Mercurial > repos > devteam > data_manager_twobit_builder

comparison data_manager/twobit_builder.py @ 7:94bffa4fdedc draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/data_managers/data_manager_twobit_builder commit 18d696a933873372ae5c77257fd31eaf2cd4ed13
author iuc
date Sun, 16 Apr 2023 08:15:04 +0000
parents ced9f4a48806
children 4a7740ee2892
comparison
equal deleted inserted replaced
6:b640d79b49cb 7:94bffa4fdedc
1 #!/usr/bin/env python 1 #!/usr/bin/env python
2 # Dan Blankenberg 2 # Dan Blankenberg
3 from __future__ import print_function 3 from __future__ import print_function
4 4
5 import json
5 import optparse 6 import optparse
6 import os 7 import os
7 import subprocess 8 import subprocess
8 import sys 9 import sys
9 import tempfile 10 import tempfile
10 from json import dumps, loads
11 11
12 CHUNK_SIZE = 2**20 # 1mb 12 CHUNK_SIZE = 2**20 # 1mb
13 13
14 14
15 def get_id_name(params, dbkey, fasta_description=None): 15 def get_id_name(params, dbkey, fasta_description=None):
24 if not sequence_name: 24 if not sequence_name:
25 sequence_name = dbkey 25 sequence_name = dbkey
26 return sequence_id, sequence_name 26 return sequence_id, sequence_name
27 27
28 28
29 def build_twobit(data_manager_dict, fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name): 29 def build_twobit(data_manager_dict, fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, long):
30 twobit_base_name = "%s.2bit" % (sequence_id) 30 twobit_base_name = "%s.2bit" % (sequence_id)
31 twobit_filename = os.path.join(target_directory, twobit_base_name) 31 twobit_filename = os.path.join(target_directory, twobit_base_name)
32 32
33 args = ['faToTwoBit', fasta_filename, twobit_filename] 33 args = ['faToTwoBit']
34 if long:
35 args.append('-long')
36 args.extend([fasta_filename, twobit_filename])
34 tmp_stderr = tempfile.NamedTemporaryFile(prefix="tmp-data-manager-twobit-builder-stderr") 37 tmp_stderr = tempfile.NamedTemporaryFile(prefix="tmp-data-manager-twobit-builder-stderr")
35 proc = subprocess.Popen(args=args, shell=False, cwd=target_directory, stderr=tmp_stderr.fileno()) 38 proc = subprocess.Popen(args=args, shell=False, cwd=target_directory, stderr=tmp_stderr.fileno())
36 return_code = proc.wait() 39 return_code = proc.wait()
37 if return_code: 40 if return_code:
38 tmp_stderr.flush() 41 tmp_stderr.flush()
40 print("Error building index:", file=sys.stderr) 43 print("Error building index:", file=sys.stderr)
41 while True: 44 while True:
42 chunk = tmp_stderr.read(CHUNK_SIZE) 45 chunk = tmp_stderr.read(CHUNK_SIZE)
43 if not chunk: 46 if not chunk:
44 break 47 break
45 sys.stderr.write(chunk) 48 sys.stderr.write(chunk.decode('utf-8'))
46 sys.exit(return_code) 49 sys.exit(return_code)
47 tmp_stderr.close() 50 tmp_stderr.close()
48 # lastz_seqs 51 # lastz_seqs
49 data_table_entry = dict(value=sequence_id, name=sequence_name, path=twobit_base_name) 52 data_table_entry = dict(value=sequence_id, name=sequence_name, path=twobit_base_name)
50 53
69 def main(): 72 def main():
70 parser = optparse.OptionParser() 73 parser = optparse.OptionParser()
71 parser.add_option('-f', '--fasta_filename', dest='fasta_filename', action='store', type="string", default=None, help='fasta_filename') 74 parser.add_option('-f', '--fasta_filename', dest='fasta_filename', action='store', type="string", default=None, help='fasta_filename')
72 parser.add_option('-d', '--fasta_dbkey', dest='fasta_dbkey', action='store', type="string", default=None, help='fasta_dbkey') 75 parser.add_option('-d', '--fasta_dbkey', dest='fasta_dbkey', action='store', type="string", default=None, help='fasta_dbkey')
73 parser.add_option('-t', '--fasta_description', dest='fasta_description', action='store', type="string", default=None, help='fasta_description') 76 parser.add_option('-t', '--fasta_description', dest='fasta_description', action='store', type="string", default=None, help='fasta_description')
77 parser.add_option('-l', '--long', dest='long', action="store_true", default=False, help='For big genomes you need to pass the -long option.')
78
74 (options, args) = parser.parse_args() 79 (options, args) = parser.parse_args()
75 80
76 filename = args[0] 81 filename = args[0]
77 82
78 params = loads(open(filename).read()) 83 with open(filename) as fh:
84 params = json.load(fh)
79 85
80 target_directory = params['output_data'][0]['extra_files_path'] 86 target_directory = params['output_data'][0]['extra_files_path']
81 os.mkdir(target_directory) 87 os.mkdir(target_directory)
82 data_manager_dict = {} 88 data_manager_dict = {}
83 89
87 raise Exception('"%s" is not a valid dbkey. You must specify a valid dbkey.' % (dbkey)) 93 raise Exception('"%s" is not a valid dbkey. You must specify a valid dbkey.' % (dbkey))
88 94
89 sequence_id, sequence_name = get_id_name(params, dbkey=dbkey, fasta_description=options.fasta_description) 95 sequence_id, sequence_name = get_id_name(params, dbkey=dbkey, fasta_description=options.fasta_description)
90 96
91 # build the index 97 # build the index
92 build_twobit(data_manager_dict, options.fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name) 98 build_twobit(data_manager_dict, options.fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, options.long)
93 99
94 # save info to json file 100 # save info to json file
95 open(filename, 'w').write(dumps(data_manager_dict, sort_keys=True)) 101 with open(filename, 'w') as fh:
102 json.dump(data_manager_dict, fh, sort_keys=True)
96 103
97 104
98 if __name__ == "__main__": 105 if __name__ == "__main__":
99 main() 106 main()