Mission 2 is to obtain the content before each number up to nth occurrence to be associated with the number extracted above.
"I require an expression which can find numbers (10) throughout a string (11) at any location (12)." >>
"I require an expression which can find numbers", "throughout a string", "at any location", "."
"It must skip numbers 32 that are not within (4) brackets in a string 2 with many (99) possible other characters." >>
"It must skip numbers 32 that are not within", "brackets in a string 2 with many", "possible other characters."
1 of 1 people found this helpful
Is it possible to pre-process the data before loading it into Tableau?
It's not so easy to achieve this transformation with string calculations. You could use a series of nested REGEXP_REPLACE calls to remove the unwanted characters (as compared to your original idea of extracting the wanted characters), but it gets messy, and covering all the edge cases can be tricky.
Thanks for your response. Yes it is possible to pre-process the data, especially if I'm using a static data snapshot. ETL layer changes would be required for an ongoing automated solution for the production database. I was hoping to at least simulate what the improvement at the database side would look like in terms of data resolution and analytics potential on live data, for the uninitiated. Conceptually simple perhaps not Tableau simple.
The context if it assists is a log of equipment downtime events that are currently recorded as a series of unstructured descriptions with manually recorded downtimes (inserted into brackets) and stored as a single entry in the source database. Presumably a hangover from when the logs were hand written and were used for context not analytics.
I'll go the pre-processing option as suggested.