Properly raise load-failure when utf-8 decoding fails (#2152)

Author: ssbarnea

.github/workflows/tox.yml

@@ -155,7 +155,7 @@ jobs:
       WSLENV: FORCE_COLOR:PYTEST_REQPASS:TOXENV:TOX_PARALLEL_NO_SPINNER
       # Number of expected test passes, safety measure for accidental skip of
       # tests. Update value if you add/remove tests.
-      PYTEST_REQPASS: 648
+      PYTEST_REQPASS: 649
 
     steps:
       - name: Activate WSL1

examples/broken/encoding.j2

@@ -0,0 +1 @@
+�-�

src/ansiblelint/file_utils.py

@@ -203,6 +203,7 @@ def get(self, key: Any, default: Any = None) -> Any:
             return default
 
     def _populate_content_cache_from_disk(self) -> None:
+        # Can raise UnicodeDecodeError
         self._content = self.path.resolve().read_text(encoding="utf-8")
         if self._original_content is None:
             self._original_content = self._content

src/ansiblelint/rules/__init__.py

@@ -385,13 +385,13 @@ def run(
             try:
                 if file.content is not None:  # loads the file content
                     pass
-            except IOError as exc:
+            except (IOError, UnicodeDecodeError) as exc:
                 return [
                     MatchError(
                         message=str(exc),
                         filename=file,
                         rule=LoadingFailureRule(),
-                        tag=exc.__class__.__name__.lower(),
+                        tag=f"{LoadingFailureRule.id}[{exc.__class__.__name__.lower()}]",
                     )
                 ]
 

test/eco/colsystem.result

@@ -16,7 +16,7 @@ warn_list:  # or 'skip_list' to silence them completely
 
 
 STDOUT:
-.ansible-lint:1: load-failure: [Errno 2] No such file or directory: '~/.cache/ansible-lint-eco/colsystem/tests/ansible-lint.yml' (filenotfounderror)
+.ansible-lint:1: load-failure: [Errno 2] No such file or directory: '~/.cache/ansible-lint-eco/colsystem/tests/ansible-lint.yml' (load-failure[filenotfounderror])
 playbooks/molecule/sudo/molecule.yml:17: yaml: line too long (576 > 160 characters) (line-length)
 roles/authorized_key/meta/main.yml:1: schema: 7 is not one of ['6.1', '7.1', '7.2', 'all'] (schema[meta])
 roles/common/meta/main.yml:1: schema: 2.8 is not of type 'string' (schema[meta])

test/eco/hardening.result

@@ -3,11 +3,12 @@ CMD: ansible-lint -f pep8 -x fqcn-builtins
 RC: 2
 
 STDERR:
-WARNING  Listing 49 violation(s) that are fatal
+WARNING  Listing 50 violation(s) that are fatal
 You can skip specific rules or tags by adding them to your configuration file:
 # .config/ansible-lint.yml
 warn_list:  # or 'skip_list' to silence them completely
   - experimental  # all rules tagged as experimental
+  - load-failure  # Failed to load or parse file.
   - schema  # Perform JSON Schema Validation for known lintable kinds.
 
 
@@ -62,3 +63,4 @@ tasks/systemdconf.yml:1: schema: 'yes' is not of type 'boolean' (schema[tasks])
 tasks/timesyncd.yml:1: schema: 'yes' is not of type 'boolean' (schema[tasks])
 tasks/umask.yml:1: schema: 'yes' is not of type 'boolean' (schema[tasks])
 tasks/users.yml:1: schema: 'yes' is not of type 'boolean' (schema[tasks])
+templates/usr/share/dict/passwords.list.j2:1: load-failure: 'utf-8' codec can't decode byte 0xc3 in position 104779: invalid continuation byte (load-failure[unicodedecodeerror])

test/test_load_failure.py

@@ -0,0 +1,13 @@
+"""Tests for LoadFailureRule."""
+from ansiblelint.rules import RulesCollection
+from ansiblelint.runner import Runner
+
+
+def test_load_failure_encoding(default_rules_collection: RulesCollection) -> None:
+    """Check that we fail when file encoding is wrong."""
+    runner = Runner("examples/broken/encoding.j2", rules=default_rules_collection)
+    matches = runner.run()
+    assert len(matches) == 1, matches
+    assert matches[0].rule.id == "load-failure"
+    assert "'utf-8' codec can't decode byte" in matches[0].message
+    assert matches[0].tag == "load-failure[unicodedecodeerror]"