seen = set()
seen_add = seen.add
return [x for x in array if not (x in seen or seen_add(x))]
# MAIN
if __name__ == "__main__":
args = interface()
input_dir = os.path.abspath(args.IN)
if not input_dir.endswith('/'):
input_dir += '/'
output_dir = os.path.abspath(args.OUT)
if not output_dir.endswith('/'):
output_dir += '/'
task_rank = args.task_rank - 1
FP = glob.glob(os.path.join(input_dir, '*.hashq.*'))
FP = [fp[fp.rfind('/') + 1:] for fp in FP]
FP = list(unique([fp[:fp.index('.')] for fp in FP]))
file_prefix = FP[task_rank % len(FP)]
print('Merging sample ' + file_prefix)
# SUPER DUMB to hardcode the fraction size
file_fraction = int(task_rank / len(FP))
hashobject = Fastq_Reader(input_dir, output_dir)
H = hashobject.hash_counts_from_hashq(file_prefix,
multi_files_fraction=file_fraction)
help_message = 'usage example: python merge_hashq_files.py -r 3 -i /project/home/hashed_reads/ -o /project/home/hashed_reads/'
if __name__ == "__main__":
try:
opts, args = getopt.getopt(sys.argv[1:],'hr:i:o:',["filerank=","inputdir=","outputdir="])
except:
print help_message
sys.exit(2)
for opt, arg in opts:
if opt in ('-h','--help'):
print help_message
sys.exit()
elif opt in ('-r','--filerank'):
fr = int(arg) - 1
elif opt in ('-i','--inputdir'):
inputdir = arg
if inputdir[-1] != '/':
inputdir += '/'
elif opt in ('-o','--outputdir'):
outputdir = arg
if outputdir[-1] != '/':
outputdir += '/'
FP = glob.glob(os.path.join(inputdir,'*.hashq.*'))
FP = [fp[fp.rfind('/')+1:] for fp in FP]
FP = list(set([fp[:fp.index('.')] for fp in FP]))
file_prefix = FP[fr%len(FP)]
# SUPER DUMB to hardcode the fraction size
file_fraction = fr/len(FP)
hashobject = Fastq_Reader(inputdir,outputdir)
H = hashobject.hash_counts_from_hashq(file_prefix,multi_files_fraction=file_fraction)
