Dynamic parsing in Magic Cleaning

Question: What is your feature/integration request?
Answer: With Magic Cleaning we can search and replace specific pre-defined characters. My suggestion is to add more dynamic features to remove unnecessary characters that we can’t define beforehand.
Two suggestions:
a) Ability to set a starting and end position by text match, such as start after “-” and end at " ". For example Account ID - 23fs8942 , and Account ID - 39DFds55. But we only want the 23fs8942 and 39DFds55. Another example, in dates Mon, 05/20/2020 we want to remove Mon by anything after ", "
b) Remove or find pre-defined entities such as email address, numbers, dates, phone numbers, url links.
(I suggest checking out mailparser.io for ideas).

Question: What problem would this feature/integration solve?
Answer: This feature would help solve removing unnecessary characters.

Question: How do you solve/workaround this problem today?
Answer: No known workaround. Unable to complete this specific project with parabola atm :frowning:

Thanks !

Hey Tim!

That is currently possible! The Magic Cleaning step is essentially a pre-set version of a step. So in your case, you should use the RegEx Step.

Regular Expressions are kind of weird to get used to, but they are super powerful. I suggest using https://regexr.com/ to build out your expression, and then use it in Parabola.

For your suggestions, here is how you do them!

  1. If you have Account ID - 23fs8942 , and Account ID - 39DFds55 as a piece of data, then first use a Find Replace step to find "Account ID - " and replace it with nothing - that should give you 23fs8942 , and 39DFds55 and then use a Column Split step to split on: " , and " which will give you your two account IDs!

  2. For the date one: “Mon, 05/20/2020” - you can either use the Date Formatter to reformat that date, or you can use a Text Extract and tell it to extract anything after a comma and a space.

  3. Pre-defined entities would be nice! You can use the RegEx for this - there is a community section on https://regexr.com/ where you can find other people’s solutions to these exact problems.

Hope this unblocks you but please reply if you are still stuck!