Oct 13, 2006
I am just getting started studying SSIS with Kirk Haselden's "Integration Services" book. The problem I am trying to solve would seem easy enough to solve in code, but I am still early in the book and would like to be able to focus on the aspects of SSIS that would help me expedite this with SSIS, or to find out early whether what I need to do cannot be easily done.
The problem itself is simple enough: I have a database of roughly 100 tables. Ignoring the poor normalization in the database for the moment, my more pressing problem is that that I need to rekey all of the main OLTP tables from a mashup of different key schemes to UNIQUEIDENTIFIERS. For example, Client table is presenrly keyed on an INT, Client Number. ClientFile table is keyed as FileType = NVARCHAR(2), ClientNumber INT, FileNumber INT (incrementing, meaningless number). Child tables to ClientFile are the same key structure as ClientFile, plus yet another (incrementing, meaningless number) INT, etc like this. I would like to know if and how or where I should be looking to convert the Client table to a UNIQUEIDENTIFIER key, and the same for the ClientFile, makes its key also a GUID, and have a reference to the new Client tables GUID key as a foreign key in the ClientFile table, and on and on like that. The Client is at the top of the food chain.
In essence, I would like to have every table's key be called ID and be a UNIQUEIDENTIFIER (GuidRow), and I would like for example, the ClientFile table to reference the Client table with a column named ClientID. I would like ClientFile's children to have a foreign key called ClientFileID, and their own keys to be ID (RowGuid).
There are also several lookups in each table where, of course, the actual string values were stored instead of a key to the value (i.e. full state or country name instead of a code from a state or country lookup table) that I need to convert to something more sensible, like replacing the state name with a state code and a country name with a country code and link to appropriate respective tables. :) In fact, some of the values I need to break out from columns could also be keyed with RowGuids as well and I would much prefer to use those than string or INT keys.
Other than those problems, most of the rest of the data in those tables could essentially ba a straight copy operation since the source database is SQL Server 2000 and moving to SQL 2005 (one notable exception is that I am converting ntext columns to nvarchar(MAX) columns.
I am assuming this is probably ridiculously simple and I just haven't found my way there yet, but I still have much of this book and the help files to go through and the Index didn't give me any comfort that this was something I will or will not be able to do easily.
The real help I am looking for is two fold: a) somebody tell me to stop reading this 700 page (very well written) tome if I would be better off writing this all in code myself, and b) if this is something that most of you could do with SSIS with both hands tied behind your back, please at least help me focus on the important transforms and tools so that I don't smend a month becoming a data warehouse wizard and ultimately not solve the problem I am most concerned with.
Please be mecriful with the heat, I have already confessed that I am new to this and am scrambling to come up to speed as fast as I can, but am beginning to think this problem is either to trivial for coverage in this book, or perhaps just not what SSIS was designed to do.
Thanks much in advance for any guidance.
View 10 Replies
View Related
Mar 31, 2008
Hi there, I have the following set of records: (DDL and data provided)
create table dbo.myNames (
FilePersonId int,
Sex varchar(20),
Title varchar(20),
Forename varchar(50),
Surname varchar(50)
)
insert into dbo.myNames (FilePersonId, Sex, Title, Forename, Surname)
select 477827, 'F', 'MISS', 'NATALIE', 'MR NIEL MC DONOUGH' union all
select 516992, 'M', 'MR', 'MARIO.', 'MISS CARLA THRINCHEIRAS' union all
select 1725198, 'M', 'MR', 'THOMAS', 'MS TAGGERT' union all
select 1907673, 'M', 'MR', 'COLIN', 'MR C PATTERSON' union all
select 2282432, 'M', 'MR', 'JOHN', 'MR GOLDSMITH' union all
select 3172847, 'M', 'MR', 'JASON', 'MR JASON MARCUS RENDALL' union all
select 3173628, 'M', 'MR', 'STUART', 'MR STUART IAN JAMES CALLOW' union all
select 3174343, 'M', 'MR', 'JOHANNES', 'MR JOHANNES ARIE JAN DE JAGER' union all
select 3176784, 'F', 'MISS', 'MONICA', 'MISS MONICA S F ALMEIDA' union all
select 3177717, 'M', 'MR', 'MOHAMMAD', 'MR MOHAMMAD NASIR KHAN SABIR' union all
select 3180017, 'M', 'MR', 'MD', 'MR MD. MAMUN UR RASHID' union all
select 3180585, 'F', 'MISS', 'KATHRYN', 'MISS KATHRYN M J GRANT' union all
select 3182306, 'M', 'MR', 'ABDUL', 'MR ABUL K M H KABIR' union all
select 3182416, 'F', 'MRS', 'CATHERINE', 'MRS CATHERINE ANN LEA DOIG' union all
select 3183759, 'M', 'MR', 'PHILIP', 'MR PHILIP ADAM MOXON' union all
select 3183790, 'F', 'MISS', 'KELLY', 'MISS KELLY NEAVE CHAPPELL' union all
select 3191895, 'F', 'MISS', 'LUCY', 'MISS LUCY ANNE MAY GAULT' union all
select 3192189, 'M', 'MR', 'JOAQUIM', 'MR JOAQUIM FILIPE ABRANTES VAZ' union all
select 3192261, 'F', 'MRS', 'C', 'MRS C D V BALBO-BELL' union all
select 3193077, 'M', 'MR', 'NIDAL', 'MR NIDAL FATHEE ABY SABEI' union all
select 3194554, 'M', 'MR', 'STUART', 'MR STUART ROBERT JAMES DOYLE' union all
select 3201437, 'M', 'MR', 'IAN', 'MR IAN' union all
select 3206600, 'M', 'MR', 'ADRIAN', 'MR ADRIAN JOHN JAMES OWEN' union all
select 3206637, 'M', 'MR', 'J', 'MR J O GARCIA CHAURAN' union all
select 3206695, 'M', 'MR', 'SYED', 'MR SYED SAFDAR SHAH' union all
select 3206744, 'M', 'MR', 'LEE', 'MR LEE ALAN JOEL MAYNARD' union all
select 3206763, 'F', 'MISS', 'TANIA', 'MISS TANIA RACHEL NASH' union all
select 3207104, 'M', 'MR', 'ALEXANDER', 'MR ALEXANDER C D HOPWOOD' union all
select 3207577, 'M', 'MR', 'J', 'MR J H S CAMPBELL' union all
select 3207611, 'F', 'MISS', 'TERRI', 'MISS TERRI LOUISE KATE FLEMING' union all
select 3209050, 'F', 'MISS', 'MELISSA', 'MISS MELISSA NEKITA MAKEMSON' union all
select 3213291, 'M', 'MR', 'MALCOLM', 'MR MALCOLM PETER FORD CROOK' union all
select 3215920, 'M', 'MR', 'IAIN', 'MR IAIN ALASTAIR SMITH DOWNEY' union all
select 3889047, 'M', 'MR', 'PEDORA', 'MR LOFARO' union all
select 3974174, 'M', 'MR', 'MICHAEL', 'MR MICHAEL J KERR' union all
select 3977713, 'M', 'MR', 'XAVIER', 'MR XAVIER DE PONTE DE SOUSA' union all
select 3977753, 'M', 'MR', 'JEAN-DANIEL', 'MR JEAN-DANIEL KENNY FLORIAN' union all
select 6815795, 'U', 'UNKNOWN OR INVALID', 'M', 'MR M WOOD' union all
select 7617439, 'F', 'MRS', 'M', 'MRS MARYMARSHALL' union all
select 7929645, 'M', 'MR', 'RICHARD', 'MR PANTER' union all
select 9092854, 'M', 'MR', 'JARED', 'MR JARED ELVANDO BLAKE' union all
select 9109880, 'M', 'MR', 'MICHAEL', 'MR MICHAEL PAUL LLOYD RAMWELL' union all
select 9165326, 'M', 'MR', 'SCOTT', 'MR SCOTT DE LA FUENTE' union all
select 9321272, 'M', 'MR', 'MPHAMMAD', 'MR MPHAMMAD HAMIDI TABAR' union all
select 9461030, 'M', 'MR', 'KELVIN', 'MR KELVIN POH ANN TAN' union all
select 9508257, 'M', 'MR', 'HARDI', 'MR HARDI M IBRAHIM' union all
select 7899594, 'F', 'MRS', 'SANDRA', 'MRSTHPOMPSON' union all
select 9616820, 'U', 'UNKNOWN OR INVALID', 'HAMMOND', 'MRJOHNHAMMOND' union all
select 9617471, 'M', 'MR', 'PAUL', 'MRPAULWAYMAN' union all
select 9817442, 'M', 'MR', 'ADAM', 'MRADAMMARLEY' union all
select 9853852, 'F', 'MISS', 'P', 'MRBARRYGILMARTIN' union all
select 9895683, 'F', 'MRS', 'REBECCA', 'MRSREBECCAWILSON' union all
select 9924241, 'M', 'MR', 'JOHN', 'MRJOHNWORDEN' union all
select 9941612, 'M', 'MR', 'BRIAN', 'MRBRIANOKEEFFE' union all
select 10214780, 'M', 'MR', 'G', 'MRJAMESWADDLE'
What I am trying to do is cleanse the Surname field. I need to remove the Title/Initials and/or Forename from the Title field. But I am struggling to come up with a neat solution to this problem.
Your help would be much appreciated.
Thanks.
View 3 Replies
View Related
