Mercurial > repos > matthias > data_manager_megan_tools

comparison data_manager/data_manager.py @ 0:1f839ba466da draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/bernt-matthias/mb-galaxy-tools/tree/master/data_managers/data_manager_megan_tools commit d7a7a198e8f8c9b95491f1520d478e7400a1f59c-dirty
author matthias
date Thu, 01 Nov 2018 12:25:53 -0400
parents
children f7ad11d31098
comparison
equal deleted inserted replaced
-1:000000000000 0:1f839ba466da
1 import argparse
2 import datetime
3 import json
4 import os
5 import shutil
6 import tarfile
7 import zipfile
8 try:
9 # For Python 3.0 and later
10 from urllib.request import Request, urlopen
11 except ImportError:
12 # Fall back to Python 2 imports
13 from urllib2 import Request, urlopen
14
15 FILE2NAME = {
16 "prot_acc2tax-June2018X1.abin.zip":"Protein accession to NCBI-taxonomy (June2018X1)",
17 "nucl_acc2tax-June2018.abin.zip":"Nucleotide accession to NCBI-taxonomy (June2018)",
18 "acc2interpro-June2018X.abin.zip":"Protein accession to InterPro (June2018X)",
19 "acc2eggnog-Oct2016X.abin.zip":"Protein accession to eggNOG (Oct2016X)",
20 "acc2seed-May2015XX.abin.zip":"Protein accession to SEED (May2015XX)",
21 "acc2kegg-Dec2017X1-ue.abin.zip":"Protein accession to KEGG (Dec2017X1). Only for use with the Ultimate Edition of MEGAN.",
22 "SSURef_Nr99_132_tax_silva_to_NCBI_synonyms.map.gz":"SSURef_Nr99_132_tax_silva_to_NCBI_synonyms.map.gz",
23 "SSURef_NR99_128_tax_silva_to_NCBI_synonyms.map.gz":"SSURef_NR99_128_tax_silva_to_NCBI_synonyms.map.gz",
24 "prot_gi2tax-Aug2016X.bin.zip":"Protein accession to NCBI-taxonomy (Aug2016X)",
25 "nucl_gi2tax-Aug2016.bin.zip":"Nucleotide accession to NCBI-taxonomy (Aug2016)",
26 "gi2eggnog-June2016X.bin.zip":"Protein accession to InterPro (June2016X)",
27 "gi2interpro-June2016X.bin.zip":"Protein accession to eggNOG (June2016X)",
28 "gi2seed-May2015X.bin.zip":"Protein accession to SEED (May2015X)",
29 "gi2kegg-Aug2016X-ue.bin.zip":"Protein accession to KEGG (Aug2016X). Only for use with the Ultimate Edition of MEGAN."
30 }
31
32 FILE2TYPE = {
33 "prot_acc2tax-June2018X1.abin.zip":"acc2tax",
34 "nucl_acc2tax-June2018.abin.zip":"acc2tax",
35 "acc2interpro-June2018X.abin.zip":"acc2interpro",
36 "acc2eggnog-Oct2016X.abin.zip":"acc2eggnog",
37 "acc2seed-May2015XX.abin.zip":"acc2seed",
38 "acc2kegg-Dec2017X1-ue.abin.zip":"acc2kegg",
39 "SSURef_Nr99_132_tax_silva_to_NCBI_synonyms.map.gz":"syn2taxa",
40 "SSURef_NR99_128_tax_silva_to_NCBI_synonyms.map.gz":"syn2taxa",
41 "prot_gi2tax-Aug2016X.bin.zip":"gi2tax",
42 "nucl_gi2tax-Aug2016.bin.zip":"gi2tax",
43 "gi2eggnog-June2016X.bin.zip":"gi2eggnog",
44 "gi2interpro-June2016X.bin.zip":"gi2interpro",
45 "gi2seed-May2015X.bin.zip":"gi2seed-",
46 "gi2kegg-Aug2016X-ue.bin.zip":"gi2kegg"
47 }
48
49 def url_download(fname, workdir):
50 file_path = os.path.join(workdir, 'download.dat')
51 if not os.path.exists(workdir):
52 os.makedirs(workdir)
53 src = None
54 dst = None
55 try:
56 req = Request("http://ab.inf.uni-tuebingen.de/data/software/megan6/download/"+fname)
57 src = urlopen(req)
58 with open(file_path, 'wb') as dst:
59 while True:
60 chunk = src.read(2**10)
61 if chunk:
62 dst.write(chunk)
63 else:
64 break
65 finally:
66 if src:
67 src.close()
68 if zipfile.is_zipfile(file_path):
69 fh = zipfile.ZipFile(file_path, 'r')
70 else:
71 return
72 fh.extractall(workdir)
73 os.remove(file_path)
74
75
76 def main(fname, outjson):
77 workdir = os.path.join(os.getcwd(), 'megan_tools')
78 url_download(fname, workdir)
79
80 data_manager_entry = {}
81 data_manager_entry['value'] = fname.split(".")[0]
82 data_manager_entry['name'] = FILE2NAME[fname]
83 data_manager_entry['type'] = FILE2TYPE[fname]
84 data_manager_entry['path'] = '.'
85
86 data_manager_json = dict(data_tables=dict(megan_tools=data_manager_entry))
87
88 params = json.loads(open(outjson).read())
89 target_directory = params['output_data'][0]['extra_files_path']
90 os.mkdir(target_directory)
91 output_path = os.path.abspath(os.path.join(os.getcwd(), 'megan_tools'))
92 for filename in os.listdir(workdir):
93 shutil.move(os.path.join(output_path, filename), target_directory)
94 file(outjson, 'w').write(json.dumps(data_manager_json))
95
96
97 if __name__ == '__main__':
98 parser = argparse.ArgumentParser(description='Create data manager json.')
99 parser.add_argument('--out', action='store', help='JSON filename')
100 parser.add_argument('--file', action='store', help='Download filename')
101 args = parser.parse_args()
102
103 main(args.file, args.out)