Mercurial > repos > iuc > stacks2_gstacks
comparison check_bcfile.py @ 2:1d839ead7ad3 draft
"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/stacks2 commit f55e2407891a3c1f73f14a77b7ddadcd6f5eb1f8"
| author | iuc |
|---|---|
| date | Wed, 15 Jul 2020 17:24:38 -0400 |
| parents | |
| children |
comparison
equal
deleted
inserted
replaced
| 1:27359c6bf3e3 | 2:1d839ead7ad3 |
|---|---|
| 1 #!/usr/bin/env python | |
| 2 | |
| 3 import argparse | |
| 4 import sys | |
| 5 | |
| 6 parser = argparse.ArgumentParser() | |
| 7 parser.add_argument('bcfile', help='barcode file') | |
| 8 args = parser.parse_args() | |
| 9 | |
| 10 barcodes = [] | |
| 11 | |
| 12 with open(args.bcfile, "r") as fh: | |
| 13 for line in fh: | |
| 14 if len(line) == 0: | |
| 15 continue | |
| 16 if line.startswith("#"): | |
| 17 continue | |
| 18 barcodes.append(line.split()) | |
| 19 | |
| 20 if len(barcodes) <= 1: | |
| 21 sys.exit("barcode file is empty") | |
| 22 | |
| 23 # check that all lines have the same number of columns | |
| 24 ncol = None | |
| 25 for bc in barcodes: | |
| 26 if ncol is None: | |
| 27 ncol = len(bc) | |
| 28 elif ncol != len(bc): | |
| 29 sys.exit("barcode file has inconsistent number of columns") | |
| 30 | |
| 31 isname = False | |
| 32 for bc in barcodes: | |
| 33 if len(bc[-1].strip("ATCGatcg")) > 0: | |
| 34 isname = True | |
| 35 break | |
| 36 | |
| 37 names = set() | |
| 38 for bc in barcodes: | |
| 39 if isname: | |
| 40 n = bc[-1] | |
| 41 else: | |
| 42 n = '-'.join(bc) | |
| 43 if n in names: | |
| 44 sys.exit("duplicate sample %s in barcode file" % n) | |
| 45 names.add(n) |