|
8
|
1 #!/usr/bin/env python
|
|
|
2
|
|
|
3 import argparse
|
|
|
4 import json
|
|
|
5 import os
|
|
|
6 import os.path
|
|
|
7 import sys
|
|
|
8
|
|
|
9 def _add_data_table_entry(data_manager_dict, data_table_entry, data_table_name):
|
|
|
10 data_manager_dict['data_tables'] = data_manager_dict.get('data_tables', {})
|
|
|
11 data_manager_dict['data_tables'][data_table_name] = data_manager_dict['data_tables'].get('ganon_databases', [])
|
|
|
12 data_manager_dict['data_tables'][data_table_name].append(data_table_entry)
|
|
|
13 return data_manager_dict
|
|
|
14
|
|
|
15 if __name__ == '__main__':
|
|
|
16 parser = argparse.ArgumentParser(description='Download RefSeq rRNA bacterial databases')
|
|
|
17 parser.add_argument('--output_directory', default='/tool-data/ganon', help='Directory to write output')
|
|
|
18 parser.add_argument('--rrna', help='rRNA sequences to download (5S, 16S, or 23S)')
|
|
|
19 args = parser.parse_args()
|
|
|
20
|
|
|
21 output_directory = args.output_directory
|
|
|
22 if not os.path.exists(output_directory):
|
|
|
23 os.mkdir(output_directory)
|
|
|
24
|
|
|
25 data_manager_dict = {}
|
|
|
26 _add_data_table_entry(data_manager_dict=data_manager_dict,
|
|
|
27 data_table_entry=dict(value="ganon"+args.rrna, dbkey="gannon"+args.rrna, name=args.rrna, path=args.output_directory),
|
|
|
28 data_table_name='ganon_databases')
|
|
|
29 open("output_file", 'w').write(json.dumps(data_manager_dict, sort_keys=True))
|
|
|
30
|
