Print UTF-8 warning only when actually encountering non-ascii filenames.

3 files changed, 31 insertions, 11 deletions

diff --git a/meson.py b/meson.py
index 57cc6dc24..13bc870e6 100755
--- a/meson.py
+++ b/meson.py
@@ -18,17 +18,6 @@ from mesonbuild import mesonmain, mesonlib
 import sys, os, locale
 
 def main():
-    # Warn if the locale is not UTF-8. This can cause various unfixable issues
-    # such as os.stat not being able to decode filenames with unicode in them.
-    # There is no way to reset both the preferred encoding and the filesystem
-    # encoding, so we can just warn about it.
-    e = locale.getpreferredencoding()
-    if e.upper() != 'UTF-8' and not mesonlib.is_windows():
-        print('Warning: You are using {!r} which is not a Unicode-compatible '
-              'locale.'.format(e), file=sys.stderr)
-        print('You might see errors if you use UTF-8 strings as '
-              'filenames, as strings, or as file contents.', file=sys.stderr)
-        print('Please switch to a UTF-8 locale for your platform.', file=sys.stderr)
     # Always resolve the command path so Ninja can find it for regen, tests, etc.
     launcher = os.path.realpath(sys.argv[0])
     return mesonmain.run(sys.argv[1:], launcher)
diff --git a/mesonbuild/interpreter.py b/mesonbuild/interpreter.py
index f33d437bf..cbf141374 100644
--- a/mesonbuild/interpreter.py
+++ b/mesonbuild/interpreter.py
@@ -2503,6 +2503,7 @@ to directly access options of other subprojects.''')
     @permittedKwargs(permitted_kwargs['subdir'])
     def func_subdir(self, node, args, kwargs):
         self.validate_arguments(args, 1, [str])
+        mesonlib.check_direntry_issues(args)
         if '..' in args[0]:
             raise InvalidArguments('Subdir contains ..')
         if self.subdir == '' and args[0] == self.subproject_dir:
@@ -2912,6 +2913,7 @@ different subdirectory.
 
     def source_strings_to_files(self, sources):
         results = []
+        mesonlib.check_direntry_issues(sources)
         for s in sources:
             if isinstance(s, (mesonlib.File, GeneratedListHolder,
                               CustomTargetHolder, CustomTargetIndexHolder)):
diff --git a/mesonbuild/mesonlib.py b/mesonbuild/mesonlib.py
index f10a1381e..9ad066821 100644
--- a/mesonbuild/mesonlib.py
+++ b/mesonbuild/mesonlib.py
@@ -19,6 +19,7 @@ import stat
 import time
 import platform, subprocess, operator, os, shutil, re
 import collections
+from mesonbuild import mlog
 
 from glob import glob
 
@@ -59,6 +60,34 @@ else:
     python_command = [sys.executable]
     meson_command = python_command + [detect_meson_py_location()]
 
+def is_ascii_string(astring):
+    try:
+        if isinstance(astring, str):
+            astring.encode('ascii')
+        if isinstance(astring, bytes):
+            astring.decode('ascii')
+    except UnicodeDecodeError:
+        return False
+    return True
+
+def check_direntry_issues(direntry_array):
+    import locale
+    # Warn if the locale is not UTF-8. This can cause various unfixable issues
+    # such as os.stat not being able to decode filenames with unicode in them.
+    # There is no way to reset both the preferred encoding and the filesystem
+    # encoding, so we can just warn about it.
+    e = locale.getpreferredencoding()
+    if e.upper() != 'UTF-8' and not is_windows():
+        if not isinstance(direntry_array, list):
+            direntry_array = [direntry_array]
+        for de in direntry_array:
+            if is_ascii_string(de):
+                continue
+            mlog.warning('''You are using {!r} which is not a Unicode-compatible '
+locale but you are trying to access a file system entry called {!r} which is
+not pure ASCII. This may cause problems.
+'''.format(e, de), file=sys.stderr)
+
 # Put this in objects that should not get dumped to pickle files
 # by accident.
 import threading