text to column non-deliminated

[becki]

i have two data sources and i need to compare a very large (150 characters in one, 160 characters in the other) field.  This field is non-delimited but i have a listing of "of this field, the first 3 characters stands for body type, next 1 character is engine type, etc..".  the problem is, a lot of them are 1 character fields and I was hoping to not have to specify 100 regex parts.  Is it possible to break a field, 1 character at a type, without delimiters (and i can visually group the fields which need to be grouped)?

thanks

becki kain

[GiuseppeC]

Hi @becki,

it would be easier if you could post some sample of the data you want to compare/parse and the listing (suppose it could be used as a look up table somehow?).

Giuseppe

[becki]

it's a VO Image:

1FTYE1CM5HKB45608 E1CHK130 C 3 AG2 01FM083 VE G 5E673 C81D 1 BX C HF 316A218 L2 YZ VK GB M1FTY5A8 101A 9M6DE

but the spacing is significant, like i can't just take all the spaces inside the string.  columns 1-3 mean something, column 4 is something else, etc...

[mmenth]

If your goal is to just split the string by each character, I suggest splitting it to rows using the Regex tool. If you use the regular expression "." this will split any character into a new row.

Best,

mmenth

[GiuseppeC]

@becki, can you also post what you'd like the output to be?

[becki]

I'm trying to compare VO images to see what the differences are, from the two data sources.

[GiuseppeC · Accepted solution]

@becki, see if the attached works. I used @mmenth's suggestion to serialise the image into rows, then added a record ID and joind the 2 images based on that, then added a formula tool to compare the 2.

Hope this helps!

Giuseppe

[becki]

yes that would work but then I figured out the data has been changed, where columns are shifted over.  i'll post that as a separate question.  thanks

[wottel]

This helped me use regex to split a field into separate columns that didn't have delimiters.. i needed each set of 5 digits to be split. The Tokenize option worked..  thanks for the post..