Parse .mojom as UTF-8.

The parser called on the changed line demands a Python |str| type, but
|contents| here can be a |unicode|. The |str| converter in Python 2
assumes an ASCII encoding and thus throws an exception if any characters
are > 127. This means that any changed .mojo files that contain non-ASCII
characters cause a presubmit error. One can't remove those characters
either because the presubmit parses the previous version of the file
too. Thus all changes to .mojo file with non-ASCII characters are
blocked without this.

Change-Id: I887890590c8ad97a4fac60d19a8bebce439abacc
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2223973
Commit-Queue: Adam Langley <agl@chromium.org>
Auto-Submit: Adam Langley <agl@chromium.org>
Reviewed-by: Ken Rockot <rockot@google.com>
Cr-Commit-Position: refs/heads/master@{#773768}

mojo/public/tools/mojom/check_stable_mojom_compatibility.py

@@ -63,7 +63,7 @@ def _ValidateDelta(root, delta):
       with open(os.path.join(root, mojom)) as f:
         contents = ''.join(f.readlines())
 
-    ast = parser.Parse(str(contents), mojom)
+    ast = parser.Parse(contents.encode('utf-8'), mojom)
     for imp in ast.import_list:
       parseMojom(imp.import_filename, file_overrides, override_modules)