Forum Discussion

BizyB's avatar
BizyB
Regular Visitor
2 years ago
Solved

Extract every string occurrence with additional characters

hello experts

have just begun having fun with the powerBI and have hit the below roadblock although sounds simple. please guide me where am going wrong!

 

problem statement: 

am trying to extract a set of strings based on a common pattern with no specific delimiter in place from a column within a table separated by columns

 

example:

Column1Expected result
AD1234 AD9833: This is sample dataAD1234, AD9833
AD6835,AD0978 This is sample dataAD6835, AD0978

 

Formula am trying to use:

 

Text.Combine(
  List.Select(
    Text.PositionOf([Column1], "AD", Occurrence.All),
    each Text.Middle([Column1], _, 6)
  ),
  ", "
)

 

Am receiving the following error

 

Expression.Error: We cannot apply field access to the type Number.
Details:
    Value=0
    Key=summary

 

 

  • I suggest:

    in the Advanced Editor:

    #"Added Custom" = Table.AddColumn(Source, "Extract", each Text.Combine(
                    List.Accumulate(
                        Text.PositionOf([Column1], "AD", Occurrence.All),
                        {},
                        (s,c)=> s & {Text.Middle([Column1],c,6)}),", "), type text)

     

    or in the UI:

    Text.Combine(
        List.Accumulate(
            Text.PositionOf([Column1], "AD", Occurrence.All),
            {},
            (s,c)=> s & {Text.Middle([Column1],c,6)}),
        ", ")

     

     

     

4 Replies

  • I suggest:

    in the Advanced Editor:

    #"Added Custom" = Table.AddColumn(Source, "Extract", each Text.Combine(
                    List.Accumulate(
                        Text.PositionOf([Column1], "AD", Occurrence.All),
                        {},
                        (s,c)=> s & {Text.Middle([Column1],c,6)}),", "), type text)

     

    or in the UI:

    Text.Combine(
        List.Accumulate(
            Text.PositionOf([Column1], "AD", Occurrence.All),
            {},
            (s,c)=> s & {Text.Middle([Column1],c,6)}),
        ", ")

     

     

     

  • BizyB's avatar
    BizyB
    Regular Visitor

    wonderful.. thanks a lot for the tip.. that worked like a charm!

  • Anonymous's avatar
    Anonymous
    Not applicable

    Why not just use this in your Table.AddColumn:

     

    each Text.Combine({Text.Range([Column1], 0, 6), Text.Range([Column1], 8, 6)}, ", ")

     

    Simpler, plus what if it's ever something besides "AD"?

     

    --Nate

    • BizyB's avatar
      BizyB
      Regular Visitor

      thanks Anonymous and apologies for coming back to you late on this. 

      the trouble with the Column1 data is that it is manually documented by hundreds of users with just a guideline and no control. So the position of "AD" is not static. Also the delimiter is not constant as well because of the varying user inputs.

       

      Because of the above conditions, think the solution from ronrsnfld works better. Also currently we are just focused on getting AD. Anything besides "AD" that comes in later will have to be dealt with it accordingly later on.. so not worries about it now 🙂

       

      thank you!