bash script to rewrite numbers sequentially - bash

Question

I'd like to 're-sequence' some variable assignment values that are within a large BASH script I'm writing. At present, I have to do this manually, and it's quite time-consuming. ;)
e.g.:
(some code here)
ab=0
(and some here too)
ab=3
(more code here)
cd=2; ab=1
(more code here)
ab=2
What I'd like to do is run a command that can re-order the assignment values of 'ab' so we get:
(some code here)
ab=0
(and some here too)
ab=1
(more code here)
cd=2; ab=2
(more code here)
ab=3
The indentations exist as these usually form part of a code block, like an 'if' or 'for' block.
The variable name will always be the same. The first occurrence in the script should be made a zero. I thought if something (like sed) could search for 'ab=' followed by an integer, then change that integer according to an incrementing value, this would be perfect.
Hoping someone out there may know of something that can do this already. I use 'Kate' for my BASH editing.
Any thoughts? Thank you.

Answer 1

$ # can also use: perl -pe 's/\bab=\K\d+/$i++/ge' file
$ perl -pe 's/(\bab=)\d+/$1.$i++/ge' file
(some code here)
ab=0
(and some here too)
ab=1
(more code here)
cd=2; ab=2
(more code here)
ab=3
(\bab=)\d+ match ab= and one or more digits. \b is word boundary marker so that words like dab=4 doesn't match
The e modifier allows to use Perl code in replacement section
$1.$i++ is string concatenation of ab= and value of $i (which is 0 by default) Then $i gets incremented
Use perl -i -pe for inplace editing

Answer 2

#teracoy:#try:
awk '/ab=/{sub(/ab=[0-9]+/,"ab="i++);print;next} 1' Input_file

Answer 3

WIth GNU awk for multi-char RS, RT, and gensub():
$ awk -v RS='\\<ab=[0-9]+' '{ORS=gensub(/[0-9]+/,i++,1,RT)}1' file
(some code here)
ab=0
(and some here too)
ab=1
(more code here)
cd=2; ab=2
(more code here)
ab=3
Use awk -i inplace ... for inplace editing if desired.