from __future__ import division
from __future__ import print_function
from builtins import range
from past.utils import old_div
from hirise_assembly import HiriseAssembly
if __name__=="__main__":
import sys
import argparse
parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('-d','--debug',default=False ,action="store_true",help="Turn on debugging ouput")
parser.add_argument('-L','--layout',default=False ,help="A file containing a layout of contigs.")
parser.add_argument('-i','--infile',default=False ,help="Filename for serialised assembly input file.")
parser.add_argument('-o','--outfile',default=False,help="Filename for writing a list of segments on the raw contigs to mask for being promiscuous in linking.")
args = parser.parse_args()
if args.infile:
asf = HiriseAssembly()
asf.load_assembly(args.infile)
if args.outfile:
f=open(args.outfile,"wt")
for contig in asf.contigs_iter():
f.write("{}\t{}\n".format(contig,asf.contig_length(contig)))
if args.infile:
asf = HiriseAssembly()
asf.load_assembly(args.infile)
for segments_file in args.mask:
asf.add_mask_regions(filename=segments_file)
asf.merge_masked_regions()
if args.contig:
contig_iter=iter([args.contig])
else:
contig_iter = asf.ocontigs_iter()
if args.outfile:
of=open(args.outfile,"wt")
else:
of=sys.stdout
for ocontig in contig_iter:
chunk = struct.unpack("<L", hashlib.md5(ocontig.encode("utf-8")).digest()[:4])[0]%args.nchunks
if (not args.contig) and (not chunk == args.chunk): continue
links={}
asf.get_links([ocontig],skipI=True,mapq=args.mapq,links=links,contigs=False,raw=False,debug=args.debug)
for c1,c2 in links.keys():
l1,l2=asf.contig_length(c1),asf.contig_length(c2)
l=links[c1,c2]
n=len(l)
of.write("\t".join(map(str,[c1,c2,l1,l2,n,l]))+"\n")
of.close()