Help With Grep Positive Look Behind

[Chris Panny]

Windows 10 64 bit, InD CC 2107

I'm working on an ePUB doc and I have to find a tab character within a specific, yet consistent text pattern and swap the tab for a space character. It always starts at the beginning of a paragraph, with a digit followed by a period mark. It may include some repetition of these characters like so:

3.4.1.4.5 tab more alpha text

I start my Grep like this just to see if it's working:

^\d\.[\d\.]*\t

which it does.

But when I try to load everything but the tab character into Positive Look Behind, but it doesn't find the tab:

(?<=^\d\.[\d\.]*)\t

What am I doing wrong?

[Mike Witherell]

Did you try leaving out the ^ carat from the positive look-behind?

[winterm]

Try use \K instead of (?<=)

About \K usage you can read here (post 14 and further)

[Chris Panny]

I tried leaving out the carrot ^, but I get: Cannot Find match.

I tried \K. It behaves the same way as ^\d\.[\d\.]*\t, meaning it finds and selects all of the characters, which isn't quite what I'm looking for. I need to isolate the tab after the digit(s) and period mark(s) and then swap it out with a space character. That's why I'm pretty sure I need Pos Look Behind - look behind the tab, confirm a match and then swap.

I have a feeling these characters that need to go inside Pos Look B-hind, need to get enclosed in something - brackets, curly braces, or parenthesis - I tried these but it didn't work. Maybe some combo of these?

[Obi-wan Kenobi · Accepted solution]

(?-s)^(\d+\.?)+\K\t

(^/)

[Chris Panny]

Obi,

thanks - it worked!

Big thanks to everyone else for their input!