[Bioperl-l] Speed ups checked in

[Ewan Birney]

i have just checked in speedups for the fasta parsing:

[localhost:~/data] birney% time perl -I ~/src/bioperl-1.1.0/
fasta2fasta.pl < rod.fa > /dev/null
38.750u 2.420s 0:49.67 82.8%    0+0k 22+0io 0pf+0w


vs


[localhost:~/data] birney% time perl -I ~/src/bioperl-live/ fasta2fasta.pl
< rod.fa > /dev/null
15.240u 1.440s 0:19.14 87.1%    0+0k 21+0io 0pf+0w




to do this speed up (factor of 2) i have had to take a pretty direct route
in creating the correct hashes, blessing them and then populating them
correctly. Rather than this being put in the main system, I have taken
advantage of the Factory system - there is now a new Seq Factory called

SeqFastaSpeedFactory

which, unsurprisingly makes sequences very fast.



I have tried to make generic object creating faster.



Next I will work on GenBank and EMBL parsing.




e.