[split] Change in behaviour in visidata 3.0 about visidata HOT 6 OPEN

Yes, now it's a single column with a list. Use ( to expand the list into multiple columns like you're used to.

from visidata.

How about =curcol[:7]?

from visidata.

Note also that the previous split-col command still exists (but is deprecated), so previous cmdlogs should still work. Or you can unbind/rebind : to that previous command. If people wind up rebinding to the older command, let us know so we can un-deprecate it.

from visidata.

/docs/columns#split

from visidata.

OK -- so I was using ; and then (.......) to take a date/time column and grab off the first 7 characters which happens to be the month, like 2023-12-01 21:32 → 2023-12. This change surprised me, but now I see I can press ( to get the first capture and then ignore/delete the cruft. Is there a better way?

from visidata.

I didn't know about curcol, that's pretty cool! I often need to run clean-names but curcol will remove some of those use cases.

Whelp. This helps the specific use-case I was most recently using this week. Just now playing with it the only other feedback/suggestion is that if the regex capture only has one group maybe it shouldn't group the column.

from visidata.