Helper tool

Author: mboudet

biotools_cleaner/cleaner.py

@@ -0,0 +1,17 @@
+#!/usr/bin/env python
+
+import argparse
+
+from .tool import Tool
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('input', type=str, help="Path to yaml file")
+    parser.add_argument('output', type=str, help="Output dir (ie, Research-software-ecosystem repository)")
+    parser.add_argument('--dry-run', action='store_true', help="Dry run")
+    parser.add_argument('--cleanup', action='store_true', help="Remove old layout files from repository")
+    args = parser.parse_args()
+
+    tool = Tool(args.input)
+    tool.write_yaml(args.output, dry_run=args.dry_run, remove_input=args.cleanup)

biotools_cleaner/cleaner_batch.py

@@ -0,0 +1,17 @@
+#!/usr/bin/env python
+
+import argparse
+import pathlib
+
+from .tool import Tool
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('rse_repo', type=str, help="Research-software-ecosystem data folder")
+    parser.add_argument('--dry-run', action='store_true', help="Dry run")
+    parser.add_argument('--cleanup', action='store_true', help="Remove old layout files from repository")
+    args = parser.parse_args()
+
+    for path in pathlib.Path(args.rse_repo).rglob("biocontainers.yaml"):
+        tool = Tool(path.name)
+        tool.write_yaml(args.rse_repo, dry_run=args.dry_run, remove_input=args.cleanup)

biotools_cleaner/tool.py

@@ -0,0 +1,35 @@
+import os
+import pathlib
+import logging
+from yaml import safe_load, dump
+
+
+class Tool:
+
+    def __init__(self, tool_yaml):
+        self.yaml_path = tool_yaml
+        self.yaml_data = {}
+
+        with open(tool_yaml, 'r') as f:
+            self.data = safe_load(f)
+
+        logging.info('Processing ' + tool_yaml)
+
+    def write_yaml(self, output_dir, dry_run=False, remove_input=False):
+        if not self.yaml_data.get('software'):
+            logging.error('"software" key not found or empty')
+        if len(self.yaml_data.get('software')) > 1:
+            logging.error('More than one software in yaml file: this should not happen')
+        tool_name = list(self.yaml_data['software'].keys())[0]
+
+        output_path = os.path.join(output_dir, tool_name, '{}.biocontainers.yaml'.format(tool_name))
+
+        logging.info("Moving {} to {}".format(self.yaml_path, output_path))
+
+        if not dry_run:
+            pathlib.Path(os.path.join(output_dir, tool_name)).mkdir(parents=True, exist_ok=True)
+            with open(output_path, 'w') as f:
+                dump(self.yaml_data, f)
+            if remove_input:
+                logging.info("Removing {}".format(self.yaml_path))
+                os.remove(self.yaml_path)