Can someone tell me, how do i write a recursive code using shell ( eg like 'for' loop in C) which outputs the record to a database table as one row per iteration?

Which OS?
Which shell?
What database?
And a for loop has nothing to do with recursion. Recursion would mean a function that calls itself, useful for Fibonacci series or XML parsing.

You mean "iterative" not "resursive".

I meant iterative...My requirement is to find text between specific strings using awk send this to csv file then move ahead and find the next string send it to csv and so on till the end of the source file...Hope i am clear in explaining..

If there is only one CSV file per execution of your script then you are likely to find many examples if you google for "echo grep shell". Perhaps you have a draft version of your code or script we can review?

No requirement is of same csv file for all iterations so for that 'echo grep shell' will work?
No i am still figuring out on how to build the code but the idea is to use only one csv output file. Let me simplify a bit.
Source: source.txt say it contains a,1,b,a,2,b,a,3,b
In code: awk '/a/,/b/' source.txt
Target: target.csv

which should get output as
1 ,one
2 ,two
3, three

I meant this: echo grep shell - Google Search. It should help you get started.