It's easier to help if you cite where you found the code. That way we can comment in context.
Basically, the DATA step takes the data from file 1 and puts it into file 2. It is different from a simple OS copy in that it can do some rudimentary code translation for different environments (depending on the additional parameters specified in the INFILE and FILE statements).
The _INFILE_ automatic variable is documented with the INFILE statement,
http://support.sas.com/documentation/cdl/en/lrdict/63026/HTML/default/viewer.htm#a000146932.htm
The way I have most commonly used this is to create a hex print of the INFILE buffer when I am trying debug some messy data problem.
Doc Muhlbaier
Duke