Mercurial > repos > pimarin > amrfinderplus_data_manager_build

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/data_manager/data_manager_build_amrfinderplus.py	Mon Nov 14 15:05:09 2022 +0000
@@ -0,0 +1,107 @@
+import argparse
+import json
+import os
+import re
+import subprocess as sp
+
+from datetime import datetime
+from pathlib import Path
+
+
+class GetDataManager:
+
+    def __init__(self):
+        self.data_table_name = "amrfinderplus_database"
+        self._db_name = "amrfinderplus-db"
+        self._db_path = Path().absolute()
+        self._today = datetime.now().strftime("%Y-%m-%d_%H:%M")
+
+    def get_data_table_format(self):
+        """
+        Skeleton of a data_table format
+        return: a data table formatted for json output
+        """
+        self.data_table_entry = {
+            "data_tables": {
+                self.data_table_name: {}
+            }
+        }
+        return self.data_table_entry
+
+    def get_data_manager(self, amrfinderplus_version):
+        self.amrfinderplus_table_list = self.get_data_table_format()
+
+        data_info = dict(value=self._today,
+                         name=amrfinderplus_version,
+                         path=self._db_name)
+        self.amrfinderplus_table_list["data_tables"][self.data_table_name] = [data_info]
+        return self.amrfinderplus_table_list
+
+    def update_amrfinderplus_db(self):
+        amrfinderplus_db_path = Path(self._db_path).joinpath(self._db_name)
+        cmd = [
+            'amrfinder_update',
+            '--database', str(amrfinderplus_db_path),
+            '--force_update'
+        ]
+        print(cmd)
+        proc = sp.run(
+            cmd,
+            stdout=sp.PIPE,
+            stderr=sp.PIPE,
+            universal_newlines=True
+        )
+        if (proc.returncode != 0):
+            print(
+                f"ERROR: AMRFinderPlus failed! command: 'amrfinder_update --force_update --database {amrfinderplus_db_path}', error code: {proc.returncode}")
+        else:
+            return amrfinderplus_db_path
+
+    def get_amrfinderplus_version(self, amrfinderplus_path):
+        version_file = Path(f'{amrfinderplus_path}/latest/version.txt')
+        with open(version_file, "r") as version:
+            version_value = version.read()
+        version_value = re.sub("\n", "", version_value)
+        return version_value
+
+    def parse_arguments(self):
+        # parse options and arguments
+        arg_parser = argparse.ArgumentParser()
+        arg_parser.add_argument("data_manager_json")
+        arg_parser.add_argument("-t", "--test", action='store_true',
+                                help="option to test the script with an lighted database")
+        return arg_parser.parse_args()
+
+    def read_json_input_file(self, json_file_path):
+        with open(json_file_path) as fh:
+            params = json.load(fh)
+        target_dir = params['output_data'][0]['extra_files_path']
+        os.makedirs(target_dir)
+        return Path(target_dir)
+
+    def write_json_infos(self, json_file_path, data_manager_infos):
+        with open(json_file_path, 'w') as fh:
+            json.dump(data_manager_infos, fh, sort_keys=True)
+
+
+def main():
+    # init the class
+    amrfinderplus_download = GetDataManager()
+    # import the arguments
+    all_args = amrfinderplus_download.parse_arguments()
+    # read the json input from galaxy to define the db path
+    path_to_download = amrfinderplus_download.read_json_input_file(json_file_path=all_args.data_manager_json)
+    # change the path to th json information
+    amrfinderplus_download._db_path = path_to_download
+    # download the last amrfinderplus database
+    amrfinder_output = amrfinderplus_download.update_amrfinderplus_db()
+    # extract the version number of the database
+    amrfinder_version = amrfinderplus_download.get_amrfinderplus_version(amrfinder_output)
+    # make a dic with database information
+    amrfinderplus_json_output = amrfinderplus_download.get_data_manager(amrfinder_version)
+    amrfinderplus_download.write_json_infos(json_file_path=all_args.data_manager_json,
+                                            data_manager_infos=amrfinderplus_json_output)
+
+
+if __name__ == '__main__':
+    main()
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/data_manager/data_manager_build_amrfinderplus.xml	Mon Nov 14 15:05:09 2022 +0000
@@ -0,0 +1,32 @@
+<tool id="data_manager_build_amrfinderplus" name="amrfinderplus_datamanager" tool_type="manage_data" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="@PROFILE@">
+    <description>AMRfinderplus database builder</description>
+    <macros>
+        <import>macro.xml</import>
+    </macros>
+    <expand macro="requirements"/>
+    <command detect_errors="exit_code">
+      <![CDATA[
+        python '$__tool_directory__/data_manager_build_amrfinderplus.py'
+        '$output_file'
+      ]]></command>
+    <inputs>
+        <param name="database_select" type="select" label="Database version">
+            <option value="latest" selected="true">Latest available version</option>
+        </param>
+    </inputs>
+    <outputs>
+        <data name="output_file" format="data_manager_json"/>
+    </outputs>
+    <tests>
+        <!-- Test download -->
+        <test expect_num_outputs="1">
+            <output name="output_file" value="amrfinderplus_test_data_manager.json" />
+        </test>
+    </tests>
+    <help><![CDATA[
+        Download the latest version of amrfinderplus database from the NCBI
+    ]]></help>
+    <citations>
+        <citation type="doi">10.1038/s41598-021-91456-0</citation>
+    </citations>
+</tool>
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/data_manager/macro.xml	Mon Nov 14 15:05:09 2022 +0000
@@ -0,0 +1,13 @@
+<?xml version="1.0"?>
+<macros>
+    <token name="@TOOL_VERSION@">3.10.45</token>
+    <token name="@PYTHON_VERSION@">3.10.6</token>
+    <token name="@VERSION_SUFFIX@">0</token>
+    <token name="@PROFILE@">21.05</token>
+    <xml name="requirements">
+        <requirements>
+            <requirement type="package" version="@TOOL_VERSION@">ncbi-amrfinderplus</requirement>
+            <requirement type="package" version="@PYTHON_VERSION@">python</requirement>
+        </requirements>
+    </xml>
+</macros>
\ No newline at end of file
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/data_manager_conf.xml	Mon Nov 14 15:05:09 2022 +0000
@@ -0,0 +1,19 @@
+<?xml version="1.0"?>
+<data_managers>
+    <data_manager tool_file="data_manager/data_manager_build_amrfinderplus.xml" id="data_manager_build_amrfinderplus" version="@TOOL_VERSION@">
+        <data_table name="amrfinderplus_database">
+            <output>
+                <column name="value" />
+                <column name="name" />
+                <column name="path" output_ref="out_file">
+                    <move type="directory" relativize_symlinks="True">
+                        <source>${path}</source>
+                        <target base="${GALAXY_DATA_MANAGER_DATA_PATH}">amrfinderplus/${value}</target>
+                    </move>
+                    <value_translation>${GALAXY_DATA_MANAGER_DATA_PATH}/amrfinderplus/${value}</value_translation>
+                    <value_translation type="function">abspath</value_translation>
+                </column>
+            </output>
+        </data_table>
+    </data_manager>
+</data_managers>
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/test-data/amrfinderplus.loc	Mon Nov 14 15:05:09 2022 +0000
@@ -0,0 +1,7 @@
+# this is a tab separated file describing the location of amrfinderplus database
+#
+# the columns are:
+# value, name, path
+#
+# for example
+2022-11-10_15:48    2022-10-11.2    ${__HERE__}
\ No newline at end of file
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tool-data/amrfinderplus.loc	Mon Nov 14 15:05:09 2022 +0000
@@ -0,0 +1,7 @@
+# this is a tab separated file describing the location of amrfinderplus database
+#
+# the columns are:
+# value, name, path
+#
+# for example
+2022-11-10_15:48    2022-10-11.2    ${__HERE__}
\ No newline at end of file
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tool_data_table_conf.xml.sample	Mon Nov 14 15:05:09 2022 +0000
@@ -0,0 +1,7 @@
+<tables>
+    <!-- Locations of amrfinderplus database in the required format -->
+    <table name="amrfinderplus_database" comment_char="#">
+        <columns>value, name, path</columns>
+        <file path="tool-data/amrfinderplus.loc" />
+    </table>
+</tables>
\ No newline at end of file
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tool_data_table_conf.xml.test	Mon Nov 14 15:05:09 2022 +0000
@@ -0,0 +1,7 @@
+<tables>
+    <!-- Locations of amrfinderplus database in the required format -->
+    <table name="amrfinderplus_database" comment_char="#">
+        <columns>value, name, path</columns>
+        <file path="${__HERE__}/test-data/amrfinderplus.loc"/>
+    </table>
+</tables>
\ No newline at end of file