Practically, you'd take the cDNA of the bacteria (or organism from which you want to extract the section of DNA you want), transcribe the section, PCR it (to amplify the # of such sections), ligate it into the vector (so your vector carries the DNA you want), transform vector into bacteria (so your bacteria has this vector), then miniprep bacteria (to check that it wasn't an empty vector that has been transformed).
O/N = overnight.
Haha. I'm bitter about the amount of this stuff I have to do. (Not really -- in a sad, routine way, I find this kind of fun.)