Josh, I tested other formula that solved it in one step. I noticed that there is space before the verse number, no matter how it ends the previous verse (exclamation, comma etc.). So I used:
Find: ([0-9]) -- [one space]([0-9])
Replace:^p\1 -- [no space]^p\1
But as you said it would create a false positive. But I think all number on this bible are written without abbreviation. At this time I am converting NT only. So it would not be a problem.
Thanks one more time. The 'use wildcards' is a very interesting feature that I did not know even existed.