Warn about non-UTF-8 files instead of crashing with UnicodeDecodeError

XML/ZCML/PT files should be UTF-8. Instead of silently skipping or
crashing, emit a warning so users can fix the encoding.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: jensens

CHANGES.md

@@ -1,5 +1,12 @@
 # Changelog
 
+## 1.0.0a5 (2026-03-02)
+
+- Warn about non-UTF-8 encoded files instead of crashing with
+  `UnicodeDecodeError`. XML/ZCML/PT files should be UTF-8; the tool
+  now prints a warning and skips the file so users can fix the encoding.
+  [jensens]
+
 ## 1.0.0a4 (2026-03-02)
 
 - Fix Phase 1 failing with "Could not find plone_codemod.import_migrator

src/plone_codemod/pt_migrator.py

@@ -13,6 +13,7 @@
 from pathlib import Path
 from typing import Any
 
+import warnings
 import yaml
 
 
@@ -59,7 +60,17 @@ def _migrate_files(
     """Walk directory, apply transformer to matching files."""
     modified = []
     for filepath in sorted(root.rglob(pattern)):
-        content = filepath.read_text(encoding="utf-8")
+        try:
+            content = filepath.read_text(encoding="utf-8")
+        except UnicodeDecodeError:
+            warnings.warn(
+                f"{filepath} is not UTF-8 encoded — skipping. "
+                f"XML/PT files should be UTF-8. Please fix the encoding.",
+                stacklevel=2,
+            )
+            continue
+        except OSError:
+            continue
         new_content = transformer(content, **kwargs)
         if new_content != content:
             modified.append(filepath)

src/plone_codemod/zcml_migrator.py

@@ -12,6 +12,7 @@
 from pathlib import Path
 from typing import Any
 
+import warnings
 import yaml
 
 
@@ -82,7 +83,17 @@ def migrate_file(
     filepath: Path, transformer: Callable[..., str], **kwargs: Any
 ) -> bool:
     """Read file, apply transformer, write back if changed. Returns True if modified."""
-    content = filepath.read_text(encoding="utf-8")
+    try:
+        content = filepath.read_text(encoding="utf-8")
+    except UnicodeDecodeError:
+        warnings.warn(
+            f"{filepath} is not UTF-8 encoded — skipping. "
+            f"XML/ZCML files should be UTF-8. Please fix the encoding.",
+            stacklevel=2,
+        )
+        return False
+    except OSError:
+        return False
     new_content = transformer(content, **kwargs)
     if new_content != content:
         filepath.write_text(new_content, encoding="utf-8")
@@ -102,7 +113,17 @@ def migrate_zcml_files(
     modified = []
     for zcml_file in sorted(root.rglob("*.zcml")):
         if dry_run:
-            content = zcml_file.read_text(encoding="utf-8")
+            try:
+                content = zcml_file.read_text(encoding="utf-8")
+            except UnicodeDecodeError:
+                warnings.warn(
+                    f"{zcml_file} is not UTF-8 encoded — skipping. "
+                    f"ZCML files should be UTF-8. Please fix the encoding.",
+                    stacklevel=2,
+                )
+                continue
+            except OSError:
+                continue
             new_content = migrate_zcml_content(content, replacements)
             if new_content != content:
                 modified.append(zcml_file)
@@ -139,7 +160,17 @@ def migrate_genericsetup_files(
             continue
 
         if dry_run:
-            content = xml_file.read_text(encoding="utf-8")
+            try:
+                content = xml_file.read_text(encoding="utf-8")
+            except UnicodeDecodeError:
+                warnings.warn(
+                    f"{xml_file} is not UTF-8 encoded — skipping. "
+                    f"XML files should be UTF-8. Please fix the encoding.",
+                    stacklevel=2,
+                )
+                continue
+            except OSError:
+                continue
             new_content = migrate_genericsetup_content(
                 content, replacements, view_replacements
             )