Unless this is a programming exercise, it will be easiest to just parse out the accessions (the first one, the 'primary accession number', suffices, the rest are old ones, kept for 'backward compatibility') and use those primary accessions to query the uniprot.org website.

To get the fasta for an accession, say 'Q94650', just construct a url like:

http://www.uniprot.org/uniprot/Q94650.fasta

and go get the fasta with wget or indeed perl LWP. It'll return:

>sp|Q94650|ARF1_PLAFA ADP-ribosylation factor 1 OS=Plasmodium falcipar
+um GN=ARF1 PE=1 SV=3
MGLYVSRLFNRLFQKKDVRILMVGLDAAGKTTILYKVKLGEVVTTIPTIGFNVETVEFRN
ISFTVWDVGGQDKIRPLWRHYYSNTDGLIFVVDSNDRERIDDAREELHRMINEEELKDAI
ILVFANKQDLPNAMSAAEVTEKLHLNTIRERNWFIQSTCATRGDGLYEGFDWLTTHLNNA
K
[download]