#!/usr/bin/env python

import arvados
import arvados_bwa
import os
import re
import sys

this_job = arvados.current_job()
this_task = arvados.current_task()
ref_dir = arvados.util.collection_extract(
    collection = this_job['script_parameters']['input'],
    path = 'reference',
    decompress = False)

ref_fasta_files = (os.path.join(ref_dir, f)
                   for f in os.listdir(ref_dir)
                   if re.search(r'\.fasta(\.gz)?$', f))

# build reference index
arvados_bwa.run('index',
                ['-a', 'bwtsw'] + list(ref_fasta_files))

# move output files to new empty directory
out_dir = os.path.join(arvados.current_task().tmpdir, 'out')
arvados.util.run_command(['rm', '-rf', out_dir], stderr=sys.stderr)
os.mkdir(out_dir)
for f in os.listdir(ref_dir):
    if re.search(r'\.(amb|ann|bwt|pac|rbwt|rpac|rsa|sa)$', f):
        sys.stderr.write("bwa output: %s (%d)\n" %
                         (f, os.stat(os.path.join(ref_dir, f)).st_size))
        os.rename(os.path.join(ref_dir, f),
                  os.path.join(out_dir, f))

# store output
out = arvados.CollectionWriter()
out.write_directory_tree(out_dir, max_manifest_depth=0)
this_task.set_output(out.finish())