why not force the collation to be case-sensitive just for the statement?

Replace displayName with (displayName collate Latin1_General_CS_AI), and this should do the trick.