De-Concatenate multi-select picklist and transform into target multi-select picklist values

  • Hi, 

    I have a requirement where I have to tranform an existing multi-select picklist column.

    The source data exists in the folllowing format:

    Source

    3G Contact::C.E.O Contact::External Newsletter Subscriber::GSMA Press::Interconnection Contact::Speaker Contact

    There are a set number of target values in the picklist and the delimiter of the target db is a single semicolon.

    Can anyone please suggest a way in which I can transform the delimiter and write a statement which checks the string for any of the source values and then maps that to it's target value?

    Thanks in advance!

    Chris

  • Sounds a little like a homework problem, so I'll ask what have you tried?   Have you considered using the REPLACE function?   As you've provided information only about a single input, and no information about the rest of the problem other than having to check this string for a particular value, I'd have to make far too many assumptions to provide any useful assistance.   How about you look up the REPLACE function and post back what you've tried and we'll help you from there?

    Steve (aka sgmunson) 🙂 🙂 🙂
    Rent Servers for Income (picks and shovels strategy)

  • aldous.chris - Monday, March 6, 2017 5:59 AM

    Hi, 

    I have a requirement where I have to tranform an existing multi-select picklist column.

    The source data exists in the folllowing format:

    Source

    3G Contact::C.E.O Contact::External Newsletter Subscriber::GSMA Press::Interconnection Contact::Speaker Contact

    There are a set number of target values in the picklist and the delimiter of the target db is a single semicolon.

    Can anyone please suggest a way in which I can transform the delimiter and write a statement which checks the string for any of the source values and then maps that to it's target value?

    Thanks in advance!

    Chris

    I'm not 100% sure what you are asking for. Are you trying to replace one delimiter with another, or is there more to it than that? Perhaps you could provide the result you are expecting, based on the sample data?

    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

  • aldous.chris - Monday, March 6, 2017 5:59 AM

    Hi, 

    I have a requirement where I have to tranform an existing multi-select picklist column.

    The source data exists in the folllowing format:

    Source

    3G Contact::C.E.O Contact::External Newsletter Subscriber::GSMA Press::Interconnection Contact::Speaker Contact

    There are a set number of target values in the picklist and the delimiter of the target db is a single semicolon.

    Can anyone please suggest a way in which I can transform the delimiter and write a statement which checks the string for any of the source values and then maps that to it's target value?

    Thanks in advance!

    Chris

    1) First thing I will say is that this is not really a thing you want SQL Server to be doing. If it is a picklist for an external app then let it do the parsing, string replacement or whatever it is you are asking for.

    2) Speaking of which, it sure would have helped if you had posted up what your expected output was for that given input.

    3) Just in case you are looking for a rudamentary string replacement, review the REPLACE function in SQL Server Books Online.

    Best,
    Kevin G. Boles
    SQL Server Consultant
    SQL MVP 2007-2012
    TheSQLGuru on googles mail service

  • Hi All, 

    Thank you for your quick responses. Apologies for the lack of detail in my previous post.

    A bit of background. I am migrating Contact data from Eloqua to SFDC. The source column is a multi select picklist and the target column in SFDC is also multi-select picklist. 

    I need to do two things: 
    1.Replace the delimiter between the values from '::' to ';' - thank you for the suggestion of utilising the REPLACE function, that would make sense.
    2. Map the values from the source column to their respective values in SFDC.

    E.g. The four rows below exist in Eloqua. I would need to map each value to it's respective value in SFDC. I.e. ' 3G Contact' = 'Contact - 3G' in SFDC. 

    Source
    1. 3G Contact::C.E.O Contact::External Newsletter Subscriber::GSMA Press::Interconnection Contact::Speaker Contact
    2. 3G Contact::C.E.O Contact::External Newsletter Subscriber::Speaker Contact::Training Contact
    3. 3G Contact::C.E.O Contact::GSMA Invoice::GSMA Main
    4. 3G Contact::C.E.O Contact::Fraud Issues Contact::GSMA Invoice::GSMA Main::Intl Gway SS7 Service Agrm & Schd IR21::IREG Tests - IR.21::Network Manager::Non-Standard IR.21::Roaming - IR.21 Administration::Roaming Main Contact::Security Issues Contact

    Target
    1. Contact - 3G;C.E.O;External Newsletter Subscription; GSMA Press;NULL; Speaker
    2. Contact - 3G;C.E.O;External Newsletter Subscription;Speaker;Training
    etc

    Is there a way I could loop through the column (948k rows) and in one statement repalce the delimiter and replace each value?

    Does that make more sense?

    Thanks

  • aldous.chris - Monday, March 6, 2017 8:17 AM

    Hi All, 

    Thank you for your quick responses. Apologies for the lack of detail in my previous post.

    A bit of background. I am migrating Contact data from Eloqua to SFDC. The source column is a multi select picklist and the target column in SFDC is also multi-select picklist. 

    I need to do two things: 
    1.Replace the delimiter between the values from '::' to ';' - thank you for the suggestion of utilising the REPLACE function, that would make sense.
    2. Map the values from the source column to their respective values in SFDC.

    E.g. The four rows below exist in Eloqua. I would need to map each value to it's respective value in SFDC. I.e. ' 3G Contact' = 'Contact - 3G' in SFDC. 

    Source
    1. 3G Contact::C.E.O Contact::External Newsletter Subscriber::GSMA Press::Interconnection Contact::Speaker Contact
    2. 3G Contact::C.E.O Contact::External Newsletter Subscriber::Speaker Contact::Training Contact
    3. 3G Contact::C.E.O Contact::GSMA Invoice::GSMA Main
    4. 3G Contact::C.E.O Contact::Fraud Issues Contact::GSMA Invoice::GSMA Main::Intl Gway SS7 Service Agrm & Schd IR21::IREG Tests - IR.21::Network Manager::Non-Standard IR.21::Roaming - IR.21 Administration::Roaming Main Contact::Security Issues Contact

    Target
    1. Contact - 3G;C.E.O;External Newsletter Subscription; GSMA Press;NULL; Speaker
    2. Contact - 3G;C.E.O;External Newsletter Subscription;Speaker;Training
    etc

    Is there a way I could loop through the column (948k rows) and in one statement repalce the delimiter and replace each value?

    Does that make more sense?

    Thanks

    That helps. There should be no need for a loop on this. You can update all the rows in one pass (or do it some N rows at a time to avoid long-duration locking/blocking if this is a production system, where N is hundreds or thousands in size and you are doing some form of quick index seek to hit the rows). Just LEFT JOIN to the lookup table and do a REPLACE on the :: and a STUFF for the looked-up value. Don't have time to write the query but that should be doable in a single pass. Obviously you can do two passes if desired.

    Best,
    Kevin G. Boles
    SQL Server Consultant
    SQL MVP 2007-2012
    TheSQLGuru on googles mail service

  • Hi Kevin, 

    Thanks for the advice. I'll give that a shot and see where I get with it!

    Cheers,
    Chris

  • one other option which may be viable depending on the number of distinct values of that field and its mapping to the new system.

    something around these lines


    define map_table with
    Source_value varchar(500) -- size as needed
    new_value varchar(500)

    populate map_table with distinct list of source_field and its corresponding mapping on new system

    insert into new_table
    select ot.field1
      ....
      , ot.fieldx --- fields as required
      , coalesce(mt.new_value, ot.source_field) as source_field
    from old_table ot
    left outer join map_table mt -- left outer to catter for nulls
    on mt.source_value = ot.source_field

Viewing 8 posts - 1 through 7 (of 7 total)

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