Closed rbuels closed 3 years ago
@rbuels we might need to do the same for gff
if I run on a gff
$ bin/flatfile-to-json.pl --gff test.gff --trackLabel testing
Unsuccessful stat on filename containing newline at /home/cdiesh/src/gmod/jbrowse/bin/../src/perl5/FeatureTrack.pm line 70.
Unsuccessful stat on filename containing newline at /home/cdiesh/src/gmod/jbrowse/bin/../src/perl5/JsonFileStorage.pm line 64.
test.gff
hi/I/have/slashes SwinePathogenAnalysisResource(SPAR) mat_peptide 1 100 . + 0 ID=GB000440/KU318406.1/A2MC2-P90/USA/2-5-2:nsp10;Name=nsp10;ORF1ab-gene;Note=full
actually nevermind, seems to work :)
maybe should escape newlines too, i'll add that
Fix works great. Thanks!
nice. we can go ahead and merge :)
just thought of something, the JS code probably needs to be updated to match, right?
looked ok to me without any additional changes, not sure what js you mean though?
the JS code probably needs to do the same regex replace, in order to be able to find the munged paths for the chunk files
didn't appear to be needed as far as I could tell, did this on this branch
bin/prepare-refseqs.pl --fasta tests/data/forward_slashes_in_refseq_name.fa
and then opening up http://localhost/jbrowse worked fine without any additional changes
I uploaded above @btski a build that includes this fix plus the fixed refseq on dev just in case it helps :)
this PR fixes formatting of reference sequences when the names have forward slashes or backslashes in them.
references gitter convo: