modules: sort meta.yml based on keys in schema

CHANGELOG.md

@@ -21,6 +21,8 @@
 
 ### Modules
 
+- sort meta.yml based on keys in schema ([#3958](https://github.com/nf-core/tools/pull/3958))
+
 ### Subworkflows
 
 - Update to new topic version handling ([#3929](https://github.com/nf-core/tools/pull/3929))

nf_core/modules/lint/__init__.py

@@ -9,7 +9,9 @@
 import logging
 import os
 import re
+from collections.abc import Mapping
 from pathlib import Path
+from typing import Any
 
 import questionary
 import rich
@@ -79,6 +81,7 @@ def __init__(
             registry=registry,
             hide_progress=hide_progress,
         )
+        self.meta_schema: Mapping[str, Any] | None = None
 
     def lint(
         self,
@@ -287,6 +290,30 @@ def lint_module(
 
             self.failed += [LintResult(mod, *m) for m in mod.failed]
 
+    def load_meta_schema(self) -> Mapping[str, Any]:
+        """
+        Load the meta.yml JSON schema from the local modules repository cache.
+        The schema is cached in self.meta_schema to avoid reloading.
+
+        Returns:
+            dict: The meta.yml JSON schema
+
+        Raises:
+            LookupError: If the local module cache is not found
+        """
+        # Return cached schema if already loaded
+        if self.meta_schema is not None:
+            return self.meta_schema
+
+        if self.modules_repo.local_repo_dir is None:
+            raise LookupError("Local module cache not found")
+
+        import json
+
+        with open(Path(self.modules_repo.local_repo_dir, "modules/meta-schema.json")) as fh:
+            self.meta_schema = json.load(fh)
+        return self.meta_schema
+
     def update_meta_yml_file(self, mod):
         """
         Update the meta.yml file with the correct inputs and outputs
@@ -324,33 +351,25 @@ def _find_meta_info(meta_yml: dict, element_name: str, is_output=False) -> dict:
             return {}
 
         def _sort_meta_yml(meta_yml: dict) -> dict:
-            """Ensure topics comes after input/output and before authors"""
-            # Early return if no topics to reorder
-            if "topics" not in meta_yml:
-                return meta_yml
-
-            result = {}
-            topics_value = meta_yml["topics"]
-            topics_added = False
-
-            for key, value in meta_yml.items():
-                if key == "topics":
-                    continue  # Skip topics, we'll add it in the right place
-
-                # Add topics before authors key (if not already added)
-                if key == "authors" and not topics_added:
-                    result["topics"] = topics_value
-                    topics_added = True
-
-                result[key] = value
-
-                # Add topics after output (preferred) or after input (if no output)
-                if key == "output":
-                    result["topics"] = topics_value
-                    topics_added = True
-                elif key == "input" and "output" not in meta_yml:
-                    result["topics"] = topics_value
-                    topics_added = True
+            """Sort meta.yml keys according to the schema's property order"""
+            # Get the schema to determine the correct key order
+            try:
+                schema = self.load_meta_schema()
+                schema_keys = list(schema["properties"].keys())
+            except (LintExceptionError, KeyError) as e:
+                raise UserWarning("Failed to load meta schema", e)
+
+            result: dict = {}
+
+            # First, add keys in the order they appear in the schema
+            for key in schema_keys:
+                if key in meta_yml:
+                    result[key] = meta_yml[key]
+
+            # Then add any keys that aren't in the schema (to preserve custom keys)
+            for key in meta_yml.keys():
+                if key not in result:
+                    result[key] = meta_yml[key]
 
             return result
 

nf_core/modules/lint/meta_yml.py

@@ -1,6 +1,8 @@
-import json
+from __future__ import annotations
+
 import logging
 from pathlib import Path
+from typing import TYPE_CHECKING
 
 import ruamel.yaml
 from jsonschema import exceptions, validators
@@ -9,10 +11,13 @@
 from nf_core.components.lint import ComponentLint, LintExceptionError
 from nf_core.components.nfcore_component import NFCoreComponent
 
+if TYPE_CHECKING:
+    from nf_core.modules.lint import ModuleLint
+
 log = logging.getLogger(__name__)
 
 
-def meta_yml(module_lint_object: ComponentLint, module: NFCoreComponent, allow_missing: bool = False) -> None:
+def meta_yml(module_lint_object: ModuleLint, module: NFCoreComponent, allow_missing: bool = False) -> None:
     """
     Lint a ``meta.yml`` file
 
@@ -76,8 +81,7 @@ def meta_yml(module_lint_object: ComponentLint, module: NFCoreComponent, allow_m
     # Confirm that the meta.yml file is valid according to the JSON schema
     valid_meta_yml = False
     try:
-        with open(Path(module_lint_object.modules_repo.local_repo_dir, "modules/meta-schema.json")) as fh:
-            schema = json.load(fh)
+        schema = module_lint_object.load_meta_schema()
         validators.validate(instance=meta_yaml, schema=schema)
         module.passed.append(("meta_yml", "meta_yml_valid", "Module `meta.yml` is valid", module.meta_yml))
         valid_meta_yml = True