view data_manager/bowtie_index_builder.py @ 5:08a3583826fa draft

"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/data_managers/data_manager_bowtie_index_builder commit 02d2967f77e3fa5a18aea63dc84aa9ab418dc165"
author iuc
date Sun, 22 Nov 2020 12:44:59 +0000
parents 86e9af693a33
children
line wrap: on
line source

#!/usr/bin/env python
from __future__ import print_function

import json
import optparse
import os
import subprocess
import sys
import tempfile

CHUNK_SIZE = 2**20

DEFAULT_DATA_TABLE_NAME = "bowtie_indexes"


def get_id_name(params, dbkey, fasta_description=None):
    # TODO: ensure sequence_id is unique and does not already appear in location file
    sequence_id = params['param_dict']['sequence_id']
    if not sequence_id:
        sequence_id = dbkey

    sequence_name = params['param_dict']['sequence_name']
    if not sequence_name:
        sequence_name = fasta_description
        if not sequence_name:
            sequence_name = dbkey
    return sequence_id, sequence_name


def build_bowtie_index(data_manager_dict, fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, data_table_name=DEFAULT_DATA_TABLE_NAME, color_space=False):
    # TODO: allow multiple FASTA input files
    fasta_base_name = os.path.split(fasta_filename)[-1]
    sym_linked_fasta_filename = os.path.join(target_directory, fasta_base_name)
    os.symlink(fasta_filename, sym_linked_fasta_filename)
    args = ['bowtie-build']
    if color_space:
        args.append('-C')
    args.append(sym_linked_fasta_filename)
    args.append(fasta_base_name)
    args.append(sym_linked_fasta_filename)
    tmp_stderr = tempfile.NamedTemporaryFile(prefix="tmp-data-manager-bowtie-index-builder-stderr")
    proc = subprocess.Popen(args=args, shell=False, cwd=target_directory, stderr=tmp_stderr.fileno())
    return_code = proc.wait()
    if return_code:
        tmp_stderr.flush()
        tmp_stderr.seek(0)
        print("Error building index:", file=sys.stderr)
        while True:
            chunk = tmp_stderr.read(CHUNK_SIZE)
            if not chunk:
                break
            sys.stderr.write(chunk)
        sys.exit(return_code)
    tmp_stderr.close()
    data_table_entry = dict(value=sequence_id, dbkey=dbkey, name=sequence_name, path=fasta_base_name)
    _add_data_table_entry(data_manager_dict, data_table_name, data_table_entry)


def _add_data_table_entry(data_manager_dict, data_table_name, data_table_entry):
    data_manager_dict['data_tables'] = data_manager_dict.get('data_tables', {})
    data_manager_dict['data_tables'][data_table_name] = data_manager_dict['data_tables'].get(data_table_name, [])
    data_manager_dict['data_tables'][data_table_name].append(data_table_entry)
    return data_manager_dict


def main():
    parser = optparse.OptionParser()
    parser.add_option('-f', '--fasta_filename', dest='fasta_filename', action='store', type="string", default=None, help='fasta_filename')
    parser.add_option('-d', '--fasta_dbkey', dest='fasta_dbkey', action='store', type="string", default=None, help='fasta_dbkey')
    parser.add_option('-t', '--fasta_description', dest='fasta_description', action='store', type="string", default=None, help='fasta_description')
    parser.add_option('-n', '--data_table_name', dest='data_table_name', action='store', type="string", default=None, help='data_table_name')
    parser.add_option('-c', '--color_space', dest='color_space', action='store_true', default=False, help='color_space')
    (options, args) = parser.parse_args()

    filename = args[0]

    with open(filename) as fh:
        params = json.load(fh)
    target_directory = params['output_data'][0]['extra_files_path']
    os.mkdir(target_directory)
    data_manager_dict = {}

    dbkey = options.fasta_dbkey

    if dbkey in [None, '', '?']:
        raise Exception('"%s" is not a valid dbkey. You must specify a valid dbkey.' % (dbkey))

    sequence_id, sequence_name = get_id_name(params, dbkey=dbkey, fasta_description=options.fasta_description)

    # build the index
    build_bowtie_index(data_manager_dict, options.fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, data_table_name=options.data_table_name or DEFAULT_DATA_TABLE_NAME, color_space=options.color_space)

    # save info to json file
    with open(filename, 'w') as fh:
        json.dump(data_manager_dict, fh, sort_keys=True)


if __name__ == "__main__":
    main()