Le Sun, Mar 09, 2014 at 10:10:28AM +0000, Aníbal Monsalve Salazar a écrit : > > The regression test fails because xx#minimal.full.sam has one extra line > at the end when it's compared to xx#minimal.full.bam.sam. > > The extra line is: > > A4 16 yy 4 1 * * 0 0 * > *
Hi Aníbal, The BAM format is the binary representation of the SAM format, for alignment of nucleotide sequences on a reference genome. I guess that the test suite checks that the tool's output is the same in both formats, by converting BAM to SAM and comparing the two SAM files. Debian also distributes the 'samtools' program, that can do this conversion as well. $ samtools view -h xx#minimal.full.bam @PG ID:scramble PN:scramble VN:1.13.3 CL:/tmp/staden-io-lib-1.13.3/progs/.libs/lt-scramble -t4 ./data/xx#minimal.sam test.out/xx#minimal.bam @PG ID:scramble.1 PN:scramble PP:scramble VN:1.13.3 CL:/tmp/staden-io-lib-1.13.3/progs/.libs/lt-scramble -t4 -r ./data/xx.fa test.out/xx#minimal.bam test.out/xx#minimal.full.cram @PG ID:scramble.2 PN:scramble PP:scramble.1 VN:1.13.3 CL:/tmp/staden-io-lib-1.13.3/progs/.libs/lt-scramble -t4 -O bam test.out/xx#minimal.full.cram @SQ SN:xx LN:20 M5:bbf4de6d8497a119dda6e074521643dc UR:/tmp/staden-io-lib-1.13.3/tests/./data/xx.fa @SQ SN:yy LN:20 M5:bbf4de6d8497a119dda6e074521643dc UR:/tmp/staden-io-lib-1.13.3/tests/./data/xx.fa a0 16 xx 4 1 10H * 0 0 * * a1 16 xx 4 1 10H * 0 0 * * a2 16 xx 4 1 5H10M5H * 0 0 AAAAAAATTT * A0 16 yy 4 1 * * 0 0 * * A1 16 yy 4 1 * * 0 0 * * A2 16 yy 4 1 * * 0 0 * * A3 16 yy 4 1 * * 0 0 * * A4 16 yy 4 1 * * 0 0 * * Missing a line on MIPS definitely looks like a bug. This said, it works on amd64 (at least the test suite) and I doubt that there are users for the staden-io-lib on MIPS. Have you tried to contact the upstream author ? Speaking of Upstream, I see a new release, version 1.13.5. Would you like me to upload it ? Cheers, Charles -- Charles Plessy Debian Med packaging team http://www.debian.org/devel/debian-med Tsurumi, Kanagawa, Japan -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
