Merge pull request #73 from dib-lab/fix/open

Replace argparse.FileType with kevlar.open package-wide

kevlar/__init__.py

@@ -12,6 +12,7 @@
 except:  # pragma: no cover
     import builtins
 from collections import namedtuple
+import sys
 from kevlar import seqio
 from kevlar.seqio import parse_augmented_fastq, print_augmented_fastq
 from kevlar import dump
@@ -35,6 +36,9 @@
 def open(filename, mode):
     if mode not in ['r', 'w']:
         raise ValueError('invalid mode "{}"'.format(mode))
+    if filename in ['-', None]:
+        filehandle = sys.stdin if mode == 'r' else sys.stdout
+        return filehandle
     openfunc = builtins.open
     if filename.endswith('.gz'):
         openfunc = gzopen

kevlar/assemble.py

@@ -30,16 +30,15 @@ def subparser(subparsers):
     subparser = subparsers.add_parser('assemble')
     subparser.add_argument('-d', '--debug', action='store_true',
                            help='print debugging output')
-    subparser.add_argument('-o', '--out', metavar='FILE', default=sys.stdout,
-                           type=argparse.FileType('w'),
+    subparser.add_argument('-o', '--out', metavar='FILE',
                            help='output file; default is terminal (stdout)')
     subparser.add_argument('--gml', metavar='FILE',
                            help='write graph to .gml file')
     subparser.add_argument('-x', '--max-abund', type=int, metavar='X',
                            default=500, help='discard interesting k-mers that '
                            'occur more than X times')
-    subparser.add_argument('augfastq', type=argparse.FileType('r'),
-                           help='annotated reads in augmented Fastq format')
+    subparser.add_argument('augfastq', help='annotated reads in augmented '
+                           'Fastq format')
 
 
 def print_read_pair(read1, pos1, read2, pos2, ksize, offset, overlap,
@@ -244,7 +243,7 @@ def main(args):
     if args.debug:
         debugout = args.logfile
 
-    reads, kmers = load_reads(args.augfastq, debugout)
+    reads, kmers = load_reads(kevlar.open(args.augfastq, 'r'), debugout)
     inputreads = list(reads.keys())
     graph = graph_init(reads, kmers, args.max_abund, debugout)
     if args.gml:
@@ -318,13 +317,14 @@ def main(args):
 
     contigcount = 0
     unassembledcount = 0
+    outstream = kevlar.open(args.out, 'w')
     for seqname in graph.nodes():
         if seqname in inputreads:
             unassembledcount += 1
             continue
         contigcount += 1
         contigrecord = reads[seqname]
-        kevlar.print_augmented_fastq(contigrecord, args.out)
+        kevlar.print_augmented_fastq(contigrecord, outstream)
 
     assembledcount = len(inputreads) - unassembledcount
     message = '[kevlar::assemble] assembled'

kevlar/dump.py

@@ -40,9 +40,8 @@ def subparser(subparsers):
                            ' is 2G')
     subparser.add_argument('--mask-k', metavar='K', default=31, type=int,
                            help='k size for genome mask')
-    subparser.add_argument('--out', metavar='FILE',
-                           type=argparse.FileType('w'),
-                           help='output file; default is terminal (stdout)')
+    subparser.add_argument('--out', metavar='FILE', help='output file; default'
+                           ' is terminal (stdout)')
     subparser.add_argument('refr', help='reference sequence in Fasta format')
     subparser.add_argument('reads', help='read alignments in BAM format')
 
@@ -66,6 +65,7 @@ def main(args):
         genomemask.consume_seqfile(args.genomemask)
 
     bam = pysam.AlignmentFile(args.reads, 'rb')
+    fastq = kevlar.open(args.out, 'w')
     for i, record in enumerate(bam):
         if i > 0 and i % 50000 == 0:
             print('...processed', i, 'records', file=args.logfile)
@@ -116,4 +116,4 @@ def main(args):
                 name += suffix
 
         print('@', name, '\n', record.seq, '\n+\n', record.qual, sep='',
-              file=args.out)
+              file=fastq)

kevlar/filter.py

@@ -92,11 +92,9 @@ def subparser(subparsers):
                            help='show this help message and exit')
     misc_args.add_argument('-k', '--ksize', type=int, default=31, metavar='K',
                            help='k-mer size; default is 31')
-    misc_args.add_argument('-o', '--out', type=argparse.FileType('w'),
-                           metavar='FILE', default=sys.stdout,
+    misc_args.add_argument('-o', '--out', metavar='FILE',
                            help='output file; default is terminal (stdout)')
-    misc_args.add_argument('--aug-out', type=argparse.FileType('w'),
-                           metavar='FILE',
+    misc_args.add_argument('--aug-out', metavar='FILE',
                            help='optional augmented Fastq output')
     misc_args.add_argument('--cc-prefix', metavar='PREFIX',
                            help='group reads by novel k-mers, and use the '
@@ -259,8 +257,10 @@ def main(args):
     timer.start('validate')
     print('[kevlar::filter] Validate k-mers and print reads',
           file=args.logfile)
+    outstream = kevlar.open(args.out, 'w')
+    augstream = kevlar.open(args.aug_out, 'w') if args.aug_out else None
     validate_and_print(readset, countgraph, refr, contam, args.min_abund,
-                       args.skip2, args.out, args.aug_out, args.logfile)
+                       args.skip2, outstream, augstream, args.logfile)
     elapsed = timer.stop('validate')
     print('[kevlar::filter] k-mers validated and reads printed',
           'in {:.2f} sec'.format(elapsed), file=args.logfile)

kevlar/find.py

@@ -74,8 +74,7 @@ def subparser(subparsers):
                            help='show this help message and exit')
     misc_args.add_argument('-k', '--ksize', type=int, default=31, metavar='K',
                            help='k-mer size; default is 31')
-    misc_args.add_argument('-o', '--out', type=argparse.FileType('w'),
-                           metavar='FILE',
+    misc_args.add_argument('-o', '--out', metavar='FILE',
                            help='output file; default is terminal (stdout)')
     misc_args.add_argument('--upint', type=float, default=1e6, metavar='INT',
                            help='update interval for log messages; default is '
@@ -167,6 +166,7 @@ def main(args):
     nkmers = 0
     nreads = 0
     unique_kmers = set()
+    outstream = kevlar.open(args.out, 'w')
     for n, record in enumerate(iter_screed(args.cases)):
         if n > 0 and n % args.upint == 0:
             elapsed = timer.probe('iter')
@@ -197,7 +197,7 @@ def main(args):
         if read_kmers > 0:
             nreads += 1
             nkmers += read_kmers
-            kevlar.print_augmented_fastq(record, args.out)
+            kevlar.print_augmented_fastq(record, outstream)
 
     elapsed = timer.stop('iter')
     message = 'Iterated over {} reads in {:.2f} seconds'.format(n, elapsed)

kevlar/mutate.py

@@ -23,11 +23,9 @@
 
 def subparser(subparsers):
     subparser = subparsers.add_parser('mutate')
-    subparser.add_argument('-o', '--out', metavar='FILE', default=sys.stdout,
-                           type=argparse.FileType('w'),
+    subparser.add_argument('-o', '--out', metavar='FILE',
                            help='output file; default is terminal (stdout)')
-    subparser.add_argument('mutations', type=argparse.FileType('r'),
-                           help='mutations file')
+    subparser.add_argument('mutations', help='mutations file')
     subparser.add_argument('genome', help='genome to mutate')
 
 
@@ -112,8 +110,8 @@ def mutate_genome(infile, mutations):
 
 def main(args):
     print('[kevlar::mutate] loading mutations', file=args.logfile)
-    mutations = load_mutations(args.mutations, args.logfile)
+    mutations = load_mutations(kevlar.open(args.mutations, 'r'), args.logfile)
 
     print('[kevlar::mutate] mutating genome', file=args.logfile)
     for record in mutate_genome(args.genome, mutations):
-        write_record(record, args.out)
+        write_record(record, kevlar.open(args.out, 'w'))

kevlar/reaugment.py

@@ -17,20 +17,20 @@
 
 def subparser(subparsers):
     subparser = subparsers.add_parser('reaugment')
-    subparser.add_argument('-o', '--out', metavar='FILE', default=sys.stdout,
-                           type=argparse.FileType('w'),
+    subparser.add_argument('-o', '--out', metavar='FILE',
                            help='output file; default is terminal (stdout)')
-    subparser.add_argument('augfastq', type=argparse.FileType('r'),
-                           help='original augmented Fastq file')
+    subparser.add_argument('augfastq', help='original augmented Fastq file')
     subparser.add_argument('fastq', help='processed Fastq file to re-annotate')
 
 
 def main(args):
     reads = dict()
-    for record in kevlar.parse_augmented_fastq(args.augfastq):
+    instream = kevlar.open(args.augfastq, 'r')
+    for record in kevlar.parse_augmented_fastq(instream):
         reads[record.name] = record
 
     reader = khmer.ReadParser(args.fastq)
+    outstream = kevlar.open(args.out, 'w')
     for read in reader:
         augrecord = reads[read.name]
         if len(read.sequence) < len(augrecord.sequence):
@@ -45,4 +45,4 @@ def main(args):
             if len(ikmers) == 0:
                 continue
             augrecord.ikmers = ikmers
-        kevlar.print_augmented_fastq(augrecord, args.out)
+        kevlar.print_augmented_fastq(augrecord, outstream)

kevlar/tests/test_find.py

@@ -60,11 +60,9 @@ def test_assumptions(kmer):
 
 @pytest.mark.parametrize('case,ctrl,mem', [
     ('trio1/case1.fq', 'trio1/ctrl[1,2].fq', '500K'),
-    ('trio1/case1.fq', 'trio1/ctrl[1,2].fq', '1M'),
     ('trio1/case2.fq', 'trio1/ctrl[1,2].fq', '1M'),
     ('trio1/case3.fq', 'trio1/ctrl[1,2].fq', '1M'),
     ('trio1/case4.fq', 'trio1/ctrl[1,2].fq', '500K'),
-    ('trio1/case4.fq', 'trio1/ctrl[1,2].fq', '1M'),
     ('trio1/case5.fq', 'trio1/ctrl[3,4].fq', '1M'),
     ('trio1/case6.fq', 'trio1/ctrl[5,6].fq', '1M'),
 ])
@@ -75,7 +73,7 @@ def test_find_single_mutation(case, ctrl, mem, capsys):
     arglist = ['find', '--ksize', '13', '--case_min', '8', '--ctrl_max', '0',
                '--memory', mem, '--cases', casestr, '--controls'] + ctrls
     args = kevlar.cli.parser().parse_args(arglist)
-    args.out = stdout
+    args.out = None
     args.err = stderr
     kevlar.find.main(args)
     out, err = capsys.readouterr()
@@ -101,7 +99,7 @@ def test_find_two_cases(capsys):
                '--case_min', '7']
     arglist += ['--cases'] + cases + ['--controls'] + ctrls
     args = kevlar.cli.parser().parse_args(arglist)
-    args.out = stdout
+    args.out = None
     args.err = stderr
     kevlar.find.main(args)
     out, err = capsys.readouterr()