Mercurial > repos > iuc > data_manager_fetch_plasmidfinder

comparison data_manager/plasmidfinder_fetch_database.py @ 0:b56071bdecad 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_fetch_plasmidfinder commit f6d40bf94e015c2690e97c4b3533517b812f3bfb
author iuc
date Thu, 15 Jun 2023 09:14:08 +0000
parents
children 3542cd220bc3
comparison
equal deleted inserted replaced
-1:000000000000 0:b56071bdecad
1 import argparse
2 import json
3 import os
4 import time
5 from pathlib import Path
6
7
8 import git
9
10
11 class GetPlasmidfinderDataManager:
12 """
13 Create the json file with database information for galaxy data manager
14 """
15
16 def __init__(self,
17 plasmidfinder_database="plasmidfinder_database",
18 db_name="plasmidfinder_database",
19 plasmidfinder_version="latest"):
20 self.data_table_name = plasmidfinder_database
21 self._db_name = db_name
22 self._plasmidfinder_version = plasmidfinder_version
23 self._plasmidfinder_date_version = None
24 self.data_table_entry = None
25 self.plasmidfinder_table_list = None
26 self._commit_number = None
27
28 def get_data_table_format(self):
29 """
30 Skeleton of a data_table format
31 return: a data table formatted for json output
32 """
33 self.data_table_entry = {
34 "data_tables": {
35 self.data_table_name: {}
36 }
37 }
38 return self.data_table_entry
39
40 def get_data_manager(self):
41 """
42 Create the empty data table format and add all the information into
43 Commit number is added if latest is required instead of version number
44 return: The data table with database information
45 """
46 self.plasmidfinder_table_list = self.get_data_table_format()
47 if self._plasmidfinder_version == "latest":
48 version_value = self._commit_number
49 else:
50 version_value = self._plasmidfinder_version
51 plasmidfinder_value = f"plasmidfinder_{self._commit_number}" \
52 f"_{self._plasmidfinder_date_version}"
53 plasmidfinder_name = f"{version_value}" \
54 f"_{self._plasmidfinder_date_version}"
55 data_info = dict(value=plasmidfinder_value,
56 name=plasmidfinder_name,
57 date=self._plasmidfinder_date_version,
58 path=self._db_name)
59 self.plasmidfinder_table_list["data_tables"][self.data_table_name] = [data_info]
60 return self.plasmidfinder_table_list
61
62
63 class DownloadPlasmidfinderDatabase(GetPlasmidfinderDataManager):
64 """
65 Download the plasmidfinder database from the bitbucket repository.
66 Build the data manager info for galaxy
67 """
68
69 def __init__(self,
70 output_dir=Path.cwd(),
71 plasmidfinder_url="https://bitbucket.org/genomicepidemiology/plasmidfinder_db/src/master",
72 db_name="plasmidfinder_database",
73 db_tmp="tmp_database",
74 plasmidfinder_version="latest",
75 json_file_path=None,
76 date_version=None):
77
78 super().__init__()
79 self.json_file_path = json_file_path
80 self._output_dir = output_dir
81 self._plasmidfinder_url = plasmidfinder_url
82 self._temporary_folder = db_tmp
83 self._db_name = db_name
84 self._db_name_tar = f'{db_name}.gz'
85 self._plasmidfinder_version = plasmidfinder_version
86 self._plasmidfinder_date_version = date_version
87 self._commit_number = None
88
89 def git_clone(self):
90 git.Repo.clone_from(url=self._plasmidfinder_url, to_path=self._output_dir)
91 self._plasmidfinder_repository = git.Repo(path=self._output_dir)
92
93 def get_commit_number(self):
94 sha = self._plasmidfinder_repository.head.commit.hexsha
95 short_sha = self._plasmidfinder_repository.git.rev_parse(sha, short=7)
96 self._commit_number = short_sha
97
98 def get_commit_date(self):
99 self._plasmidfinder_date_version = time.strftime("%Y_%m_%d", time.gmtime(self._plasmidfinder_repository.head.commit.committed_date))
100
101 def download_database(self):
102 """
103 Download the plasmidfinder database using git lib
104 Extract commit and commit date
105 """
106 self._output_dir = Path(self._output_dir)
107 self.git_clone()
108 if self._plasmidfinder_version != "latest":
109 self._plasmidfinder_repository.git.checkout(self._plasmidfinder_version)
110 self.get_commit_number()
111 self.get_commit_date()
112
113 def read_json_input_file(self):
114 """
115 Import the json file
116 """
117 with open(self.json_file_path) as fh:
118 params = json.load(fh)
119 target_dir = params['output_data'][0]['extra_files_path']
120 os.makedirs(target_dir)
121 self._output_dir = target_dir
122
123 def write_json_infos(self):
124 """
125 Write in the imported json file
126 """
127 with open(self.json_file_path, 'w') as fh:
128 json.dump(self.get_data_manager(), fh, sort_keys=True)
129
130
131 def parse_arguments():
132 """
133 List of arguments provided by the user
134 return: parsed arguments
135 """
136 # parse options and arguments
137 arg_parser = argparse.ArgumentParser()
138 arg_parser.add_argument("data_manager_json",
139 help="json file from galaxy")
140 arg_parser.add_argument("-v", "--db_version",
141 help="version of the plasmidfinder (latest or 2.1)")
142 return arg_parser.parse_args()
143
144
145 def main():
146 all_args = parse_arguments()
147 plasmidfinder_download = DownloadPlasmidfinderDatabase(json_file_path=all_args.data_manager_json, plasmidfinder_version=all_args.db_version)
148 plasmidfinder_download.read_json_input_file()
149 plasmidfinder_download.download_database()
150 plasmidfinder_download.write_json_infos()
151
152
153 if __name__ == '__main__':
154 main()