OK, I did as Neef had suggested, with some modification. My effort was far from perfect , but here it is for reference for anyone who requires to do this in future :
Added this line to it:
R n 60000 read.csv
It basically seems to tell scalpel to look for 60 kb blocks starting with the text read.csv . ( I knew my file to have read.csv near top so used it as a pseudo header of my R files ) Similarly, I Did not supply any footer as there is no definite end marking on the plain text source code files.
It took 2500 secs to work for the 98 GB partition and created the directory ./recovered_myRfiles and dumped a number of R files to it.
- It had gleaned all sorts of files pieces which started with read.csv and dumped them into numbered files with .R extension, each 50 kbs in size. Some were R files, some html and some other sorts of documentation. I had to do a find to figure out if my file was amongst them.
Hope this helps someone in future. Please comment to add improvements.
And many thanks, berf for clarifying the bits and pointing me to the relevant section of that article. Can't accept is as a working answer but highly appreciated.