Capture python warnings and report them as matches

This change introduces a new way to generate warnings in linter, one
that uses the warnings support in Python. The benefit is that this
allows use to generate warnings from anywhere inside the code without
having to pass them from function to function.

For start we will be using this for internal rules, like ability to
report outdated noqa comments.

src/ansiblelint/runner.py

@@ -5,6 +5,7 @@
 import multiprocessing
 import multiprocessing.pool
 import os
+import warnings
 from collections.abc import Generator
 from dataclasses import dataclass
 from fnmatch import fnmatch
@@ -113,6 +114,27 @@ def is_excluded(self, lintable: Lintable) -> bool:
 
     def run(self) -> list[MatchError]:  # noqa: C901
         """Execute the linting process."""
+        matches: list[MatchError] = []
+        with warnings.catch_warnings(record=True) as captured_warnings:
+            warnings.simplefilter("always")
+            matches = self._run()
+            for warn in captured_warnings:
+                # For the moment we are ignoring deprecation warnings as Ansible
+                # modules outside current content can generate them and user
+                # might not be able to do anything about them.
+                if warn.category is DeprecationWarning:
+                    continue
+                _logger.warning(
+                    "%s:%s %s %s",
+                    warn.filename,
+                    warn.lineno or 1,
+                    warn.category.__name__,
+                    warn.message,
+                )
+        return matches
+
+    def _run(self) -> list[MatchError]:  # noqa: C901
+        """Internal implementation of run."""
         files: list[Lintable] = []
         matches: list[MatchError] = []